setup_64.c 20.5 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

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/mmu_context.h>
67
#include <asm/code-patching.h>
68
#include <asm/kvm_ppc.h>
69
#include <asm/hugetlb.h>
70
#include <asm/epapr_hcalls.h>
71 72 73 74 75 76 77

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

78
int spinning_secondaries;
79 80
u64 ppc64_pft_size;

81 82 83 84
/* Pick defaults since we might want to patch instructions
 * before we've read this from the device tree.
 */
struct ppc64_caches ppc64_caches = {
85 86 87 88
	.dline_size = 0x40,
	.log_dline_size = 6,
	.iline_size = 0x40,
	.log_iline_size = 6
89
};
90 91 92 93 94 95 96 97 98 99
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;

100 101 102 103 104
#if defined(CONFIG_PPC_BOOK3E) && defined(CONFIG_SMP)
static void setup_tlb_core_data(void)
{
	int cpu;

105 106
	BUILD_BUG_ON(offsetof(struct tlb_core_data, lock) != 0);

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
	for_each_possible_cpu(cpu) {
		int first = cpu_first_thread_sibling(cpu);

		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

132 133
#ifdef CONFIG_SMP

134
static char *smt_enabled_cmdline;
135 136 137 138 139

/* Look for ibm,smt-enabled OF option */
static void check_smt_enabled(void)
{
	struct device_node *dn;
140
	const char *smt_option;
141

142 143
	/* Default to enabling all threads */
	smt_enabled_at_boot = threads_per_core;
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
	/* 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 {
			long smt;
			int rc;

			rc = strict_strtol(smt_enabled_cmdline, 10, &smt);
			if (!rc)
				smt_enabled_at_boot =
					min(threads_per_core, (int)smt);
		}
	} 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);
		}
	}
176 177 178 179 180
}

/* Look for smt-enabled= cmdline option */
static int __init early_smt_enabled(char *p)
{
181
	smt_enabled_cmdline = p;
182 183 184 185
	return 0;
}
early_param("smt-enabled", early_smt_enabled);

P
Paul Mackerras 已提交
186 187
#else
#define check_smt_enabled()
188 189
#endif /* CONFIG_SMP */

190 191 192 193 194 195 196 197 198
/** 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;
}

199 200 201 202 203 204
static void cpu_ready_for_interrupts(void)
{
	/* Set IR and DR in PACA MSR */
	get_paca()->kernel_msr = MSR_KERNEL;

	/* Enable AIL if supported */
205 206
	if (cpu_has_feature(CPU_FTR_HVMODE) &&
	    cpu_has_feature(CPU_FTR_ARCH_207S)) {
207 208 209 210 211
		unsigned long lpcr = mfspr(SPRN_LPCR);
		mtspr(SPRN_LPCR, lpcr | LPCR_AIL_3);
	}
}

212 213 214 215 216 217
/*
 * 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
218
 * some early parsing of the device-tree to setup out MEMBLOCK
219 220 221 222 223 224 225 226 227 228 229 230 231 232
 * 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)
{
233 234
	static __initdata struct paca_struct boot_paca;

235 236
	/* -------- printk is _NOT_ safe to use here ! ------- */

237
	/* Identify CPU type */
238
	identify_cpu(0, mfspr(SPRN_PVR));
239

240
	/* Assume we're on cpu 0 for now. Don't write to the paca yet! */
241 242
	initialise_paca(&boot_paca, 0);
	setup_paca(&boot_paca);
243
	fixup_boot_paca();
244

245 246 247
	/* Initialize lockdep early or else spinlocks will blow */
	lockdep_init();

248 249
	/* -------- printk is now safe to use ------- */

250 251 252
	/* Enable early debugging if any specified (see udbg.h) */
	udbg_early_init();

253
 	DBG(" -> early_setup(), dt_ptr: 0x%lx\n", dt_ptr);
254 255

	/*
256 257 258
	 * Do early initialization using the flattened device
	 * tree, such as retrieving the physical memory map or
	 * calculating/retrieving the hash table size.
259 260 261
	 */
	early_init_devtree(__va(dt_ptr));

