kernel-fxtec-pro1x/arch/parisc/include/asm/system.h
Helge Deller 7d17e27631 parisc: fix ldcw inline assembler
There are two reasons to expose the memory *a in the asm:

1) To prevent the compiler from discarding a preceeding write to *a, and
2) to prevent it from caching *a in a register over the asm.

The change has had a few days testing with a SMP build of 2.6.22.19
running on a rp3440.

This patch is about the correctness of the __ldcw() macro itself.
The use of the macro should be confined to small inline functions
to try to limit the effect of clobbering memory on GCC's optimization
of loads and stores.

Signed-off-by: Dave Anglin <dave.anglin@nrc-cnrc.gc.ca>
Signed-off-by: Helge Deller <deller@gmx.de>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
2009-07-03 03:34:07 +00:00

182 lines
5.3 KiB
C

#ifndef __PARISC_SYSTEM_H
#define __PARISC_SYSTEM_H
#include <asm/psw.h>
/* The program status word as bitfields. */
struct pa_psw {
unsigned int y:1;
unsigned int z:1;
unsigned int rv:2;
unsigned int w:1;
unsigned int e:1;
unsigned int s:1;
unsigned int t:1;
unsigned int h:1;
unsigned int l:1;
unsigned int n:1;
unsigned int x:1;
unsigned int b:1;
unsigned int c:1;
unsigned int v:1;
unsigned int m:1;
unsigned int cb:8;
unsigned int o:1;
unsigned int g:1;
unsigned int f:1;
unsigned int r:1;
unsigned int q:1;
unsigned int p:1;
unsigned int d:1;
unsigned int i:1;
};
#ifdef CONFIG_64BIT
#define pa_psw(task) ((struct pa_psw *) ((char *) (task) + TASK_PT_PSW + 4))
#else
#define pa_psw(task) ((struct pa_psw *) ((char *) (task) + TASK_PT_PSW))
#endif
struct task_struct;
extern struct task_struct *_switch_to(struct task_struct *, struct task_struct *);
#define switch_to(prev, next, last) do { \
(last) = _switch_to(prev, next); \
} while(0)
/* interrupt control */
#define local_save_flags(x) __asm__ __volatile__("ssm 0, %0" : "=r" (x) : : "memory")
#define local_irq_disable() __asm__ __volatile__("rsm %0,%%r0\n" : : "i" (PSW_I) : "memory" )
#define local_irq_enable() __asm__ __volatile__("ssm %0,%%r0\n" : : "i" (PSW_I) : "memory" )
#define local_irq_save(x) \
__asm__ __volatile__("rsm %1,%0" : "=r" (x) :"i" (PSW_I) : "memory" )
#define local_irq_restore(x) \
__asm__ __volatile__("mtsm %0" : : "r" (x) : "memory" )
#define irqs_disabled() \
({ \
unsigned long flags; \
local_save_flags(flags); \
(flags & PSW_I) == 0; \
})
#define mfctl(reg) ({ \
unsigned long cr; \
__asm__ __volatile__( \
"mfctl " #reg ",%0" : \
"=r" (cr) \
); \
cr; \
})
#define mtctl(gr, cr) \
__asm__ __volatile__("mtctl %0,%1" \
: /* no outputs */ \
: "r" (gr), "i" (cr) : "memory")
/* these are here to de-mystefy the calling code, and to provide hooks */
/* which I needed for debugging EIEM problems -PB */
#define get_eiem() mfctl(15)
static inline void set_eiem(unsigned long val)
{
mtctl(val, 15);
}
#define mfsp(reg) ({ \
unsigned long cr; \
__asm__ __volatile__( \
"mfsp " #reg ",%0" : \
"=r" (cr) \
); \
cr; \
})
#define mtsp(gr, cr) \
__asm__ __volatile__("mtsp %0,%1" \
: /* no outputs */ \
: "r" (gr), "i" (cr) : "memory")
/*
** This is simply the barrier() macro from linux/kernel.h but when serial.c
** uses tqueue.h uses smp_mb() defined using barrier(), linux/kernel.h
** hasn't yet been included yet so it fails, thus repeating the macro here.
**
** PA-RISC architecture allows for weakly ordered memory accesses although
** none of the processors use it. There is a strong ordered bit that is
** set in the O-bit of the page directory entry. Operating systems that
** can not tolerate out of order accesses should set this bit when mapping
** pages. The O-bit of the PSW should also be set to 1 (I don't believe any
** of the processor implemented the PSW O-bit). The PCX-W ERS states that
** the TLB O-bit is not implemented so the page directory does not need to
** have the O-bit set when mapping pages (section 3.1). This section also
** states that the PSW Y, Z, G, and O bits are not implemented.
** So it looks like nothing needs to be done for parisc-linux (yet).
** (thanks to chada for the above comment -ggg)
**
** The __asm__ op below simple prevents gcc/ld from reordering
** instructions across the mb() "call".
*/
#define mb() __asm__ __volatile__("":::"memory") /* barrier() */
#define rmb() mb()
#define wmb() mb()
#define smp_mb() mb()
#define smp_rmb() mb()
#define smp_wmb() mb()
#define smp_read_barrier_depends() do { } while(0)
#define read_barrier_depends() do { } while(0)
#define set_mb(var, value) do { var = value; mb(); } while (0)
#ifndef CONFIG_PA20
/* Because kmalloc only guarantees 8-byte alignment for kmalloc'd data,
and GCC only guarantees 8-byte alignment for stack locals, we can't
be assured of 16-byte alignment for atomic lock data even if we
specify "__attribute ((aligned(16)))" in the type declaration. So,
we use a struct containing an array of four ints for the atomic lock
type and dynamically select the 16-byte aligned int from the array
for the semaphore. */
#define __PA_LDCW_ALIGNMENT 16
#define __ldcw_align(a) ({ \
unsigned long __ret = (unsigned long) &(a)->lock[0]; \
__ret = (__ret + __PA_LDCW_ALIGNMENT - 1) \
& ~(__PA_LDCW_ALIGNMENT - 1); \
(volatile unsigned int *) __ret; \
})
#define __LDCW "ldcw"
#else /*CONFIG_PA20*/
/* From: "Jim Hull" <jim.hull of hp.com>
I've attached a summary of the change, but basically, for PA 2.0, as
long as the ",CO" (coherent operation) completer is specified, then the
16-byte alignment requirement for ldcw and ldcd is relaxed, and instead
they only require "natural" alignment (4-byte for ldcw, 8-byte for
ldcd). */
#define __PA_LDCW_ALIGNMENT 4
#define __ldcw_align(a) ((volatile unsigned int *)a)
#define __LDCW "ldcw,co"
#endif /*!CONFIG_PA20*/
/* LDCW, the only atomic read-write operation PA-RISC has. *sigh*. */
#define __ldcw(a) ({ \
unsigned __ret; \
__asm__ __volatile__(__LDCW " 0(%2),%0" \
: "=r" (__ret), "+m" (*(a)) : "r" (a)); \
__ret; \
})
#ifdef CONFIG_SMP
# define __lock_aligned __attribute__((__section__(".data.lock_aligned")))
#endif
#define arch_align_stack(x) (x)
#endif