lkml.org 
[lkml]   [2021]   [Dec]   [2]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
SubjectRe: [PATCH V1] mmc: sdhci-msm: Add eMMC and SD card err_stat sysfs entry
From
Date
On 02/12/2021 08:42, Sajida Bhanu (Temp) (QUIC) wrote:
> Gentle Reminder.
>
> Thanks,
> Sajida
> -----Original Message-----
> From: Sajida Bhanu (Temp) (QUIC) <quic_c_sbhanu@quicinc.com>
> Sent: Friday, November 26, 2021 10:54 AM
> To: Ulf Hansson <ulf.hansson@linaro.org>; Sajida Bhanu (Temp) (QUIC) <quic_c_sbhanu@quicinc.com>
> Cc: adrian.hunter@intel.com; riteshh@codeaurora.org; Asutosh Das (asd) <asutoshd@quicinc.com>; agross@kernel.org; bjorn.andersson@linaro.org; linux-mmc@vger.kernel.org; linux-arm-msm@vger.kernel.org; linux-kernel@vger.kernel.org; stummala@codeaurora.org; vbadigan@codeaurora.org; Ram Prakash Gupta (QUIC) <quic_rampraka@quicinc.com>; Pradeep Pragallapati (QUIC) <quic_pragalla@quicinc.com>; sartgarg@codeaurora.org; nitirawa@codeaurora.org; sayalil@codeaurora.org
> Subject: RE: [PATCH V1] mmc: sdhci-msm: Add eMMC and SD card err_stat sysfs entry
>
>
>
> -----Original Message-----
> From: Ulf Hansson <ulf.hansson@linaro.org>
> Sent: Thursday, November 25, 2021 5:01 PM
> To: Sajida Bhanu (Temp) (QUIC) <quic_c_sbhanu@quicinc.com>
> Cc: adrian.hunter@intel.com; riteshh@codeaurora.org; Asutosh Das (asd) <asutoshd@quicinc.com>; agross@kernel.org; bjorn.andersson@linaro.org; linux-mmc@vger.kernel.org; linux-arm-msm@vger.kernel.org; linux-kernel@vger.kernel.org; stummala@codeaurora.org; vbadigan@codeaurora.org; Ram Prakash Gupta (QUIC) <quic_rampraka@quicinc.com>; Pradeep Pragallapati (QUIC) <quic_pragalla@quicinc.com>; sartgarg@codeaurora.org; nitirawa@codeaurora.org; sayalil@codeaurora.org
> Subject: Re: [PATCH V1] mmc: sdhci-msm: Add eMMC and SD card err_stat sysfs entry
>
> On Wed, 17 Nov 2021 at 07:20, Shaik Sajida Bhanu <quic_c_sbhanu@quicinc.com> wrote:
>>
>> Add sysfs entry to query eMMC and SD card errors statistics.
>> This feature is useful for debug and testing.
>>
>> Signed-off-by: Shaik Sajida Bhanu <quic_c_sbhanu@quicinc.com>
>> ---
>> drivers/mmc/host/cqhci.h | 1 +
>> drivers/mmc/host/sdhci-msm.c | 192 +++++++++++++++++++++++++++++++++++++++++++
>> drivers/mmc/host/sdhci.c | 17 ++++
>> drivers/mmc/host/sdhci.h | 1 +
>> include/linux/mmc/host.h | 1 +
>> 5 files changed, 212 insertions(+)
>
> To allow an easier review, I strongly suggest splitting up the changes in smaller pieces. Maybe in three steps: add interfaces, implement them, add sysfs - or something along those lines.
>
> I am also trying to understand the benefit of having these stats available. Can you perhaps share a little bit of background on how they are usable for you? Is it for debug purpose or does it have other purposes too?
>
> If it turns out that we agree on finding these stats valuable for us, then I am inclined to think that this should be integrated closer with the mmc core.
>
> The error codes that are propagated to the core from failed requests are common error codes, so we should be able to use that information from the core itself, I think.
>
> Kind regards
> Uffe
>
> Hi Ulf,
>
> Thanks for the review
>
> I am also trying to understand the benefit of having these stats available. Can you perhaps share a little bit of background on how they are usable for you? Is it for debug purpose or does it have other purposes too?
>
>>>>>>>>>>>>> These changes for debug purpose only .. if any errors occurred while testing eMMC and SD card those will be captured in these sysfs entries , we can directly go and check the sysfs entries and get to know what is the error occurred in driver level.

