xref: /linux/init/main.c (revision 8fa5723aa7e053d498336b48448b292fc2e0458b)
1 /*
2  *  linux/init/main.c
3  *
4  *  Copyright (C) 1991, 1992  Linus Torvalds
5  *
6  *  GK 2/5/95  -  Changed to support mounting root fs via NFS
7  *  Added initrd & change_root: Werner Almesberger & Hans Lermen, Feb '96
8  *  Moan early if gcc is old, avoiding bogus kernels - Paul Gortmaker, May '96
9  *  Simplified starting of init:  Michael A. Griffith <grif@acm.org>
10  */
11 
12 #include <linux/types.h>
13 #include <linux/module.h>
14 #include <linux/proc_fs.h>
15 #include <linux/kernel.h>
16 #include <linux/syscalls.h>
17 #include <linux/string.h>
18 #include <linux/ctype.h>
19 #include <linux/delay.h>
20 #include <linux/utsname.h>
21 #include <linux/ioport.h>
22 #include <linux/init.h>
23 #include <linux/smp_lock.h>
24 #include <linux/initrd.h>
25 #include <linux/bootmem.h>
26 #include <linux/tty.h>
27 #include <linux/gfp.h>
28 #include <linux/percpu.h>
29 #include <linux/kmod.h>
30 #include <linux/vmalloc.h>
31 #include <linux/kernel_stat.h>
32 #include <linux/start_kernel.h>
33 #include <linux/security.h>
34 #include <linux/smp.h>
35 #include <linux/workqueue.h>
36 #include <linux/profile.h>
37 #include <linux/rcupdate.h>
38 #include <linux/moduleparam.h>
39 #include <linux/kallsyms.h>
40 #include <linux/writeback.h>
41 #include <linux/cpu.h>
42 #include <linux/cpuset.h>
43 #include <linux/cgroup.h>
44 #include <linux/efi.h>
45 #include <linux/tick.h>
46 #include <linux/interrupt.h>
47 #include <linux/taskstats_kern.h>
48 #include <linux/delayacct.h>
49 #include <linux/unistd.h>
50 #include <linux/rmap.h>
51 #include <linux/mempolicy.h>
52 #include <linux/key.h>
53 #include <linux/unwind.h>
54 #include <linux/buffer_head.h>
55 #include <linux/page_cgroup.h>
56 #include <linux/debug_locks.h>
57 #include <linux/debugobjects.h>
58 #include <linux/lockdep.h>
59 #include <linux/pid_namespace.h>
60 #include <linux/device.h>
61 #include <linux/kthread.h>
62 #include <linux/sched.h>
63 #include <linux/signal.h>
64 #include <linux/idr.h>
65 #include <linux/ftrace.h>
66 
67 #include <asm/io.h>
68 #include <asm/bugs.h>
69 #include <asm/setup.h>
70 #include <asm/sections.h>
71 #include <asm/cacheflush.h>
72 
73 #ifdef CONFIG_X86_LOCAL_APIC
74 #include <asm/smp.h>
75 #endif
76 
77 /*
78  * This is one of the first .c files built. Error out early if we have compiler
79  * trouble.
80  */
81 
82 #if __GNUC__ == 4 && __GNUC_MINOR__ == 1 && __GNUC_PATCHLEVEL__ == 0
83 #warning gcc-4.1.0 is known to miscompile the kernel.  A different compiler version is recommended.
