mmu-hash.h 20.5 KB
Newer Older
1 2
#ifndef _ASM_POWERPC_BOOK3S_64_MMU_HASH_H_
#define _ASM_POWERPC_BOOK3S_64_MMU_HASH_H_
3 4 5 6 7 8 9 10 11 12 13 14 15 16
/*
 * PowerPC64 memory management structures
 *
 * Dave Engebretsen & Mike Corrigan <{engebret|mikejc}@us.ibm.com>
 *   PPC64 rework.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; either version
 * 2 of the License, or (at your option) any later version.
 */

#include <asm/asm-compat.h>
#include <asm/page.h>
17
#include <asm/bug.h>
18

19 20 21 22 23
/*
 * This is necessary to get the definition of PGTABLE_RANGE which we
 * need for various slices related matters. Note that this isn't the
 * complete pgtable.h but only a portion of it.
 */
24
#include <asm/book3s/64/pgtable.h>
25
#include <asm/bug.h>
26
#include <asm/processor.h>
27
#include <asm/cpu_has_feature.h>
28

29 30 31 32 33 34
/*
 * SLB
 */

#define SLB_NUM_BOLTED		3
#define SLB_CACHE_ENTRIES	8
35
#define SLB_MIN_SIZE		32
36 37 38 39 40 41

/* Bits in the SLB ESID word */
#define SLB_ESID_V		ASM_CONST(0x0000000008000000) /* valid */

/* Bits in the SLB VSID word */
#define SLB_VSID_SHIFT		12
P
Paul Mackerras 已提交
42 43
#define SLB_VSID_SHIFT_1T	24
#define SLB_VSID_SSIZE_SHIFT	62
44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62
#define SLB_VSID_B		ASM_CONST(0xc000000000000000)
#define SLB_VSID_B_256M		ASM_CONST(0x0000000000000000)
#define SLB_VSID_B_1T		ASM_CONST(0x4000000000000000)
#define SLB_VSID_KS		ASM_CONST(0x0000000000000800)
#define SLB_VSID_KP		ASM_CONST(0x0000000000000400)
#define SLB_VSID_N		ASM_CONST(0x0000000000000200) /* no-execute */
#define SLB_VSID_L		ASM_CONST(0x0000000000000100)
#define SLB_VSID_C		ASM_CONST(0x0000000000000080) /* class */
#define SLB_VSID_LP		ASM_CONST(0x0000000000000030)
#define SLB_VSID_LP_00		ASM_CONST(0x0000000000000000)
#define SLB_VSID_LP_01		ASM_CONST(0x0000000000000010)
#define SLB_VSID_LP_10		ASM_CONST(0x0000000000000020)
#define SLB_VSID_LP_11		ASM_CONST(0x0000000000000030)
#define SLB_VSID_LLP		(SLB_VSID_L|SLB_VSID_LP)

#define SLB_VSID_KERNEL		(SLB_VSID_KP)
#define SLB_VSID_USER		(SLB_VSID_KP|SLB_VSID_KS|SLB_VSID_C)

#define SLBIE_C			(0x08000000)
P
Paul Mackerras 已提交
63
#define SLBIE_SSIZE_SHIFT	25
64 65 66 67 68 69 70

/*
 * Hash table
 */

#define HPTES_PER_GROUP 8

71
#define HPTE_V_SSIZE_SHIFT	62
72
#define HPTE_V_AVPN_SHIFT	7
73
#define HPTE_V_AVPN		ASM_CONST(0x3fffffffffffff80)
74
#define HPTE_V_AVPN_VAL(x)	(((x) & HPTE_V_AVPN) >> HPTE_V_AVPN_SHIFT)
75
#define HPTE_V_COMPARE(x,y)	(!(((x) ^ (y)) & 0xffffffffffffff80UL))
76 77 78 79 80 81
#define HPTE_V_BOLTED		ASM_CONST(0x0000000000000010)
#define HPTE_V_LOCK		ASM_CONST(0x0000000000000008)
#define HPTE_V_LARGE		ASM_CONST(0x0000000000000004)
#define HPTE_V_SECONDARY	ASM_CONST(0x0000000000000002)
#define HPTE_V_VALID		ASM_CONST(0x0000000000000001)

