setup_64.c 18.6 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12
/*
 * 
 * Common boot and setup code.
 *
 * Copyright (C) 2001 PPC64 Team, IBM Corp
 *
 *      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.
 */

13
#define DEBUG
14

15
#include <linux/export.h>
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
#include <linux/string.h>
#include <linux/sched.h>
#include <linux/init.h>
#include <linux/kernel.h>
#include <linux/reboot.h>
#include <linux/delay.h>
#include <linux/initrd.h>
#include <linux/seq_file.h>
#include <linux/ioport.h>
#include <linux/console.h>
#include <linux/utsname.h>
#include <linux/tty.h>
#include <linux/root_dev.h>
#include <linux/notifier.h>
#include <linux/cpu.h>
#include <linux/unistd.h>
#include <linux/serial.h>
#include <linux/serial_8250.h>
34
#include <linux/bootmem.h>
35
#include <linux/pci.h>
36
#include <linux/lockdep.h>
Y
Yinghai Lu 已提交
37
#include <linux/memblock.h>
38
#include <linux/memory.h>
39
#include <linux/nmi.h>
40

41
#include <asm/io.h>
42
#include <asm/kdump.h>
43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62
#include <asm/prom.h>
#include <asm/processor.h>
#include <asm/pgtable.h>
#include <asm/smp.h>
#include <asm/elf.h>
#include <asm/machdep.h>
#include <asm/paca.h>
#include <asm/time.h>
#include <asm/cputable.h>
#include <asm/sections.h>
#include <asm/btext.h>
#include <asm/nvram.h>
#include <asm/setup.h>
#include <asm/rtas.h>
#include <asm/iommu.h>
#include <asm/serial.h>
#include <asm/cache.h>
#include <asm/page.h>
#include <asm/mmu.h>
#include <asm/firmware.h>
P
Paul Mackerras 已提交
63
#include <asm/xmon.h>
D
David Gibson 已提交
64
#include <asm/udbg.h>
65
#include <asm/kexec.h>
66
#include <asm/code-patching.h>
67
#include <asm/livepatch.h>
68
#include <asm/opal.h>
69
#include <asm/cputhreads.h>
70 71 72 73 74 75 76

#ifdef DEBUG
#define DBG(fmt...) udbg_printf(fmt)
#else
#define DBG(fmt...)
#endif

77
int spinning_secondaries;
78 79
u64 ppc64_pft_size;

80
struct ppc64_caches ppc64_caches = {
81 82 83 84 85 86 87 88
	.l1d = {
		.block_size = 0x40,
		.log_block_size = 6,
	},
	.l1i = {
		.block_size = 0x40,
		.log_block_size = 6
	},
89
};
90 91
EXPORT_SYMBOL_GPL(ppc64_caches);

92
#if defined(CONFIG_PPC_BOOK3E) && defined(CONFIG_SMP)
93
void __init setup_tlb_core_data(void)
94 95 96
{
	int cpu;

97 98
	BUILD_BUG_ON(offsetof(struct tlb_core_data, lock) != 0);

99 100 101
	for_each_possible_cpu(cpu) {
		int first = cpu_first_thread_sibling(cpu);

102 103 104 105 106 107 108 109
		/*
		 * If we boot via kdump on a non-primary thread,
		 * make sure we point at the thread that actually
		 * set up this TLB.
		 */
		if (cpu_first_thread_sibling(boot_cpuid) == first)
			first = boot_cpuid;

110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
		paca[cpu].tcd_ptr = &paca[first].tcd;

		/*
		 * If we have threads, we need either tlbsrx.
		 * or e6500 tablewalk mode, or else TLB handlers
		 * will be racy and could produce duplicate entries.
		 */
		if (smt_enabled_at_boot >= 2 &&
		    !mmu_has_feature(MMU_FTR_USE_TLBRSRV) &&
		    book3e_htw_mode != PPC_HTW_E6500) {
			/* Should we panic instead? */
			WARN_ONCE("%s: unsupported MMU configuration -- expect problems\n",
				  __func__);
		}
	}
}
#endif

128 129
#ifdef CONFIG_SMP

130
static char *smt_enabled_cmdline;
131 132

/* Look for ibm,smt-enabled OF option */
133
void __init check_smt_enabled(void)
134 135
{
	struct device_node *dn;
136
	const char *smt_option;
137

138 139
	/* Default to enabling all threads */
	smt_enabled_at_boot = threads_per_core;
140

