diff mbox

[2/5] mmc: sdhci-msm: Fix HW issue with power IRQ handling during reset

Message ID 1503033582-48703-3-git-send-email-vviswana@codeaurora.org (mailing list archive)
State New, archived
Headers show

Commit Message

Vijay Viswanath Aug. 18, 2017, 5:19 a.m. UTC 
From: Sahitya Tummala <stummala@codeaurora.org>

There is a rare scenario in HW, where the first clear pulse could
be lost when the actual reset and clear/read of status register
are happening at the same time. Fix this by retrying upto 10 times
to ensure the status register gets cleared. Otherwise, this will
lead to a spurious power IRQ which results in system instability.

Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
Signed-off-by: Vijay Viswanath <vviswana@codeaurora.org>
---
 drivers/mmc/host/sdhci-msm.c | 43 ++++++++++++++++++++++++++++++++++++++++---
 1 file changed, 40 insertions(+), 3 deletions(-)

Comments

Adrian Hunter Aug. 24, 2017, 7:42 a.m. UTC | #1 
On 18/08/17 08:19, Vijay Viswanath wrote:
> From: Sahitya Tummala <stummala@codeaurora.org>
> 
> There is a rare scenario in HW, where the first clear pulse could
> be lost when the actual reset and clear/read of status register
> are happening at the same time. Fix this by retrying upto 10 times
> to ensure the status register gets cleared. Otherwise, this will
> lead to a spurious power IRQ which results in system instability.
> 
> Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
> Signed-off-by: Vijay Viswanath <vviswana@codeaurora.org>
> ---
>  drivers/mmc/host/sdhci-msm.c | 43 ++++++++++++++++++++++++++++++++++++++++---
>  1 file changed, 40 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/mmc/host/sdhci-msm.c b/drivers/mmc/host/sdhci-msm.c
> index 0957199..f3e0489 100644
> --- a/drivers/mmc/host/sdhci-msm.c
> +++ b/drivers/mmc/host/sdhci-msm.c
> @@ -995,17 +995,51 @@ static void sdhci_msm_set_uhs_signaling(struct sdhci_host *host,
>  		sdhci_msm_hs400(host, &mmc->ios);
>  }
>  
> -static void sdhci_msm_voltage_switch(struct sdhci_host *host)
> +static void sdhci_msm_dump_pwr_ctrl_regs(struct sdhci_host *host)
> +{
> +	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
> +	struct sdhci_msm_host *msm_host = sdhci_pltfm_priv(pltfm_host);
> +
> +	pr_err("%s: PWRCTL_STATUS: 0x%08x | PWRCTL_MASK: 0x%08x | PWRCTL_CTL: 0x%08x\n",
> +			mmc_hostname(host->mmc),
> +			readl_relaxed(msm_host->core_mem + CORE_PWRCTL_STATUS),
> +			readl_relaxed(msm_host->core_mem + CORE_PWRCTL_MASK),
> +			readl_relaxed(msm_host->core_mem + CORE_PWRCTL_CTL));
> +}
> +
> +static void sdhci_msm_handle_pwr_irq(struct sdhci_host *host, int irq)
>  {
>  	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>  	struct sdhci_msm_host *msm_host = sdhci_pltfm_priv(pltfm_host);
>  	u32 irq_status, irq_ack = 0;
> +	int retry = 10;
>  
>  	irq_status = readl_relaxed(msm_host->core_mem + CORE_PWRCTL_STATUS);
>  	irq_status &= INT_MASK;
>  
>  	writel_relaxed(irq_status, msm_host->core_mem + CORE_PWRCTL_CLEAR);
>  
> +	/*
> +	 * There is a rare HW scenario where the first clear pulse could be
> +	 * lost when actual reset and clear/read of status register is
> +	 * happening at a time. Hence, retry for at least 10 times to make
> +	 * sure status register is cleared. Otherwise, this will result in
> +	 * a spurious power IRQ resulting in system instability.
> +	 */
> +	while (irq_status & readl_relaxed(msm_host->core_mem +
> +				CORE_PWRCTL_STATUS)) {
> +		if (retry == 0) {
> +			pr_err("%s: Timedout clearing (0x%x) pwrctl status register\n",
> +					mmc_hostname(host->mmc), irq_status);
> +			sdhci_msm_dump_pwr_ctrl_regs(host);
> +			WARN_ON(1);

Is it your intention to loop forever here?