84 #endif
85 
86 static int kernel_init(void *);
87 
88 extern void init_IRQ(void);
89 extern void fork_init(unsigned long);
90 extern void mca_init(void);
91 extern void sbus_init(void);
92 extern void prio_tree_init(void);
93 extern void radix_tree_init(void);
94 extern void free_initmem(void);
95 #ifdef	CONFIG_ACPI
96 extern void acpi_early_init(void);
97 #else
98 static inline void acpi_early_init(void) { }
99 #endif
100 #ifndef CONFIG_DEBUG_RODATA
101 static inline void mark_rodata_ro(void) { }
102 #endif
103 
104 #ifdef CONFIG_TC
105 extern void tc_init(void);
106 #endif
107 
108 enum system_states system_state;
109 EXPORT_SYMBOL(system_state);
110 
111 /*
112  * Boot command-line arguments
113  */
114 #define MAX_INIT_ARGS CONFIG_INIT_ENV_ARG_LIMIT
115 #define MAX_INIT_ENVS CONFIG_INIT_ENV_ARG_LIMIT
116 
117 extern void time_init(void);
118 /* Default late time init is NULL. archs can override this later. */
119 void (*late_time_init)(void);
120 extern void softirq_init(void);
121 
122 /* Untouched command line saved by arch-specific code. */
123 char __initdata boot_command_line[COMMAND_LINE_SIZE];
124 /* Untouched saved command line (eg. for /proc) */
125 char *saved_command_line;
126 /* Command line for parameter parsing */
127 static char *static_command_line;
128 
129 static char *execute_command;
130 static char *ramdisk_execute_command;
131 
132 #ifdef CONFIG_SMP
133 /* Setup configured maximum number of CPUs to activate */
134 unsigned int __initdata setup_max_cpus = NR_CPUS;
135 
136 /*
137  * Setup routine for controlling SMP activation
138  *
139  * Command-line option of "nosmp" or "maxcpus=0" will disable SMP
140  * activation entirely (the MPS table probe still happens, though).
141  *
142  * Command-line option of "maxcpus=<NUM>", where <NUM> is an integer
143  * greater than 0, limits the maximum number of CPUs activated in
144  * SMP mode to <NUM>.
145  */
146 #ifndef CONFIG_X86_IO_APIC
147 static inline void disable_ioapic_setup(void) {};
148 #endif
149 
150 static int __init nosmp(char *str)
151 {
152 	setup_max_cpus = 0;
153 	disable_ioapic_setup();
154 	return 0;
155 }
156 
157 early_param("nosmp", nosmp);
158 
159 static int __init maxcpus(char *str)
160 {
161 	get_option(&str, &setup_max_cpus);
162 	if (setup_max_cpus == 0)
163 		disable_ioapic_setup();
164 
165 	return 0;
166 }
167 
168 early_param("maxcpus", maxcpus);
169 #else
170 #define setup_max_cpus NR_CPUS
171 #endif
172 
173 /*
174  * If set, this is an indication to the drivers that reset the underlying
175  * device before going ahead with the initialization otherwise driver might
176  * rely on the BIOS and skip the reset operation.
177  *
178  * This is useful if kernel is booting in an unreliable environment.
179  * For ex. kdump situaiton where previous kernel has crashed, BIOS has been
180  * skipped and devices will be in unknown state.
181  */
182 unsigned int reset_devices;
183 EXPORT_SYMBOL(reset_devices);
184 
185 static int __init set_reset_devices(char *str)
186 {
187 	reset_devices = 1;
188 	return 1;
189 }
190 
191 __setup("reset_devices", set_reset_devices);
192 
193 static char * argv_init[MAX_INIT_ARGS+2] = { "init", NULL, };
194 char * envp_init[MAX_INIT_ENVS+2] = { "HOME=/", "TERM=linux", NULL, };
195 static const char *panic_later, *panic_param;
196 
197 extern struct obs_kernel_param __setup_start[], __setup_end[];
198 
199 static int __init obsolete_checksetup(char *line)
200 {
201 	struct obs_kernel_param *p;
202 	int had_early_param = 0;
203 
204 	p = __setup_start;
205 	do {
206 		int n = strlen(p->str);
207 		if (!strncmp(line, p->str, n)) {
208 			if (p->early) {
209 				/* Already done in parse_early_param?
210 				 * (Needs exact match on param part).