141 142 143 144 145 146 147
	/* Allow the command line to overrule the OF option */
	if (smt_enabled_cmdline) {
		if (!strcmp(smt_enabled_cmdline, "on"))
			smt_enabled_at_boot = threads_per_core;
		else if (!strcmp(smt_enabled_cmdline, "off"))
			smt_enabled_at_boot = 0;
		else {
148
			int smt;
149 150
			int rc;

151
			rc = kstrtoint(smt_enabled_cmdline, 10, &smt);
152 153
			if (!rc)
				smt_enabled_at_boot =
154
					min(threads_per_core, smt);
155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171
		}
	} else {
		dn = of_find_node_by_path("/options");
		if (dn) {
			smt_option = of_get_property(dn, "ibm,smt-enabled",
						     NULL);

			if (smt_option) {
				if (!strcmp(smt_option, "on"))
					smt_enabled_at_boot = threads_per_core;
				else if (!strcmp(smt_option, "off"))
					smt_enabled_at_boot = 0;
			}

			of_node_put(dn);
		}
	}
172 173 174 175 176
}

/* Look for smt-enabled= cmdline option */
static int __init early_smt_enabled(char *p)
{
177
	smt_enabled_cmdline = p;
178 179 180 181 182 183
	return 0;
}
early_param("smt-enabled", early_smt_enabled);

#endif /* CONFIG_SMP */

184
/** Fix up paca fields required for the boot cpu */
185
static void __init fixup_boot_paca(void)
186 187 188 189 190 191 192
{
	/* The boot cpu is started */
	get_paca()->cpu_start = 1;
	/* Allow percpu accesses to work until we setup percpu data */
	get_paca()->data_offset = 0;
}

193
static void __init configure_exceptions(void)
194
{
195
	/*
196 197
	 * Setup the trampolines from the lowmem exception vectors
	 * to the kdump kernel when not using a relocatable kernel.
198
	 */
199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221
	setup_kdump_trampoline();

	/* Under a PAPR hypervisor, we need hypercalls */
	if (firmware_has_feature(FW_FEATURE_SET_MODE)) {
		/* Enable AIL if possible */
		pseries_enable_reloc_on_exc();

		/*
		 * Tell the hypervisor that we want our exceptions to
		 * be taken in little endian mode.
		 *
		 * We don't call this for big endian as our calling convention
		 * makes us always enter in BE, and the call may fail under
		 * some circumstances with kdump.
		 */
#ifdef __LITTLE_ENDIAN__
		pseries_little_endian_exceptions();
#endif
	} else {
		/* Set endian mode using OPAL */
		if (firmware_has_feature(FW_FEATURE_OPAL))
			opal_configure_cores();

222
		/* AIL on native is done in cpu_ready_for_interrupts() */
223 224 225
	}
}

226 227
static void cpu_ready_for_interrupts(void)
{
228 229 230 231 232 233 234 235 236 237 238 239 240
	/*
	 * Enable AIL if supported, and we are in hypervisor mode. This
	 * is called once for every processor.
	 *
	 * If we are not in hypervisor mode the job is done once for
	 * the whole partition in configure_exceptions().
	 */
	if (early_cpu_has_feature(CPU_FTR_HVMODE) &&
	    early_cpu_has_feature(CPU_FTR_ARCH_207S)) {
		unsigned long lpcr = mfspr(SPRN_LPCR);
		mtspr(SPRN_LPCR, lpcr | LPCR_AIL_3);
	}

241 242 243 244
	/* Set IR and DR in PACA MSR */
	get_paca()->kernel_msr = MSR_KERNEL;
}

245 246 247 248 249 250
/*
 * Early initialization entry point. This is called by head.S
 * with MMU translation disabled. We rely on the "feature" of
 * the CPU that ignores the top 2 bits of the address in real
 * mode so we can access kernel globals normally provided we
 * only toy with things in the RMO region. From here, we do
Y
Yinghai Lu 已提交
251
 * some early parsing of the device-tree to setup out MEMBLOCK
252 253 254 255 256 257 258 259 260 261 262 263 264 265
 * data structures, and allocate & initialize the hash table
 * and segment tables so we can start running with translation
 * enabled.
 *
 * It is this function which will call the probe() callback of
 * the various platform types and copy the matching one to the
 * global ppc_md structure. Your platform can eventually do
 * some very early initializations from the probe() routine, but
 * this is not recommended, be very careful as, for example, the
 * device-tree is not accessible via normal means at this point.
 */

void __init early_setup(unsigned long dt_ptr)
{
266 267
	static __initdata struct paca_struct boot_paca;

