Re: [PATCH 4/4] [SCSI] ufshcd: SCSI error handling

[Santosh Y <santoshsy@gmail.com>]

On Sun, Feb 5, 2012 at 1:07 PM, Namjae Jeon <linkinjeon@gmail.com> wrote:
>> +
>> +/**
>> + * 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);
> Hi.
> You called spin_lock_irqsave in ufshcd_issue_tm_cmd() without
> spin_unlock_irqrestore.

Thanks for reviewing the code. We'll correct it.
Please let us know your comments on the other patches too.

>
>> +       if (!err) {
>> +               spin_lock_irqsave(host->host_lock, flags);
> This case is same also.
> Thanks.
>> +               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;
>> +}
>> +

-- 
~Santosh