211 				 * Keep iterating, as we can have early
212 				 * params and __setups of same names 8( */
213 				if (line[n] == '\0' || line[n] == '=')
214 					had_early_param = 1;
215 			} else if (!p->setup_func) {
216 				printk(KERN_WARNING "Parameter %s is obsolete,"
217 				       " ignored\n", p->str);
218 				return 1;
219 			} else if (p->setup_func(line + n))
220 				return 1;
221 		}
222 		p++;
223 	} while (p < __setup_end);
224 
225 	return had_early_param;
226 }
227 
228 /*
229  * This should be approx 2 Bo*oMips to start (note initial shift), and will
230  * still work even if initially too large, it will just take slightly longer
231  */
232 unsigned long loops_per_jiffy = (1<<12);
233 
234 EXPORT_SYMBOL(loops_per_jiffy);
235 
236 static int __init debug_kernel(char *str)
237 {
238 	console_loglevel = 10;
239 	return 0;
240 }
241 
242 static int __init quiet_kernel(char *str)
243 {
244 	console_loglevel = 4;
245 	return 0;
246 }
247 
248 early_param("debug", debug_kernel);
249 early_param("quiet", quiet_kernel);
250 
251 static int __init loglevel(char *str)
252 {
253 	get_option(&str, &console_loglevel);
254 	return 0;
255 }
256 
257 early_param("loglevel", loglevel);
258 
259 /*
260  * Unknown boot options get handed to init, unless they look like
261  * failed parameters
262  */
263 static int __init unknown_bootoption(char *param, char *val)
264 {
265 	/* Change NUL term back to "=", to make "param" the whole string. */
266 	if (val) {
267 		/* param=val or param="val"? */
268 		if (val == param+strlen(param)+1)
269 			val[-1] = '=';
270 		else if (val == param+strlen(param)+2) {
271 			val[-2] = '=';
272 			memmove(val-1, val, strlen(val)+1);
273 			val--;
274 		} else
275 			BUG();
276 	}
277 
278 	/* Handle obsolete-style parameters */
279 	if (obsolete_checksetup(param))
280 		return 0;
281 
282 	/*
283 	 * Preemptive maintenance for "why didn't my misspelled command
284 	 * line work?"
285 	 */
286 	if (strchr(param, '.') && (!val || strchr(param, '.') < val)) {
287 		printk(KERN_ERR "Unknown boot option `%s': ignoring\n", param);
288 		return 0;
289 	}
290 
291 	if (panic_later)
292 		return 0;
293 
294 	if (val) {
295 		/* Environment option */
296 		unsigned int i;
297 		for (i = 0; envp_init[i]; i++) {
298 			if (i == MAX_INIT_ENVS) {
299 				panic_later = "Too many boot env vars at `%s'";
300 				panic_param = param;
301 			}
302 			if (!strncmp(param, envp_init[i], val - param))
303 				break;
304 		}
305 		envp_init[i] = param;
306 	} else {
307 		/* Command line option */
308 		unsigned int i;
309 		for (i = 0; argv_init[i]; i++) {
310 			if (i == MAX_INIT_ARGS) {
311 				panic_later = "Too many boot init vars at `%s'";
312 				panic_param = param;
313 			}
314 		}
315 		argv_init[i] = param;
316 	}
317 	return 0;
318 }
319 
320 #ifdef CONFIG_DEBUG_PAGEALLOC
321 int __read_mostly debug_pagealloc_enabled = 0;
322 #endif
323 
324 static int __init init_setup(char *str)
325 {
326 	unsigned int i;
327 
328 	execute_command = str;
329 	/*
330 	 * In case LILO is going to boot us with default command line,
331 	 * it prepends "auto" before the whole cmdline which makes
332 	 * the shell think it should execute a script with such name.
333 	 * So we ignore all arguments entered _before_ init=... [MJ]
334 	 */
335 	for (i = 1; i < MAX_INIT_ARGS; i++)
336 		argv_init[i] = NULL;
337 	return 1;
338 }
339 __setup("init=", init_setup);
340 
341 static int __init rdinit_setup(char *str)
342 {
343 	unsigned int i;
344 
