Local atomic operations are fast and highly reentrant per CPU counters. Used for percpu variable updates. Local atomic operations only guarantee variable modification atomicity wrt the CPU which owns the data and these needs to be executed in a preemption safe way.
Here is the design of this patch. Since local_* operations are only need to be atomic to interrupts (IIUC), we have two options. Either replay the "op" if interrupted or replay the interrupt after the "op". Initial patchset posted was based on implementing local_* operation based on CR5 which replay's the "op". Patchset had issues in case of rewinding the address pointor from an array. This make the slow path really slow. Since CR5 based implementation proposed using __ex_table to find the rewind addressr, this rasied concerns about size of __ex_table and vmlinux. https://lists.ozlabs.org/pipermail/linuxppc-dev/2014-December/123115.html But this patch uses, powerpc_local_irq_pmu_save to soft_disable interrupts (including PMIs). After finishing the "op", powerpc_local_irq_pmu_restore() called and correspondingly interrupts are replayed if any occured. patch re-write the current local_* functions to use arch_local_irq_disbale. Base flow for each function is { powerpc_local_irq_pmu_save(flags) load .. store powerpc_local_irq_pmu_restore(flags) } Reason for the approach is that, currently l[w/d]arx/st[w/d]cx. instruction pair is used for local_* operations, which are heavy on cycle count and they dont support a local variant. So to see whether the new implementation helps, used a modified version of Rusty's benchmark code on local_t. https://lkml.org/lkml/2008/12/16/450 Modifications to Rusty's benchmark code: - Executed only local_t test Here are the values with the patch. Time in ns per iteration Local_t Without Patch With Patch _inc 38 10 _add 38 10 _read 4 4 _add_return 38 10 Currently only asm/local.h has been rewritten, and also the entire change is tested only in PPC64 (pseries guest) and PPC64 host (LE) Patch convert the inline asm implementation of the local_t to C. Also removed the local_dec_if_positive since there are no users for this. Reviewed-by: Nicholas Piggin <npig...@gmail.com> Signed-off-by: Madhavan Srinivasan <ma...@linux.vnet.ibm.com> --- arch/powerpc/include/asm/local.h | 200 +++++++++++++++++---------------------- 1 file changed, 86 insertions(+), 114 deletions(-) diff --git a/arch/powerpc/include/asm/local.h b/arch/powerpc/include/asm/local.h index b8da91363864..7795359b3663 100644 --- a/arch/powerpc/include/asm/local.h +++ b/arch/powerpc/include/asm/local.h @@ -3,74 +3,62 @@ #include <linux/percpu.h> #include <linux/atomic.h> +#include <linux/irqflags.h> + +#include <asm/hw_irq.h> + +#ifdef CONFIG_PPC64 typedef struct { - atomic_long_t a; + long v; } local_t; -#define LOCAL_INIT(i) { ATOMIC_LONG_INIT(i) } - -#define local_read(l) atomic_long_read(&(l)->a) -#define local_set(l,i) atomic_long_set(&(l)->a, (i)) +#define LOCAL_INIT(i) { (i) } -#define local_add(i,l) atomic_long_add((i),(&(l)->a)) -#define local_sub(i,l) atomic_long_sub((i),(&(l)->a)) -#define local_inc(l) atomic_long_inc(&(l)->a) -#define local_dec(l) atomic_long_dec(&(l)->a) - -static __inline__ long local_add_return(long a, local_t *l) +static __inline__ long local_read(local_t *l) { - long t; - - __asm__ __volatile__( -"1:" PPC_LLARX(%0,0,%2,0) " # local_add_return\n\ - add %0,%1,%0\n" - PPC405_ERR77(0,%2) - PPC_STLCX "%0,0,%2 \n\ - bne- 1b" - : "=&r" (t) - : "r" (a), "r" (&(l->a.counter)) - : "cc", "memory"); - - return t; + return READ_ONCE(l->v); } -#define local_add_negative(a, l) (local_add_return((a), (l)) < 0) - -static __inline__ long local_sub_return(long a, local_t *l) +static __inline__ void local_set(local_t *l, long i) { - long t; + WRITE_ONCE(l->v, i); +} - __asm__ __volatile__( -"1:" PPC_LLARX(%0,0,%2,0) " # local_sub_return\n\ - subf %0,%1,%0\n" - PPC405_ERR77(0,%2) - PPC_STLCX "%0,0,%2 \n\ - bne- 1b" - : "=&r" (t) - : "r" (a), "r" (&(l->a.counter)) - : "cc", "memory"); +#define LOCAL_OP(op, c_op) \ +static __inline__ void local_##op(long i, local_t *l) \ +{ \ + unsigned long flags; \ + \ + powerpc_local_irq_pmu_save(flags); \ + l->v c_op i; \ + powerpc_local_irq_pmu_restore(flags); \ +} - return t; +#define LOCAL_OP_RETURN(op, c_op) \ +static __inline__ long local_##op##_return(long a, local_t *l) \ +{ \ + long t; \ + unsigned long flags; \ + \ + powerpc_local_irq_pmu_save(flags); \ + t = (l->v c_op a); \ + powerpc_local_irq_pmu_restore(flags); \ + \ + return t; \ } -static __inline__ long local_inc_return(local_t *l) -{ - long t; +#define LOCAL_OPS(op, c_op) \ + LOCAL_OP(op, c_op) \ + LOCAL_OP_RETURN(op, c_op) - __asm__ __volatile__( -"1:" PPC_LLARX(%0,0,%1,0) " # local_inc_return\n\ - addic %0,%0,1\n" - PPC405_ERR77(0,%1) - PPC_STLCX "%0,0,%1 \n\ - bne- 1b" - : "=&r" (t) - : "r" (&(l->a.counter)) - : "cc", "xer", "memory"); +LOCAL_OPS(add, +=) +LOCAL_OPS(sub, -=) - return t; -} +#define local_add_negative(a, l) (local_add_return((a), (l)) < 0) +#define local_inc_return(l) local_add_return(1LL, l) +#define local_inc(l) local_inc_return(l) /* * local_inc_and_test - increment and test @@ -80,28 +68,39 @@ static __inline__ long local_inc_return(local_t *l) * and returns true if the result is zero, or false for all * other cases. */ -#define local_inc_and_test(l) (local_inc_return(l) == 0) +#define local_inc_and_test(l) (local_inc_return(l) == 0) -static __inline__ long local_dec_return(local_t *l) +#define local_dec_return(l) local_sub_return(1LL, l) +#define local_dec(l) local_dec_return(l) +#define local_sub_and_test(a, l) (local_sub_return((a), (l)) == 0) +#define local_dec_and_test(l) (local_dec_return((l)) == 0) + +static __inline__ long local_cmpxchg(local_t *l, long o, long n) { long t; + unsigned long flags; - __asm__ __volatile__( -"1:" PPC_LLARX(%0,0,%1,0) " # local_dec_return\n\ - addic %0,%0,-1\n" - PPC405_ERR77(0,%1) - PPC_STLCX "%0,0,%1\n\ - bne- 1b" - : "=&r" (t) - : "r" (&(l->a.counter)) - : "cc", "xer", "memory"); + powerpc_local_irq_pmu_save(flags); + t = l->v; + if (t == o) + l->v = n; + powerpc_local_irq_pmu_restore(flags); return t; } -#define local_cmpxchg(l, o, n) \ - (cmpxchg_local(&((l)->a.counter), (o), (n))) -#define local_xchg(l, n) (xchg_local(&((l)->a.counter), (n))) +static __inline__ long local_xchg(local_t *l, long n) +{ + long t; + unsigned long flags; + + powerpc_local_irq_pmu_save(flags); + t = l->v; + l->v = n; + powerpc_local_irq_pmu_restore(flags); + + return t; +} /** * local_add_unless - add unless the number is a given value @@ -114,62 +113,35 @@ static __inline__ long local_dec_return(local_t *l) */ static __inline__ int local_add_unless(local_t *l, long a, long u) { - long t; - - __asm__ __volatile__ ( -"1:" PPC_LLARX(%0,0,%1,0) " # local_add_unless\n\ - cmpw 0,%0,%3 \n\ - beq- 2f \n\ - add %0,%2,%0 \n" - PPC405_ERR77(0,%2) - PPC_STLCX "%0,0,%1 \n\ - bne- 1b \n" -" subf %0,%2,%0 \n\ -2:" - : "=&r" (t) - : "r" (&(l->a.counter)), "r" (a), "r" (u) - : "cc", "memory"); - - return t != u; -} - -#define local_inc_not_zero(l) local_add_unless((l), 1, 0) - -#define local_sub_and_test(a, l) (local_sub_return((a), (l)) == 0) -#define local_dec_and_test(l) (local_dec_return((l)) == 0) - -/* - * Atomically test *l and decrement if it is greater than 0. - * The function returns the old value of *l minus 1. - */ -static __inline__ long local_dec_if_positive(local_t *l) -{ - long t; + unsigned long flags; + int ret = 0; - __asm__ __volatile__( -"1:" PPC_LLARX(%0,0,%1,0) " # local_dec_if_positive\n\ - cmpwi %0,1\n\ - addi %0,%0,-1\n\ - blt- 2f\n" - PPC405_ERR77(0,%1) - PPC_STLCX "%0,0,%1\n\ - bne- 1b" - "\n\ -2:" : "=&b" (t) - : "r" (&(l->a.counter)) - : "cc", "memory"); + powerpc_local_irq_pmu_save(flags); + if (l->v != u) { + l->v += a; + ret = 1; + } + powerpc_local_irq_pmu_restore(flags); - return t; + return ret; } +#define local_inc_not_zero(l) local_add_unless((l), 1, 0) + /* Use these for per-cpu local_t variables: on some archs they are * much more efficient than these naive implementations. Note they take * a variable, not an address. */ -#define __local_inc(l) ((l)->a.counter++) -#define __local_dec(l) ((l)->a.counter++) -#define __local_add(i,l) ((l)->a.counter+=(i)) -#define __local_sub(i,l) ((l)->a.counter-=(i)) +#define __local_inc(l) ((l)->v++) +#define __local_dec(l) ((l)->v++) +#define __local_add(i,l) ((l)->v+=(i)) +#define __local_sub(i,l) ((l)->v-=(i)) + +#else /* CONFIG_PPC64 */ + +#include <asm-generic/local.h> + +#endif /* CONFIG_PPC64 */ #endif /* _ARCH_POWERPC_LOCAL_H */ -- 2.7.4