> +		}
> +		writel_relaxed(irq_status,
> +				msm_host->core_mem + CORE_PWRCTL_CLEAR);
> +		retry--;
> +		udelay(10);
> +	}
> +
>  	if (irq_status & (CORE_PWRCTL_BUS_ON | CORE_PWRCTL_BUS_OFF))
>  		irq_ack |= CORE_PWRCTL_BUS_SUCCESS;
>  	if (irq_status & (CORE_PWRCTL_IO_LOW | CORE_PWRCTL_IO_HIGH))
> @@ -1017,13 +1051,17 @@ static void sdhci_msm_voltage_switch(struct sdhci_host *host)
>  	 * switches are handled by the sdhci core, so just report success.
>  	 */
>  	writel_relaxed(irq_ack, msm_host->core_mem + CORE_PWRCTL_CTL);
> +
> +	pr_debug("%s: %s: Handled IRQ(%d), irq_status=0x%x, ack=0x%x\n",
> +		mmc_hostname(msm_host->mmc), __func__, irq, irq_status,
> +		irq_ack);
>  }
>  
>  static irqreturn_t sdhci_msm_pwr_irq(int irq, void *data)
>  {
>  	struct sdhci_host *host = (struct sdhci_host *)data;
>  
> -	sdhci_msm_voltage_switch(host);
> +	sdhci_msm_handle_pwr_irq(host, irq);
>  
>  	return IRQ_HANDLED;
>  }
> @@ -1106,7 +1144,6 @@ static void sdhci_msm_set_clock(struct sdhci_host *host, unsigned int clock)
>  	.get_max_clock = sdhci_msm_get_max_clock,
>  	.set_bus_width = sdhci_set_bus_width,
>  	.set_uhs_signaling = sdhci_msm_set_uhs_signaling,
> -	.voltage_switch = sdhci_msm_voltage_switch,
>  };
>  
>  static const struct sdhci_pltfm_data sdhci_msm_pdata = {
>
Vijay Viswanath Aug. 28, 2017, 12:34 p.m. UTC | #2 
On 8/24/2017 1:12 PM, Adrian Hunter wrote:
> On 18/08/17 08:19, Vijay Viswanath wrote:
>> From: Sahitya Tummala <stummala@codeaurora.org>
>>
>> There is a rare scenario in HW, where the first clear pulse could
>> be lost when the actual reset and clear/read of status register
>> are happening at the same time. Fix this by retrying upto 10 times
>> to ensure the status register gets cleared. Otherwise, this will
>> lead to a spurious power IRQ which results in system instability.
>>
>> Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
>> Signed-off-by: Vijay Viswanath <vviswana@codeaurora.org>
>> ---
>>   drivers/mmc/host/sdhci-msm.c | 43 ++++++++++++++++++++++++++++++++++++++++---
>>   1 file changed, 40 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/mmc/host/sdhci-msm.c b/drivers/mmc/host/sdhci-msm.c
>> index 0957199..f3e0489 100644
>> --- a/drivers/mmc/host/sdhci-msm.c
>> +++ b/drivers/mmc/host/sdhci-msm.c
>> @@ -995,17 +995,51 @@ static void sdhci_msm_set_uhs_signaling(struct sdhci_host *host,
>>   		sdhci_msm_hs400(host, &mmc->ios);
>>   }
>>   
>> -static void sdhci_msm_voltage_switch(struct sdhci_host *host)
>> +static void sdhci_msm_dump_pwr_ctrl_regs(struct sdhci_host *host)
>> +{
>> +	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>> +	struct sdhci_msm_host *msm_host = sdhci_pltfm_priv(pltfm_host);
>> +
>> +	pr_err("%s: PWRCTL_STATUS: 0x%08x | PWRCTL_MASK: 0x%08x | PWRCTL_CTL: 0x%08x\n",
>> +			mmc_hostname(host->mmc),