82 83 84 85
/*
 * ISA 3.0 have a different HPTE format.
 */
#define HPTE_R_3_0_SSIZE_SHIFT	58
86 87
#define HPTE_R_PP0		ASM_CONST(0x8000000000000000)
#define HPTE_R_TS		ASM_CONST(0x4000000000000000)
88
#define HPTE_R_KEY_HI		ASM_CONST(0x3000000000000000)
89
#define HPTE_R_RPN_SHIFT	12
90
#define HPTE_R_RPN		ASM_CONST(0x0ffffffffffff000)
91
#define HPTE_R_PP		ASM_CONST(0x0000000000000003)
92
#define HPTE_R_PPP		ASM_CONST(0x8000000000000003)
93
#define HPTE_R_N		ASM_CONST(0x0000000000000004)
94 95 96 97 98
#define HPTE_R_G		ASM_CONST(0x0000000000000008)
#define HPTE_R_M		ASM_CONST(0x0000000000000010)
#define HPTE_R_I		ASM_CONST(0x0000000000000020)
#define HPTE_R_W		ASM_CONST(0x0000000000000040)
#define HPTE_R_WIMG		ASM_CONST(0x0000000000000078)
99 100
#define HPTE_R_C		ASM_CONST(0x0000000000000080)
#define HPTE_R_R		ASM_CONST(0x0000000000000100)
101
#define HPTE_R_KEY_LO		ASM_CONST(0x0000000000000e00)
102

103 104 105
#define HPTE_V_1TB_SEG		ASM_CONST(0x4000000000000000)
#define HPTE_V_VRMA_MASK	ASM_CONST(0x4001ffffff000000)

106 107 108 109 110
/* Values for PP (assumes Ks=0, Kp=1) */
#define PP_RWXX	0	/* Supervisor read/write, User none */
#define PP_RWRX 1	/* Supervisor read/write, User read */
#define PP_RWRW 2	/* Supervisor read/write, User read/write */
#define PP_RXRX 3	/* Supervisor read,       User read */
111
#define PP_RXXX	(HPTE_R_PP0 | 2)	/* Supervisor read, user none */
112

113 114 115 116 117 118 119 120 121
/* Fields for tlbiel instruction in architecture 2.06 */
#define TLBIEL_INVAL_SEL_MASK	0xc00	/* invalidation selector */
#define  TLBIEL_INVAL_PAGE	0x000	/* invalidate a single page */
#define  TLBIEL_INVAL_SET_LPID	0x800	/* invalidate a set for current LPID */
#define  TLBIEL_INVAL_SET	0xc00	/* invalidate a set for all LPIDs */
#define TLBIEL_INVAL_SET_MASK	0xfff000	/* set number to inval. */
#define TLBIEL_INVAL_SET_SHIFT	12

#define POWER7_TLB_SETS		128	/* # sets in POWER7 TLB */
122
#define POWER8_TLB_SETS		512	/* # sets in POWER8 TLB */
123
#define POWER9_TLB_SETS_HASH	256	/* # sets in POWER9 TLB Hash mode */
124
#define POWER9_TLB_SETS_RADIX	128	/* # sets in POWER9 TLB Radix mode */
125