268 269
	/* -------- printk is _NOT_ safe to use here ! ------- */

270
	/* Identify CPU type */
271
	identify_cpu(0, mfspr(SPRN_PVR));
272

273
	/* Assume we're on cpu 0 for now. Don't write to the paca yet! */
274 275
	initialise_paca(&boot_paca, 0);
	setup_paca(&boot_paca);
276
	fixup_boot_paca();
277

278 279
	/* -------- printk is now safe to use ------- */

280 281 282
	/* Enable early debugging if any specified (see udbg.h) */
	udbg_early_init();

283
 	DBG(" -> early_setup(), dt_ptr: 0x%lx\n", dt_ptr);
284 285

	/*
286 287 288
	 * Do early initialization using the flattened device
	 * tree, such as retrieving the physical memory map or
	 * calculating/retrieving the hash table size.
289 290 291
	 */
	early_init_devtree(__va(dt_ptr));

292
	/* Now we know the logical id of our boot cpu, setup the paca. */
293
	setup_paca(&paca[boot_cpuid]);
294
	fixup_boot_paca();
295

296
	/*
297 298
	 * Configure exception handlers. This include setting up trampolines
	 * if needed, setting exception endian mode, etc...
299
	 */
300
	configure_exceptions();
301

302 303
	/* Apply all the dynamic patching */
	apply_feature_fixups();
304
	setup_feature_keys();
305

306 307 308
	/* Initialize the hash table or TLB handling */
	early_init_mmu();

309 310 311
	/*
	 * At this point, we can let interrupts switch to virtual mode
	 * (the MMU has been setup), so adjust the MSR in the PACA to
312
	 * have IR and DR set and enable AIL if it exists
313
	 */
314
	cpu_ready_for_interrupts();
315

316
	DBG(" <- early_setup()\n");
317 318 319 320 321 322 323 324 325 326 327 328

#ifdef CONFIG_PPC_EARLY_DEBUG_BOOTX
	/*
	 * This needs to be done *last* (after the above DBG() even)
	 *
	 * Right after we return from this function, we turn on the MMU
	 * which means the real-mode access trick that btext does will
	 * no longer work, it needs to switch to using a real MMU
	 * mapping. This call will ensure that it does
	 */
	btext_map();
#endif /* CONFIG_PPC_EARLY_DEBUG_BOOTX */
329 330
}

331 332 333
#ifdef CONFIG_SMP
void early_setup_secondary(void)
{
334
	/* Mark interrupts disabled in PACA */
335
	get_paca()->soft_enabled = 0;
336

337 338
	/* Initialize the hash table or TLB handling */
	early_init_mmu_secondary();
339 340 341 342 343 344

	/*
	 * At this point, we can let interrupts switch to virtual mode
	 * (the MMU has been setup), so adjust the MSR in the PACA to
	 * have IR and DR set.
	 */
345
	cpu_ready_for_interrupts();
346 347 348
}

#endif /* CONFIG_SMP */
349

350
#if defined(CONFIG_SMP) || defined(CONFIG_KEXEC_CORE)
351 352 353 354 355 356 357 358 359 360 361 362
static bool use_spinloop(void)
{
	if (!IS_ENABLED(CONFIG_PPC_BOOK3E))
		return true;

	/*
	 * When book3e boots from kexec, the ePAPR spin table does
	 * not get used.
	 */
	return of_property_read_bool(of_chosen, "linux,booted-from-kexec");
}

363 364
void smp_release_cpus(void)
{
365
	unsigned long *ptr;
366
	int i;
367

368 369 370
	if (!use_spinloop())
		return;

371 372 373 374 375 376
	DBG(" -> smp_release_cpus()\n");

	/* All secondary cpus are spinning on a common spinloop, release them
	 * all now so they can start to spin on their individual paca
	 * spinloops. For non SMP kernels, the secondary cpus never get out
	 * of the common spinloop.
377
	 */
378

379 380
	ptr  = (unsigned long *)((unsigned long)&__secondary_hold_spinloop
			- PHYSICAL_START);
381
	*ptr = ppc_function_entry(generic_secondary_smp_init);
382 383 384 385 386

