[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <56D6C05A.9020503@denx.de>
Date: Wed, 02 Mar 2016 11:28:42 +0100
From: Marek Vasut <marex@...x.de>
To: Oliver Hartkopp <socketcan@...tkopp.net>, linux-can@...r.kernel.org
CC: netdev@...r.kernel.org, Marc Kleine-Budde <mkl@...gutronix.de>,
Mark Rutland <mark.rutland@....com>,
Wolfgang Grandegger <wg@...ndegger.com>
Subject: Re: [PATCH 3/4] net: can: ifi: Fix RX and TX ID mask
On 03/02/2016 07:10 AM, Oliver Hartkopp wrote:
> Hi Marek,
>
> On 03/01/2016 10:23 PM, Marek Vasut wrote:
>> On 03/01/2016 06:49 PM, Oliver Hartkopp wrote:
>
>
>>>> -#define IFI_CANFD_RXFIFO_ID_ID_STD_MASK 0x3ff
>>>> +#define IFI_CANFD_RXFIFO_ID_ID_STD_MASK 0x7ff
>>>> #define IFI_CANFD_RXFIFO_ID_ID_XTD_MASK 0x1fffffff
>>>
>>> You should use the CAN_SFF_MASK and CAN_EFF_MASK in your code instead of
>>> defining you private IFI_CANFD_RXFIFO_ID_ID_?TD_MASK definitions.
>>>
>>> You won't have trapped into this problem then :-)
>>
>> These are register bitfield definitions, so should I really ?
>>
>> My OCD kicks in and tells me it'd be odd and inconsistent with the rest
>> of the bitfields, but if you prefer it that way, I'll just send an
>> updated patch.
>>
>
> Your bit mask is masking the CAN ID out of a given variable.
> That's what CAN_SFF_MASK and CAN_EFF_MASK is made for.
>
> So at least it should be:
>
> #define IFI_CANFD_RXFIFO_ID_ID_STD_MASK CAN_SFF_MASK
> #define IFI_CANFD_RXFIFO_ID_ID_XTD_MASK CAN_EFF_MASK
This is good, I will do this. Thanks!
> Btw. These defines are _never_ referenced in ifi_canfd.c so they should be
> removed anyway.
The documentation for this core is not available, so if you don't mind,
I'd like to keep those.
--
Best regards,
Marek Vasut
Powered by blists - more mailing lists