setup_64.c 21.0 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/hugetlb.h>
39
#include <linux/memory.h>
40
#include <linux/nmi.h>
41

42
#include <asm/io.h>
43
#include <asm/kdump.h>
44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
#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
64
#include <asm/xmon.h>
D
David Gibson 已提交
65
#include <asm/udbg.h>
66
#include <asm/kexec.h>
67
#include <asm/mmu_context.h>
68
#include <asm/code-patching.h>
69
#include <asm/kvm_ppc.h>
70
#include <asm/hugetlb.h>
71
#include <asm/livepatch.h>
72
#include <asm/opal.h>
73 74 75 76 77 78 79

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

80
int spinning_secondaries;
81 82
u64 ppc64_pft_size;

83 84 85 86
/* Pick defaults since we might want to patch instructions
 * before we've read this from the device tree.
 */
struct ppc64_caches ppc64_caches = {
87 88 89 90
	.dline_size = 0x40,
	.log_dline_size = 6,
	.iline_size = 0x40,
	.log_iline_size = 6
91
};
92 93 94 95 96 97 98 99 100 101
EXPORT_SYMBOL_GPL(ppc64_caches);

/*
 * These are used in binfmt_elf.c to put aux entries on the stack
 * for each elf executable being started.
 */
int dcache_bsize;
int icache_bsize;
int ucache_bsize;

102 103 104 105 106
#if defined(CONFIG_PPC_BOOK3E) && defined(CONFIG_SMP)
static void setup_tlb_core_data(void)
{
	int cpu;

107 108
	BUILD_BUG_ON(offsetof(struct tlb_core_data, lock) != 0);

109 110 111
	for_each_possible_cpu(cpu) {
		int first = cpu_first_thread_sibling(cpu);

112 113 114 115 116 117 118 119
		/*
		 * 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;

120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141
		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__);
		}
	}
}
#else
static void setup_tlb_core_data(void)
{
}
#endif

142 143
#ifdef CONFIG_SMP

144
static char *smt_enabled_cmdline;
145 146 147 148 149

/* Look for ibm,smt-enabled OF option */
static void check_smt_enabled(void)
{
	struct device_node *dn;
150
	const char *smt_option;
151

152 153
	/* Default to enabling all threads */
	smt_enabled_at_boot = threads_per_core;
154

155 156 157 158 159 160 161
	/* 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 {
162
			int smt;
163 164
			int rc;

165
			rc = kstrtoint(smt_enabled_cmdline, 10, &smt);
166 167
			if (!rc)
				smt_enabled_at_boot =
168
					min(threads_per_core, smt);
169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185
		}
	} 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);
		}
	}
186 187 188 189 190
}

/* Look for smt-enabled= cmdline option */
static int __init early_smt_enabled(char *p)
{
191
	smt_enabled_cmdline = p;
192 193 194 195
	return 0;
}
early_param("smt-enabled", early_smt_enabled);

P
Paul Mackerras 已提交
196 197
#else
#define check_smt_enabled()
198 199
#endif /* CONFIG_SMP */

200 201 202 203 204 205 206 207 208
/** Fix up paca fields required for the boot cpu */
static void fixup_boot_paca(void)
{
	/* 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;
}

209
static void configure_exceptions(void)
210
{
211
	/*
212 213
	 * Setup the trampolines from the lowmem exception vectors
	 * to the kdump kernel when not using a relocatable kernel.
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
	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();

		/* Enable AIL if supported, and we are in hypervisor mode */
		if (cpu_has_feature(CPU_FTR_HVMODE) &&
		    cpu_has_feature(CPU_FTR_ARCH_207S)) {
			unsigned long lpcr = mfspr(SPRN_LPCR);
			mtspr(SPRN_LPCR, lpcr | LPCR_AIL_3);
		}
244 245 246
	}
}

247 248 249 250 251 252
static void cpu_ready_for_interrupts(void)
{
	/* Set IR and DR in PACA MSR */
	get_paca()->kernel_msr = MSR_KERNEL;
}

