Re: [PATCH V17 1/3] dmaengine: qcom_hidma: implement lower level hardware interface

From: Sinan Kaya
Date: Tue Apr 26 2016 - 11:23:21 EST


On 4/26/2016 11:10 AM, Andy Shevchenko wrote:
> On Tue, Apr 26, 2016 at 6:04 PM, Sinan Kaya <okaya@xxxxxxxxxxxxxx> wrote:
>> On 4/25/2016 11:28 PM, Vinod Koul wrote:
>>> On Mon, Apr 11, 2016 at 10:21:11AM -0400, Sinan Kaya wrote:
>
>>>> + while (cause) {
>>>> + if ((cause & BIT(HIDMA_IRQ_TR_CH_INVALID_TRE_BIT_POS)) ||
>>>> + (cause & BIT(HIDMA_IRQ_TR_CH_TRE_RD_RSP_ER_BIT_POS)) ||
>>>> + (cause & BIT(HIDMA_IRQ_EV_CH_WR_RESP_BIT_POS)) ||
>>>> + (cause & BIT(HIDMA_IRQ_TR_CH_DATA_RD_ER_BIT_POS)) ||
>>>> + (cause & BIT(HIDMA_IRQ_TR_CH_DATA_WR_ER_BIT_POS))) {
>>>
>>> Switch please
>>
>> Cause is a combined status register. Let's say it contains 0x41. I need to check
>> if bit 0 or bit 6 is set in this value for each case condition. The value is not 0x40
>> or 0x1.
>>
>> I created macro like this instead.
>>
>> +#define HIDMA_IS_ERR_INTERRUPT(cause) \
>> + (cause & BIT(HIDMA_IRQ_TR_CH_INVALID_TRE_BIT_POS)) || \
>> + (cause & BIT(HIDMA_IRQ_TR_CH_TRE_RD_RSP_ER_BIT_POS)) || \
>> + (cause & BIT(HIDMA_IRQ_EV_CH_WR_RESP_BIT_POS)) || \
>> + (cause & BIT(HIDMA_IRQ_TR_CH_DATA_RD_ER_BIT_POS)) || \
>> + (cause & BIT(HIDMA_IRQ_TR_CH_DATA_WR_ER_BIT_POS))
>
> This looks overheaded.
>
> #define HIDMA_XXX (BIT(a) | BIT (b) ... BIT(n))
>
>>
>> and replaced the if statement as follows
>>
>> if (HIDMA_IS_ERR_INTERRUPT(cause)) {
>
> if (cause & HIDMA_XXX) {
>

This is even better.

--
Sinan Kaya
Qualcomm Technologies, Inc. on behalf of Qualcomm Innovation Center, Inc.
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project