Re: [PATCH V2 2/2] mailbox: Introduce TI message manager driver

From: Nishanth Menon
Date: Wed Mar 16 2016 - 20:30:27 EST


Hi Jassi,

On 03/16/2016 12:16 AM, Jassi Brar wrote:

[...]
>> Alright, i will drop this check since it is causing a lot more
>> confusion
>>
> It's confusing because you check ti_msgmgr_queue_get_num_messages()
> also in ti_msgmgr_last_tx_done() which doesn't make sense because the
> former accounts for messages from other senders also (as you say there
> could be multiple senders).


True -> I will drop it for now. I will see if the case I was trying to
protect is actually possible to be hit in the first place. And if proven
to be required, I will introduce it back with a better explanation and
the usecase where this is needed.

>> that that is worth. we can introduce it when we finally do
>> hit an issue eventually with multiple processors trying to transmit on
>> the same queue manager. that is not a concern at the very immediate
>> time, so we should be good to drop.
>>
>> please let me know if you are ok with this.
>>
> I am ok with whatever you assert is needed for your platform. I just
> point out what I think are inconsistencies in your assumptions. I'll
> pick the next revision however it is.


Thanks for your patience and guidance with this series. I have tried to
incorporate all the alignment we have had on this thread as part of
V3[1] of the series.


[1] http://marc.info/?l=linux-arm-kernel&m=145817434531691&w=2

--
Regards,
Nishanth Menon