262 263
	epapr_paravirt_early_init();

264
	/* Now we know the logical id of our boot cpu, setup the paca. */
265
	setup_paca(&paca[boot_cpuid]);
266
	fixup_boot_paca();
267

268 269
	/* Probe the machine type */
	probe_machine();
270

271
	setup_kdump_trampoline();
272

273 274
	DBG("Found, Initializing memory management...\n");

275 276
	/* Initialize the hash table or TLB handling */
	early_init_mmu();
277

278 279 280
	/*
	 * At this point, we can let interrupts switch to virtual mode
	 * (the MMU has been setup), so adjust the MSR in the PACA to
281
	 * have IR and DR set and enable AIL if it exists
282
	 */
283
	cpu_ready_for_interrupts();
284 285

	/* Reserve large chunks of memory for use by CMA for KVM */
286 287
	kvm_cma_reserve();

288 289 290 291 292 293 294
	/*
	 * 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();

295
	DBG(" <- early_setup()\n");
296 297 298 299 300 301 302 303 304 305 306 307

#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 */
308 309
}

310 311 312
#ifdef CONFIG_SMP
void early_setup_secondary(void)
{
313
	/* Mark interrupts enabled in PACA */
314
	get_paca()->soft_enabled = 0;
315

316 317
	/* Initialize the hash table or TLB handling */
	early_init_mmu_secondary();
318 319 320 321 322 323

	/*
	 * 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.
	 */
324
	cpu_ready_for_interrupts();
325 326 327
}

#endif /* CONFIG_SMP */
328

329 330 331
#if defined(CONFIG_SMP) || defined(CONFIG_KEXEC)
void smp_release_cpus(void)
{
332
	unsigned long *ptr;
333
	int i;
334 335 336 337 338 339 340

	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.
341
	 */
342

343 344
	ptr  = (unsigned long *)((unsigned long)&__secondary_hold_spinloop
			- PHYSICAL_START);
345
	*ptr = ppc_function_entry(generic_secondary_smp_init);
346 347 348 349 350

	/* And wait a bit for them to catch up */
	for (i = 0; i < 100000; i++) {
		mb();
		HMT_low();
351
		if (spinning_secondaries == 0)
352 353 354
			break;
		udelay(1);
	}
355
	DBG("spinning_secondaries = %d\n", spinning_secondaries);
356 357 358 359 360

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

361
/*
362 363
 * Initialize some remaining members of the ppc64_caches and systemcfg
 * structures
364 365 366 367 368 369 370 371 372 373 374
 * (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");

375
	for_each_node_by_type(np, "cpu") {
376 377
		num_cpus += 1;

A
Anton Blanchard 已提交
378 379
		/*
		 * We're assuming *all* of the CPUs have the same
380 381
		 * d-cache and i-cache sizes... -Peter
		 */
A
Anton Blanchard 已提交
382
		if (num_cpus == 1) {
383
			const __be32 *sizep, *lsizep;
384 385 386 387
			u32 size, lsize;

			size = 0;
			lsize = cur_cpu_spec->dcache_bsize;
388
			sizep = of_get_property(np, "d-cache-size", NULL);
389
			if (sizep != NULL)
390
				size = be32_to_cpu(*sizep);
A
Anton Blanchard 已提交
391 392
			lsizep = of_get_property(np, "d-cache-block-size",
						 NULL);
393 394
			/* fallback if block size missing */
			if (lsizep == NULL)
A
Anton Blanchard 已提交
395 396 397
				lsizep = of_get_property(np,
							 "d-cache-line-size",
							 NULL);
398
			if (lsizep != NULL)
399
				lsize = be32_to_cpu(*lsizep);
400
			if (sizep == NULL || lsizep == NULL)
401 402 403
				DBG("Argh, can't find dcache properties ! "
				    "sizep: %p, lsizep: %p\n", sizep, lsizep);