I would suggest using debugfs and adding support in mmc host debugfs
e.g.

static inline void mmc_debugfs_err_stats_enable(struct mmc_host *mmc)
{
mmc->err_stats_enabled = true;
}

enum mmc_err_stat {
MMC_ERR_CMD_TIMEOUT,
MMC_ERR_CMD_CRC,
MMC_ERR_DAT_TIMEOUT,
MMC_ERR_DAT_CRC,
MMC_ERR_AUTO_CMD,
MMC_ERR_ADMA,
MMC_ERR_TUNING,
MMC_ERR_CMDQ_RED,
MMC_ERR_CMDQ_GCE,
MMC_ERR_CMDQ_ICCE,
MMC_ERR_REQ_TIMEOUT,
MMC_ERR_CMDQ_REQ_TIMEOUT,
MMC_ERR_ICE_CFG,
MMC_ERR_MAX,
};

static inline void mmc_debugfs_err_stats_inc(struct mmc_host *mmc, enum mmc_err_stat stat)
{
mmc->err_stats[stat] += 1;
}

And amend mmc_add_host_debugfs() to create the err_stats file etc.

sdhci.c could call mmc_debugfs_err_stats_enable() and mmc_debugfs_err_stats_inc() as needed.
cqhci.c could call mmc_debugfs_err_stats_inc() as needed.

Ulf, does that sound reasonable?

