lkml.org 
[lkml]   [2012]   [Feb]   [6]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
From
Date
SubjectRe: [PATCH 4/4] [SCSI] ufshcd: SCSI error handling
On Mon, Feb 6, 2012 at 12:46 PM, Amit Sahrawat
<amit.sahrawat83@gmail.com> wrote:
> Hi,
>
> In function:
> +static int ufshcd_abort(struct scsi_cmnd *cmd)
> +{...
> -       int err;
> +       int err = -1;
> ...
> +       spin_lock_irqsave(host->host_lock, flags);
> +       pos = (1 << tag);
> +
> +       /* check if command is still pending */
> +       if (!(hba->outstanding_reqs & pos)) {
> -               err = -1;
> -               spin_unlock_irqrestore(host->host_lock, flags);
> +               goto out;
> +       }
> ...
> ...
> +out:
> +       spin_unlock_irqrestore(host->host_lock, flags);
> +       return err;
> +}
> this will also take of matching out
> spin_lock_irqsave()->spin_unlock_irqrestore() before exiting the
> function.

That will not work again.
In case "if (!(hba->outstanding_reqs & pos))" is false,
i.e if the command for which the abort is issued still exists,

ufshcd_issue_tm_cmd() will be called and if no error returned
by the function, the code after "out:" will be executed. Which will
try to "spin_unlock_irqrestore",
even though no lock has been acquired.

>
> In function:
> +static int
> +ufshcd_issue_tm_cmd(struct ufs_hba *hba,
> +                   struct ufshcd_lrb *lrbp,
> +                   u8 tm_function)
> ...
> +       spin_lock_irqsave(host->host_lock, flags);
> +
> +       /* If task management queue is full */
> +       if (!ufshcd_is_tmq_full(hba->outstanding_tasks)) {
> +               dev_err(&hba->pdev->dev, "Task management queue full\n");
> I think this will map to a Buffered printf - which can potentially
> result in BUG: Schedule while atomic()...
> +               spin_unlock_irqrestore(host->host_lock, flags);
> +               return err;
> +       }
> So, it could be like this:

ok, Thanks, I'll look into it.

> +       spin_lock_irqsave(host->host_lock, flags);
> +
> +       /* If task management queue is full */
> +       if (!ufshcd_is_tmq_full(hba->outstanding_tasks)) {
> +               spin_unlock_irqrestore(host->host_lock, flags);
> +               dev_err(&hba->pdev->dev, "Task management queue full\n");
> +               return err;
> +       }
>
> Thanks & Regards,
> Amit Sahrawat
>
>
> On Thu, Feb 2, 2012 at 10:27 AM, Vinayak Holikatti
> <vinholikatti@gmail.com> wrote:
>> From: Santosh Yaraganavi <santoshsy@gmail.com>
>>
>> UFSHCD SCSI error handling includes following implementations,
>>  - Abort task
>>  - Device reset
>>  - Host reset
>>
>> Signed-off-by: Santosh Yaraganavi <santoshsy@gmail.com>
>> Signed-off-by: Vinayak Holikatti <vinholikatti@gmail.com>
>> Reviewed-by: Arnd Bergmann <arnd@arndb.de>
>> Reviewed-by: Saugata Das <saugata.das@linaro.org>
>> Reviewed-by: Vishak G <vishak.g@samsung.com>
>> Reviewed-by: Girish K S <girish.shivananjappa@linaro.org>
>> ---
>>  drivers/scsi/ufs/ufshcd.c |  312 +++++++++++++++++++++++++++++++++++++++++++++
>>  1 files changed, 312 insertions(+), 0 deletions(-)
>>
>> diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
>> index 1bfae84..7589dd1 100644
>> --- a/drivers/scsi/ufs/ufshcd.c
>> +++ b/drivers/scsi/ufs/ufshcd.c
>> @@ -76,6 +76,8 @@
>>  #define NULL 0
>>  #endif  /* NULL */
>>
>> +#define UFSHCD_MAX_TM_SLOTS    0xFF
>> +
>>  #define BYTES_TO_DWORDS(p)     (p >> 2)
>>  #define UFSHCD_MMIO_BASE       (hba->mmio_base)
>>
>> @@ -83,6 +85,7 @@ enum {
>>        UFSHCD_MAX_CHANNEL      = 1,
>>        UFSHCD_MAX_ID           = 1,
>>        UFSHCD_MAX_LUNS         = 8,
>> +       UFSHCD_MAX_TM_CMDS      = 8,
>>        UFSHCD_CMD_PER_LUN      = 16,
