On Tue, Sep 25, 2018 at 12:45:41PM +0200, Peter Zijlstra wrote:
> On Tue, Sep 25, 2018 at 12:08:03PM +0200, Andrea Parri wrote:
> > Hi Guo,
> >
> > > +/*
> > > + * set_bit - Atomically set a bit in memory
> > > + * @nr: the bit to set
> > > + * @addr: the address to start counting from
> > > + *
> > > + * This function is atomic and may not be reordered. See __set_bit()
> > > + * if you do not require the atomic guarantees.
> > > + *
> > > + * Note: there are no guarantees that this function will not be reordered
> > > + * on non x86 architectures, so if you are writing portable code,
> > > + * make sure not to rely on its reordering guarantees.
> > > + *
> > > + * Note that @nr may be almost arbitrarily large; this function is not
> > > + * restricted to acting on a single-word quantity.
> > > + */
> > > +static inline void set_bit(int nr, volatile unsigned long *addr)
> > > +{
> > > + unsigned long mask = BIT_MASK(nr);
> > > + unsigned long *p = ((unsigned long *)addr) + BIT_WORD(nr);
> > > + unsigned long tmp;
> > > +
> > > + /* *p |= mask; */
> > > + smp_mb();
> > > + asm volatile (
> > > + "1: ldex.w %0, (%2) \n"
> > > + " or32 %0, %0, %1 \n"
> > > + " stex.w %0, (%2) \n"
> > > + " bez %0, 1b \n"
> > > + : "=&r"(tmp)
> > > + : "r"(mask), "r"(p)
> > > + : "memory");
> > > + smp_mb();
> > > +}
> > > +
> > > +/**
> > > + * clear_bit - Clears a bit in memory
> > > + * @nr: Bit to clear
> > > + * @addr: Address to start counting from
> > > + *
> > > + * clear_bit() is atomic and may not be reordered. However, it does
> > > + * not contain a memory barrier, so if it is used for locking purposes,
> > > + * you should call smp_mb__before_atomic() and/or smp_mb__after_atomic()
> > > + * in order to ensure changes are visible on other processors.
> > > + */
> > > +static inline void clear_bit(int nr, volatile unsigned long *addr)
> > > +{
> > > + unsigned long mask = BIT_MASK(nr);
> > > + unsigned long *p = ((unsigned long *)addr) + BIT_WORD(nr);
> > > + unsigned long tmp;
> > > +
> > > + /* *p &= ~mask; */
> > > + mask = ~mask;
> > > + smp_mb();
> > > + asm volatile (
> > > + "1: ldex.w %0, (%2) \n"
> > > + " and32 %0, %0, %1 \n"
> > > + " stex.w %0, (%2) \n"
> > > + " bez %0, 1b \n"
> > > + : "=&r"(tmp)
> > > + : "r"(mask), "r"(p)
> > > + : "memory");
> > > + smp_mb();
> > > +}
> > > +
> > > +/**
> > > + * change_bit - Toggle a bit in memory
> > > + * @nr: Bit to change
> > > + * @addr: Address to start counting from
> > > + *
> > > + * change_bit() is atomic and may not be reordered. It may be
> > > + * reordered on other architectures than x86.
> > > + * Note that @nr may be almost arbitrarily large; this function is not
> > > + * restricted to acting on a single-word quantity.
> > > + */
> > > +static inline void change_bit(int nr, volatile unsigned long *addr)
> > > +{
> > > + unsigned long mask = BIT_MASK(nr);
> > > + unsigned long *p = ((unsigned long *)addr) + BIT_WORD(nr);
> > > + unsigned long tmp;
> > > +
> > > + /* *p ^= mask; */
> > > + smp_mb();
> > > + asm volatile (
> > > + "1: ldex.w %0, (%2) \n"
> > > + " xor32 %0, %0, %1 \n"
> > > + " stex.w %0, (%2) \n"
> > > + " bez %0, 1b \n"
> > > + : "=&r"(tmp)
> > > + : "r"(mask), "r"(p)
> > > + : "memory");
> > > + smp_mb();
> > > +}
> >
> > The {set,clear,change}_bit() operations don't have to be ordered: you
> > might want to remove the above smp_mb()s (and adjust the comments).
I confused it with cmpxchg, seems cmpxchg need smp_mb() before&after.
See: https://lkml.org/lkml/2018/7/6/383
> Better yet, you can entirely delete all that and use
> asm-generic/bitops/atomic.h instead.
Yes, approve. Seems I only need care about asm/atomic.h, it's good for
me.
Best Regards
Guo Ren