>
> Thanks,
> Sajida
>>
>> diff --git a/drivers/mmc/host/cqhci.h b/drivers/mmc/host/cqhci.h index
>> ba9387e..f72a1d6 100644
>> --- a/drivers/mmc/host/cqhci.h
>> +++ b/drivers/mmc/host/cqhci.h
>> @@ -286,6 +286,7 @@ struct cqhci_host_ops {
>> u64 *data);
>> void (*pre_enable)(struct mmc_host *mmc);
>> void (*post_disable)(struct mmc_host *mmc);
>> + void (*err_stats)(struct mmc_host *mmc, unsigned long flags);
>> #ifdef CONFIG_MMC_CRYPTO
>> int (*program_key)(struct cqhci_host *cq_host,
>> const union cqhci_crypto_cfg_entry *cfg,
>> int slot); diff --git a/drivers/mmc/host/sdhci-msm.c
>> b/drivers/mmc/host/sdhci-msm.c index 50c71e0..e1dcd2d 100644
>> --- a/drivers/mmc/host/sdhci-msm.c
>> +++ b/drivers/mmc/host/sdhci-msm.c
>> @@ -242,6 +242,23 @@ struct sdhci_msm_variant_ops {
>> u32 offset);
>> };
>>
>> +enum {
>> + MMC_ERR_CMD_TIMEOUT,
>> + MMC_ERR_CMD_CRC,
>> + MMC_ERR_DAT_TIMEOUT,
>> + MMC_ERR_DAT_CRC,
>> + MMC_ERR_AUTO_CMD,
>> + MMC_ERR_ADMA,
>> + MMC_ERR_TUNING,
>> + MMC_ERR_CMDQ_RED,
>> + MMC_ERR_CMDQ_GCE,
>> + MMC_ERR_CMDQ_ICCE,
>> + MMC_ERR_REQ_TIMEOUT,
>> + MMC_ERR_CMDQ_REQ_TIMEOUT,
>> + MMC_ERR_ICE_CFG,
>> + MMC_ERR_MAX,
>> +};
>> +
>> /*
>> * From V5, register spaces have changed. Wrap this info in a structure
>> * and choose the data_structure based on version info mentioned in DT.
>> @@ -285,6 +302,8 @@ struct sdhci_msm_host {
>> u32 dll_config;
>> u32 ddr_config;
>> bool vqmmc_enabled;
>> + bool err_occurred;
>> + u32 err_stats[MMC_ERR_MAX];
>> };
>>
>> static const struct sdhci_msm_offset *sdhci_priv_msm_offset(struct
>> sdhci_host *host) @@ -2106,9 +2125,20 @@ static void sdhci_msm_set_timeout(struct sdhci_host *host, struct mmc_command *c
>> host->data_timeout = 22LL * NSEC_PER_SEC; }
>>
>> +void sdhci_msm_cqe_err_stats(struct mmc_host *mmc, unsigned long
>> +flags) {
>> + struct sdhci_host *host = mmc_priv(mmc);
>> + struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>> + struct sdhci_msm_host *msm_host =
>> +sdhci_pltfm_priv(pltfm_host);
>> +
>> + if (flags & BIT(0))
>> + msm_host->err_stats[MMC_ERR_CMDQ_REQ_TIMEOUT]++;
>> +}
>> +
>> static const struct cqhci_host_ops sdhci_msm_cqhci_ops = {
>> .enable = sdhci_msm_cqe_enable,
>> .disable = sdhci_msm_cqe_disable,
>> + .err_stats = sdhci_msm_cqe_err_stats,
>> #ifdef CONFIG_MMC_CRYPTO
>> .program_key = sdhci_msm_program_key,
>> #endif
>> @@ -2403,6 +2433,46 @@ static void sdhci_msm_dump_vendor_regs(struct sdhci_host *host)
>> readl_relaxed(host->ioaddr +
>> msm_offset->core_vendor_spec_func2),
>> readl_relaxed(host->ioaddr +
>> msm_offset->core_vendor_spec3));
>> + msm_host->err_occurred = true; }
>> +
>> +void sdhci_msm_err_stats(struct sdhci_host *host, u32 intmask) {
>> + int command;
>> + struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>> + struct sdhci_msm_host *msm_host =
>> +sdhci_pltfm_priv(pltfm_host);
>> +
>> + if (!msm_host->err_occurred)
>> + msm_host->err_stats[MMC_ERR_CMD_TIMEOUT] = 0;
>> +
>> + if (host && host->mmc && host->mmc->timer) {
>> + msm_host->err_stats[MMC_ERR_REQ_TIMEOUT]++;
>> + host->mmc->timer = false;
>> + }
>> +
>> + if (intmask & SDHCI_INT_CRC) {
>> + command = SDHCI_GET_CMD(sdhci_readw(host, SDHCI_COMMAND));
>> + if (command != MMC_SEND_TUNING_BLOCK ||
>> + command != MMC_SEND_TUNING_BLOCK_HS200)
>> + msm_host->err_stats[MMC_ERR_CMD_CRC]++;
>> + } else if ((intmask & SDHCI_INT_TIMEOUT) ||
>> + (intmask & SDHCI_INT_DATA_TIMEOUT))
>> + msm_host->err_stats[MMC_ERR_CMD_TIMEOUT]++;
>> + else if (intmask & SDHCI_INT_DATA_CRC) {
>> + command = SDHCI_GET_CMD(sdhci_readw(host, SDHCI_COMMAND));
>> + if (command != MMC_SEND_TUNING_BLOCK ||
>> + command != MMC_SEND_TUNING_BLOCK_HS200)
>> + msm_host->err_stats[MMC_ERR_DAT_CRC]++;
>> + } else if (intmask & SDHCI_INT_DATA_TIMEOUT)
>> + msm_host->err_stats[MMC_ERR_DAT_TIMEOUT]++;