	/* And wait a bit for them to catch up */
	for (i = 0; i < 100000; i++) {
		mb();
		HMT_low();
387
		if (spinning_secondaries == 0)
388 389 390
			break;
		udelay(1);
	}
391
	DBG("spinning_secondaries = %d\n", spinning_secondaries);
392 393 394

	DBG(" <- smp_release_cpus()\n");
}
395
#endif /* CONFIG_SMP || CONFIG_KEXEC_CORE */
396

397
/*
398 399
 * Initialize some remaining members of the ppc64_caches and systemcfg
 * structures
400 401 402 403
 * (at least until we get rid of them completely). This is mostly some
 * cache informations about the CPU that will be used by cache flush
 * routines and/or provided to userland
 */
404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472

static void init_cache_info(struct ppc_cache_info *info, u32 size, u32 lsize,
			    u32 bsize, u32 sets)
{
	info->size = size;
	info->sets = sets;
	info->line_size = lsize;
	info->block_size = bsize;
	info->log_block_size = __ilog2(bsize);
	info->blocks_per_page = PAGE_SIZE / bsize;
}

static bool __init parse_cache_info(struct device_node *np,
				    bool icache,
				    struct ppc_cache_info *info)
{
	static const char *ipropnames[] __initdata = {
		"i-cache-size",
		"i-cache-sets",
		"i-cache-block-size",
		"i-cache-line-size",
	};
	static const char *dpropnames[] __initdata = {
		"d-cache-size",
		"d-cache-sets",
		"d-cache-block-size",
		"d-cache-line-size",
	};
	const char **propnames = icache ? ipropnames : dpropnames;
	const __be32 *sizep, *lsizep, *bsizep, *setsp;
	u32 size, lsize, bsize, sets;
	bool success = true;

	size = 0;
	sets = -1u;
	lsize = bsize = cur_cpu_spec->dcache_bsize;
	sizep = of_get_property(np, propnames[0], NULL);
	if (sizep != NULL)
		size = be32_to_cpu(*sizep);
	setsp = of_get_property(np, propnames[1], NULL);
	if (setsp != NULL)
		sets = be32_to_cpu(*setsp);
	bsizep = of_get_property(np, propnames[2], NULL);
	lsizep = of_get_property(np, propnames[3], NULL);
	if (bsizep == NULL)
		bsizep = lsizep;
	if (lsizep != NULL)
		lsize = be32_to_cpu(*lsizep);
	if (bsizep != NULL)
		bsize = be32_to_cpu(*bsizep);
	if (sizep == NULL || bsizep == NULL || lsizep == NULL)
		success = false;

	/*
	 * OF is weird .. it represents fully associative caches
	 * as "1 way" which doesn't make much sense and doesn't
	 * leave room for direct mapped. We'll assume that 0
	 * in OF means direct mapped for that reason.
	 */
	if (sets == 1)
		sets = 0;
	else if (sets == 0)
		sets = 1;

	init_cache_info(info, size, lsize, bsize, sets);

	return success;
}

473
void __init initialize_cache_info(void)
474
{
475 476
	struct device_node *cpu = NULL, *l2, *l3 = NULL;
	u32 pvr;
477 478 479

	DBG(" -> initialize_cache_info()\n");

480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495
	/*
	 * All shipping POWER8 machines have a firmware bug that
	 * puts incorrect information in the device-tree. This will
	 * be (hopefully) fixed for future chips but for now hard
	 * code the values if we are running on one of these
	 */
	pvr = PVR_VER(mfspr(SPRN_PVR));
	if (pvr == PVR_POWER8 || pvr == PVR_POWER8E ||
	    pvr == PVR_POWER8NVL) {
						/* size    lsize   blk  sets */
		init_cache_info(&ppc64_caches.l1i, 0x8000,   128,  128, 32);
		init_cache_info(&ppc64_caches.l1d, 0x10000,  128,  128, 64);
		init_cache_info(&ppc64_caches.l2,  0x80000,  128,  0,   512);
		init_cache_info(&ppc64_caches.l3,  0x800000, 128,  0,   8192);
	} else
		cpu = of_find_node_by_type(NULL, "cpu");
496

497 498 499 500
	/*
	 * We're assuming *all* of the CPUs have the same
	 * d-cache and i-cache sizes... -Peter
	 */
501 502
	if (cpu) {
		if (!parse_cache_info(cpu, false, &ppc64_caches.l1d))
503 504
			DBG("Argh, can't find dcache properties !\n");