404 405
			ppc64_caches.dsize = size;
			ppc64_caches.dline_size = lsize;
406 407 408 409 410
			ppc64_caches.log_dline_size = __ilog2(lsize);
			ppc64_caches.dlines_per_page = PAGE_SIZE / lsize;

			size = 0;
			lsize = cur_cpu_spec->icache_bsize;
411
			sizep = of_get_property(np, "i-cache-size", NULL);
412
			if (sizep != NULL)
413
				size = be32_to_cpu(*sizep);
A
Anton Blanchard 已提交
414 415
			lsizep = of_get_property(np, "i-cache-block-size",
						 NULL);
416
			if (lsizep == NULL)
A
Anton Blanchard 已提交
417 418 419
				lsizep = of_get_property(np,
							 "i-cache-line-size",
							 NULL);
420
			if (lsizep != NULL)
421
				lsize = be32_to_cpu(*lsizep);
422
			if (sizep == NULL || lsizep == NULL)
423 424 425
				DBG("Argh, can't find icache properties ! "
				    "sizep: %p, lsizep: %p\n", sizep, lsizep);

426 427
			ppc64_caches.isize = size;
			ppc64_caches.iline_size = lsize;
428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444
			ppc64_caches.log_iline_size = __ilog2(lsize);
			ppc64_caches.ilines_per_page = PAGE_SIZE / lsize;
		}
	}

	DBG(" <- initialize_cache_info()\n");
}


/*
 * Do some initial setup of the system.  The parameters are those which 
 * were passed in from the bootloader.
 */
void __init setup_system(void)
{
	DBG(" -> setup_system()\n");

445 446
	/* Apply the CPUs-specific and firmware specific fixups to kernel
	 * text (nop out sections not relevant to this CPU or this firmware)
447
	 */
448
	do_feature_fixups(cur_cpu_spec->cpu_features,
449
			  &__start___ftr_fixup, &__stop___ftr_fixup);
450 451
	do_feature_fixups(cur_cpu_spec->mmu_features,
			  &__start___mmu_ftr_fixup, &__stop___mmu_ftr_fixup);
452 453
	do_feature_fixups(powerpc_firmware_features,
			  &__start___fw_ftr_fixup, &__stop___fw_ftr_fixup);
K
Kumar Gala 已提交
454 455
	do_lwsync_fixups(cur_cpu_spec->cpu_features,
			 &__start___lwsync_fixup, &__stop___lwsync_fixup);
456
	do_final_fixups();
457

458 459 460 461 462 463 464
	/*
	 * Unflatten the device-tree passed by prom_init or kexec
	 */
	unflatten_device_tree();

	/*
	 * Fill the ppc64_caches & systemcfg structures with informations
465
 	 * retrieved from the device-tree.
466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485
	 */
	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();

	/*
	 * Do some platform specific early initializations, that includes
	 * setting up the hash table pointers. It also sets up some interrupt-mapping
	 * related options that will be used by finish_device_tree()
	 */
486 487
	if (ppc_md.init_early)
		ppc_md.init_early();
488

489 490 491 492 493 494 495
 	/*
	 * 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();

496 497 498 499 500
	/*
	 * Register early console
	 */
	register_early_udbg_console();

501 502 503 504
	/*
	 * Initialize xmon
	 */
	xmon_setup();
505

P
Paul Mackerras 已提交
506
	smp_setup_cpu_maps();
507
	check_smt_enabled();
508
	setup_tlb_core_data();
509

510
#ifdef CONFIG_SMP
511 512 513 514
	/* Release secondary cpus out of their spinloops at 0x60 now that
	 * we can map physical -> logical CPU ids
	 */
	smp_release_cpus();
515
#endif
516