>> + else if (intmask & CQHCI_IS_RED)
>> + msm_host->err_stats[MMC_ERR_CMDQ_RED]++;
>> + else if (intmask & CQHCI_IS_GCE)
>> + msm_host->err_stats[MMC_ERR_CMDQ_GCE]++;
>> + else if (intmask & CQHCI_IS_ICCE)
>> + msm_host->err_stats[MMC_ERR_CMDQ_ICCE]++;
>> + else if (intmask & SDHCI_INT_ADMA_ERROR)
>> + msm_host->err_stats[MMC_ERR_ADMA]++;
>> }
>>
>> static const struct sdhci_msm_variant_ops mci_var_ops = { @@ -2456,6
>> +2526,7 @@ static const struct sdhci_ops sdhci_msm_ops = {
>> .dump_vendor_regs = sdhci_msm_dump_vendor_regs,
>> .set_power = sdhci_set_power_noreg,
>> .set_timeout = sdhci_msm_set_timeout,
>> + .err_stats = sdhci_msm_err_stats,
>> };
>>
>> static const struct sdhci_pltfm_data sdhci_msm_pdata = { @@ -2482,6
>> +2553,125 @@ static inline void sdhci_msm_get_of_property(struct
>> +platform_device *pdev,
>> of_property_read_u32(node, "qcom,dll-config",
>> &msm_host->dll_config); }
>>
>> +static ssize_t err_state_show(struct device *dev,
>> + struct device_attribute *attr, char *buf) {
>> + struct sdhci_host *host = dev_get_drvdata(dev);
>> + struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>> + struct sdhci_msm_host *msm_host =
>> +sdhci_pltfm_priv(pltfm_host);
>> +
>> + if (!host || !host->mmc)
>> + return -EINVAL;
>> +
>> + return scnprintf(buf, PAGE_SIZE, "%d\n",
>> +!!msm_host->err_occurred); }
>> +
>> +static ssize_t err_state_store(struct device *dev,
>> + struct device_attribute *attr,
>> + const char *buf, size_t count) {
>> + struct sdhci_host *host = dev_get_drvdata(dev);
>> + struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>> + struct sdhci_msm_host *msm_host = sdhci_pltfm_priv(pltfm_host);
>> + unsigned int val;
>> +
>> + if (kstrtouint(buf, 0, &val))
>> + return -EINVAL;
>> + if (!host || !host->mmc)
>> + return -EINVAL;
>> +
>> + msm_host->err_occurred = !!val;
>> + if (!val)
>> + memset(msm_host->err_stats, 0,
>> + sizeof(msm_host->err_stats));
>> +
>> + return count;
>> +}
>> +static DEVICE_ATTR_RW(err_state);
>> +
>> +static ssize_t err_stats_show(struct device *dev,
>> + struct device_attribute *attr, char
>> +*buf) {
>> + struct sdhci_host *host = dev_get_drvdata(dev);
>> + struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>> + struct sdhci_msm_host *msm_host = sdhci_pltfm_priv(pltfm_host);
>> + char tmp[100];
>> +
>> + if (!host || !host->mmc)
>> + return -EINVAL;
>> +
>> + scnprintf(tmp, sizeof(tmp), "# Command Timeout Error: %d\n",
>> + msm_host->err_stats[MMC_ERR_CMD_TIMEOUT]);
>> + strlcpy(buf, tmp, PAGE_SIZE);
>> +
>> + scnprintf(tmp, sizeof(tmp), "# Command CRC Error: %d\n",
>> + msm_host->err_stats[MMC_ERR_CMD_CRC]);
>> + strlcat(buf, tmp, PAGE_SIZE);
>> +
>> + scnprintf(tmp, sizeof(tmp), "# Data Timeout Error: %d\n",
>> + msm_host->err_stats[MMC_ERR_DAT_TIMEOUT]);
>> + strlcat(buf, tmp, PAGE_SIZE);
>> +
>> + scnprintf(tmp, sizeof(tmp), "# Data CRC Error: %d\n",
>> + msm_host->err_stats[MMC_ERR_DAT_CRC]);
>> + strlcat(buf, tmp, PAGE_SIZE);
>> +
>> + scnprintf(tmp, sizeof(tmp), "# Auto-Cmd Error: %d\n",
>> + msm_host->err_stats[MMC_ERR_ADMA]);
>> + strlcat(buf, tmp, PAGE_SIZE);
>> +
>> + scnprintf(tmp, sizeof(tmp), "# ADMA Error: %d\n",
>> + msm_host->err_stats[MMC_ERR_ADMA]);
>> + strlcat(buf, tmp, PAGE_SIZE);
>> +
>> + scnprintf(tmp, sizeof(tmp), "# Tuning Error: %d\n",
>> + msm_host->err_stats[MMC_ERR_TUNING]);
>> + strlcat(buf, tmp, PAGE_SIZE);
>> +
>> + scnprintf(tmp, sizeof(tmp), "# CMDQ RED Errors: %d\n",
>> + msm_host->err_stats[MMC_ERR_CMDQ_RED]);
>> + strlcat(buf, tmp, PAGE_SIZE);
>> +
>> + scnprintf(tmp, sizeof(tmp), "# CMDQ GCE Errors: %d\n",
>> + msm_host->err_stats[MMC_ERR_CMDQ_GCE]);
>> + strlcat(buf, tmp, PAGE_SIZE);
>> +
>> + scnprintf(tmp, sizeof(tmp), "# CMDQ ICCE Errors: %d\n",
>> + msm_host->err_stats[MMC_ERR_CMDQ_ICCE]);
