Re: mmc: dw_mmc: release hardware lock error

[Dongjin Kim <dongjin.kim@agreeyamobility.net>]

Hi Jaehoon,

Sorry for the inconvenience, here is the patch line-wrapping fixed.

-------------------->8-------------------->8--------------------
diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
index 2f7f5ee..05ccee8 100644
--- a/drivers/mmc/host/dw_mmc.c
+++ b/drivers/mmc/host/dw_mmc.c
@@ -520,15 +520,33 @@ static void mci_send_cmd(struct dw_mci_slot *slot,
 	struct dw_mci *host = slot->host;
 	unsigned long timeout = jiffies + msecs_to_jiffies(500);
 	unsigned int cmd_status = 0;
+	u32 ctrl;
 
+resend:
 	mci_writel(host, CMDARG, arg);
 	wmb();
 	mci_writel(host, CMD, SDMMC_CMD_START | cmd);
+
 	while (time_before(jiffies, timeout)) {
 		cmd_status = mci_readl(host, CMD);
 		if (!(cmd_status & SDMMC_CMD_START))
 			return;
+
+		if (host->cmd_status & SDMMC_INT_HLE) {
+			dev_err(&slot->mmc->class_dev,
+			"Hardware locked error (cmd %#x arg %#x)\n",
+			cmd, arg);
+
+			ctrl = mci_readl(host, CTRL);
+			ctrl |= SDMMC_CTRL_RESET;
+			mci_writel(host, CTRL, ctrl);
+
+			host->cmd_status &= ~SDMMC_INT_HLE;
+
+			goto resend;
+		}
 	}
+
 	dev_err(&slot->mmc->class_dev,
 		"Timeout sending command (cmd %#x arg %#x status %#x)\n",
 		cmd, arg, cmd_status);
@@ -1429,6 +1447,12 @@ static irqreturn_t dw_mci_interrupt(int irq, void
 		if (!pending)
 			break;
 
+		if (pending & SDMMC_INT_HLE) {
+			mci_writel(host, RINTSTS, SDMMC_INT_HLE);
+			host->cmd_status = status;
+			smp_wmb();
+		}
+
 		if (pending & DW_MCI_CMD_ERROR_FLAGS) {
 			mci_writel(host, RINTSTS, DW_MCI_CMD_ERROR_FLAGS);
 			host->cmd_status = status;
-------------------->8-------------------->8--------------------

On Tue, 28 Feb 2012 16:44:08 +0900, Jaehoon Chung wrote:

