android_kernel_motorola_sm6225/include/asm-s390/system.h
Martin Schwidefsky 94c12cc7d1 [S390] Inline assembly cleanup.
Major cleanup of all s390 inline assemblies. They now have a common
coding style. Quite a few have been shortened, mainly by using register
asm variables. Use of the EX_TABLE macro helps  as well. The atomic ops,
bit ops and locking inlines new use the Q-constraint if a newer gcc
is used.  That results in slightly better code.

Thanks to Christian Borntraeger for proof reading the changes.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2006-09-28 16:56:43 +02:00

391 lines
9.2 KiB
C

/*
* include/asm-s390/system.h
*
* S390 version
* Copyright (C) 1999 IBM Deutschland Entwicklung GmbH, IBM Corporation
* Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com),
*
* Derived from "include/asm-i386/system.h"
*/
#ifndef __ASM_SYSTEM_H
#define __ASM_SYSTEM_H
#include <linux/kernel.h>
#include <asm/types.h>
#include <asm/ptrace.h>
#include <asm/setup.h>
#include <asm/processor.h>
#ifdef __KERNEL__
struct task_struct;
extern struct task_struct *__switch_to(void *, void *);
static inline void save_fp_regs(s390_fp_regs *fpregs)
{
asm volatile(
" std 0,8(%1)\n"
" std 2,24(%1)\n"
" std 4,40(%1)\n"
" std 6,56(%1)"
: "=m" (*fpregs) : "a" (fpregs), "m" (*fpregs) : "memory");
if (!MACHINE_HAS_IEEE)
return;
asm volatile(
" stfpc 0(%1)\n"
" std 1,16(%1)\n"
" std 3,32(%1)\n"
" std 5,48(%1)\n"
" std 7,64(%1)\n"
" std 8,72(%1)\n"
" std 9,80(%1)\n"
" std 10,88(%1)\n"
" std 11,96(%1)\n"
" std 12,104(%1)\n"
" std 13,112(%1)\n"
" std 14,120(%1)\n"
" std 15,128(%1)\n"
: "=m" (*fpregs) : "a" (fpregs), "m" (*fpregs) : "memory");
}
static inline void restore_fp_regs(s390_fp_regs *fpregs)
{
asm volatile(
" ld 0,8(%0)\n"
" ld 2,24(%0)\n"
" ld 4,40(%0)\n"
" ld 6,56(%0)"
: : "a" (fpregs), "m" (*fpregs));
if (!MACHINE_HAS_IEEE)
return;
asm volatile(
" lfpc 0(%0)\n"
" ld 1,16(%0)\n"
" ld 3,32(%0)\n"
" ld 5,48(%0)\n"
" ld 7,64(%0)\n"
" ld 8,72(%0)\n"
" ld 9,80(%0)\n"
" ld 10,88(%0)\n"
" ld 11,96(%0)\n"
" ld 12,104(%0)\n"
" ld 13,112(%0)\n"
" ld 14,120(%0)\n"
" ld 15,128(%0)\n"
: : "a" (fpregs), "m" (*fpregs));
}
static inline void save_access_regs(unsigned int *acrs)
{
asm volatile("stam 0,15,0(%0)" : : "a" (acrs) : "memory");
}
static inline void restore_access_regs(unsigned int *acrs)
{
asm volatile("lam 0,15,0(%0)" : : "a" (acrs));
}
#define switch_to(prev,next,last) do { \
if (prev == next) \
break; \
save_fp_regs(&prev->thread.fp_regs); \
restore_fp_regs(&next->thread.fp_regs); \
save_access_regs(&prev->thread.acrs[0]); \
restore_access_regs(&next->thread.acrs[0]); \
prev = __switch_to(prev,next); \
} while (0)
/*
* On SMP systems, when the scheduler does migration-cost autodetection,
* it needs a way to flush as much of the CPU's caches as possible.
*
* TODO: fill this in!
