kernel_optimize_test/include/asm-arm/system.h
Russell King 6d9b37a3a8 [PATCH] ARM SMP: Add ARMv6 memory barriers
Convert explicit gcc asm-based memory barriers into smp_mb() calls.
These change between barrier() and the ARMv6 data memory barrier
instruction depending on whether ARMv6 SMP is enabled.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2005-07-26 19:44:26 +01:00

415 lines
10 KiB
C

#ifndef __ASM_ARM_SYSTEM_H
#define __ASM_ARM_SYSTEM_H
#ifdef __KERNEL__
#include <linux/config.h>
#define CPU_ARCH_UNKNOWN 0
#define CPU_ARCH_ARMv3 1
#define CPU_ARCH_ARMv4 2
#define CPU_ARCH_ARMv4T 3
#define CPU_ARCH_ARMv5 4
#define CPU_ARCH_ARMv5T 5
#define CPU_ARCH_ARMv5TE 6
#define CPU_ARCH_ARMv5TEJ 7
#define CPU_ARCH_ARMv6 8
/*
* CR1 bits (CP#15 CR1)
*/
#define CR_M (1 << 0) /* MMU enable */
#define CR_A (1 << 1) /* Alignment abort enable */
#define CR_C (1 << 2) /* Dcache enable */
#define CR_W (1 << 3) /* Write buffer enable */
#define CR_P (1 << 4) /* 32-bit exception handler */
#define CR_D (1 << 5) /* 32-bit data address range */
#define CR_L (1 << 6) /* Implementation defined */
#define CR_B (1 << 7) /* Big endian */
#define CR_S (1 << 8) /* System MMU protection */
#define CR_R (1 << 9) /* ROM MMU protection */
#define CR_F (1 << 10) /* Implementation defined */
#define CR_Z (1 << 11) /* Implementation defined */
#define CR_I (1 << 12) /* Icache enable */
#define CR_V (1 << 13) /* Vectors relocated to 0xffff0000 */
#define CR_RR (1 << 14) /* Round Robin cache replacement */
#define CR_L4 (1 << 15) /* LDR pc can set T bit */
#define CR_DT (1 << 16)
#define CR_IT (1 << 18)
#define CR_ST (1 << 19)
#define CR_FI (1 << 21) /* Fast interrupt (lower latency mode) */
#define CR_U (1 << 22) /* Unaligned access operation */
#define CR_XP (1 << 23) /* Extended page tables */
#define CR_VE (1 << 24) /* Vectored interrupts */
#define CPUID_ID 0
#define CPUID_CACHETYPE 1
#define CPUID_TCM 2
#define CPUID_TLBTYPE 3
#define read_cpuid(reg) \
({ \
unsigned int __val; \
asm("mrc p15, 0, %0, c0, c0, " __stringify(reg) \
: "=r" (__val) \
: \
: "cc"); \
__val; \
})
/*
* This is used to ensure the compiler did actually allocate the register we
* asked it for some inline assembly sequences. Apparently we can't trust
* the compiler from one version to another so a bit of paranoia won't hurt.
* This string is meant to be concatenated with the inline asm string and
* will cause compilation to stop on mismatch.