> Hi Dongjin.
> 
> This patch is corrupt due to line-wrapping.
> 
> Best Regards,
> Jaehoon Chung
> 
> On 02/28/2012 04:31 PM, Dongjin Kim wrote:
> 
>> Jaehoon Chung <jh80.chung <at> samsung.com> writes:
>> 
>> 
>>> On 02/28/2012 02:32 PM, Dongjin Kim wrote:
>>>
>>>> Hello,
>>>>
>>>> I encountered the problem that my hardware is frozen while switching
>>>> clock
>> from
>>>> 400kHz to others. This is caused that hardware locked error was
>>>> triggered
>> when
>>>> CMD register was written and interrupt was pended. Tested IP version
>>>> is
>> 2.10a
>>>>
>>>>
>>> Hi Dongjin.
>>>
>>> It seems like this problem. Can you check this patch?
>>> http://www.spinics.net/lists/linux-mmc/msg09895.html
>>>
>>> Best Regards,
>>> Jaehoon Chung
>>>
>> Hi Jaehoon,
>> 
>> Thank you for your feedback.
>> I had tested the patch from the link but the result is same on my
>> hardware.
>> 
>> This is the log message what I had before, the hardware hanged after
>> this log when mmc card is inserted.
>> 
>> [   62.450000] end_request: I/O error, dev mmcblk0, sector 7585 [  
>> 62.460000] mmcblk0: error -123 sending read/write command, response
>> 0x0, card status 0x0
>> [   62.470000] end_request: I/O error, dev mmcblk0, sector 7586 [  
>> 62.470000] mmcblk0: error -123 sending read/write command, response
>> 0x0, card status 0x0
>> [   62.480000] end_request: I/O error, dev mmcblk0, sector 7587 [  
>> 62.490000] mmcblk0: error -123 sending read/write command, response
>> 0x0, card status 0x0
>> [   62.490000] end_request: I/O error, dev mmcblk0, sector 7588 [  
>> 62.500000] mmcblk0: error -123 sending read/write command, response
>> 0x0, card status 0x0
>> [   62.510000] end_request: I/O error, dev mmcblk0, sector 7589 [  
>> 62.510000] Buffer I/O error on device mmcblk0p1, logical block 940 [  
>> 63.770000] mmc_host mmc1: Bus speed (slot 0) = 50000000Hz (slot req
>> 400000Hz, actual 403225HZ div = 62)
>> [   64.270000] mmc_host mmc1: Timeout sending command (cmd 0x202000 arg
>> 0x0 status 0x80202000)
>> 
>> In normal case, the log message have to go through as below and usually
>> it is, except on certain case.
>> 
>> [   80.620000] mmcblk0: error -123 sending read/write command, response
>> 0x0, card status 0x0
>> [   80.620000] mmcblk0: error -123 requesting status [   80.620000]
>> end_request: I/O error, dev mmcblk0, sector 7614 [   82.930000]
>> mmc_host mmc1: Bus speed (slot 0) = 50000000Hz (slot req 400000Hz,
>> actual 403225HZ div = 62)
>> [   83.730000] mmc_host mmc1: Bus speed (slot 0) = 50000000Hz (slot req
>> 50000000Hz, actual 25000000HZ div = 1) [   83.750000] mmc1: new high
>> speed SD card at address 0007 [   83.760000] mmcblk0: mmc1:0007 SD02G
>> 1.84 GiB [   83.780000]  mmcblk0: p1
>> 
>> Below is the log message with my change to reset the bit, Hardware
>> Locked Error. This error is not happening always, I was testing the
>> driver removing mmc card without unmount.
>> 
>> [  118.310000] end_request: I/O error, dev mmcblk1, sector 1108 [ 
>> 118.310000] mmcblk1: error -123 sending read/write command, response
>> 0x0, card status 0x0
>> [  118.310000] end_request: I/O error, dev mmcblk1, sector 1109 [ 
>> 124.190000] mmc_host mmc1: Bus speed (slot 0) = 50000000Hz (slot req
>> 400000Hz, actual 403225HZ div = 62)
>> [  124.690000] mmc_host mmc1: Timeout sending command (cmd 0x202000 arg
>> 0x0 status 0x80202000)
>> [  124.690000] mmc_host mmc1: Hardware locked error (cmd 0x202000 arg
>> 0x0) [  125.550000] mmc_host mmc1: Bus speed (slot 0) = 50000000Hz
>> (slot req 50000000Hz, actual 25000000HZ div = 1) [  125.560000] mmc1:
>> new high speed SD card at address 0007 [  125.560000] mmcblk0:
>> mmc1:0007 SD02G 1.84 GiB [  125.570000]  mmcblk0: p1
>> 
>> Cheers,
>> Dongjin.
>> 
>>>> Cheers,
>>>> Dongjin.
>>>>
>>>> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>>>> index 2f7f5ee..05ccee8 100644
>>>> --- a/drivers/mmc/host/dw_mmc.c
>>>> +++ b/drivers/mmc/host/dw_mmc.c
>>>> @@ -520,15 +520,33 @@ static void mci_send_cmd(struct dw_mci_slot
>>>> *slot, u32 cmd, u32 arg)
>>>>  	struct dw_mci *host = slot->host;
>>>>  	unsigned long timeout = jiffies + msecs_to_jiffies(500); unsigned
>>>>  	int cmd_status = 0;
>>>> +	u32 ctrl;
>>>>  
>>>> +resend:
>>>>  	mci_writel(host, CMDARG, arg);
>>>>  	wmb();
>>>>  	mci_writel(host, CMD, SDMMC_CMD_START | cmd);
>>>> +
>>>>  	while (time_before(jiffies, timeout)) {
>>>>  		cmd_status = mci_readl(host, CMD);
>>>>  		if (!(cmd_status & SDMMC_CMD_START))
>>>>  			return;
>>>> +
>>>> +		if (host->cmd_status & SDMMC_INT_HLE) { +		
>>>> dev_err(&slot->mmc->class_dev,
>>>> +				"Hardware locked error (cmd %#x arg %#x)
\n", +				cmd, arg);
>>>> +
>>>> +			ctrl = mci_readl(host, CTRL);
>>>> +			ctrl |= SDMMC_CTRL_RESET;
>>>> +			mci_writel(host, CTRL, ctrl);
>>>> +
>>>> +			host->cmd_status &= ~SDMMC_INT_HLE; +
>>>> +			goto resend;
>>>> +		}
>>>>  	}
>>>> +
>>>>  	dev_err(&slot->mmc->class_dev,
>>>>  		"Timeout sending command (cmd %#x arg %#x status %#x)\n", 
cmd,
>>>>  		arg, cmd_status);
>>>> @@ -1429,6 +1447,12 @@ static irqreturn_t dw_mci_interrupt(int irq,
>>>> void *dev_id)
>>>>  		if (!pending)
>>>>  			break;
>>>>  
>>>> +		if (pending & SDMMC_INT_HLE) {
>>>> +			mci_writel(host, RINTSTS, SDMMC_INT_HLE); 
+			host->cmd_status =
>>>> status;
>>>> +			smp_wmb();
>>>> +		}
>>>> +
>>>>  		if (pending & DW_MCI_CMD_ERROR_FLAGS) {
>>>>  			mci_writel(host, RINTSTS, DW_MCI_CMD_ERROR_FLAGS);
>>>>  			host->cmd_status = status;
>>>>
>>>> --
>>>> To unsubscribe from this list: send the line "unsubscribe linux-mmc"
>>>> in the body of a message to majordomo <at> vger.kernel.org More
>>>> majordomo info at  http://vger.kernel.org/majordomo-info.html
>>>>
>>>>
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux-mmc"
>>> in the body of a message to majordomo <at> vger.kernel.org More
>>> majordomo info at  http://vger.kernel.org/majordomo-info.html
>>>
>>>
>>>
>> 
>> 
>> 
>> --
>> 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
>>