253 254 255 256 257 258
/*
 * 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
259
 * some early parsing of the device-tree to setup out MEMBLOCK
260 261 262 263 264 265 266 267 268 269 270 271 272 273
 * 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)
{
274 275
	static __initdata struct paca_struct boot_paca;

276 277
	/* -------- printk is _NOT_ safe to use here ! ------- */

278
	/* Identify CPU type */
279
	identify_cpu(0, mfspr(SPRN_PVR));
280

281
	/* Assume we're on cpu 0 for now. Don't write to the paca yet! */
282 283
	initialise_paca(&boot_paca, 0);
	setup_paca(&boot_paca);
284
	fixup_boot_paca();
285

286 287
	/* -------- printk is now safe to use ------- */

288 289 290
	/* Enable early debugging if any specified (see udbg.h) */
	udbg_early_init();

291
 	DBG(" -> early_setup(), dt_ptr: 0x%lx\n", dt_ptr);
292 293

	/*
294 295 296
	 * Do early initialization using the flattened device
	 * tree, such as retrieving the physical memory map or
	 * calculating/retrieving the hash table size.
297 298 299
	 */
	early_init_devtree(__va(dt_ptr));

300
	/* Now we know the logical id of our boot cpu, setup the paca. */
301
	setup_paca(&paca[boot_cpuid]);
302
	fixup_boot_paca();
303

304
	/*
305 306
	 * Configure exception handlers. This include setting up trampolines
	 * if needed, setting exception endian mode, etc...
307
	 */
308
	configure_exceptions();
309

310 311
	/* Initialize the hash table or TLB handling */
	early_init_mmu();
312

313 314 315
	/* Apply all the dynamic patching */
	apply_feature_fixups();

316 317 318
	/*
	 * At this point, we can let interrupts switch to virtual mode
	 * (the MMU has been setup), so adjust the MSR in the PACA to
319
	 * have IR and DR set and enable AIL if it exists
320
	 */
321
	cpu_ready_for_interrupts();
322

323
	DBG(" <- early_setup()\n");
324 325 326 327 328 329 330 331 332 333 334 335

#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 */
336 337
}

338 339 340
#ifdef CONFIG_SMP
void early_setup_secondary(void)
{
341
	/* Mark interrupts disabled in PACA */
342
	get_paca()->soft_enabled = 0;
343

344 345
	/* Initialize the hash table or TLB handling */
	early_init_mmu_secondary();
346 347 348 349 350 351

	/*
	 * 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.
	 */
352
	cpu_ready_for_interrupts();
353 354 355
}

#endif /* CONFIG_SMP */
356

357
#if defined(CONFIG_SMP) || defined(CONFIG_KEXEC)
358 359 360 361 362 363 364 365 366 367 368 369
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");
}

370 371
void smp_release_cpus(void)
{
372
	unsigned long *ptr;
373
	int i;
374

375 376 377
	if (!use_spinloop())
		return;

378 379 380 381 382 383
	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.
384
	 */
385

386 387
	ptr  = (unsigned long *)((unsigned long)&__secondary_hold_spinloop
			- PHYSICAL_START);
388
	*ptr = ppc_function_entry(generic_secondary_smp_init);
389 390 391 392 393

	/* And wait a bit for them to catch up */
	for (i = 0; i < 100000; i++) {
		mb();
		HMT_low();
394
		if (spinning_secondaries == 0)
395 396 397
			break;
		udelay(1);
	}
398
	DBG("spinning_secondaries = %d\n", spinning_secondaries);
399 400 401 402 403

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

404
/*
405 406
 * Initialize some remaining members of the ppc64_caches and systemcfg
 * structures
407 408 409 410 411 412 413 414 415 416 417
 * (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
 */
static void __init initialize_cache_info(void)
{
	struct device_node *np;
	unsigned long num_cpus = 0;

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

418
	for_each_node_by_type(np, "cpu") {
419 420
		num_cpus += 1;