>>        UFSHCD_CAN_QUEUE        = 32,
>>        BYTES_128               = 128,
>> @@ -149,6 +152,7 @@ struct uic_command {
>>  * @host: Scsi_Host instance of the driver
>>  * @pdev: PCI device handle
>>  * @lrb: local reference block
>> + * @outstanding_tasks: Bits representing outstanding task requests
>>  * @outstanding_reqs: Bits representing outstanding transfer requests
>>  * @capabilities: UFS Controller Capabilities
>>  * @nutrs: Transfer Request Queue depth supported by controller
>> @@ -192,6 +196,7 @@ struct ufs_hba {
>>
>>        struct ufshcd_lrb *lrb;
>>
>> +       u32 outstanding_tasks;
>>        u32 outstanding_reqs;
>>
>>        u32 capabilities;
>> @@ -200,6 +205,8 @@ struct ufs_hba {
>>        u32 ufs_version;
>>
>>        struct uic_command active_uic_cmd;
>> +       wait_queue_head_t ufshcd_tm_wait_queue;
>> +       u8 tm_condition[UFSHCD_MAX_TM_CMDS];
>>
>>        u32 ufshcd_state;
>>        u32 int_enable_mask;
>> @@ -278,6 +285,30 @@ static inline int ufshcd_get_tr_ocs(struct ufshcd_lrb *lrbp)
>>  }
>>
>>  /**
>> + * ufshcd_get_tmr_ocs - Get the UTMRD Overall Command Status
>> + * @task_req_descp: pointer to utp_task_req_desc structure
>> + *
>> + * This function is used to get the OCS field from UTMRD
>> + * Returns the OCS field in the UTMRD
>> + */
>> +static inline int
>> +ufshcd_get_tmr_ocs(struct utp_task_req_desc *task_req_descp)
>> +{
>> +       return task_req_descp->header.dword_2 & MASK_OCS;
>> +}
>> +
>> +/**
>> + * ufshcd_is_tmq_full - checks if the task management slots are full
>> + * @outstanding_tasks: contains the task management doorbell value
>> + *
>> + * Returns 1 if a free slot available, 0 if task slots are full
>> + */
>> +static inline int ufshcd_is_tmq_full(u32 outstanding_tasks)
>> +{
>> +       return (UFSHCD_MAX_TM_SLOTS == outstanding_tasks) ? 0 : 1;
>> +}
>> +
>> +/**
>>  * ufshcd_get_lists_status - Check UCRDY, UTRLRDY and UTMRLRDY
>>  * @reg: Register value of host controller status
>>  *
>> @@ -1098,6 +1129,45 @@ static void ufshcd_slave_destroy(struct scsi_device *sdev)
>>  }
>>
>>  /**
>> + * ufshcd_task_req_compl - handle task management request completion
>> + * @hba: per adapter instance
>> + * @index: index of the completed request
>> + *
>> + * Returns 0 on success, non-zero value on failure
>> + */
>> +static int ufshcd_task_req_compl(struct ufs_hba *hba, u32 index)
>> +{
>> +       struct utp_upiu_task_rsp *task_rsp_upiup;
>> +       struct utp_task_req_desc *task_req_descp;
>> +       unsigned long flags;
>> +       int ocs_value;
>> +       int task_response = -1;
>> +
>> +       spin_lock_irqsave(hba->host->host_lock, flags);
>> +
>> +       /* clear completed tasks from outstanding_tasks */
>> +       hba->outstanding_tasks ^= index;
>> +
>> +       task_req_descp =
>> +               (struct utp_task_req_desc *)hba->utmrdl_virt_addr_aligned;
>> +       ocs_value = ufshcd_get_tmr_ocs(&task_req_descp[index]);
>> +
>> +       if (OCS_SUCCESS == ocs_value) {
>> +
>> +               task_rsp_upiup = (struct utp_upiu_task_rsp *)
>> +                                task_req_descp[index].task_rsp_upiu;
>> +               task_response = be32_to_cpu(task_rsp_upiup->header.dword_1);
>> +               task_response = ((task_response & MASK_TASK_RESPONSE) >> 8);
>> +
>> +               /* clear task response */
>> +               memset(task_rsp_upiup, 0, sizeof(struct utp_upiu_task_rsp));
>> +       }
>> +       spin_unlock_irqrestore(hba->host->host_lock, flags);
>> +
>> +       return task_response;
>> +}
>> +
>> +/**
>>  * ufshcd_scsi_cmd_status - Update SCSI command result based on scsi status
>>  * @lrb: pointer to local reference block of completed command
>>  * @scsi_status: SCSI command status
