On Tue, Feb 23, 2021 at 02:39:49PM -0700, Dave Jiang wrote:
On 2/23/2021 2:14 PM, ramesh.thomas(a)intel.com wrote:
> From: Ramesh Thomas <ramesh.thomas(a)intel.com>
>
> Library functions do their own checking and return errors without
> issuing commands to the hardware. Command status register retrieved
> from "cmd_status" sysfs attribute may not be set in such cases. Return
> NULL from function that retrieves the error status to indicate this.
>
> Signed-off-by: Ramesh Thomas <ramesh.thomas(a)intel.com>
> ---
> accfg/lib/libaccfg.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/accfg/lib/libaccfg.c b/accfg/lib/libaccfg.c
> index ae862f8..7d5e9f4 100644
> --- a/accfg/lib/libaccfg.c
> +++ b/accfg/lib/libaccfg.c
> @@ -1469,7 +1469,7 @@ ACCFG_EXPORT const char *
accfg_device_get_cmd_status_str(struct accfg_device *d
> const char *stat_str;
>
> status = accfg_device_get_cmd_status(device);
> - if (status < 0)
> + if (status <= 0)
Would it be better to just return CMD_SUCCESS or something?
This function returns a string error message. The calling places assume
that if the string is not NULL then it is an error. They don't display
the message if it is NULL. I do like the idea of it returning 0 =
Success. I can make this function return only the integer error code and
create another function to return the string. That way the caller can
decide whether or not to display the error message if it is "Success".
> return NULL;
>
> if (status > ACCFG_CMD_STATUS_MAX)