*/
static inline void sched_cacheflush(void)
{
}
#ifdef CONFIG_VIRT_CPU_ACCOUNTING
extern void account_vtime(struct task_struct *);
extern void account_tick_vtime(struct task_struct *);
extern void account_system_vtime(struct task_struct *);
#else
#define account_vtime(x) do { /* empty */ } while (0)
#endif
#define finish_arch_switch(prev) do { \
set_fs(current->thread.mm_segment); \
account_vtime(prev); \
} while (0)
#define nop() asm volatile("nop")
#define xchg(ptr,x) \
({ \
__typeof__(*(ptr)) __ret; \
__ret = (__typeof__(*(ptr))) \
__xchg((unsigned long)(x), (void *)(ptr),sizeof(*(ptr))); \
__ret; \
})
static inline unsigned long __xchg(unsigned long x, void * ptr, int size)
{
unsigned long addr, old;
int shift;
switch (size) {
case 1:
addr = (unsigned long) ptr;
shift = (3 ^ (addr & 3)) << 3;
addr ^= addr & 3;
asm volatile(
" l %0,0(%4)\n"
"0: lr 0,%0\n"
" nr 0,%3\n"
" or 0,%2\n"
" cs %0,0,0(%4)\n"
" jl 0b\n"
: "=&d" (old), "=m" (*(int *) addr)
: "d" (x << shift), "d" (~(255 << shift)), "a" (addr),
"m" (*(int *) addr) : "memory", "cc", "0");
x = old >> shift;
break;
case 2:
addr = (unsigned long) ptr;
shift = (2 ^ (addr & 2)) << 3;
addr ^= addr & 2;
asm volatile(
" l %0,0(%4)\n"
"0: lr 0,%0\n"
" nr 0,%3\n"
" or 0,%2\n"
" cs %0,0,0(%4)\n"
" jl 0b\n"
: "=&d" (old), "=m" (*(int *) addr)
: "d" (x << shift), "d" (~(65535 << shift)), "a" (addr),
"m" (*(int *) addr) : "memory", "cc", "0");
x = old >> shift;
break;
case 4:
asm volatile(
" l %0,0(%3)\n"
"0: cs %0,%2,0(%3)\n"
" jl 0b\n"
: "=&d" (old), "=m" (*(int *) ptr)
: "d" (x), "a" (ptr), "m" (*(int *) ptr)
: "memory", "cc");
x = old;
break;
#ifdef __s390x__
case 8:
asm volatile(
" lg %0,0(%3)\n"
"0: csg %0,%2,0(%3)\n"
" jl 0b\n"
: "=&d" (old), "=m" (*(long *) ptr)
: "d" (x), "a" (ptr), "m" (*(long *) ptr)
: "memory", "cc");
x = old;
break;
#endif /* __s390x__ */
}
return x;
}
/*
* Atomic compare and exchange. Compare OLD with MEM, if identical,
* store NEW in MEM. Return the initial value in MEM. Success is
* indicated by comparing RETURN with OLD.
*/
#define __HAVE_ARCH_CMPXCHG 1
#define cmpxchg(ptr,o,n)\
((__typeof__(*(ptr)))__cmpxchg((ptr),(unsigned long)(o),\
(unsigned long)(n),sizeof(*(ptr))))
static inline unsigned long
__cmpxchg(volatile void *ptr, unsigned long old, unsigned long new, int size)
{
unsigned long addr, prev, tmp;
int shift;
switch (size) {
case 1:
addr = (unsigned long) ptr;
shift = (3 ^ (addr & 3)) << 3;
addr ^= addr & 3;
asm volatile(
" l %0,0(%4)\n"
"0: nr %0,%5\n"
" lr %1,%0\n"
" or %0,%2\n"
" or %1,%3\n"
" cs %0,%1,0(%4)\n"
" jnl 1f\n"
" xr %1,%0\n"
" nr %1,%5\n"
" jnz 0b\n"
"1:"
: "=&d" (prev), "=&d" (tmp)
: "d" (old << shift), "d" (new << shift), "a" (ptr),
"d" (~(255 << shift))
: "memory", "cc");
return prev >> shift;
case 2:
addr = (unsigned long) ptr;
shift = (2 ^ (addr & 2)) << 3;
addr ^= addr & 2;
asm volatile(
" l %0,0(%4)\n"
"0: nr %0,%5\n"
" lr %1,%0\n"
" or %0,%2\n"
" or %1,%3\n"
" cs %0,%1,0(%4)\n"
" jnl 1f\n"
" xr %1,%0\n"
" nr %1,%5\n"
" jnz 0b\n"
"1:"
: "=&d" (prev), "=&d" (tmp)
: "d" (old << shift), "d" (new << shift), "a" (ptr),
"d" (~(65535 << shift))
: "memory", "cc");
return prev >> shift;
case 4:
asm volatile(
" cs %0,%2,0(%3)\n"
: "=&d" (prev) : "0" (old), "d" (new), "a" (ptr)
: "memory", "cc");
return prev;
#ifdef __s390x__
case 8:
asm volatile(
" csg %0,%2,0(%3)\n"
: "=&d" (prev) : "0" (old), "d" (new), "a" (ptr)
: "memory", "cc");
return prev;
#endif /* __s390x__ */
}
return old;
}
/*
* Force strict CPU ordering.