126 127
#ifndef __ASSEMBLY__

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
struct mmu_hash_ops {
	void            (*hpte_invalidate)(unsigned long slot,
					   unsigned long vpn,
					   int bpsize, int apsize,
					   int ssize, int local);
	long		(*hpte_updatepp)(unsigned long slot,
					 unsigned long newpp,
					 unsigned long vpn,
					 int bpsize, int apsize,
					 int ssize, unsigned long flags);
	void            (*hpte_updateboltedpp)(unsigned long newpp,
					       unsigned long ea,
					       int psize, int ssize);
	long		(*hpte_insert)(unsigned long hpte_group,
				       unsigned long vpn,
				       unsigned long prpn,
				       unsigned long rflags,
				       unsigned long vflags,
				       int psize, int apsize,
				       int ssize);
	long		(*hpte_remove)(unsigned long hpte_group);
	int             (*hpte_removebolted)(unsigned long ea,
					     int psize, int ssize);
	void		(*flush_hash_range)(unsigned long number, int local);
	void		(*hugepage_invalidate)(unsigned long vsid,
					       unsigned long addr,
					       unsigned char *hpte_slot_array,
					       int psize, int ssize, int local);
	/*
	 * Special for kexec.
	 * To be called in real mode with interrupts disabled. No locks are
	 * taken as such, concurrent access on pre POWER5 hardware could result
	 * in a deadlock.
	 * The linear mapping is destroyed as well.
	 */
	void		(*hpte_clear_all)(void);
};
extern struct mmu_hash_ops mmu_hash_ops;

167
struct hash_pte {
168 169
	__be64 v;
	__be64 r;
170
};
171

172
extern struct hash_pte *htab_address;
173 174 175
extern unsigned long htab_size_bytes;
extern unsigned long htab_hash_mask;

176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192

static inline int shift_to_mmu_psize(unsigned int shift)
{
	int psize;

	for (psize = 0; psize < MMU_PAGE_COUNT; ++psize)
		if (mmu_psize_defs[psize].shift == shift)
			return psize;
	return -1;
}

static inline unsigned int mmu_psize_to_shift(unsigned int mmu_psize)
{
	if (mmu_psize_defs[mmu_psize].shift)
		return mmu_psize_defs[mmu_psize].shift;
	BUG();
}
193

194 195 196 197 198 199 200 201 202
static inline unsigned long get_sllp_encoding(int psize)
{
	unsigned long sllp;

	sllp = ((mmu_psize_defs[psize].sllp & SLB_VSID_L) >> 6) |
		((mmu_psize_defs[psize].sllp & SLB_VSID_LP) >> 4);
	return sllp;
}

203 204
#endif /* __ASSEMBLY__ */

205 206 207 208 209 210 211 212 213
/*
 * Segment sizes.
 * These are the values used by hardware in the B field of
 * SLB entries and the first dword of MMU hashtable entries.
 * The B field is 2 bits; the values 2 and 3 are unused and reserved.
 */
#define MMU_SEGSIZE_256M	0
#define MMU_SEGSIZE_1T		1

214 215 216 217 218 219 220 221 222
/*
 * encode page number shift.
 * in order to fit the 78 bit va in a 64 bit variable we shift the va by
 * 12 bits. This enable us to address upto 76 bit va.
 * For hpt hash from a va we can ignore the page size bits of va and for
 * hpte encoding we ignore up to 23 bits of va. So ignoring lower 12 bits ensure
 * we work in all cases including 4k page size.
 */
#define VPN_SHIFT	12
P
Paul Mackerras 已提交
223

224 225 226 227 228 229 230
/*
 * HPTE Large Page (LP) details
 */
#define LP_SHIFT	12
#define LP_BITS		8
#define LP_MASK(i)	((0xFF >> (i)) << LP_SHIFT)

231 232
#ifndef __ASSEMBLY__

233 234 235 236 237 238 239
static inline int slb_vsid_shift(int ssize)
{
	if (ssize == MMU_SEGSIZE_256M)
		return SLB_VSID_SHIFT;
	return SLB_VSID_SHIFT_1T;
}

240 241 242 243 244 245 246
static inline int segment_shift(int ssize)
{
	if (ssize == MMU_SEGSIZE_256M)
		return SID_SHIFT;
	return SID_SHIFT_1T;
}

