Re: [PATCH V19 07/13] MIPS: Loongson 3: Add IRQ init and dispatch support

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

 



> On Sun, Feb 16, 2014 at 04:01:24PM +0800, Huacai Chen wrote:
>
>> +#define LOONGSON_HT1_INT_VECTOR_BASE	(LOONGSON_HT1_CFG_BASE + 0x80)
>> +#define LOONGSON_HT1_INT_EN_BASE	(LOONGSON_HT1_CFG_BASE + 0xa0)
>> +#define LOONGSON_HT1_INT_VECTOR(n)	\
>> +		LOONGSON3_REG32(LOONGSON_HT1_INT_VECTOR_BASE, 4 * n)
>> +#define LOONGSON_HT1_INTN_EN(n)		\
>> +		LOONGSON3_REG32(LOONGSON_HT1_INT_EN_BASE, 4 * n)
>
> Doing math on macro arguments without parenthesis is dangerous,  please
> use (n) instead.
>
>> +
>> +#define LOONGSON_INT_ROUTER_OFFSET	0x1400
>> +#define LOONGSON_INT_ROUTER_INTEN	\
>> +	  LOONGSON3_REG32(LOONGSON3_REG_BASE, LOONGSON_INT_ROUTER_OFFSET +
>> 0x24)
>> +#define LOONGSON_INT_ROUTER_INTENSET	\
>> +	  LOONGSON3_REG32(LOONGSON3_REG_BASE, LOONGSON_INT_ROUTER_OFFSET +
>> 0x28)
>> +#define LOONGSON_INT_ROUTER_INTENCLR	\
>> +	  LOONGSON3_REG32(LOONGSON3_REG_BASE, LOONGSON_INT_ROUTER_OFFSET +
>> 0x2c)
>> +#define LOONGSON_INT_ROUTER_ENTRY(n)	\
>> +	  LOONGSON3_REG8(LOONGSON3_REG_BASE, LOONGSON_INT_ROUTER_OFFSET + n)
>> +#define LOONGSON_INT_ROUTER_LPC		LOONGSON_INT_ROUTER_ENTRY(0x0a)
>> +#define LOONGSON_INT_ROUTER_HT1(n)	LOONGSON_INT_ROUTER_ENTRY(n + 0x18)
>
> These two also.
>
>> +static void ht_irqdispatch(void)
>> +{
>> +	unsigned int i, irq;
>> +	unsigned int ht_irq[] = {1, 3, 4, 5, 6, 7, 8, 12, 14, 15};
>> +
>> +	irq = LOONGSON_HT1_INT_VECTOR(0);
>> +	LOONGSON_HT1_INT_VECTOR(0) = irq; /* Acknowledge the IRQs */
>> +
>> +	for (i = 0; i < ARRAY_SIZE(ht_irq); i++) {
>> +		if (irq & (0x1 << ht_irq[i]))
>> +			do_IRQ(ht_irq[i]);
>> +	}
>> +}
>
> Ouch.
>
> Initializing an array like this in C will generate code which at runtime
> initializes the ht_irq[] array each time ht_irqdispatch is invoked
> and slowing this function to a crawl.
>
> You want to make either move the definition of this array out of the
> function body or make it "static const unsigned int ht_irq[] ..." to
> avoid this.
Since there are several patches have problem, and the first one is
merged, I will send a V20 patchset without the first one ASAP.

>
>   Ralf
>




[Index of Archives]     [Linux MIPS Home]     [LKML Archive]     [Linux ARM Kernel]     [Linux ARM]     [Linux]     [Git]     [Yosemite News]     [Linux SCSI]     [Linux Hams]

  Powered by Linux