505
		if (!parse_cache_info(cpu, true, &ppc64_caches.l1i))
506
			DBG("Argh, can't find icache properties !\n");
507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522

		/*
		 * Try to find the L2 and L3 if any. Assume they are
		 * unified and use the D-side properties.
		 */
		l2 = of_find_next_cache_node(cpu);
		of_node_put(cpu);
		if (l2) {
			parse_cache_info(l2, false, &ppc64_caches.l2);
			l3 = of_find_next_cache_node(l2);
			of_node_put(l2);
		}
		if (l3) {
			parse_cache_info(l3, false, &ppc64_caches.l3);
			of_node_put(l3);
		}
523 524
	}

525
	/* For use by binfmt_elf */
526 527
	dcache_bsize = ppc64_caches.l1d.block_size;
	icache_bsize = ppc64_caches.l1i.block_size;
528

529 530 531
	DBG(" <- initialize_cache_info()\n");
}

532 533 534 535 536
/* This returns the limit below which memory accesses to the linear
 * mapping are guarnateed not to cause a TLB or SLB miss. This is
 * used to allocate interrupt or emergency stacks for which our
 * exception entry path doesn't deal with being interrupted.
 */
537
static __init u64 safe_stack_limit(void)
538
{
539 540 541 542 543 544 545 546 547
#ifdef CONFIG_PPC_BOOK3E
	/* Freescale BookE bolts the entire linear mapping */
	if (mmu_has_feature(MMU_FTR_TYPE_FSL_E))
		return linear_map_top;
	/* Other BookE, we assume the first GB is bolted */
	return 1ul << 30;
#else
	/* BookS, the first segment is bolted */
	if (mmu_has_feature(MMU_FTR_1T_SEGMENT))
548 549
		return 1UL << SID_SHIFT_1T;
	return 1UL << SID_SHIFT;
550
#endif
551 552
}

553
void __init irqstack_early_init(void)
554
{
555
	u64 limit = safe_stack_limit();
556 557 558
	unsigned int i;

	/*
559 560
	 * Interrupt stacks must be in the first segment since we
	 * cannot afford to take SLB misses on them.
561
	 */
562
	for_each_possible_cpu(i) {
563
		softirq_ctx[i] = (struct thread_info *)
Y
Yinghai Lu 已提交
564
			__va(memblock_alloc_base(THREAD_SIZE,
565
					    THREAD_SIZE, limit));
566
		hardirq_ctx[i] = (struct thread_info *)
Y
Yinghai Lu 已提交
567
			__va(memblock_alloc_base(THREAD_SIZE,
568
					    THREAD_SIZE, limit));
569 570 571
	}
}

572
#ifdef CONFIG_PPC_BOOK3E
573
void __init exc_lvl_early_init(void)
574 575
{
	unsigned int i;
576
	unsigned long sp;
577 578

	for_each_possible_cpu(i) {
579 580 581 582 583 584 585 586 587 588 589
		sp = memblock_alloc(THREAD_SIZE, THREAD_SIZE);
		critirq_ctx[i] = (struct thread_info *)__va(sp);
		paca[i].crit_kstack = __va(sp + THREAD_SIZE);

		sp = memblock_alloc(THREAD_SIZE, THREAD_SIZE);
		dbgirq_ctx[i] = (struct thread_info *)__va(sp);
		paca[i].dbg_kstack = __va(sp + THREAD_SIZE);

		sp = memblock_alloc(THREAD_SIZE, THREAD_SIZE);
		mcheckirq_ctx[i] = (struct thread_info *)__va(sp);
		paca[i].mc_kstack = __va(sp + THREAD_SIZE);
590
	}
591 592

	if (cpu_has_feature(CPU_FTR_DEBUG_LVL_EXC))
593
		patch_exception(0x040, exc_debug_debug_book3e);
594 595 596
}
#endif

597 598
/*
 * Stack space used when we detect a bad kernel stack pointer, and
599 600
 * early in SMP boots before relocation is enabled. Exclusive emergency
 * stack for machine checks.
601
 */
602
void __init emergency_stack_init(void)
603
{
604
	u64 limit;
605 606 607 608 609 610 611 612 613 614 615
	unsigned int i;