247
/*
P
Paul Mackerras 已提交
248
 * The current system page and segment sizes
249
 */
P
Paul Mackerras 已提交
250 251
extern int mmu_kernel_ssize;
extern int mmu_highuser_ssize;
252
extern u16 mmu_slb_size;
253
extern unsigned long tce_alloc_start, tce_alloc_end;
254 255 256 257 258 259 260 261 262

/*
 * If the processor supports 64k normal pages but not 64k cache
 * inhibited pages, we have to be prepared to switch processes
 * to use 4k pages when they create cache-inhibited mappings.
 * If this is the case, mmu_ci_restrictions will be set to 1.
 */
extern int mmu_ci_restrictions;

263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281
/*
 * This computes the AVPN and B fields of the first dword of a HPTE,
 * for use when we want to match an existing PTE.  The bottom 7 bits
 * of the returned value are zero.
 */
static inline unsigned long hpte_encode_avpn(unsigned long vpn, int psize,
					     int ssize)
{
	unsigned long v;
	/*
	 * The AVA field omits the low-order 23 bits of the 78 bits VA.
	 * These bits are not needed in the PTE, because the
	 * low-order b of these bits are part of the byte offset
	 * into the virtual page and, if b < 23, the high-order
	 * 23-b of these bits are always used in selecting the
	 * PTEGs to be searched
	 */
	v = (vpn >> (23 - VPN_SHIFT)) & ~(mmu_psize_defs[psize].avpnm);
	v <<= HPTE_V_AVPN_SHIFT;
282 283
	if (!cpu_has_feature(CPU_FTR_ARCH_300))
		v |= ((unsigned long) ssize) << HPTE_V_SSIZE_SHIFT;
284 285 286
	return v;
}

287 288
/*
 * This function sets the AVPN and L fields of the HPTE  appropriately
289
 * using the base page size and actual page size.
290
 */
291 292
static inline unsigned long hpte_encode_v(unsigned long vpn, int base_psize,
					  int actual_psize, int ssize)
293
{
P
Paul Mackerras 已提交
294
	unsigned long v;
295 296
	v = hpte_encode_avpn(vpn, base_psize, ssize);
	if (actual_psize != MMU_PAGE_4K)
297 298 299 300 301 302 303 304 305
		v |= HPTE_V_LARGE;
	return v;
}

/*
 * This function sets the ARPN, and LP fields of the HPTE appropriately
 * for the page size. We assume the pa is already "clean" that is properly
 * aligned for the requested page size
 */
306
static inline unsigned long hpte_encode_r(unsigned long pa, int base_psize,
307
					  int actual_psize, int ssize)
308
{
309 310 311 312

	if (cpu_has_feature(CPU_FTR_ARCH_300))
		pa |= ((unsigned long) ssize) << HPTE_R_3_0_SSIZE_SHIFT;

313
	/* A 4K page needs no special encoding */
314
	if (actual_psize == MMU_PAGE_4K)
315 316
		return pa & HPTE_R_RPN;
	else {
317 318 319
		unsigned int penc = mmu_psize_defs[base_psize].penc[actual_psize];
		unsigned int shift = mmu_psize_defs[actual_psize].shift;
		return (pa & ~((1ul << shift) - 1)) | (penc << LP_SHIFT);
320 321 322 323
	}
}

/*
324
 * Build a VPN_SHIFT bit shifted va given VSID, EA and segment size.
325
 */
326 327
static inline unsigned long hpt_vpn(unsigned long ea,
				    unsigned long vsid, int ssize)
P
Paul Mackerras 已提交
328
{
329 330 331 332 333
	unsigned long mask;
	int s_shift = segment_shift(ssize);

	mask = (1ul << (s_shift - VPN_SHIFT)) - 1;
	return (vsid << (s_shift - VPN_SHIFT)) | ((ea >> VPN_SHIFT) & mask);
P
Paul Mackerras 已提交
334
}
335

