[openib-general] [PATCH] osm: adding max_lid_ho field to osm_switch_t

[Yevgeny Kliteynik]

Sasha Khapyorsky wrote:
> On 23:03 Tue 19 Dec     , Yevgeny Kliteynik wrote:
>>  
>>>> @@ -809,6 +815,37 @@ osm_switch_get_max_lid_ho(
>>>>  * SEE ALSO
>>>>  *********/
>>>>  
>>>> +/****f* OpenSM: Switch/osm_switch_set_max_lid_ho
>>>> +* NAME
>>>> +*	osm_switch_set_max_lid_ho
>>>> +*
>>>> +* DESCRIPTION
>>>> +*	Set the maximum LID (host order) value accessed from this switch
>>>> +* SYNOPSIS
>>>> +*/
>>>> +static inline void
>>>> +osm_switch_set_max_lid_ho(
>>>> +	IN osm_switch_t* const p_sw,
>>>> +	IN uint16_t max_lid_ho )
>>>> +{
>>>> +	p_sw->max_lid_ho = max_lid_ho;
>>>> +}
>>>> +/*
>>>> +* PARAMETERS
>>>> +*	p_sw
>>>> +*		[in] Pointer to a switch object.
>>>> +*
>>>> +*	max_lid_ho
>>>> +*		Max LID (host order) value accessed from this switch
>>>> +*
>>>> +* RETURN VALUES
>>>> +*	None.
>>>> +*
>>>> +* NOTES
>>>> +*
>>>> +* SEE ALSO
>>>> +*********/
>>>> +
>>> Do we need those +31 lines of code instead of just
>>> p_sw->max_lid_ho = N; ?
>> Since there are access functions for the rest of the fields,
>> I didn't want to make an exception in this case either.
> 
> I think you did anyway - there is no full set of access methods. I'm
> perfectly fine with it. And don't call you to cleanup the rest, just to
> not add new ones.

You're right - setter is not needed.
I will issue a V2 series of patches that will address this and Hal's 
comments.
 
-- Yevgeny

> Sasha
>