>> @@ -1314,6 +1384,31 @@ fatal_eh:
>>  }
>>
>>  /**
>> + * ufshcd_tmc_handler - handle task management function completion
>> + * @hba: per adapter instance
>> + */
>> +static void ufshcd_tmc_handler(struct ufs_hba *hba)
>> +{
>> +       u32 completed_reqs;
>> +       u32 tm_doorbell;
>> +       int i;
>> +
>> +       tm_doorbell = readl(hba->mmio_base + REG_UTP_TASK_REQ_DOOR_BELL);
>> +       completed_reqs = tm_doorbell ^ hba->outstanding_tasks;
>> +       for (i = 0; i < hba->nutmrs; i++) {
>> +               if (completed_reqs & (1 << i)) {
>> +
>> +                       /*
>> +                        * Change the default value 0xFF to 0 to indicate
>> +                        * completion of respective task management command
>> +                        */
>> +                       hba->tm_condition[i] = 0;
>> +                       wake_up_interruptible(&hba->ufshcd_tm_wait_queue);
>> +               }
>> +       }
>> +}
>> +
>> +/**
>>  * ufshcd_sl_intr - Interrupt service routine
>>  * @hba: per adapter instance
>>  * @intr_status: contains interrupts generated by the controller
>> @@ -1327,6 +1422,9 @@ static void ufshcd_sl_intr(struct ufs_hba *hba, u32 intr_status)
>>        if (intr_status & UIC_COMMAND_COMPL)
>>                schedule_work(&hba->uic_workq);
>>
>> +       if (intr_status & UTP_TASK_REQ_COMPL)
>> +               ufshcd_tmc_handler(hba);
>> +
>>        if (intr_status & UTP_TRANSFER_REQ_COMPL)
>>                ufshcd_transfer_req_compl(hba);
>>  }
>> @@ -1362,6 +1460,209 @@ static irqreturn_t ufshcd_intr(int irq, void *__hba)
>>        return retval;
>>  }
>>
>> +/**
>> + * ufshcd_issue_tm_cmd - issues task management commands to controller
>> + * @hba: per adapter instance
>> + * @lrbp: pointer to local reference block
>> + *
>> + * Returns 0 on success, non-zero value on failure
>> + */
>> +static int
>> +ufshcd_issue_tm_cmd(struct ufs_hba *hba,
>> +                   struct ufshcd_lrb *lrbp,
>> +                   u8 tm_function)
>> +{
>> +       struct utp_task_req_desc *task_req_descp;
>> +       struct utp_upiu_task_req *task_req_upiup;
>> +       struct Scsi_Host *host;
>> +       unsigned long flags;
>> +       int i;
>> +       int free_slot = 0;
>> +       int err = -1;
>> +
>> +       host = hba->host;
>> +
>> +       spin_lock_irqsave(host->host_lock, flags);
>> +
>> +       /* If task management queue is full */
>> +       if (!ufshcd_is_tmq_full(hba->outstanding_tasks)) {
>> +               dev_err(&hba->pdev->dev, "Task management queue full\n");
>> +               spin_unlock_irqrestore(host->host_lock, flags);
>> +               return err;
>> +       }
>> +
>> +       for (i = 0; i < hba->nutmrs; i++) {
>> +               if (!(hba->outstanding_tasks & (1 << i))) {
>> +                       free_slot = i;
>> +                       break;
>> +               }
>> +       }
>> +
>> +       /* Configure task request descriptor */
>> +       task_req_descp =
>> +               (struct utp_task_req_desc *) hba->utmrdl_virt_addr_aligned;
>> +       task_req_descp += free_slot;
>> +
>> +       memset(task_req_descp, 0, sizeof(struct utp_task_req_desc));
>> +       task_req_descp->header.dword_0 = cpu_to_le32(UTP_REQ_DESC_INT_CMD);
>> +       task_req_descp->header.dword_2 =
>> +                       cpu_to_le32(OCS_INVALID_COMMAND_STATUS);
>> +
>> +       /* Configure task request UPIU */
>> +       task_req_upiup =
>> +               (struct utp_upiu_task_req *) task_req_descp->task_req_upiu;
>> +       task_req_upiup->header.dword_0 =
>> +               cpu_to_be32(UPIU_HEADER_DWORD(UPIU_TRANSACTION_TASK_REQ, 0,
>> +                                             lrbp->lun, lrbp->task_tag));