517
	printk("Starting Linux PPC64 %s\n", init_utsname()->version);
518 519

	printk("-----------------------------------------------------\n");
520
	printk("ppc64_pft_size                = 0x%llx\n", ppc64_pft_size);
Y
Yinghai Lu 已提交
521
	printk("physicalMemorySize            = 0x%llx\n", memblock_phys_mem_size());
522 523 524 525 526 527
	if (ppc64_caches.dline_size != 0x80)
		printk("ppc64_caches.dcache_line_size = 0x%x\n",
		       ppc64_caches.dline_size);
	if (ppc64_caches.iline_size != 0x80)
		printk("ppc64_caches.icache_line_size = 0x%x\n",
		       ppc64_caches.iline_size);
528
#ifdef CONFIG_PPC_STD_MMU_64
529 530
	if (htab_address)
		printk("htab_address                  = 0x%p\n", htab_address);
531
	printk("htab_hash_mask                = 0x%lx\n", htab_hash_mask);
532
#endif /* CONFIG_PPC_STD_MMU_64 */
533
	if (PHYSICAL_START > 0)
534 535
		printk("physical_start                = 0x%llx\n",
		       (unsigned long long)PHYSICAL_START);
536 537 538 539 540
	printk("-----------------------------------------------------\n");

	DBG(" <- setup_system()\n");
}

541 542 543 544 545 546
/* 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)
547
{
548 549 550 551 552 553 554 555 556
#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))
557 558
		return 1UL << SID_SHIFT_1T;
	return 1UL << SID_SHIFT;
559
#endif
560 561
}

562 563
static void __init irqstack_early_init(void)
{
564
	u64 limit = safe_stack_limit();
565 566 567
	unsigned int i;

	/*
568 569
	 * Interrupt stacks must be in the first segment since we
	 * cannot afford to take SLB misses on them.
570
	 */
571
	for_each_possible_cpu(i) {
572
		softirq_ctx[i] = (struct thread_info *)
Y
Yinghai Lu 已提交
573
			__va(memblock_alloc_base(THREAD_SIZE,
574
					    THREAD_SIZE, limit));
575
		hardirq_ctx[i] = (struct thread_info *)
Y
Yinghai Lu 已提交
576
			__va(memblock_alloc_base(THREAD_SIZE,
577
					    THREAD_SIZE, limit));
578 579 580
	}
}

581 582 583 584
#ifdef CONFIG_PPC_BOOK3E
static void __init exc_lvl_early_init(void)
{
	unsigned int i;
585
	unsigned long sp;
586 587

	for_each_possible_cpu(i) {
588 589 590 591 592 593 594 595 596 597 598
		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);
599
	}
600 601

	if (cpu_has_feature(CPU_FTR_DEBUG_LVL_EXC))
602
		patch_exception(0x040, exc_debug_debug_book3e);
603 604 605 606 607
}
#else
#define exc_lvl_early_init()
#endif

608 609
/*
 * Stack space used when we detect a bad kernel stack pointer, and
610 611
 * early in SMP boots before relocation is enabled. Exclusive emergency
 * stack for machine checks.
612 613 614
 */
static void __init emergency_stack_init(void)
{
615
	u64 limit;
616 617 618 619 620 621 622 623 624 625 626
	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.
	 */
627
	limit = min(safe_stack_limit(), ppc64_rma_size);
628

629 630
	for_each_possible_cpu(i) {
		unsigned long sp;
Y
Yinghai Lu 已提交
631
		sp  = memblock_alloc_base(THREAD_SIZE, THREAD_SIZE, limit);
632 633
		sp += THREAD_SIZE;
		paca[i].emergency_sp = __va(sp);
634 635 636 637 638 639 640

#ifdef CONFIG_PPC_BOOK3S_64
		/* emergency stack for machine check exception handling. */
		sp  = memblock_alloc_base(THREAD_SIZE, THREAD_SIZE, limit);
		sp += THREAD_SIZE;
		paca[i].mc_emergency_sp = __va(sp);
#endif
641
	}
642 643 644
}

