[Skiboot] [PATCH V3 3/6] opal/errorlog : Properly raising/clearing the event variable OPAL_EVENT_ERROR_LOG_AVAIL

[Stewart Smith]

Mukesh Ojha <mukesh02 at linux.vnet.ibm.com> writes:
> --- a/include/errorlog.h
> +++ b/include/errorlog.h
> @@ -22,6 +22,19 @@
>  #include <stdint.h>
>  #include <ccan/list/list.h>
>  
> +/* Following variables are used to indicate state of the
> + * head log entry which is being fetched from FSP/OPAL and
> + * these variables are not overwritten until next log is
> + * retrieved from FSP/OPAL.
> + */
> +enum elog_head_state {
> +	ELOG_STATE_FETCHING,    /*In the process of reading log from FSP. */
> +	ELOG_STATE_FETCHED_INFO,/* Indicates reading log info is completed */
> +	ELOG_STATE_FETCHED_DATA,/* Indicates reading log is completed */
> +	ELOG_STATE_NONE,        /* Indicates to fetch next log */
> +	ELOG_STATE_REJECTED,    /* resend all pending logs to linux */
> +};

Cranking a state machine for what the FSP requires shouldn't leak into
platform independent code.
> @@ -358,6 +371,8 @@ void log_commit(struct errorlog *elog);
>  void opal_elog_complete(struct errorlog *elog, bool success);
>  
>  void opal_elog_init(void);
> +void elog_set_head_state(bool platform_log, enum elog_head_state old_state,
> +		enum elog_head_state new_state);

This doesn't look obvious as to what it does.

-- 
Stewart Smith
OPAL Architect, IBM.