* (for details, see gcc PR 15089)
*/
#define __asmeq(x, y) ".ifnc " x "," y " ; .err ; .endif\n\t"
#ifndef __ASSEMBLY__
#include <linux/linkage.h>
struct thread_info;
struct task_struct;
/* information about the system we're running on */
extern unsigned int system_rev;
extern unsigned int system_serial_low;
extern unsigned int system_serial_high;
extern unsigned int mem_fclk_21285;
struct pt_regs;
void die(const char *msg, struct pt_regs *regs, int err)
__attribute__((noreturn));
struct siginfo;
void notify_die(const char *str, struct pt_regs *regs, struct siginfo *info,
unsigned long err, unsigned long trap);
void hook_fault_code(int nr, int (*fn)(unsigned long, unsigned int,
struct pt_regs *),
int sig, const char *name);
#include <asm/proc-fns.h>
#define xchg(ptr,x) \
((__typeof__(*(ptr)))__xchg((unsigned long)(x),(ptr),sizeof(*(ptr))))
#define tas(ptr) (xchg((ptr),1))
extern asmlinkage void __backtrace(void);
extern asmlinkage void c_backtrace(unsigned long fp, int pmode);
extern void show_pte(struct mm_struct *mm, unsigned long addr);
extern void __show_regs(struct pt_regs *);
extern int cpu_architecture(void);
extern void cpu_init(void);
#define set_cr(x) \
__asm__ __volatile__( \
"mcr p15, 0, %0, c1, c0, 0 @ set CR" \
: : "r" (x) : "cc")
#define get_cr() \
({ \
unsigned int __val; \
__asm__ __volatile__( \
"mrc p15, 0, %0, c1, c0, 0 @ get CR" \
: "=r" (__val) : : "cc"); \
__val; \
})
extern unsigned long cr_no_alignment; /* defined in entry-armv.S */
extern unsigned long cr_alignment; /* defined in entry-armv.S */
#define UDBG_UNDEFINED (1 << 0)
#define UDBG_SYSCALL (1 << 1)
#define UDBG_BADABORT (1 << 2)
#define UDBG_SEGV (1 << 3)
#define UDBG_BUS (1 << 4)
extern unsigned int user_debug;
#if __LINUX_ARM_ARCH__ >= 4
#define vectors_high() (cr_alignment & CR_V)
#else
#define vectors_high() (0)
#endif
#if __LINUX_ARM_ARCH__ >= 6
#define mb() __asm__ __volatile__ ("mcr p15, 0, %0, c7, c10, 5" \
: : "r" (0) : "memory")
#else
#define mb() __asm__ __volatile__ ("" : : : "memory")
#endif
#define rmb() mb()
#define wmb() mb()
#define read_barrier_depends() do { } while(0)
#define set_mb(var, value) do { var = value; mb(); } while (0)
#define set_wmb(var, value) do { var = value; wmb(); } while (0)
#define nop() __asm__ __volatile__("mov\tr0,r0\t@ nop\n\t");
/*
* switch_mm() may do a full cache flush over the context switch,
* so enable interrupts over the context switch to avoid high
* latency.
*/
#define __ARCH_WANT_INTERRUPTS_ON_CTXSW
/*
* switch_to(prev, next) should switch from task `prev' to `next'
* `prev' will never be the same as `next'. schedule() itself
* contains the memory barrier to tell GCC not to cache `current'.
*/
extern struct task_struct *__switch_to(struct task_struct *, struct thread_info *, struct thread_info *);
#define switch_to(prev,next,last) \
do { \
last = __switch_to(prev,prev->thread_info,next->thread_info); \
} while (0)
/*
* CPU interrupt mask handling.
*/
#if __LINUX_ARM_ARCH__ >= 6
#define local_irq_save(x) \
({ \
__asm__ __volatile__( \
"mrs %0, cpsr @ local_irq_save\n" \
"cpsid i" \
: "=r" (x) : : "memory", "cc"); \
})
#define local_irq_enable() __asm__("cpsie i @ __sti" : : : "memory", "cc")
#define local_irq_disable() __asm__("cpsid i @ __cli" : : : "memory", "cc")
#define local_fiq_enable() __asm__("cpsie f @ __stf" : : : "memory", "cc")
#define local_fiq_disable() __asm__("cpsid f @ __clf" : : : "memory", "cc")
#else
/*
* Save the current interrupt enable state & disable IRQs
*/
#define local_irq_save(x) \
({ \
unsigned long temp; \
(void) (&temp == &x); \
__asm__ __volatile__( \
"mrs %0, cpsr @ local_irq_save\n" \
" orr %1, %0, #128\n" \
" msr cpsr_c, %1" \
: "=r" (x), "=r" (temp) \
: \
: "memory", "cc"); \
})
/*
* Enable IRQs
*/
#define local_irq_enable() \
({ \
unsigned long temp; \
__asm__ __volatile__( \
"mrs %0, cpsr @ local_irq_enable\n" \
" bic %0, %0, #128\n" \
" msr cpsr_c, %0" \
: "=r" (temp) \
: \
: "memory", "cc"); \
})
/*
* Disable IRQs
*/
#define local_irq_disable() \
({ \
unsigned long temp; \
__asm__ __volatile__( \
"mrs %0, cpsr @ local_irq_disable\n" \
" orr %0, %0, #128\n" \
" msr cpsr_c, %0" \
: "=r" (temp) \
: \
: "memory", "cc"); \
})
/*
* Enable FIQs
*/
#define local_fiq_enable() \
({ \
unsigned long temp; \
__asm__ __volatile__( \
"mrs %0, cpsr @ stf\n" \
" bic %0, %0, #64\n" \
" msr cpsr_c, %0" \
: "=r" (temp) \
: \
: "memory", "cc"); \
})
/*
* Disable FIQs
*/
#define local_fiq_disable() \
({ \
unsigned long temp; \
__asm__ __volatile__( \
"mrs %0, cpsr @ clf\n" \
" orr %0, %0, #64\n" \
" msr cpsr_c, %0" \
: "=r" (temp) \
: \
: "memory", "cc"); \
})
#endif
/*
* Save the current interrupt enable state.
