Re: [PATCH 07/21] lpfc: Fix setting of EQ delay Multiplier

[James Smart <james.smart@emulex.com>]

Tomas,

This one I'm going to leave as is.  bf_set is a macro and I'd prefer we 
kept the variable with a real data type and not be subject to any 
compiler-based type conversions in the equations.  I'd rather not risk 
any disruption to the macro to add all the paren'd type declarations.

-- james


On 2/6/2015 7:18 AM, Tomas Henzl wrote:
> @@ -12959,11 +12959,8 @@ lpfc_eq_create(struct lpfc_hba *phba, struct lpfc_queue *eq, uint32_t imax)
>   	bf_set(lpfc_eq_context_size, &eq_create->u.request.context,
>   	       LPFC_EQE_SIZE);
>   	bf_set(lpfc_eq_context_valid, &eq_create->u.request.context, 1);
> -	/* Calculate delay multiper from maximum interrupt per second */
> -	if (imax > LPFC_DMULT_CONST)
> -		dmult = 0;
> -	else
> -		dmult = LPFC_DMULT_CONST/imax - 1;
> +	/* don't setup delay multiplier using EQ_CREATE */
> +	dmult = 0;
>   	bf_set(lpfc_eq_context_delay_multi, &eq_create->u.request.context,
>   	       dmult);
> This is the only use of 'dmult' in this function, please remove the variable completely
> and use this instead - 	bf_set(lpfc_eq_context_delay_multi, &eq_create->u.request.context, 0);
> Thanks, Tomas
>
>