RE: [PATCH 4/4] staging: comedi: ni_labpc: remove *_ai_gain_bits tables

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Wednesday, April 24, 2013 2:09 AM, Ian Abbott wrote:
> On 2013-04-23 20:58, H Hartley Sweeten wrote:
>> The bits needed to set the analog input gain can be simply calculated
>> based on the 'range'.
>>
>> The LabPC versions of the board do not have the '0x10' gain that the
>> LabPC+ board supports. By incrementing the range appropriately the
>> correct gain bits can still be calculated.
>>
>> This allows removing the two gain tables, as well as the export, along
>> with the 'ai_range_code' data in the boardinfo.
>>
>> Signed-off-by: H Hartley Sweeten <hsweeten@xxxxxxxxxxxxxxxxxxx>
>> Cc: Ian Abbott <abbbotti@xxxxxxxxx>
>> Cc: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
>> ---
>>   drivers/staging/comedi/drivers/ni_labpc.c     | 30 +++++++++++++--------------
>>   drivers/staging/comedi/drivers/ni_labpc.h     |  3 ---
>>   drivers/staging/comedi/drivers/ni_labpc_cs.c  |  1 -
>>   drivers/staging/comedi/drivers/ni_labpc_pci.c |  1 -
>>   4 files changed, 14 insertions(+), 21 deletions(-)
>>
>> diff --git a/drivers/staging/comedi/drivers/ni_labpc.c b/drivers/staging/comedi/drivers/ni_labpc.c
>> index 5ed9a6f..773368d 100644
>> --- a/drivers/staging/comedi/drivers/ni_labpc.c
>> +++ b/drivers/staging/comedi/drivers/ni_labpc.c
>> @@ -88,7 +88,7 @@
>>   #define CMD1_REG		0x00	/* W: Command 1 reg */
>>   #define CMD1_MA(x)		(((x) & 0x7) << 0)
>>   #define CMD1_TWOSCMP		(1 << 3)
>> -#define CMD1_GAIN_MASK		(7 << 4)
>> +#define CMD1_GAIN(x)		(((x) & 0x7) << 4)
>>   #define CMD1_SCANEN		(1 << 7)
>>   #define CMD2_REG		0x01	/* W: Command 2 reg */
>>   #define CMD2_PRETRIG		(1 << 0)
>> @@ -153,11 +153,6 @@ enum scan_mode {
>>   	MODE_MULT_CHAN_DOWN,
>>   };
>>
>> -static const int labpc_plus_ai_gain_bits[] = {
>> -	0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
>> -	0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
>> -};
>> -
>>   static const struct comedi_lrange range_labpc_plus_ai = {
>>   	16, {
>>   		BIP_RANGE(5),
>> @@ -179,12 +174,6 @@ static const struct comedi_lrange range_labpc_plus_ai = {
>>   	}
>>   };
>>
>> -const int labpc_1200_ai_gain_bits[] = {
>> -	0x00, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
>> -	0x00, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
>> -};
>> -EXPORT_SYMBOL_GPL(labpc_1200_ai_gain_bits);
>> -
>>   static const struct comedi_lrange range_labpc_1200_ai = {
>>   	14, {
>>   		BIP_RANGE(5),
>> @@ -237,20 +226,17 @@ static inline void labpc_writeb(unsigned int byte, unsigned long address)
>>   static const struct labpc_boardinfo labpc_boards[] = {
>>   	{
>>   		.name			= "lab-pc-1200",
>> -		.ai_range_code		= labpc_1200_ai_gain_bits,
>>   		.ai_speed		= 10000,
>>   		.ai_scan_up		= 1,
>>   		.has_ao			= 1,
>>   		.is_labpc1200		= 1,
>>   	}, {
>>   		.name			= "lab-pc-1200ai",
>> -		.ai_range_code		= labpc_1200_ai_gain_bits,
>>   		.ai_speed		= 10000,
>>   		.ai_scan_up		= 1,
>>   		.is_labpc1200		= 1,
>>   	}, {
>>   		.name			= "lab-pc+",
>> -		.ai_range_code		= labpc_plus_ai_gain_bits,
>>   		.ai_speed		= 12000,
>>   		.has_ao			= 1,
>>   	},
>> @@ -321,12 +307,24 @@ static void labpc_ai_set_chan_and_gain(struct comedi_device *dev,
>>   	const struct labpc_boardinfo *board = comedi_board(dev);
>>   	struct labpc_private *devpriv = dev->private;
>>
>> +	if (board->is_labpc1200) {
>> +		/*
>> +		 * The LabPC-1200 boards do not have a gain
>> +		 * of '0x10'. Skip the range values that would
>> +		 * result in this gain.
>> +		 */
>> +		if (range > 1)
>> +			range++;
>> +		if (range > 8)
>> +			range++;
>
> Off-by one error.  That would change the sequence 0 to 13 to the following:
>
>	0, 1, 3, 4, 5, 6, 7, 8, 10, 11, 12, 13, 14, 15
>
> I think you meant:
>
>		if (range > 0)
>			range++;
>		if (range > 8)
>			range++;
>
> which maps the sequence 0 to 13 to the following:
>
>	0, 2, 3, 4, 5, 6, 7, 8, 10, 11, 12, 13, 14, 15

Ugh.. You are correct. Brain fart with that.

> You could also use:
>
>		range += (range > 0) + (range > 7);
>
> which I think is easier to understand (and doesn't need the curly braces).

I like that one better. I'll update the patch.

Greg,

Do you want all four patchs again as v2 or just an update of this one?

> Probably harder to understand, and possibly not as efficient, but this 
> also works:
>
>		range = (((range + 6) * 8) / 7) - 6;

That one is nasty... ;-)

Thanks,
Hartley

_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/devel




[Index of Archives]     [Linux Driver Backports]     [DMA Engine]     [Linux GPIO]     [Linux SPI]     [Video for Linux]     [Linux USB Devel]     [Linux Coverity]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [Yosemite Backpacking]
  Powered by Linux