345 	ramdisk_execute_command = str;
346 	/* See "auto" comment in init_setup */
347 	for (i = 1; i < MAX_INIT_ARGS; i++)
348 		argv_init[i] = NULL;
349 	return 1;
350 }
351 __setup("rdinit=", rdinit_setup);
352 
353 #ifndef CONFIG_SMP
354 
355 #ifdef CONFIG_X86_LOCAL_APIC
356 static void __init smp_init(void)
357 {
358 	APIC_init_uniprocessor();
359 }
360 #else
361 #define smp_init()	do { } while (0)
362 #endif
363 
364 static inline void setup_per_cpu_areas(void) { }
365 static inline void setup_nr_cpu_ids(void) { }
366 static inline void smp_prepare_cpus(unsigned int maxcpus) { }
367 
368 #else
369 
370 #if NR_CPUS > BITS_PER_LONG
371 cpumask_t cpu_mask_all __read_mostly = CPU_MASK_ALL;
372 EXPORT_SYMBOL(cpu_mask_all);
373 #endif
374 
375 /* Setup number of possible processor ids */
376 int nr_cpu_ids __read_mostly = NR_CPUS;
377 EXPORT_SYMBOL(nr_cpu_ids);
378 
379 /* An arch may set nr_cpu_ids earlier if needed, so this would be redundant */
380 static void __init setup_nr_cpu_ids(void)
381 {
382 	int cpu, highest_cpu = 0;
383 
384 	for_each_possible_cpu(cpu)
385 		highest_cpu = cpu;
386 
387 	nr_cpu_ids = highest_cpu + 1;
388 }
389 
390 #ifndef CONFIG_HAVE_SETUP_PER_CPU_AREA
391 unsigned long __per_cpu_offset[NR_CPUS] __read_mostly;
392 
393 EXPORT_SYMBOL(__per_cpu_offset);
394 
395 static void __init setup_per_cpu_areas(void)
396 {
397 	unsigned long size, i;
398 	char *ptr;
399 	unsigned long nr_possible_cpus = num_possible_cpus();
400 
401 	/* Copy section for each CPU (we discard the original) */
402 	size = ALIGN(PERCPU_ENOUGH_ROOM, PAGE_SIZE);
403 	ptr = alloc_bootmem_pages(size * nr_possible_cpus);
404 
405 	for_each_possible_cpu(i) {
406 		__per_cpu_offset[i] = ptr - __per_cpu_start;
407 		memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start);
408 		ptr += size;
409 	}
410 }
411 #endif /* CONFIG_HAVE_SETUP_PER_CPU_AREA */
412 
413 /* Called by boot processor to activate the rest. */
414 static void __init smp_init(void)
415 {
416 	unsigned int cpu;
417 
418 	/*
419 	 * Set up the current CPU as possible to migrate to.
420 	 * The other ones will be done by cpu_up/cpu_down()
421 	 */
422 	cpu = smp_processor_id();
423 	cpu_set(cpu, cpu_active_map);
424 
425 	/* FIXME: This should be done in userspace --RR */
426 	for_each_present_cpu(cpu) {
427 		if (num_online_cpus() >= setup_max_cpus)
428 			break;
429 		if (!cpu_online(cpu))
430 			cpu_up(cpu);
431 	}
432 
433 	/* Any cleanup work */
434 	printk(KERN_INFO "Brought up %ld CPUs\n", (long)num_online_cpus());
435 	smp_cpus_done(setup_max_cpus);
436 }
437 
438 #endif
439 
440 /*
441  * We need to store the untouched command line for future reference.
442  * We also need to store the touched command line since the parameter
443  * parsing is performed in place, and we should allow a component to
444  * store reference of name/value for future reference.
445  */
446 static void __init setup_command_line(char *command_line)
447 {
448 	saved_command_line = alloc_bootmem(strlen (boot_command_line)+1);
449 	static_command_line = alloc_bootmem(strlen (command_line)+1);
450 	strcpy (saved_command_line, boot_command_line);
451 	strcpy (static_command_line, command_line);
452 }
453 
454 /*
455  * We need to finalize in a non-__init function or else race conditions
456  * between the root thread and the init thread may cause start_kernel to
457  * be reaped by free_initmem before the root thread has proceeded to
458  * cpu_idle.
459  *
460  * gcc-3.4 accidentally inlines this function, so use noinline.