A
Anton Blanchard 已提交
421 422
		/*
		 * We're assuming *all* of the CPUs have the same
423 424
		 * d-cache and i-cache sizes... -Peter
		 */
A
Anton Blanchard 已提交
425
		if (num_cpus == 1) {
426
			const __be32 *sizep, *lsizep;
427 428 429 430
			u32 size, lsize;

			size = 0;
			lsize = cur_cpu_spec->dcache_bsize;
431
			sizep = of_get_property(np, "d-cache-size", NULL);
432
			if (sizep != NULL)
433
				size = be32_to_cpu(*sizep);
A
Anton Blanchard 已提交
434 435
			lsizep = of_get_property(np, "d-cache-block-size",
						 NULL);
436 437
			/* fallback if block size missing */
			if (lsizep == NULL)
A
Anton Blanchard 已提交
438 439 440
				lsizep = of_get_property(np,
							 "d-cache-line-size",
							 NULL);
441
			if (lsizep != NULL)
442
				lsize = be32_to_cpu(*lsizep);
443
			if (sizep == NULL || lsizep == NULL)
444 445 446
				DBG("Argh, can't find dcache properties ! "
				    "sizep: %p, lsizep: %p\n", sizep, lsizep);

447 448
			ppc64_caches.dsize = size;
			ppc64_caches.dline_size = lsize;
449 450 451 452 453
			ppc64_caches.log_dline_size = __ilog2(lsize);
			ppc64_caches.dlines_per_page = PAGE_SIZE / lsize;

			size = 0;
			lsize = cur_cpu_spec->icache_bsize;
454
			sizep = of_get_property(np, "i-cache-size", NULL);
455
			if (sizep != NULL)
456
				size = be32_to_cpu(*sizep);
A
Anton Blanchard 已提交
457 458
			lsizep = of_get_property(np, "i-cache-block-size",
						 NULL);
459
			if (lsizep == NULL)
A
Anton Blanchard 已提交
460 461 462
				lsizep = of_get_property(np,
							 "i-cache-line-size",
							 NULL);
463
			if (lsizep != NULL)
464
				lsize = be32_to_cpu(*lsizep);
465
			if (sizep == NULL || lsizep == NULL)
466 467 468
				DBG("Argh, can't find icache properties ! "
				    "sizep: %p, lsizep: %p\n", sizep, lsizep);

469 470
			ppc64_caches.isize = size;
			ppc64_caches.iline_size = lsize;
471 472 473 474 475
			ppc64_caches.log_iline_size = __ilog2(lsize);
			ppc64_caches.ilines_per_page = PAGE_SIZE / lsize;
		}
	}

476 477 478 479
	/* For use by binfmt_elf */
	dcache_bsize = ppc64_caches.dline_size;
	icache_bsize = ppc64_caches.iline_size;

480 481 482
	DBG(" <- initialize_cache_info()\n");
}

483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513
static __init void print_system_info(void)
{
	pr_info("-----------------------------------------------------\n");
	pr_info("ppc64_pft_size    = 0x%llx\n", ppc64_pft_size);
	pr_info("phys_mem_size     = 0x%llx\n", memblock_phys_mem_size());

	if (ppc64_caches.dline_size != 0x80)
		pr_info("dcache_line_size  = 0x%x\n", ppc64_caches.dline_size);
	if (ppc64_caches.iline_size != 0x80)
		pr_info("icache_line_size  = 0x%x\n", ppc64_caches.iline_size);

	pr_info("cpu_features      = 0x%016lx\n", cur_cpu_spec->cpu_features);
	pr_info("  possible        = 0x%016lx\n", CPU_FTRS_POSSIBLE);
	pr_info("  always          = 0x%016lx\n", CPU_FTRS_ALWAYS);
	pr_info("cpu_user_features = 0x%08x 0x%08x\n", cur_cpu_spec->cpu_user_features,
		cur_cpu_spec->cpu_user_features2);
	pr_info("mmu_features      = 0x%08x\n", cur_cpu_spec->mmu_features);
	pr_info("firmware_features = 0x%016lx\n", powerpc_firmware_features);

