Re: [Qemu-devel] [PATCH v3 1/2] ARM: exynos4210: CMU support

[Maksim Kozlov <m.kozlov@samsung.com>]

04.07.2012 18:14, Dunrong Huang пишет:
>>
>>>> +
>>>> +static void exynos4210_cmu_set_pll(void *opaque, Exynos4210ClockState
>>>> *pll)
>>>> +{
>>>> +    Exynos4210CmuState *s = opaque;
>>>> +    Exynos4210ClockState *source;
>>>> +    target_phys_addr_t offset = pll->div_reg;
>>>> +    ClockChangeEntry *cce;
>>>> +    uint32_t pdiv, mdiv, sdiv, enable;
>>>> +
>>>> +    source = exynos4210_clock_find(pll->src_id);
>>>> +
>>>> +    if (source == NULL) {
>>>> +        hw_error("We haven't find source clock %d (requested for %s)\n",
>>>> +                 pll->src_id, pll->name);
>>>> +    }
>>>> +
>>>> +    /*
>>>> +     * FOUT = MDIV * FIN / (PDIV * 2^(SDIV-1))
>>>> +     */
>>>> +
>>>> +    enable = (s->reg[I_(offset)]&   PLL_ENABLE_MASK)>>   PLL_ENABLE_SHIFT;
>>>> +    mdiv   = (s->reg[I_(offset)]&   PLL_MDIV_MASK)>>   PLL_MDIV_SHIFT;
>>>> +    pdiv   = (s->reg[I_(offset)]&   PLL_PDIV_MASK)>>   PLL_PDIV_SHIFT;
>>>> +    sdiv   = (s->reg[I_(offset)]&   PLL_SDIV_MASK)>>   PLL_SDIV_SHIFT;
>>>> +
>>>> +    if (source) {
>>>> +        if (enable) {
>>>> +            pll->rate = mdiv * source->rate / (pdiv * (1<<   (sdiv-1)));
>>>> +        } else {
>>>> +            pll->rate = 0;
>>>> +        }
>>>> +    } else {
>>>> +        hw_error("%s: Source undefined for %s\n", __func__, pll->name);
>>>> +    }
>>>> +
>>>> +    QTAILQ_FOREACH(cce,&pll->clock_change_handler, entry) {
>>>>
>>>> +        cce->func(cce->opaque);
>>>> +    }
>>>> +
>>>> +    PRINT_DEBUG("%s rate: %llu\n", pll->name, pll->rate);
>>>
>>> pll->rate is of type uint64_t incompatible with "%llu"
>>
>>
>> Type uint64_t is included from /usr/include/stdint.h as
>>
>> typedef unsigned long long int  uint64_t;
>  From  /usr/include/stdint.h ,the uint64_t is defined by:
> #if __WORDSIZE == 64
> typedef unsigned long int	uint64_t;
> #else
> __extension__
> typedef unsigned long long int	uint64_t;
> #endif
> On my machine(64 bit), there will be a incompatibility error.

Yes, I have understood now. I don't use neither 64 bit system nor 
compiler and have not ever faced with this problem. Thanks for remarks

>
> As Peter said, you should use PRIu64 instead of llu
>>
>> and 'll' specifies that a following 'u' conversion specifier applies to a
>> unsigned long long argument
>>
>> Why do you think they incompatible?
>>
>>
>> --
>> MK
>
>
>