461  */
462 
463 static void noinline __init_refok rest_init(void)
464 	__releases(kernel_lock)
465 {
466 	int pid;
467 
468 	kernel_thread(kernel_init, NULL, CLONE_FS | CLONE_SIGHAND);
469 	numa_default_policy();
470 	pid = kernel_thread(kthreadd, NULL, CLONE_FS | CLONE_FILES);
471 	kthreadd_task = find_task_by_pid_ns(pid, &init_pid_ns);
472 	unlock_kernel();
473 
474 	/*
475 	 * The boot idle thread must execute schedule()
476 	 * at least once to get things moving:
477 	 */
478 	init_idle_bootup_task(current);
479 	preempt_enable_no_resched();
480 	schedule();
481 	preempt_disable();
482 
483 	/* Call into cpu_idle with preempt disabled */
484 	cpu_idle();
485 }
486 
487 /* Check for early params. */
488 static int __init do_early_param(char *param, char *val)
489 {
490 	struct obs_kernel_param *p;
491 
492 	for (p = __setup_start; p < __setup_end; p++) {
493 		if ((p->early && strcmp(param, p->str) == 0) ||
494 		    (strcmp(param, "console") == 0 &&
495 		     strcmp(p->str, "earlycon") == 0)
496 		) {
497 			if (p->setup_func(val) != 0)
498 				printk(KERN_WARNING
499 				       "Malformed early option '%s'\n", param);
500 		}
501 	}
502 	/* We accept everything at this stage. */
503 	return 0;
504 }
505 
506 /* Arch code calls this early on, or if not, just before other parsing. */
507 void __init parse_early_param(void)
508 {
509 	static __initdata int done = 0;
510 	static __initdata char tmp_cmdline[COMMAND_LINE_SIZE];
511 
512 	if (done)
513 		return;
514 
515 	/* All fall through to do_early_param. */
516 	strlcpy(tmp_cmdline, boot_command_line, COMMAND_LINE_SIZE);
517 	parse_args("early options", tmp_cmdline, NULL, 0, do_early_param);
518 	done = 1;
519 }
520 
521 /*
522  *	Activate the first processor.
523  */
524 
525 static void __init boot_cpu_init(void)
526 {
527 	int cpu = smp_processor_id();
528 	/* Mark the boot cpu "present", "online" etc for SMP and UP case */
529 	cpu_set(cpu, cpu_online_map);
530 	cpu_set(cpu, cpu_present_map);
531 	cpu_set(cpu, cpu_possible_map);
532 }
533 
534 void __init __weak smp_setup_processor_id(void)
535 {
536 }
537 
538 void __init __weak thread_info_cache_init(void)
539 {
540 }
541 
542 asmlinkage void __init start_kernel(void)
543 {
544 	char * command_line;
545 	extern struct kernel_param __start___param[], __stop___param[];
546 
547 	smp_setup_processor_id();
548 
549 	/*
550 	 * Need to run as early as possible, to initialize the
551 	 * lockdep hash:
552 	 */
553 	unwind_init();
554 	lockdep_init();
555 	debug_objects_early_init();
556 	cgroup_init_early();
557 
558 	local_irq_disable();
559 	early_boot_irqs_off();
560 	early_init_irq_lock_class();
561 
562 /*
563  * Interrupts are still disabled. Do necessary setups, then
564  * enable them
565  */
566 	lock_kernel();
567 	tick_init();
568 	boot_cpu_init();
569 	page_address_init();
570 	printk(KERN_NOTICE);
571 	printk(linux_banner);
572 	setup_arch(&command_line);
573 	mm_init_owner(&init_mm, &init_task);
574 	setup_command_line(command_line);
575 	unwind_setup();
576 	setup_per_cpu_areas();
577 	setup_nr_cpu_ids();
578 	smp_prepare_boot_cpu();	/* arch-specific boot-cpu hooks */
579 
580 	/*
581 	 * Set up the scheduler prior starting any interrupts (such as the
582 	 * timer interrupt). Full topology setup happens at smp_init()
583 	 * time - but meanwhile we still have a functioning scheduler.
584 	 */
585 	sched_init();
586 	/*
587 	 * Disable preemption - early bootup scheduling is extremely
588 	 * fragile until we cpu_idle() for the first time.
589 	 */
590 	preempt_disable();
591 	build_all_zonelists();
592 	page_alloc_init();
593 	printk(KERN_NOTICE "Kernel command line: %s\n", boot_command_line);
594 	parse_early_param();
595 	parse_args("Booting kernel", static_command_line, __start___param,
596 		   __stop___param - __start___param,
597 		   &unknown_bootoption);
598 	if (!irqs_disabled()) {
599 		printk(KERN_WARNING "start_kernel(): bug: interrupts were "
600 				"enabled *very* early, fixing it\n");
601 		local_irq_disable();
602 	}
603 	sort_main_extable();
604 	trap_init();
605 	rcu_init();
606 	init_IRQ();
607 	pidhash_init();
608 	init_timers();
609 	hrtimers_init();
610 	softirq_init();
611 	timekeeping_init();
612 	time_init();
613 	sched_clock_init();
614 	profile_init();
615 	if (!irqs_disabled())
616 		printk("start_kernel(): bug: interrupts were enabled early\n");
617 	early_boot_irqs_on();
618 	local_irq_enable();
619 
620 	/*
621 	 * HACK ALERT! This is early. We're enabling the console before
622 	 * we've done PCI setups etc, and console_init() must be aware of
623 	 * this. But we do want output early, in case something goes wrong.
