On Wed, Nov 19, 2008 at 2:56 PM, Andrew Morton
<akpm@linux-foundation.org> wrote:
quoted text
> On Tue, 18 Nov 2008 17:05:05 +0800 Bryan Wu <cooloney@kernel.org> wrote:
>
>> From: Graf Yang <graf.yang@analog.com>
>>
>> Blackfin dual core BF561 processor can support SMP like features.
>> https://docs.blackfin.uclinux.org/doku.php?id=linux-kernel:smp-like
>>
>> In this patch, we provide SMP extend to Blackfin header files
>> and machine common code
>>
>>
>> ...
>>
>> +#define atomic_add_unless(v, a, u)                           \
>> +({                                                           \
>> +     int c, old;                                             \
>> +     c = atomic_read(v);                                     \
>> +     while (c != (u) && (old = atomic_cmpxchg((v), c, c + (a))) != c) \
>> +             c = old;                                        \
>> +     c != (u);                                               \
>> +})
>
> The macro references its args multiple times and will do weird or
> inefficient things when called with expressions which have
> side-effects, or which do slow things.
>

Right, I think we can replace them to inline functions


quoted text
>>
>> ...
>>
>> +#include <asm/system.h>              /* save_flags */
>> +
>> +static inline void set_bit(int nr, volatile unsigned long *addr)
>>  {
>>       int *a = (int *)addr;
>>       int mask;
>> @@ -57,21 +91,23 @@ static __inline__ void clear_bit(int nr, volatile unsigned long *addr)
>>       a += nr >> 5;
>>       mask = 1 << (nr & 0x1f);
>>       local_irq_save(flags);
>> -     *a &= ~mask;
>> +     *a |= mask;
>
> I think you just broke clear_bit().  Maybe I'm misreading the diff.
>
>>       local_irq_restore(flags);
>>  }
>>
>>
>> ...
>>
>> +#define smp_mb__before_clear_bit()   barrier()
>> +#define smp_mb__after_clear_bit()    barrier()
>> +
>> +static inline void __set_bit(int nr, volatile unsigned long *addr)
>> +{
>> +     int *a = (int *)addr;
>> +     int mask;
>> +
>> +     a += nr >> 5;
>> +     mask = 1 << (nr & 0x1f);
>> +     *a |= mask;
>> +}
>> +
>> +static inline void __clear_bit(int nr, volatile unsigned long *addr)
>> +{
>> +     int *a = (int *)addr;
>> +     int mask;
>> +
>> +     a += nr >> 5;
>> +     mask = 1 << (nr & 0x1f);
>> +     *a &= ~mask;
>> +}
>> +
>> +static inline void __change_bit(int nr, volatile unsigned long *addr)
>> +{
>> +     int mask;
>> +     unsigned long *ADDR = (unsigned long *)addr;
>> +
>> +     ADDR += nr >> 5;
>> +     mask = 1 << (nr & 31);
>> +     *ADDR ^= mask;
>> +}
>
> I'm surprised there isn't any generic code which can be used for the above.
>

As Nick said, include/asm-generic/bitops/non-atomic.h is the generic code.
We will try it.

quoted text
>>
>> ...
>>
>
> Gad what a lot of code.  I don't think I have time to read it all, sorry.
>

Thanks a lot for the review, I will forward this patchset to linux-arch.
-Bryan
--
unsubscribe notice
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/