* And yes, this is required on UP too when we're talking
* to devices.
*
* This is very similar to the ppc eieio/sync instruction in that is
* does a checkpoint syncronisation & makes sure that
* all memory ops have completed wrt other CPU's ( see 7-15 POP DJB ).
*/
#define eieio() asm volatile("bcr 15,0" : : : "memory")
#define SYNC_OTHER_CORES(x) eieio()
#define mb() eieio()
#define rmb() eieio()
#define wmb() eieio()
#define read_barrier_depends() do { } while(0)
#define smp_mb() mb()
#define smp_rmb() rmb()
#define smp_wmb() wmb()
#define smp_read_barrier_depends() read_barrier_depends()
#define smp_mb__before_clear_bit() smp_mb()
#define smp_mb__after_clear_bit() smp_mb()
#define set_mb(var, value) do { var = value; mb(); } while (0)
#ifdef __s390x__
#define __ctl_load(array, low, high) ({ \
typedef struct { char _[sizeof(array)]; } addrtype; \
asm volatile( \
" lctlg %1,%2,0(%0)\n" \
: : "a" (&array), "i" (low), "i" (high), \
"m" (*(addrtype *)(array))); \
})
#define __ctl_store(array, low, high) ({ \
typedef struct { char _[sizeof(array)]; } addrtype; \
asm volatile( \
" stctg %2,%3,0(%1)\n" \
: "=m" (*(addrtype *)(array)) \
: "a" (&array), "i" (low), "i" (high)); \
})
#else /* __s390x__ */
#define __ctl_load(array, low, high) ({ \
typedef struct { char _[sizeof(array)]; } addrtype; \
asm volatile( \
" lctl %1,%2,0(%0)\n" \
: : "a" (&array), "i" (low), "i" (high), \
"m" (*(addrtype *)(array))); \
})
#define __ctl_store(array, low, high) ({ \
typedef struct { char _[sizeof(array)]; } addrtype; \
asm volatile( \
" stctl %2,%3,0(%1)\n" \
: "=m" (*(addrtype *)(array)) \
: "a" (&array), "i" (low), "i" (high)); \
})
#endif /* __s390x__ */
#define __ctl_set_bit(cr, bit) ({ \
unsigned long __dummy; \
__ctl_store(__dummy, cr, cr); \
__dummy |= 1UL << (bit); \
__ctl_load(__dummy, cr, cr); \
})
#define __ctl_clear_bit(cr, bit) ({ \
unsigned long __dummy; \
__ctl_store(__dummy, cr, cr); \
__dummy &= ~(1UL << (bit)); \
__ctl_load(__dummy, cr, cr); \
})
#include <linux/irqflags.h>
/*
* Use to set psw mask except for the first byte which
* won't be changed by this function.
*/
static inline void
__set_psw_mask(unsigned long mask)
{
__load_psw_mask(mask | (__raw_local_irq_stosm(0x00) & ~(-1UL >> 8)));
}
#define local_mcck_enable() __set_psw_mask(PSW_KERNEL_BITS)
#define local_mcck_disable() __set_psw_mask(PSW_KERNEL_BITS & ~PSW_MASK_MCHECK)
#ifdef CONFIG_SMP
extern void smp_ctl_set_bit(int cr, int bit);
extern void smp_ctl_clear_bit(int cr, int bit);
#define ctl_set_bit(cr, bit) smp_ctl_set_bit(cr, bit)
#define ctl_clear_bit(cr, bit) smp_ctl_clear_bit(cr, bit)
#else
#define ctl_set_bit(cr, bit) __ctl_set_bit(cr, bit)
#define ctl_clear_bit(cr, bit) __ctl_clear_bit(cr, bit)
#endif /* CONFIG_SMP */
extern void (*_machine_restart)(char *command);
extern void (*_machine_halt)(void);
extern void (*_machine_power_off)(void);
#define arch_align_stack(x) (x)
#endif /* __KERNEL__ */
#endif