Re: [PATCH V5 for-next 1/4] IB/core: Introduce capabilitymask2 field in ClassPortInfo mad

[Hal Rosenstock <hal-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org>]

On 5/23/2016 11:26 AM, Bart Van Assche wrote:
> On 05/23/2016 04:49 AM, Erez Shitrit wrote:
>> On Mon, May 23, 2016 at 2:33 PM, Hal Rosenstock
>> <hal-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org> wrote:
>>> On 5/23/2016 4:14 AM, Erez Shitrit wrote:
>>>> diff --git a/drivers/infiniband/hw/qib/qib_mad.c
>>>> b/drivers/infiniband/hw/qib/qib_mad.c
>>>> index 0bd1837..a872865 100644
>>>> --- a/drivers/infiniband/hw/qib/qib_mad.c
>>>> +++ b/drivers/infiniband/hw/qib/qib_mad.c
>>>> @@ -1172,11 +1172,11 @@ static int pma_get_classportinfo(struct
>>>> ib_pma_mad *pmp,
>>>>        * Set the most significant bit of CM2 to indicate support for
>>>>        * congestion statistics
>>>>        */
>>>> -     p->reserved[0] = dd->psxmitwait_supported << 7;
>>>> +     ib_set_cpi_capmask2(p, dd->psxmitwait_supported << 7);
>>>
>>> Is this really equivalent to what it's replacing ?
>>
>> Yes. the previous code sets the msb bit in the first byte of the
>> capmask2 (where it was with 3 bytes), no other setting for capmask2
>> in that flow, so it is the same.
> 
> Hello Erez,
> 
> I think you need the following to set the highest bit of
> cap_mask2_resp_time:
> 
> ib_set_cpi_capmask2(p, dd->psxmitwait_supported << (31 - 5))

Agreed. I confirmed this with some test code which I ran on both LE and
BE archs.

-- Hal

> Bart.
> 
> 
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html