P
Paul Mackerras 已提交
336 337 338
/*
 * This hashes a virtual address
 */
339 340
static inline unsigned long hpt_hash(unsigned long vpn,
				     unsigned int shift, int ssize)
341
{
342
	int mask;
P
Paul Mackerras 已提交
343 344
	unsigned long hash, vsid;

345
	/* VPN_SHIFT can be atmost 12 */
P
Paul Mackerras 已提交
346
	if (ssize == MMU_SEGSIZE_256M) {
347 348 349
		mask = (1ul << (SID_SHIFT - VPN_SHIFT)) - 1;
		hash = (vpn >> (SID_SHIFT - VPN_SHIFT)) ^
			((vpn & mask) >> (shift - VPN_SHIFT));
P
Paul Mackerras 已提交
350
	} else {
351 352 353 354
		mask = (1ul << (SID_SHIFT_1T - VPN_SHIFT)) - 1;
		vsid = vpn >> (SID_SHIFT_1T - VPN_SHIFT);
		hash = vsid ^ (vsid << 25) ^
			((vpn & mask) >> (shift - VPN_SHIFT)) ;
P
Paul Mackerras 已提交
355 356
	}
	return hash & 0x7fffffffffUL;
357 358
}

359 360 361
#define HPTE_LOCAL_UPDATE	0x1
#define HPTE_NOHPTE_UPDATE	0x2

362 363
extern int __hash_page_4K(unsigned long ea, unsigned long access,
			  unsigned long vsid, pte_t *ptep, unsigned long trap,
364
			  unsigned long flags, int ssize, int subpage_prot);
365 366
extern int __hash_page_64K(unsigned long ea, unsigned long access,
			   unsigned long vsid, pte_t *ptep, unsigned long trap,
367
			   unsigned long flags, int ssize);
368
struct mm_struct;
369
unsigned int hash_page_do_lazy_icache(unsigned int pp, pte_t pte, int trap);
370 371 372 373 374
extern int hash_page_mm(struct mm_struct *mm, unsigned long ea,
			unsigned long access, unsigned long trap,
			unsigned long flags);
extern int hash_page(unsigned long ea, unsigned long access, unsigned long trap,
		     unsigned long dsisr);
375
int __hash_page_huge(unsigned long ea, unsigned long access, unsigned long vsid,
376 377
		     pte_t *ptep, unsigned long trap, unsigned long flags,
		     int ssize, unsigned int shift, unsigned int mmu_psize);
378 379 380
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
extern int __hash_page_thp(unsigned long ea, unsigned long access,
			   unsigned long vsid, pmd_t *pmdp, unsigned long trap,
381
			   unsigned long flags, int ssize, unsigned int psize);
382 383 384
#else
static inline int __hash_page_thp(unsigned long ea, unsigned long access,
				  unsigned long vsid, pmd_t *pmdp,
385
				  unsigned long trap, unsigned long flags,
386 387 388
				  int ssize, unsigned int psize)
{
	BUG();
389
	return -1;
390 391
}
#endif
392 393
extern void hash_failure_debug(unsigned long ea, unsigned long access,
			       unsigned long vsid, unsigned long trap,
394 395
			       int ssize, int psize, int lpsize,
			       unsigned long pte);
396
extern int htab_bolt_mapping(unsigned long vstart, unsigned long vend,
397
			     unsigned long pstart, unsigned long prot,
P
Paul Mackerras 已提交
398
			     int psize, int ssize);
399 400
int htab_remove_mapping(unsigned long vstart, unsigned long vend,
			int psize, int ssize);
B
Becky Bruce 已提交
401
extern void add_gpage(u64 addr, u64 page_size, unsigned long number_of_pages);
402
extern void demote_segment_4k(struct mm_struct *mm, unsigned long addr);
403

