summaryrefslogtreecommitdiffstats
path: root/include/asm-sparc/system_32.h
blob: b4b024445fc984f0d419f5f997690b83a8efd733 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
#ifndef __SPARC_SYSTEM_H
#define __SPARC_SYSTEM_H

#include <linux/kernel.h>
#include <linux/threads.h>	/* NR_CPUS */
#include <linux/thread_info.h>

#include <asm/page.h>
#include <asm/psr.h>
#include <asm/ptrace.h>
#include <asm/btfixup.h>
#include <asm/smp.h>

#ifndef __ASSEMBLY__

#include <linux/irqflags.h>

/*
 * Sparc (general) CPU types
 */
enum sparc_cpu {
  sun4        = 0x00,
  sun4c       = 0x01,
  sun4m       = 0x02,
  sun4d       = 0x03,
  sun4e       = 0x04,
  sun4u       = 0x05, /* V8 ploos ploos */
  sun_unknown = 0x06,
  ap1000      = 0x07, /* almost a sun4m */
};

/* Really, userland should not be looking at any of this... */
#ifdef __KERNEL__

extern enum sparc_cpu sparc_cpu_model;

#ifndef CONFIG_SUN4
#define ARCH_SUN4C_SUN4 (sparc_cpu_model==sun4c)
#define ARCH_SUN4 0
#else
#define ARCH_SUN4C_SUN4 1
#define ARCH_SUN4 1
#endif

#define SUN4M_NCPUS            4              /* Architectural limit of sun4m. */

extern char reboot_command[];

extern struct thread_info *current_set[NR_CPUS];

extern unsigned long empty_bad_page;
extern unsigned long empty_bad_page_table;
extern unsigned long empty_zero_page;

extern void sun_do_break(void);
extern int serial_console;
extern int stop_a_enabled;

static inline int con_is_present(void)
{
	return serial_console ? 0 : 1;
}

/* When a context switch happens we must flush all user windows so that
 * the windows of the current process are flushed onto its stack. This
 * way the windows are all clean for the next process and the stack
 * frames are up to date.
 */
extern void flush_user_windows(void);
extern void kill_user_windows(void);
extern void synchronize_user_stack(void);
extern void fpsave(unsigned long *fpregs, unsigned long *fsr,
		   void *fpqueue, unsigned long *fpqdepth);

#ifdef CONFIG_SMP
#define SWITCH_ENTER(prv) \
	do {			\
	if (test_tsk_thread_flag(prv, TIF_USEDFPU)) { \
		put_psr(get_psr() | PSR_EF); \
		fpsave(&(prv)->thread.float_regs[0], &(prv)->thread.fsr, \
		       &(prv)->thread.fpqueue[0], &(prv)->thread.fpqdepth); \
		clear_tsk_thread_flag(prv, TIF_USEDFPU); \
		(prv)->thread.kregs->psr &= ~PSR_EF; \
	} \
	} while(0)

#define SWITCH_DO_LAZY_FPU(next)	/* */
#else
#define SWITCH_ENTER(prv)		/* */
#define SWITCH_DO_LAZY_FPU(nxt)	\
	do {			\
	if (last_task_used_math != (nxt))		\
		(nxt)->thread.kregs->psr&=~PSR_EF;	\
	} while(0)
#endif

extern void flushw_all(void);

/*
 * Flush windows so that the VM switch which follows
 * would not pull the stack from under us.
 *
 * SWITCH_ENTER and SWITH_DO_LAZY_FPU do not work yet (e.g. SMP does not work)
 * XXX WTF is the above comment? Found in late teen 2.4.x.
 */
#define prepare_arch_switch(next) do { \
	__asm__ __volatile__( \
	".globl\tflush_patch_switch\nflush_patch_switch:\n\t" \
	"save %sp, -0x40, %sp; save %sp, -0x40, %sp; save %sp, -0x40, %sp\n\t" \
	"save %sp, -0x40, %sp; save %sp, -0x40, %sp; save %sp, -0x40, %sp\n\t" \
	"save %sp, -0x40, %sp\n\t" \
	"restore; restore; restore; restore; restore; restore; restore"); \
} while(0)