>> +			readl_relaxed(msm_host->core_mem + CORE_PWRCTL_STATUS),
>> +			readl_relaxed(msm_host->core_mem + CORE_PWRCTL_MASK),
>> +			readl_relaxed(msm_host->core_mem + CORE_PWRCTL_CTL));
>> +}
>> +
>> +static void sdhci_msm_handle_pwr_irq(struct sdhci_host *host, int irq)
>>   {
>>   	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>>   	struct sdhci_msm_host *msm_host = sdhci_pltfm_priv(pltfm_host);
>>   	u32 irq_status, irq_ack = 0;
>> +	int retry = 10;
>>   
>>   	irq_status = readl_relaxed(msm_host->core_mem + CORE_PWRCTL_STATUS);
>>   	irq_status &= INT_MASK;
>>   
>>   	writel_relaxed(irq_status, msm_host->core_mem + CORE_PWRCTL_CLEAR);
>>   
>> +	/*
>> +	 * There is a rare HW scenario where the first clear pulse could be
>> +	 * lost when actual reset and clear/read of status register is
>> +	 * happening at a time. Hence, retry for at least 10 times to make
>> +	 * sure status register is cleared. Otherwise, this will result in
>> +	 * a spurious power IRQ resulting in system instability.
>> +	 */
>> +	while (irq_status & readl_relaxed(msm_host->core_mem +
>> +				CORE_PWRCTL_STATUS)) {
>> +		if (retry == 0) {
>> +			pr_err("%s: Timedout clearing (0x%x) pwrctl status register\n",
>> +					mmc_hostname(host->mmc), irq_status);
>> +			sdhci_msm_dump_pwr_ctrl_regs(host);
>> +			WARN_ON(1);
> 
> Is it your intention to loop forever here?
> 
A bad mistake from my side. Will add break here.
>> +		}
>> +		writel_relaxed(irq_status,
>> +				msm_host->core_mem + CORE_PWRCTL_CLEAR);
>> +		retry--;
>> +		udelay(10);
>> +	}
>> +
>>   	if (irq_status & (CORE_PWRCTL_BUS_ON | CORE_PWRCTL_BUS_OFF))
>>   		irq_ack |= CORE_PWRCTL_BUS_SUCCESS;
>>   	if (irq_status & (CORE_PWRCTL_IO_LOW | CORE_PWRCTL_IO_HIGH))
>> @@ -1017,13 +1051,17 @@ static void sdhci_msm_voltage_switch(struct sdhci_host *host)
>>   	 * switches are handled by the sdhci core, so just report success.
>>   	 */
>>   	writel_relaxed(irq_ack, msm_host->core_mem + CORE_PWRCTL_CTL);
>> +
>> +	pr_debug("%s: %s: Handled IRQ(%d), irq_status=0x%x, ack=0x%x\n",
>> +		mmc_hostname(msm_host->mmc), __func__, irq, irq_status,
>> +		irq_ack);
>>   }
>>   
>>   static irqreturn_t sdhci_msm_pwr_irq(int irq, void *data)
>>   {
>>   	struct sdhci_host *host = (struct sdhci_host *)data;
>>   
>> -	sdhci_msm_voltage_switch(host);
>> +	sdhci_msm_handle_pwr_irq(host, irq);
>>   
>>   	return IRQ_HANDLED;
>>   }
>> @@ -1106,7 +1144,6 @@ static void sdhci_msm_set_clock(struct sdhci_host *host, unsigned int clock)
>>   	.get_max_clock = sdhci_msm_get_max_clock,
>>   	.set_bus_width = sdhci_set_bus_width,
>>   	.set_uhs_signaling = sdhci_msm_set_uhs_signaling,
>> -	.voltage_switch = sdhci_msm_voltage_switch,
>>   };
>>   
>>   static const struct sdhci_pltfm_data sdhci_msm_pdata = {
>>
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
diff mbox