#ifdef CONFIG_PPC_STD_MMU_64
	if (htab_address)
		pr_info("htab_address      = 0x%p\n", htab_address);

	pr_info("htab_hash_mask    = 0x%lx\n", htab_hash_mask);
#endif

	if (PHYSICAL_START > 0)
		pr_info("physical_start    = 0x%llx\n",
		       (unsigned long long)PHYSICAL_START);
	pr_info("-----------------------------------------------------\n");
}
514

515 516 517 518 519 520
/* 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.
 */
static u64 safe_stack_limit(void)
521
{
522 523 524 525 526 527 528 529 530
#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))
531 532
		return 1UL << SID_SHIFT_1T;
	return 1UL << SID_SHIFT;
533
#endif
534 535
}

536 537
static void __init irqstack_early_init(void)
{
538
	u64 limit = safe_stack_limit();
539 540 541
	unsigned int i;

	/*
542 543
	 * Interrupt stacks must be in the first segment since we
	 * cannot afford to take SLB misses on them.
544
	 */
545
	for_each_possible_cpu(i) {
546
		softirq_ctx[i] = (struct thread_info *)
Y
Yinghai Lu 已提交
547
			__va(memblock_alloc_base(THREAD_SIZE,
548
					    THREAD_SIZE, limit));
549
		hardirq_ctx[i] = (struct thread_info *)
Y
Yinghai Lu 已提交
550
			__va(memblock_alloc_base(THREAD_SIZE,
551
					    THREAD_SIZE, limit));
552 553 554
	}
}

555 556 557 558
#ifdef CONFIG_PPC_BOOK3E
static void __init exc_lvl_early_init(void)
{
	unsigned int i;
559
	unsigned long sp;
560 561

	for_each_possible_cpu(i) {
562 563 564 565 566 567 568 569 570 571 572
		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);
573
	}
574 575

	if (cpu_has_feature(CPU_FTR_DEBUG_LVL_EXC))
576
		patch_exception(0x040, exc_debug_debug_book3e);
577 578 579 580 581
}
#else
#define exc_lvl_early_init()
#endif

582 583
/*
 * Stack space used when we detect a bad kernel stack pointer, and
584 585
 * early in SMP boots before relocation is enabled. Exclusive emergency
 * stack for machine checks.
586 587 588
 */
static void __init emergency_stack_init(void)
{
589
	u64 limit;
590 591 592 593 594 595 596 597 598 599 600
	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.
	 */
601
	limit = min(safe_stack_limit(), ppc64_rma_size);
602

603
	for_each_possible_cpu(i) {
604 605 606 607
		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;
608 609 610

#ifdef CONFIG_PPC_BOOK3S_64
		/* emergency stack for machine check exception handling. */
611 612 613
		ti = __va(memblock_alloc_base(THREAD_SIZE, THREAD_SIZE, limit));
		klp_init_thread_info(ti);
		paca[i].mc_emergency_sp = (void *)ti + THREAD_SIZE;
614
#endif
615
	}
616 617 618
}

/*
619
 * Called into from start_kernel this initializes memblock, which is used
620
 * to manage page allocation until mem_init is called.
621 622 623
 */
void __init setup_arch(char **cmdline_p)
{
624
	*cmdline_p = boot_command_line;
625

626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663
	/*
	 * Unflatten the device-tree passed by prom_init or kexec
	 */
	unflatten_device_tree();

	/*
	 * Fill the ppc64_caches & systemcfg structures with informations
	 * retrieved from the device-tree.
	 */
	initialize_cache_info();

#ifdef CONFIG_PPC_RTAS
	/*
	 * Initialize RTAS if available
	 */
	rtas_initialize();
#endif /* CONFIG_PPC_RTAS */

	/*
	 * Check if we have an initrd provided via the device-tree
	 */
	check_for_initrd();

	/* Probe the machine type */
	probe_machine();