624 	 */
625 	console_init();
626 	if (panic_later)
627 		panic(panic_later, panic_param);
628 
629 	lockdep_info();
630 
631 	/*
632 	 * Need to run this when irqs are enabled, because it wants
633 	 * to self-test [hard/soft]-irqs on/off lock inversion bugs
634 	 * too:
635 	 */
636 	locking_selftest();
637 
638 #ifdef CONFIG_BLK_DEV_INITRD
639 	if (initrd_start && !initrd_below_start_ok &&
640 	    page_to_pfn(virt_to_page((void *)initrd_start)) < min_low_pfn) {
641 		printk(KERN_CRIT "initrd overwritten (0x%08lx < 0x%08lx) - "
642 		    "disabling it.\n",
643 		    page_to_pfn(virt_to_page((void *)initrd_start)),
644 		    min_low_pfn);
645 		initrd_start = 0;
646 	}
647 #endif
648 	vmalloc_init();
649 	vfs_caches_init_early();
650 	cpuset_init_early();
651 	page_cgroup_init();
652 	mem_init();
653 	enable_debug_pagealloc();
654 	cpu_hotplug_init();
655 	kmem_cache_init();
656 	debug_objects_mem_init();
657 	idr_init_cache();
658 	setup_per_cpu_pageset();
659 	numa_policy_init();
660 	if (late_time_init)
661 		late_time_init();
662 	calibrate_delay();
663 	pidmap_init();
664 	pgtable_cache_init();
665 	prio_tree_init();
666 	anon_vma_init();
667 #ifdef CONFIG_X86
668 	if (efi_enabled)
669 		efi_enter_virtual_mode();
670 #endif
671 	thread_info_cache_init();
672 	fork_init(num_physpages);
673 	proc_caches_init();
674 	buffer_init();
675 	key_init();
676 	security_init();
677 	vfs_caches_init(num_physpages);
678 	radix_tree_init();
679 	signals_init();
680 	/* rootfs populating might need page-writeback */
681 	page_writeback_init();
682 #ifdef CONFIG_PROC_FS
683 	proc_root_init();
684 #endif
685 	cgroup_init();
686 	cpuset_init();
687 	taskstats_init_early();
688 	delayacct_init();
689 
690 	check_bugs();
691 
692 	acpi_early_init(); /* before LAPIC and SMP init */
693 
694 	ftrace_init();
695 
696 	/* Do the rest non-__init'ed, we're now alive */
697 	rest_init();
698 }
699 
700 static int initcall_debug;
701 core_param(initcall_debug, initcall_debug, bool, 0644);
702 
703 int do_one_initcall(initcall_t fn)
704 {
705 	int count = preempt_count();
706 	ktime_t delta;
707 	char msgbuf[64];
708 	struct boot_trace it;
709 
710 	if (initcall_debug) {
711 		it.caller = task_pid_nr(current);
712 		printk("calling  %pF @ %i\n", fn, it.caller);
713 		it.calltime = ktime_get();
714 	}
715 
716 	it.result = fn();
717 
718 	if (initcall_debug) {
719 		it.rettime = ktime_get();
720 		delta = ktime_sub(it.rettime, it.calltime);
721 		it.duration = (unsigned long long) delta.tv64 >> 10;
722 		printk("initcall %pF returned %d after %Ld usecs\n", fn,
723 			it.result, it.duration);
724 		trace_boot(&it, fn);
725 	}
726 
727 	msgbuf[0] = 0;
728 
729 	if (it.result && it.result != -ENODEV && initcall_debug)
730 		sprintf(msgbuf, "error code %d ", it.result);
731 
732 	if (preempt_count() != count) {
733 		strlcat(msgbuf, "preemption imbalance ", sizeof(msgbuf));
734 		preempt_count() = count;
735 	}
736 	if (irqs_disabled()) {
737 		strlcat(msgbuf, "disabled interrupts ", sizeof(msgbuf));
738 		local_irq_enable();
739 	}
740 	if (msgbuf[0]) {
741 		printk("initcall %pF returned with %s\n", fn, msgbuf);
742 	}
743 
744 	return it.result;
745 }
746 
747 
748 extern initcall_t __initcall_start[], __initcall_end[], __early_initcall_end[];
749 
750 static void __init do_initcalls(void)
751 {
752 	initcall_t *call;
753 
754 	for (call = __early_initcall_end; call < __initcall_end; call++)
755 		do_one_initcall(*call);
756 
757 	/* Make sure there is no pending stuff from the initcall sequence */
758 	flush_scheduled_work();
759 }
760 
761 /*
762  * Ok, the machine is now initialized. None of the devices
763  * have been touched yet, but the CPU subsystem is up and
764  * running, and memory and process management works.