404 405 406 407 408 409
#ifdef CONFIG_PPC_PSERIES
void hpte_init_pseries(void);
#else
static inline void hpte_init_pseries(void) { }
#endif

410 411 412 413 414
extern void hpte_init_native(void);

extern void slb_initialize(void);
extern void slb_flush_and_rebolt(void);

415
extern void slb_vmalloc_update(void);
416
extern void slb_set_size(u16 size);
417 418 419
#endif /* __ASSEMBLY__ */

/*
420
 * VSID allocation (256MB segment)
421
 *
422 423
 * We first generate a 37-bit "proto-VSID". Proto-VSIDs are generated
 * from mmu context id and effective segment id of the address.
424
 *
425 426 427 428 429 430 431
 * For user processes max context id is limited to ((1ul << 19) - 5)
 * for kernel space, we use the top 4 context ids to map address as below
 * NOTE: each context only support 64TB now.
 * 0x7fffc -  [ 0xc000000000000000 - 0xc0003fffffffffff ]
 * 0x7fffd -  [ 0xd000000000000000 - 0xd0003fffffffffff ]
 * 0x7fffe -  [ 0xe000000000000000 - 0xe0003fffffffffff ]
 * 0x7ffff -  [ 0xf000000000000000 - 0xf0003fffffffffff ]
432 433 434 435 436 437
 *
 * The proto-VSIDs are then scrambled into real VSIDs with the
 * multiplicative hash:
 *
 *	VSID = (proto-VSID * VSID_MULTIPLIER) % VSID_MODULUS
 *
438
 * VSID_MULTIPLIER is prime, so in particular it is
439 440
 * co-prime to VSID_MODULUS, making this a 1:1 scrambling function.
 * Because the modulus is 2^n-1 we can compute it efficiently without
441 442 443
 * a divide or extra multiply (see below). The scramble function gives
 * robust scattering in the hash table (at least based on some initial
 * results).
444
 *
445 446 447
 * We also consider VSID 0 special. We use VSID 0 for slb entries mapping
 * bad address. This enables us to consolidate bad address handling in
 * hash_page.
448
 *
449 450 451 452 453
 * We also need to avoid the last segment of the last context, because that
 * would give a protovsid of 0x1fffffffff. That will result in a VSID 0
 * because of the modulo operation in vsid scramble. But the vmemmap
 * (which is what uses region 0xf) will never be close to 64TB in size
 * (it's 56 bytes per page of system memory).
454 455
 */

456
#define CONTEXT_BITS		19
457 458
#define ESID_BITS		18
#define ESID_BITS_1T		6
459

460 461
/*
 * 256MB segment
462
 * The proto-VSID space has 2^(CONTEX_BITS + ESID_BITS) - 1 segments
463 464 465 466 467 468 469
 * available for user + kernel mapping. The top 4 contexts are used for
 * kernel mapping. Each segment contains 2^28 bytes. Each
 * context maps 2^46 bytes (64TB) so we can support 2^19-1 contexts
 * (19 == 37 + 28 - 46).
 */
#define MAX_USER_CONTEXT	((ASM_CONST(1) << CONTEXT_BITS) - 5)

A
Aneesh Kumar K.V 已提交
470 471 472 473 474
/*
 * This should be computed such that protovosid * vsid_mulitplier
 * doesn't overflow 64 bits. It should also be co-prime to vsid_modulus
 */
#define VSID_MULTIPLIER_256M	ASM_CONST(12538073)	/* 24-bit prime */
475
#define VSID_BITS_256M		(CONTEXT_BITS + ESID_BITS)
P
Paul Mackerras 已提交
476
#define VSID_MODULUS_256M	((1UL<<VSID_BITS_256M)-1)
477

P
Paul Mackerras 已提交
478
#define VSID_MULTIPLIER_1T	ASM_CONST(12538073)	/* 24-bit prime */
479
#define VSID_BITS_1T		(CONTEXT_BITS + ESID_BITS_1T)
P
Paul Mackerras 已提交
480 481
#define VSID_MODULUS_1T		((1UL<<VSID_BITS_1T)-1)