	/*
	 * Emergency stacks must be under 256MB, we cannot afford to take
	 * SLB misses on them. The ABI also requires them to be 128-byte
	 * aligned.
	 *
	 * Since we use these as temporary stacks during secondary CPU
	 * bringup, we need to get at them in real mode. This means they
	 * must also be within the RMO region.
	 */
616
	limit = min(safe_stack_limit(), ppc64_rma_size);
617

618
	for_each_possible_cpu(i) {
619 620 621 622
		struct thread_info *ti;
		ti = __va(memblock_alloc_base(THREAD_SIZE, THREAD_SIZE, limit));
		klp_init_thread_info(ti);
		paca[i].emergency_sp = (void *)ti + THREAD_SIZE;
623 624 625

#ifdef CONFIG_PPC_BOOK3S_64
		/* emergency stack for machine check exception handling. */
626 627 628
		ti = __va(memblock_alloc_base(THREAD_SIZE, THREAD_SIZE, limit));
		klp_init_thread_info(ti);
		paca[i].mc_emergency_sp = (void *)ti + THREAD_SIZE;
629
#endif
630
	}
631 632
}

633
#ifdef CONFIG_SMP
634 635 636
#define PCPU_DYN_SIZE		()

static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size, size_t align)
637
{
638 639 640
	return __alloc_bootmem_node(NODE_DATA(cpu_to_node(cpu)), size, align,
				    __pa(MAX_DMA_ADDRESS));
}
641

642 643 644 645
static void __init pcpu_fc_free(void *ptr, size_t size)
{
	free_bootmem(__pa(ptr), size);
}
646

647 648 649 650 651 652 653 654
static int pcpu_cpu_distance(unsigned int from, unsigned int to)
{
	if (cpu_to_node(from) == cpu_to_node(to))
		return LOCAL_DISTANCE;
	else
		return REMOTE_DISTANCE;
}

655 656 657
unsigned long __per_cpu_offset[NR_CPUS] __read_mostly;
EXPORT_SYMBOL(__per_cpu_offset);

658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681
void __init setup_per_cpu_areas(void)
{
	const size_t dyn_size = PERCPU_MODULE_RESERVE + PERCPU_DYNAMIC_RESERVE;
	size_t atom_size;
	unsigned long delta;
	unsigned int cpu;
	int rc;

	/*
	 * Linear mapping is one of 4K, 1M and 16M.  For 4K, no need
	 * to group units.  For larger mappings, use 1M atom which
	 * should be large enough to contain a number of units.
	 */
	if (mmu_linear_psize == MMU_PAGE_4K)
		atom_size = PAGE_SIZE;
	else
		atom_size = 1 << 20;

	rc = pcpu_embed_first_chunk(0, dyn_size, atom_size, pcpu_cpu_distance,
				    pcpu_fc_alloc, pcpu_fc_free);
	if (rc < 0)
		panic("cannot initialize percpu area (err=%d)", rc);

	delta = (unsigned long)pcpu_base_addr - (unsigned long)__per_cpu_start;
682 683 684 685
	for_each_possible_cpu(cpu) {
                __per_cpu_offset[cpu] = delta + pcpu_unit_offsets[cpu];
		paca[cpu].data_offset = __per_cpu_offset[cpu];
	}
686 687
}
#endif
688

689 690 691 692 693 694 695 696 697
#ifdef CONFIG_MEMORY_HOTPLUG_SPARSE
unsigned long memory_block_size_bytes(void)
{
	if (ppc_md.memory_block_size)
		return ppc_md.memory_block_size();

	return MIN_MEMORY_BLOCK_SIZE;
}
#endif
698

699
#if defined(CONFIG_PPC_INDIRECT_PIO) || defined(CONFIG_PPC_INDIRECT_MMIO)
700 701
struct ppc_pci_io ppc_pci_io;
EXPORT_SYMBOL(ppc_pci_io);
702
#endif
703 704 705 706 707 708 709 710 711 712 713 714 715

#ifdef CONFIG_HARDLOCKUP_DETECTOR
u64 hw_nmi_get_sample_period(int watchdog_thresh)
{
	return ppc_proc_freq * watchdog_thresh;
}

/*
 * The hardlockup detector breaks PMU event based branches and is likely
 * to get false positives in KVM guests, so disable it by default.
 */
static int __init disable_hardlockup_detector(void)
{
716
	hardlockup_detector_disable();
717 718 719 720 721

	return 0;
}
early_initcall(disable_hardlockup_detector);
#endif