>> + strlcat(buf, tmp, PAGE_SIZE);
>> +
>> + scnprintf(tmp, sizeof(tmp), "# CMDQ Request Timedout: %d\n",
>> + msm_host->err_stats[MMC_ERR_CMDQ_REQ_TIMEOUT]);
>> + strlcat(buf, tmp, PAGE_SIZE);
>> +
>> + scnprintf(tmp, sizeof(tmp), "# Request Timedout Error: %d\n",
>> + msm_host->err_stats[MMC_ERR_REQ_TIMEOUT]);
>> + strlcat(buf, tmp, PAGE_SIZE);
>> +
>> + return strlen(buf);
>> +}
>> +static DEVICE_ATTR_RO(err_stats);
>> +
>> +static struct attribute *sdhci_msm_sysfs_attrs[] = {
>> + &dev_attr_err_state.attr,
>> + &dev_attr_err_stats.attr,
>> + NULL
>> +};
>> +
>> +static const struct attribute_group sdhci_msm_sysfs_group = {
>> + .name = "qcom",
>> + .attrs = sdhci_msm_sysfs_attrs, };
>> +
>> +static int sdhci_msm_init_sysfs(struct platform_device *pdev) {
>> + int ret;
>> +
>> + ret = sysfs_create_group(&pdev->dev.kobj, &sdhci_msm_sysfs_group);
>> + if (ret)
>> + dev_err(&pdev->dev, "%s: Failed sdhci_msm sysfs group err=%d\n",
>> + __func__, ret);
>> + return ret;
>> +}
>>
>> static int sdhci_msm_probe(struct platform_device *pdev) { @@
>> -2734,6 +2924,8 @@ static int sdhci_msm_probe(struct platform_device *pdev)
>> if (ret)
>> goto pm_runtime_disable;
>>
>> + sdhci_msm_init_sysfs(pdev);
>> +
>> pm_runtime_mark_last_busy(&pdev->dev);
>> pm_runtime_put_autosuspend(&pdev->dev);
>>
>> diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c index
>> 07c6da1..f82a3eff 100644
>> --- a/drivers/mmc/host/sdhci.c
>> +++ b/drivers/mmc/host/sdhci.c
>> @@ -3159,6 +3159,13 @@ static void sdhci_timeout_timer(struct timer_list *t)
>> spin_lock_irqsave(&host->lock, flags);
>>
>> if (host->cmd && !sdhci_data_line_cmd(host->cmd)) {
>> + if (host->ops->err_stats) {
>> + u32 intmask;
>> +
>> + host->mmc->timer = true;
>> + intmask = sdhci_readl(host, SDHCI_INT_STATUS);
>> + host->ops->err_stats(host, intmask);
>> + }
>> pr_err("%s: Timeout waiting for hardware cmd interrupt.\n",
>> mmc_hostname(host->mmc));
>> sdhci_dumpregs(host);
>> @@ -3181,6 +3188,13 @@ static void sdhci_timeout_data_timer(struct
>> timer_list *t)
>>
>> if (host->data || host->data_cmd ||
>> (host->cmd && sdhci_data_line_cmd(host->cmd))) {
>> + if (host->ops->err_stats) {
>> + u32 intmask;
>> +
>> + host->mmc->timer = true;
>> + intmask = sdhci_readl(host, SDHCI_INT_STATUS);
>> + host->ops->err_stats(host, intmask);
>> + }
>> pr_err("%s: Timeout waiting for hardware interrupt.\n",
>> mmc_hostname(host->mmc));
>> sdhci_dumpregs(host);
>> @@ -3466,6 +3480,9 @@ static irqreturn_t sdhci_irq(int irq, void *dev_id)
>> }
>>
>> intmask = sdhci_readl(host, SDHCI_INT_STATUS);
>> + if (host->ops->err_stats)
>> + host->ops->err_stats(host, intmask);
>> +
>> if (!intmask || intmask == 0xffffffff) {
>> result = IRQ_NONE;
>> goto out;
>> diff --git a/drivers/mmc/host/sdhci.h b/drivers/mmc/host/sdhci.h index
>> d7929d7..a1546b3 100644
>> --- a/drivers/mmc/host/sdhci.h
>> +++ b/drivers/mmc/host/sdhci.h
>> @@ -658,6 +658,7 @@ struct sdhci_ops {
>> void (*request_done)(struct sdhci_host *host,
>> struct mmc_request *mrq);
>> void (*dump_vendor_regs)(struct sdhci_host *host);
>> + void (*err_stats)(struct sdhci_host *host, u32 intmask);
>> };
>>
>> #ifdef CONFIG_MMC_SDHCI_IO_ACCESSORS
>> diff --git a/include/linux/mmc/host.h b/include/linux/mmc/host.h index
>> 7afb57c..33186ff 100644
>> --- a/include/linux/mmc/host.h
>> +++ b/include/linux/mmc/host.h
>> @@ -492,6 +492,7 @@ struct mmc_host {
>> int cqe_qdepth;
>> bool cqe_enabled;
>> bool cqe_on;
>> + bool timer;
>>
>> /* Inline encryption support */ #ifdef CONFIG_MMC_CRYPTO
>> --
>> QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a
>> member of Code Aurora Forum, hosted by The Linux Foundation
>>
>

\
 
 \ /
  Last update: 2021-12-02 08:29    [W:0.311 / U:0.920 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site