Patch

diff --git a/drivers/mmc/host/sdhci-msm.c b/drivers/mmc/host/sdhci-msm.c
index 0957199..f3e0489 100644
--- a/drivers/mmc/host/sdhci-msm.c
+++ b/drivers/mmc/host/sdhci-msm.c
@@ -995,17 +995,51 @@  static void sdhci_msm_set_uhs_signaling(struct sdhci_host *host,
 		sdhci_msm_hs400(host, &mmc->ios);
 }
 
-static void sdhci_msm_voltage_switch(struct sdhci_host *host)
+static void sdhci_msm_dump_pwr_ctrl_regs(struct sdhci_host *host)
+{
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct sdhci_msm_host *msm_host = sdhci_pltfm_priv(pltfm_host);
+
+	pr_err("%s: PWRCTL_STATUS: 0x%08x | PWRCTL_MASK: 0x%08x | PWRCTL_CTL: 0x%08x\n",
+			mmc_hostname(host->mmc),
+			readl_relaxed(msm_host->core_mem + CORE_PWRCTL_STATUS),
+			readl_relaxed(msm_host->core_mem + CORE_PWRCTL_MASK),
+			readl_relaxed(msm_host->core_mem + CORE_PWRCTL_CTL));
+}
+
+static void sdhci_msm_handle_pwr_irq(struct sdhci_host *host, int irq)
 {
 	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
 	struct sdhci_msm_host *msm_host = sdhci_pltfm_priv(pltfm_host);
 	u32 irq_status, irq_ack = 0;
+	int retry = 10;
 
 	irq_status = readl_relaxed(msm_host->core_mem + CORE_PWRCTL_STATUS);
 	irq_status &= INT_MASK;
 
 	writel_relaxed(irq_status, msm_host->core_mem + CORE_PWRCTL_CLEAR);
 
+	/*
+	 * There is a rare HW scenario where the first clear pulse could be
+	 * lost when actual reset and clear/read of status register is
+	 * happening at a time. Hence, retry for at least 10 times to make
+	 * sure status register is cleared. Otherwise, this will result in
+	 * a spurious power IRQ resulting in system instability.
+	 */
+	while (irq_status & readl_relaxed(msm_host->core_mem +
+				CORE_PWRCTL_STATUS)) {
+		if (retry == 0) {
+			pr_err("%s: Timedout clearing (0x%x) pwrctl status register\n",
+					mmc_hostname(host->mmc), irq_status);
+			sdhci_msm_dump_pwr_ctrl_regs(host);
+			WARN_ON(1);
+		}
+		writel_relaxed(irq_status,
+				msm_host->core_mem + CORE_PWRCTL_CLEAR);
+		retry--;
+		udelay(10);
+	}
+
 	if (irq_status & (CORE_PWRCTL_BUS_ON | CORE_PWRCTL_BUS_OFF))
 		irq_ack |= CORE_PWRCTL_BUS_SUCCESS;
 	if (irq_status & (CORE_PWRCTL_IO_LOW | CORE_PWRCTL_IO_HIGH))
@@ -1017,13 +1051,17 @@  static void sdhci_msm_voltage_switch(struct sdhci_host *host)
 	 * switches are handled by the sdhci core, so just report success.
 	 */
 	writel_relaxed(irq_ack, msm_host->core_mem + CORE_PWRCTL_CTL);
+
+	pr_debug("%s: %s: Handled IRQ(%d), irq_status=0x%x, ack=0x%x\n",
+		mmc_hostname(msm_host->mmc), __func__, irq, irq_status,
+		irq_ack);
 }
 
 static irqreturn_t sdhci_msm_pwr_irq(int irq, void *data)
 {
 	struct sdhci_host *host = (struct sdhci_host *)data;
 
-	sdhci_msm_voltage_switch(host);
+	sdhci_msm_handle_pwr_irq(host, irq);
 
 	return IRQ_HANDLED;
 }
@@ -1106,7 +1144,6 @@  static void sdhci_msm_set_clock(struct sdhci_host *host, unsigned int clock)
 	.get_max_clock = sdhci_msm_get_max_clock,
 	.set_bus_width = sdhci_set_bus_width,
 	.set_uhs_signaling = sdhci_msm_set_uhs_signaling,
-	.voltage_switch = sdhci_msm_voltage_switch,
 };
 
 static const struct sdhci_pltfm_data sdhci_msm_pdata = {