>> +       task_req_upiup->header.dword_1 =
>> +       cpu_to_be32(UPIU_HEADER_DWORD(0, tm_function, 0, 0));
>> +
>> +       task_req_upiup->input_param1 = lrbp->lun;
>> +       task_req_upiup->input_param1 =
>> +               cpu_to_be32(task_req_upiup->input_param1);
>> +       task_req_upiup->input_param2 = lrbp->task_tag;
>> +       task_req_upiup->input_param2 =
>> +               cpu_to_be32(task_req_upiup->input_param2);
>> +
>> +       /* send command to the controller */
>> +       hba->outstanding_tasks |= (1 << free_slot);
>> +       writel((1 << free_slot),
>> +              (UFSHCD_MMIO_BASE + REG_UTP_TASK_REQ_DOOR_BELL));
>> +
>> +       spin_unlock_irqrestore(host->host_lock, flags);
>> +
>> +       /* wait until the task management command is completed */
>> +       err = wait_event_interruptible_timeout(hba->ufshcd_tm_wait_queue,
>> +                                        (hba->tm_condition[free_slot] != 0),
>> +                                        60 * HZ);
>> +       hba->tm_condition[free_slot] = 0xFF;
>> +       if (!err) {
>> +               dev_err(&hba->pdev->dev,
>> +                       "Task management command timed-out\n");
>> +               return err;
>> +       }
>> +       return ufshcd_task_req_compl(hba, free_slot);
>> +}
>> +
>> +/**
>> + * ufshcd_device_reset - reset device and abort all the pending commands
>> + * @cmd: SCSI command pointer
>> + *
>> + * Returns 0 on success, non-zero value on failure
>> + */
>> +static int ufshcd_device_reset(struct scsi_cmnd *cmd)
>> +{
>> +       struct Scsi_Host *host;
>> +       struct ufs_hba *hba;
>> +       unsigned long flags;
>> +       int reset_lun;
>> +       unsigned int tag;
>> +       u32 i;
>> +       u32 pos; /* pos: represents a bit in a register */
>> +       int err = -1;
>> +
>> +       host = cmd->device->host;
>> +       hba = (struct ufs_hba *)host->hostdata;
>> +       tag = cmd->request->tag;
>> +
>> +       if ((hba->lrb[tag].cmd == cmd))
>> +               reset_lun = hba->lrb[tag].lun;
>> +       else
>> +               goto out;
>> +
>> +       err = ufshcd_issue_tm_cmd(hba, &hba->lrb[tag], UFS_LOGICAL_RESET);
>> +       if (!err) {
>> +               spin_lock_irqsave(host->host_lock, flags);
>> +               for (i = 0; i < hba->nutrs; i++) {
>> +                       pos = (1 << i);
>> +                       if ((pos & hba->outstanding_reqs) &&
>> +                           (reset_lun == hba->lrb[i].lun)) {
>> +
>> +                               /* clear the respective UTRLCLR bit */
>> +                               writel(~pos, (UFSHCD_MMIO_BASE +
>> +                               REG_UTP_TRANSFER_REQ_LIST_CLEAR));
>> +
>> +                               hba->outstanding_reqs ^= pos;
>> +
>> +                               if (hba->lrb[i].cmd) {
>> +                                       scsi_dma_unmap(hba->lrb[i].cmd);
>> +                                       hba->lrb[i].cmd->result =
>> +                                                       DID_ABORT << 16;
>> +                                       hba->lrb[i].cmd->scsi_done(cmd);
>> +                                       hba->lrb[i].cmd = NULL;
>> +                               }
>> +                       }
>> +               } /* end of for */
>> +               spin_unlock_irqrestore(host->host_lock, flags);
>> +       } /*end of if */
>> +out:
>> +       return err;
>> +}
>> +
>> +/**
>> + * ufshcd_host_reset - Main reset function registered with scsi layer
>> + * @cmd: SCSI command pointer
>> + *
>> + * Returns 0 on success, non-zero value on failure
>> + */
>> +static int ufshcd_host_reset(struct scsi_cmnd *cmd)
>> +{