	/*
	 * We can discover serial ports now since the above did setup the
	 * hash table management for us, thus ioremap works. We do that early
	 * so that further code can be debugged
	 */
	find_legacy_serial_ports();

	/*
	 * Register early console
	 */
	register_early_udbg_console();

664 665
	smp_setup_cpu_maps();

666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688
	/*
	 * Initialize xmon
	 */
	xmon_setup();

	check_smt_enabled();
	setup_tlb_core_data();

	/*
	 * Freescale Book3e parts spin in a loop provided by firmware,
	 * so smp_release_cpus() does nothing for them
	 */
#if defined(CONFIG_SMP)
	/*
	 * Release secondary cpus out of their spinloops at 0x60 now that
	 * we can map physical -> logical CPU ids
	 */
	smp_release_cpus();
#endif

	/* Print various info about the machine that has been gathered so far. */
	print_system_info();

689 690 691 692 693 694 695 696 697 698
	/* Reserve large chunks of memory for use by CMA for KVM */
	kvm_cma_reserve();

	/*
	 * Reserve any gigantic pages requested on the command line.
	 * memblock needs to have been initialized by the time this is
	 * called since this will reserve memory.
	 */
	reserve_hugetlb_gpages();

699
	if (ppc_md.panic)
700
		setup_panic();
701

702 703
	klp_init_thread_info(&init_thread_info);

704
	init_mm.start_code = (unsigned long)_stext;
705 706 707
	init_mm.end_code = (unsigned long) _etext;
	init_mm.end_data = (unsigned long) _edata;
	init_mm.brk = klimit;
708 709
#ifdef CONFIG_PPC_64K_PAGES
	init_mm.context.pte_frag = NULL;
710 711 712
#endif
#ifdef CONFIG_SPAPR_TCE_IOMMU
	mm_iommu_init(&init_mm.context);
713
#endif
714
	irqstack_early_init();
715
	exc_lvl_early_init();
716 717
	emergency_stack_init();

718
	initmem_init();
719

720 721 722
#ifdef CONFIG_DUMMY_CONSOLE
	conswitchp = &dummy_con;
#endif
723 724
	if (ppc_md.setup_arch)
		ppc_md.setup_arch();
725 726

	paging_init();
727 728 729 730

	/* Initialize the MMU context management stuff */
	mmu_context_init();

731 732 733 734
	/* Interrupt code needs to be 64K-aligned */
	if ((unsigned long)_stext & 0xffff)
		panic("Kernelbase not 64K-aligned (0x%lx)!\n",
		      (unsigned long)_stext);
735 736
}

737
#ifdef CONFIG_SMP
738 739 740
#define PCPU_DYN_SIZE		()

static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size, size_t align)
741
{
742 743 744
	return __alloc_bootmem_node(NODE_DATA(cpu_to_node(cpu)), size, align,
				    __pa(MAX_DMA_ADDRESS));
}
745

746 747 748 749
static void __init pcpu_fc_free(void *ptr, size_t size)
{
	free_bootmem(__pa(ptr), size);
}
750

751 752 753 754 755 756 757 758
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;
}

759 760 761
unsigned long __per_cpu_offset[NR_CPUS] __read_mostly;
EXPORT_SYMBOL(__per_cpu_offset);

762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785
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;
786 787 788 789
	for_each_possible_cpu(cpu) {
                __per_cpu_offset[cpu] = delta + pcpu_unit_offsets[cpu];
		paca[cpu].data_offset = __per_cpu_offset[cpu];
	}
790 791
}
#endif
792

793 794 795 796 797 798 799 800 801
#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
802

803
#if defined(CONFIG_PPC_INDIRECT_PIO) || defined(CONFIG_PPC_INDIRECT_MMIO)
804 805
struct ppc_pci_io ppc_pci_io;
EXPORT_SYMBOL(ppc_pci_io);
806
#endif
807 808 809 810 811 812 813 814 815 816 817 818 819

#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)
{
820
	hardlockup_detector_disable();
821 822 823 824 825

	return 0;
}
early_initcall(disable_hardlockup_detector);
#endif