*/
#define local_save_flags(x) \
({ \
__asm__ __volatile__( \
"mrs %0, cpsr @ local_save_flags" \
: "=r" (x) : : "memory", "cc"); \
})
/*
* restore saved IRQ & FIQ state
*/
#define local_irq_restore(x) \
__asm__ __volatile__( \
"msr cpsr_c, %0 @ local_irq_restore\n" \
: \
: "r" (x) \
: "memory", "cc")
#define irqs_disabled() \
({ \
unsigned long flags; \
local_save_flags(flags); \
(int)(flags & PSR_I_BIT); \
})
#ifdef CONFIG_SMP
#define smp_mb() mb()
#define smp_rmb() rmb()
#define smp_wmb() wmb()
#define smp_read_barrier_depends() read_barrier_depends()
#else
#define smp_mb() barrier()
#define smp_rmb() barrier()
#define smp_wmb() barrier()
#define smp_read_barrier_depends() do { } while(0)
#endif /* CONFIG_SMP */
#if defined(CONFIG_CPU_SA1100) || defined(CONFIG_CPU_SA110)
/*
* On the StrongARM, "swp" is terminally broken since it bypasses the
* cache totally. This means that the cache becomes inconsistent, and,
* since we use normal loads/stores as well, this is really bad.
* Typically, this causes oopsen in filp_close, but could have other,
* more disasterous effects. There are two work-arounds:
* 1. Disable interrupts and emulate the atomic swap
* 2. Clean the cache, perform atomic swap, flush the cache
*
* We choose (1) since its the "easiest" to achieve here and is not
* dependent on the processor type.
*
* NOTE that this solution won't work on an SMP system, so explcitly
* forbid it here.
*/
#define swp_is_buggy
#endif
static inline unsigned long __xchg(unsigned long x, volatile void *ptr, int size)
{
extern void __bad_xchg(volatile void *, int);
unsigned long ret;
#ifdef swp_is_buggy
unsigned long flags;
#endif
#if __LINUX_ARM_ARCH__ >= 6
unsigned int tmp;
#endif
switch (size) {
#if __LINUX_ARM_ARCH__ >= 6
case 1:
asm volatile("@ __xchg1\n"
"1: ldrexb %0, [%3]\n"
" strexb %1, %2, [%3]\n"
" teq %1, #0\n"
" bne 1b"
: "=&r" (ret), "=&r" (tmp)
: "r" (x), "r" (ptr)
: "memory", "cc");
break;
case 4:
asm volatile("@ __xchg4\n"
"1: ldrex %0, [%3]\n"
" strex %1, %2, [%3]\n"
" teq %1, #0\n"
" bne 1b"
: "=&r" (ret), "=&r" (tmp)
: "r" (x), "r" (ptr)
: "memory", "cc");
break;
#elif defined(swp_is_buggy)
#ifdef CONFIG_SMP
#error SMP is not supported on this platform
#endif
case 1:
local_irq_save(flags);
ret = *(volatile unsigned char *)ptr;
*(volatile unsigned char *)ptr = x;
local_irq_restore(flags);
break;
case 4:
local_irq_save(flags);
ret = *(volatile unsigned long *)ptr;
*(volatile unsigned long *)ptr = x;
local_irq_restore(flags);
break;
#else
case 1:
asm volatile("@ __xchg1\n"
" swpb %0, %1, [%2]"
: "=&r" (ret)
: "r" (x), "r" (ptr)
: "memory", "cc");
break;
case 4:
asm volatile("@ __xchg4\n"
" swp %0, %1, [%2]"
: "=&r" (ret)
: "r" (x), "r" (ptr)
: "memory", "cc");
break;
#endif
default:
__bad_xchg(ptr, size), ret = 0;
break;
}
return ret;
}
#endif /* __ASSEMBLY__ */
#define arch_align_stack(x) (x)
#endif /* __KERNEL__ */
#endif