482

483
#define USER_VSID_RANGE	(1UL << (ESID_BITS + SID_SHIFT))
484 485 486 487 488 489

/*
 * This macro generates asm code to compute the VSID scramble
 * function.  Used in slb_allocate() and do_stab_bolted.  The function
 * computed is: (protovsid*VSID_MULTIPLIER) % VSID_MODULUS
 *
M
Michael Ellerman 已提交
490
 *	rt = register containing the proto-VSID and into which the
491 492 493 494
 *		VSID will be stored
 *	rx = scratch register (clobbered)
 *
 * 	- rt and rx must be different registers
P
Paul Mackerras 已提交
495
 * 	- The answer will end up in the low VSID_BITS bits of rt.  The higher
496 497 498
 * 	  bits may contain other garbage, so you may need to mask the
 * 	  result.
 */
P
Paul Mackerras 已提交
499 500 501
#define ASM_VSID_SCRAMBLE(rt, rx, size)					\
	lis	rx,VSID_MULTIPLIER_##size@h;				\
	ori	rx,rx,VSID_MULTIPLIER_##size@l;				\
502 503
	mulld	rt,rt,rx;		/* rt = rt * MULTIPLIER */	\
									\
P
Paul Mackerras 已提交
504 505
	srdi	rx,rt,VSID_BITS_##size;					\
	clrldi	rt,rt,(64-VSID_BITS_##size);				\
506
	add	rt,rt,rx;		/* add high and low bits */	\
507 508
	/* NOTE: explanation based on VSID_BITS_##size = 36		\
	 * Now, r3 == VSID (mod 2^36-1), and lies between 0 and		\
509 510 511 512 513 514
	 * 2^36-1+2^28-1.  That in particular means that if r3 >=	\
	 * 2^36-1, then r3+1 has the 2^36 bit set.  So, if r3+1 has	\
	 * the bit clear, r3 already has the answer we want, if it	\
	 * doesn't, the answer is the low 36 bits of r3+1.  So in all	\
	 * cases the answer is the low 36 bits of (r3 + ((r3+1) >> 36))*/\
	addi	rx,rt,1;						\
P
Paul Mackerras 已提交
515
	srdi	rx,rx,VSID_BITS_##size;	/* extract 2^VSID_BITS bit */	\
516 517
	add	rt,rt,rx

518
/* 4 bits per slice and we have one slice per 1TB */
519
#define SLICE_ARRAY_SIZE  (H_PGTABLE_RANGE >> 41)
520 521 522

#ifndef __ASSEMBLY__

523 524 525 526 527 528 529 530 531 532 533 534 535 536
#ifdef CONFIG_PPC_SUBPAGE_PROT
/*
 * For the sub-page protection option, we extend the PGD with one of
 * these.  Basically we have a 3-level tree, with the top level being
 * the protptrs array.  To optimize speed and memory consumption when
 * only addresses < 4GB are being protected, pointers to the first
 * four pages of sub-page protection words are stored in the low_prot
 * array.
 * Each page of sub-page protection words protects 1GB (4 bytes
 * protects 64k).  For the 3-level tree, each page of pointers then
 * protects 8TB.
 */
struct subpage_prot_table {
	unsigned long maxaddr;	/* only addresses < this are protected */
537
	unsigned int **protptrs[(TASK_SIZE_USER64 >> 43)];
538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554
	unsigned int *low_prot[4];
};