>> +       struct ufs_hba *hba = NULL;
>> +
>> +       hba = (struct ufs_hba *) &cmd->device->host->hostdata[0];
>> +
>> +       if (UFSHCD_STATE_RESET == hba->ufshcd_state)
>> +               return 0;
>> +
>> +       return ufshcd_do_reset(hba) ? -1 : 0;
>> +}
>> +
>> +/**
>> + * ufshcd_abort - abort a specific command
>> + * @cmd: SCSI command pointer
>> + *
>> + * Returns 0 on success, non-zero value on failure
>> + */
>> +static int ufshcd_abort(struct scsi_cmnd *cmd)
>> +{
>> +       struct Scsi_Host *host;
>> +       struct ufs_hba *hba;
>> +       unsigned long flags;
>> +       unsigned int tag;
>> +       unsigned int pos;
>> +       int err;
>> +
>> +       host = cmd->device->host;
>> +       hba = (struct ufs_hba *) host->hostdata;
>> +       tag = cmd->request->tag;
>> +
>> +       spin_lock_irqsave(host->host_lock, flags);
>> +       pos = (1 << tag);
>> +
>> +       /* check if command is still pending */
>> +       if (!(hba->outstanding_reqs & pos)) {
>> +               err = -1;
>> +               spin_unlock_irqrestore(host->host_lock, flags);
>> +               goto out;
>> +       }
>> +
>> +       err = ufshcd_issue_tm_cmd(hba, &hba->lrb[tag], UFS_ABORT_TASK);
>> +       if (!err) {
>> +               spin_lock_irqsave(host->host_lock, flags);
>> +               scsi_dma_unmap(cmd);
>> +
>> +               /* clear the respective UTRLCLR bit */
>> +               writel(~pos,
>> +                       (UFSHCD_MMIO_BASE +
>> +                        REG_UTP_TRANSFER_REQ_LIST_CLEAR));
>> +               hba->outstanding_reqs &= ~pos;
>> +               hba->lrb[tag].cmd = NULL;
>> +               spin_unlock_irqrestore(host->host_lock, flags);
>> +       }
>> +out:
>> +       return err;
>> +}
>> +
>>  static struct scsi_host_template ufshcd_driver_template = {
>>        .module                 = THIS_MODULE,
>>        .name                   = UFSHCD,
>> @@ -1369,6 +1670,9 @@ static struct scsi_host_template ufshcd_driver_template = {
>>        .queuecommand           = ufshcd_queuecommand,
>>        .slave_alloc            = ufshcd_slave_alloc,
>>        .slave_destroy          = ufshcd_slave_destroy,
>> +       .eh_abort_handler       = ufshcd_abort,
>> +       .eh_device_reset_handler = ufshcd_device_reset,
>> +       .eh_host_reset_handler  = ufshcd_host_reset,
>>        .this_id                = -1,
>>        .sg_tablesize           = SG_ALL,
>>        .cmd_per_lun            = UFSHCD_CMD_PER_LUN,
>> @@ -1483,6 +1787,7 @@ ufshcd_probe(struct pci_dev *pdev, const struct pci_device_id *id)
>>        struct Scsi_Host *host;
>>        struct ufs_hba *hba;
>>        int ufs_hba_len;
>> +       int index;
>>        int err;
>>
>>        ufs_hba_len = sizeof(struct ufs_hba);
>> @@ -1547,6 +1852,13 @@ ufshcd_probe(struct pci_dev *pdev, const struct pci_device_id *id)
>>        host->unique_id = host->host_no;
>>        host->max_cmd_len = MAX_CDB_SIZE;
>>
>> +       /* Initailze wait queue for task management */
>> +       init_waitqueue_head(&hba->ufshcd_tm_wait_queue);
>> +
>> +       /* Initially assign invalid value to tm_condition */
>> +       for (index = 0; index < hba->nutmrs; index++)
>> +               hba->tm_condition[index] = 0xFF;
>> +
>>        /* Initialize work queues */
>>        INIT_WORK(&hba->uic_workq, ufshcd_uic_cc_handler);
>>        INIT_WORK(&hba->feh_workq, ufshcd_fatal_err_handler);
>> --
>> 1.7.5.4
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>> Please read the FAQ at  http://www.tux.org/lkml/



--
~Santosh
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
\
 
 \ /
  Last update: 2012-02-06 20:03    [W:0.226 / U:0.832 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site