/*
645 646
 * Called into from start_kernel this initializes bootmem, which is used
 * to manage page allocation until mem_init is called.
647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662
 */
void __init setup_arch(char **cmdline_p)
{
	ppc64_boot_msg(0x12, "Setup Arch");

	*cmdline_p = cmd_line;

	/*
	 * Set cache line size based on type of cpu as a default.
	 * Systems with OF can look in the properties on the cpu node(s)
	 * for a possibly more accurate value.
	 */
	dcache_bsize = ppc64_caches.dline_size;
	icache_bsize = ppc64_caches.iline_size;

	if (ppc_md.panic)
663
		setup_panic();
664

665
	init_mm.start_code = (unsigned long)_stext;
666 667 668
	init_mm.end_code = (unsigned long) _etext;
	init_mm.end_data = (unsigned long) _edata;
	init_mm.brk = klimit;
669 670 671
#ifdef CONFIG_PPC_64K_PAGES
	init_mm.context.pte_frag = NULL;
#endif
672
	irqstack_early_init();
673
	exc_lvl_early_init();
674 675
	emergency_stack_init();

676
#ifdef CONFIG_PPC_STD_MMU_64
677
	stabs_alloc();
678
#endif
679 680 681 682
	/* set up the bootmem stuff with available memory */
	do_init_bootmem();
	sparse_init();

683 684 685 686
#ifdef CONFIG_DUMMY_CONSOLE
	conswitchp = &dummy_con;
#endif

687 688
	if (ppc_md.setup_arch)
		ppc_md.setup_arch();
689 690

	paging_init();
691 692 693 694

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

695 696 697 698 699
	/* Interrupt code needs to be 64K-aligned */
	if ((unsigned long)_stext & 0xffff)
		panic("Kernelbase not 64K-aligned (0x%lx)!\n",
		      (unsigned long)_stext);

700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727
	ppc64_boot_msg(0x15, "Setup Done");
}


/* ToDo: do something useful if ppc_md is not yet setup. */
#define PPC64_LINUX_FUNCTION 0x0f000000
#define PPC64_IPL_MESSAGE 0xc0000000
#define PPC64_TERM_MESSAGE 0xb0000000

static void ppc64_do_msg(unsigned int src, const char *msg)
{
	if (ppc_md.progress) {
		char buf[128];

		sprintf(buf, "%08X\n", src);
		ppc_md.progress(buf, 0);
		snprintf(buf, 128, "%s", msg);
		ppc_md.progress(buf, 0);
	}
}

/* Print a boot progress message. */
void ppc64_boot_msg(unsigned int src, const char *msg)
{
	ppc64_do_msg(PPC64_LINUX_FUNCTION|PPC64_IPL_MESSAGE|src, msg);
	printk("[boot]%04x %s\n", src, msg);
}

728
#ifdef CONFIG_SMP
729 730 731
#define PCPU_DYN_SIZE		()

static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size, size_t align)
732
{
733 734 735
	return __alloc_bootmem_node(NODE_DATA(cpu_to_node(cpu)), size, align,
				    __pa(MAX_DMA_ADDRESS));
}
736

737 738 739 740
static void __init pcpu_fc_free(void *ptr, size_t size)
{
	free_bootmem(__pa(ptr), size);
}
741

742 743 744 745 746 747 748 749
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;
}

750 751 752
unsigned long __per_cpu_offset[NR_CPUS] __read_mostly;
EXPORT_SYMBOL(__per_cpu_offset);

753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776
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;
777 778 779 780
	for_each_possible_cpu(cpu) {
                __per_cpu_offset[cpu] = delta + pcpu_unit_offsets[cpu];
		paca[cpu].data_offset = __per_cpu_offset[cpu];
	}
781 782
}
#endif
783

784 785 786 787 788 789 790 791 792
#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
793

794
#if defined(CONFIG_PPC_INDIRECT_PIO) || defined(CONFIG_PPC_INDIRECT_MMIO)
795 796
struct ppc_pci_io ppc_pci_io;
EXPORT_SYMBOL(ppc_pci_io);
797
#endif