	/* Much care has gone into this code, do not touch it.
	 *
	 * We need to loadup regs l0/l1 for the newly forked child
	 * case because the trap return path relies on those registers
	 * holding certain values, gcc is told that they are clobbered.
	 * Gcc needs registers for 3 values in and 1 value out, so we
	 * clobber every non-fixed-usage register besides l2/l3/o4/o5.  -DaveM
	 *
	 * Hey Dave, that do not touch sign is too much of an incentive
	 * - Anton & Pete
	 */
#define switch_to(prev, next, last) do {						\
	SWITCH_ENTER(prev);								\
	SWITCH_DO_LAZY_FPU(next);							\
	cpu_set(smp_processor_id(), next->active_mm->cpu_vm_mask);			\
	__asm__ __volatile__(								\
	"sethi	%%hi(here - 0x8), %%o7\n\t"						\
	"mov	%%g6, %%g3\n\t"								\
	"or	%%o7, %%lo(here - 0x8), %%o7\n\t"					\
	"rd	%%psr, %%g4\n\t"							\
	"std	%%sp, [%%g6 + %4]\n\t"							\
	"rd	%%wim, %%g5\n\t"							\
	"wr	%%g4, 0x20, %%psr\n\t"							\
	"nop\n\t"									\
	"std	%%g4, [%%g6 + %3]\n\t"							\
	"ldd	[%2 + %3], %%g4\n\t"							\
	"mov	%2, %%g6\n\t"								\
	".globl	patchme_store_new_current\n"						\
"patchme_store_new_current:\n\t"							\
	"st	%2, [%1]\n\t"								\
	"wr	%%g4, 0x20, %%psr\n\t"							\
	"nop\n\t"									\
	"nop\n\t"									\
	"nop\n\t"	/* LEON needs all 3 nops: load to %sp depends on CWP. */		\
	"ldd	[%%g6 + %4], %%sp\n\t"							\
	"wr	%%g5, 0x0, %%wim\n\t"							\
	"ldd	[%%sp + 0x00], %%l0\n\t"						\
	"ldd	[%%sp + 0x38], %%i6\n\t"						\
	"wr	%%g4, 0x0, %%psr\n\t"							\
	"nop\n\t"									\
	"nop\n\t"									\
	"jmpl	%%o7 + 0x8, %%g0\n\t"							\
	" ld	[%%g3 + %5], %0\n\t"							\
	"here:\n"									\
        : "=&r" (last)									\
        : "r" (&(current_set[hard_smp_processor_id()])),	\
	  "r" (task_thread_info(next)),				\
	  "i" (TI_KPSR),					\
	  "i" (TI_KSP),						\
	  "i" (TI_TASK)						\
	:       "g1", "g2", "g3", "g4", "g5",       "g7",	\
	  "l0", "l1",       "l3", "l4", "l5", "l6", "l7",	\
	  "i0", "i1", "i2", "i3", "i4", "i5",			\
	  "o0", "o1", "o2", "o3",                   "o7");	\
	} while(0)

/* XXX Change this if we ever use a PSO mode kernel. */
#define mb()	__asm__ __volatile__ ("" : : : "memory")
#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 smp_mb()	__asm__ __volatile__("":::"memory")
#define smp_rmb()	__asm__ __volatile__("":::"memory")
#define smp_wmb()	__asm__ __volatile__("":::"memory")
#define smp_read_barrier_depends()	do { } while(0)

#define nop() __asm__ __volatile__ ("nop")

/* This has special calling conventions */
#ifndef CONFIG_SMP
BTFIXUPDEF_CALL(void, ___xchg32, void)
#endif

static inline unsigned long xchg_u32(__volatile__ unsigned long *m, unsigned long val)
{
#ifdef CONFIG_SMP
	__asm__ __volatile__("swap [%2], %0"
			     : "=&r" (val)
			     : "0" (val), "r" (m)
			     : "memory");
	return val;
#else
	register unsigned long *ptr asm("g1");
	register unsigned long ret asm("g2");

	ptr = (unsigned long *) m;
	ret = val;

	/* Note: this is magic and the nop there is
	   really needed. */
	__asm__ __volatile__(
	"mov	%%o7, %%g4\n\t"
	"call	___f____xchg32\n\t"
	" nop\n\t"
	: "=&r" (ret)
	: "0" (ret), "r" (ptr)
	: "g3", "g4", "g7", "memory", "cc");

	return ret;
#endif
}

#define xchg(ptr,x) ((__typeof__(*(ptr)))__xchg((unsigned long)(x),(ptr),sizeof(*(ptr))))

extern void __xchg_called_with_bad_pointer(void);

static inline unsigned long __xchg(unsigned long x, __volatile__ void * ptr, int size)
{
	switch (size) {
	case 4:
		return xchg_u32(ptr, x);
	};
	__xchg_called_with_bad_pointer();
	return x;
}

/* Emulate cmpxchg() the same way we emulate atomics,
 * by hashing the object address and indexing into an array
 * of spinlocks to get a bit of performance...
 *
 * See arch/sparc/lib/atomic32.c for implementation.
 *
 * Cribbed from <asm-parisc/atomic.h>
 */
#define __HAVE_ARCH_CMPXCHG	1

/* bug catcher for when unsupported size is used - won't link */
extern void __cmpxchg_called_with_bad_pointer(void);
/* we only need to support cmpxchg of a u32 on sparc */
extern unsigned long __cmpxchg_u32(volatile u32 *m, u32 old, u32 new_);

/* don't worry...optimizer will get rid of most of this */
static inline unsigned long
__cmpxchg(volatile void *ptr, unsigned long old, unsigned long new_, int size)
{
	switch (size) {
	case 4:
		return __cmpxchg_u32((u32 *)ptr, (u32)old, (u32)new_);
	default:
		__cmpxchg_called_with_bad_pointer();
		break;
	}
	return old;
}

#define cmpxchg(ptr, o, n)						\
({									\
	__typeof__(*(ptr)) _o_ = (o);					\
	__typeof__(*(ptr)) _n_ = (n);					\
	(__typeof__(*(ptr))) __cmpxchg((ptr), (unsigned long)_o_,	\
			(unsigned long)_n_, sizeof(*(ptr)));		\
})

#include <asm-generic/cmpxchg-local.h>

/*
 * cmpxchg_local and cmpxchg64_local are atomic wrt current CPU. Always make
 * them available.
 */
#define cmpxchg_local(ptr, o, n)				  	       \
	((__typeof__(*(ptr)))__cmpxchg_local_generic((ptr), (unsigned long)(o),\
			(unsigned long)(n), sizeof(*(ptr))))
#define cmpxchg64_local(ptr, o, n) __cmpxchg64_local_generic((ptr), (o), (n))

extern void die_if_kernel(char *str, struct pt_regs *regs) __attribute__ ((noreturn));

#endif /* __KERNEL__ */

#endif /* __ASSEMBLY__ */

#define arch_align_stack(x) (x)

#endif /* !(__SPARC_SYSTEM_H) */