[bug report] scsi: ufs-qcom: dump additional testbus registers

[bug report] scsi: ufs-qcom: dump additional testbus registers

[Dan Carpenter]

Hello Venkat Gopalakrishnan,

The patch 9c46b8676271: "scsi: ufs-qcom: dump additional testbus
registers" from Feb 3, 2017, leads to the following static checker
warning:

	drivers/scsi/ufs/ufs-qcom.c:1531 ufs_qcom_testbus_cfg_is_ok()
	warn: impossible condition '(host->testbus.select_minor > 255) => (0-255 > 255)'

drivers/scsi/ufs/ufs-qcom.c
  1517  static bool ufs_qcom_testbus_cfg_is_ok(struct ufs_qcom_host *host)
  1518  {
  1519          if (host->testbus.select_major >= TSTBUS_MAX) {
  1520                  dev_err(host->hba->dev,
  1521                          "%s: UFS_CFG1[TEST_BUS_SEL} may not equal 0x%05X\n",
  1522                          __func__, host->testbus.select_major);
  1523                  return false;
  1524          }
  1525  
  1526          /*
  1527           * Not performing check for each individual select_major
  1528           * mappings of select_minor, since there is no harm in
  1529           * configuring a non-existent select_minor
  1530           */
  1531          if (host->testbus.select_minor > 0xFF) {
                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

It might make sense to keep this check.  I don't know.  But it's
confusing that 0xFF is a magic number.  Better to make it a define.

  1532                  dev_err(host->hba->dev,
  1533                          "%s: 0x%05X is not a legal testbus option\n",
  1534                          __func__, host->testbus.select_minor);
  1535                  return false;
  1536          }
  1537  
  1538          return true;
  1539  }

regards,
dan carpenter

Re: [bug report] scsi: ufs-qcom: dump additional testbus registers

[Subhash Jadavani]

On 2017-02-13 23:58, Dan Carpenter wrote:
> Hello Venkat Gopalakrishnan,
> 
> The patch 9c46b8676271: "scsi: ufs-qcom: dump additional testbus
> registers" from Feb 3, 2017, leads to the following static checker
> warning:
> 
> 	drivers/scsi/ufs/ufs-qcom.c:1531 ufs_qcom_testbus_cfg_is_ok()
> 	warn: impossible condition '(host->testbus.select_minor > 255) =>
> (0-255 > 255)'
> 
> drivers/scsi/ufs/ufs-qcom.c
>   1517  static bool ufs_qcom_testbus_cfg_is_ok(struct ufs_qcom_host 
> *host)
>   1518  {
>   1519          if (host->testbus.select_major >= TSTBUS_MAX) {
>   1520                  dev_err(host->hba->dev,
>   1521                          "%s: UFS_CFG1[TEST_BUS_SEL} may not
> equal 0x%05X\n",
>   1522                          __func__, host->testbus.select_major);
>   1523                  return false;
>   1524          }
>   1525
>   1526          /*
>   1527           * Not performing check for each individual 
> select_major
>   1528           * mappings of select_minor, since there is no harm in
>   1529           * configuring a non-existent select_minor
>   1530           */
>   1531          if (host->testbus.select_minor > 0xFF) {
>                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> 
> It might make sense to keep this check.  I don't know.  But it's
> confusing that 0xFF is a magic number.  Better to make it a define.

Yes, i agree, this check is redundant. I will post a fix to remove this.

> 
>   1532                  dev_err(host->hba->dev,
>   1533                          "%s: 0x%05X is not a legal testbus 
> option\n",
>   1534                          __func__, host->testbus.select_minor);
>   1535                  return false;
>   1536          }
>   1537
>   1538          return true;
>   1539  }
> 
> regards,
> dan carpenter

-- 
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project