[PATCH 1/3] OMAP2+: UART: Remove cpu checks for populating errata flags

[Raja, Govindraj]

On Wed, Mar 28, 2012 at 10:13 PM, Jon Hunter <jon-hunter at ti.com> wrote:
> Hi Govindraj,
>
>
> On 3/28/2012 6:09, Raja, Govindraj wrote:
>>
>> On Wed, Mar 28, 2012 at 2:38 AM, Jon Hunter<jon-hunter at ti.com>  wrote:
>>>
>>> Hi Govindraj,
>>>
>>>

[...]

>>>> +       u32 mvr, scheme;
>>>> +       u16 revision, major, minor;
>>>> +
>>>> +       mvr = serial_in(up, UART_OMAP_MVER);
>>>> +
>>>> +       /* Check revision register scheme */
>>>> +       scheme = mvr&    (0x03<<    30);
>>>> +       scheme>>= 30;
>>>
>>>
>>>
>>> Minor nit-pick, why not ...
>>>
>>>        scheme = mvr>>  30;
>>>
>>
>> yes will correct it.
>
>
> Thinking some more, could be better to add some #defines for
> OMAP_MVR_VERSION_MASK/SHIFT here.
>

Yes sure.

>
>>>
>>>> +       switch (scheme) {
>>>> +       case 0: /* Legacy Scheme: OMAP2/3 */
>>>> +               /* MINOR_REV[0:4], MAJOR_REV[4:7] */
>>>
>>>
>>>
>>> This scheme is also true from OMAP1 devices. Maybe we could include in
>>> the
>>> comment.
>>
>>
>> No omap1 from $SUBJECT also reason,
>>
>> mach-omap1/serial.c =>  8250.c
>> mach-omap2/serial.c =>  omap-serial.c
>
>
> Got it! Thanks.
>
>>>> +               major = (mvr&    0xf0)>>    4;
>>>> +               minor = (mvr&    0x0f);
>>>>
>>>> +               break;
>>>> +       case 1:
>>>> +               /* New Scheme: OMAP4+ */
>>>> +               /* MINOR_REV[0:5], MAJOR_REV[8:10] */
>>>> +               major = (mvr&    0x7<<    8)>>    8;
>>>
>>>
>>>
>>> Nit-pick ...
>>>
>>>        major = (mvr>>  8)&  0x7;
>>
>>
>> yes fine will correct this.
>
>
> May be we should add #defines here to for OMAP_UART_MVR1_MAJ_MASK/SHIFT,
> OMAP_UART_MVR1_MIN_MASK/SHIFT, OMAP_UART_MVR2_MAJ_MASK/SHIFT and
> OMAP_UART_MVR2_MIN_MASK/SHIFT.
>

okay will add this.

--
Thanks,
Govindraj.R