#define SBP_L1_BITS		(PAGE_SHIFT - 2)
#define SBP_L2_BITS		(PAGE_SHIFT - 3)
#define SBP_L1_COUNT		(1 << SBP_L1_BITS)
#define SBP_L2_COUNT		(1 << SBP_L2_BITS)
#define SBP_L2_SHIFT		(PAGE_SHIFT + SBP_L1_BITS)
#define SBP_L3_SHIFT		(SBP_L2_SHIFT + SBP_L2_BITS)

extern void subpage_prot_free(struct mm_struct *mm);
extern void subpage_prot_init_new_context(struct mm_struct *mm);
#else
static inline void subpage_prot_free(struct mm_struct *mm) {}
static inline void subpage_prot_init_new_context(struct mm_struct *mm) { }
#endif /* CONFIG_PPC_SUBPAGE_PROT */

555
#if 0
P
Paul Mackerras 已提交
556 557 558 559 560 561
/*
 * The code below is equivalent to this function for arguments
 * < 2^VSID_BITS, which is all this should ever be called
 * with.  However gcc is not clever enough to compute the
 * modulus (2^n-1) without a second multiply.
 */
562
#define vsid_scramble(protovsid, size) \
P
Paul Mackerras 已提交
563
	((((protovsid) * VSID_MULTIPLIER_##size) % VSID_MODULUS_##size))
564

P
Paul Mackerras 已提交
565 566 567 568 569 570 571 572
#else /* 1 */
#define vsid_scramble(protovsid, size) \
	({								 \
		unsigned long x;					 \
		x = (protovsid) * VSID_MULTIPLIER_##size;		 \
		x = (x >> VSID_BITS_##size) + (x & VSID_MODULUS_##size); \
		(x + ((x+1) >> VSID_BITS_##size)) & VSID_MODULUS_##size; \
	})
573 574
#endif /* 1 */

P
Paul Mackerras 已提交
575 576
/* Returns the segment size indicator for a user address */
static inline int user_segment_size(unsigned long addr)
577
{
P
Paul Mackerras 已提交
578 579 580 581
	/* Use 1T segments if possible for addresses >= 1T */
	if (addr >= (1UL << SID_SHIFT_1T))
		return mmu_highuser_ssize;
	return MMU_SEGSIZE_256M;
582 583
}

P
Paul Mackerras 已提交
584 585 586
static inline unsigned long get_vsid(unsigned long context, unsigned long ea,
				     int ssize)
{
587 588 589
	/*
	 * Bad address. We return VSID 0 for that
	 */
590
	if ((ea & ~REGION_MASK) >= H_PGTABLE_RANGE)
591 592
		return 0;

P
Paul Mackerras 已提交
593
	if (ssize == MMU_SEGSIZE_256M)
594
		return vsid_scramble((context << ESID_BITS)
P
Paul Mackerras 已提交
595
				     | (ea >> SID_SHIFT), 256M);
596
	return vsid_scramble((context << ESID_BITS_1T)
P
Paul Mackerras 已提交
597 598 599
			     | (ea >> SID_SHIFT_1T), 1T);
}

600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618
/*
 * This is only valid for addresses >= PAGE_OFFSET
 *
 * For kernel space, we use the top 4 context ids to map address as below
 * 0x7fffc -  [ 0xc000000000000000 - 0xc0003fffffffffff ]
 * 0x7fffd -  [ 0xd000000000000000 - 0xd0003fffffffffff ]
 * 0x7fffe -  [ 0xe000000000000000 - 0xe0003fffffffffff ]
 * 0x7ffff -  [ 0xf000000000000000 - 0xf0003fffffffffff ]
 */
static inline unsigned long get_kernel_vsid(unsigned long ea, int ssize)
{
	unsigned long context;

	/*
	 * kernel take the top 4 context from the available range
	 */
	context = (MAX_USER_CONTEXT) + ((ea >> 60) - 0xc) + 1;
	return get_vsid(context, ea, ssize);
}
619 620 621

unsigned htab_shift_for_mem_size(unsigned long mem_size);

622 623
#endif /* __ASSEMBLY__ */

624
#endif /* _ASM_POWERPC_BOOK3S_64_MMU_HASH_H_ */