765  *
766  * Now we can finally start doing some real work..
767  */
768 static void __init do_basic_setup(void)
769 {
770 	rcu_init_sched(); /* needed by module_init stage. */
771 	init_workqueues();
772 	usermodehelper_init();
773 	driver_init();
774 	init_irq_proc();
775 	do_initcalls();
776 }
777 
778 static void __init do_pre_smp_initcalls(void)
779 {
780 	initcall_t *call;
781 
782 	for (call = __initcall_start; call < __early_initcall_end; call++)
783 		do_one_initcall(*call);
784 }
785 
786 static void run_init_process(char *init_filename)
787 {
788 	argv_init[0] = init_filename;
789 	kernel_execve(init_filename, argv_init, envp_init);
790 }
791 
792 /* This is a non __init function. Force it to be noinline otherwise gcc
793  * makes it inline to init() and it becomes part of init.text section
794  */
795 static int noinline init_post(void)
796 {
797 	free_initmem();
798 	unlock_kernel();
799 	mark_rodata_ro();
800 	system_state = SYSTEM_RUNNING;
801 	numa_default_policy();
802 
803 	if (sys_open((const char __user *) "/dev/console", O_RDWR, 0) < 0)
804 		printk(KERN_WARNING "Warning: unable to open an initial console.\n");
805 
806 	(void) sys_dup(0);
807 	(void) sys_dup(0);
808 
809 	current->signal->flags |= SIGNAL_UNKILLABLE;
810 
811 	if (ramdisk_execute_command) {
812 		run_init_process(ramdisk_execute_command);
813 		printk(KERN_WARNING "Failed to execute %s\n",
814 				ramdisk_execute_command);
815 	}
816 
817 	/*
818 	 * We try each of these until one succeeds.
819 	 *
820 	 * The Bourne shell can be used instead of init if we are
821 	 * trying to recover a really broken machine.
822 	 */
823 	if (execute_command) {
824 		run_init_process(execute_command);
825 		printk(KERN_WARNING "Failed to execute %s.  Attempting "
826 					"defaults...\n", execute_command);
827 	}
828 	run_init_process("/sbin/init");
829 	run_init_process("/etc/init");
830 	run_init_process("/bin/init");
831 	run_init_process("/bin/sh");
832 
833 	panic("No init found.  Try passing init= option to kernel.");
834 }
835 
836 static int __init kernel_init(void * unused)
837 {
838 	lock_kernel();
839 	/*
840 	 * init can run on any cpu.
841 	 */
842 	set_cpus_allowed_ptr(current, CPU_MASK_ALL_PTR);
843 	/*
844 	 * Tell the world that we're going to be the grim
845 	 * reaper of innocent orphaned children.
846 	 *
847 	 * We don't want people to have to make incorrect
848 	 * assumptions about where in the task array this
849 	 * can be found.
850 	 */
851 	init_pid_ns.child_reaper = current;
852 
853 	cad_pid = task_pid(current);
854 
855 	smp_prepare_cpus(setup_max_cpus);
856 
857 	do_pre_smp_initcalls();
858 	start_boot_trace();
859 
860 	smp_init();
861 	sched_init_smp();
862 
863 	cpuset_init_smp();
864 
865 	do_basic_setup();
866 
867 	/*
868 	 * check if there is an early userspace init.  If yes, let it do all
869 	 * the work
870 	 */
871 
872 	if (!ramdisk_execute_command)
873 		ramdisk_execute_command = "/init";
874 
875 	if (sys_access((const char __user *) ramdisk_execute_command, 0) != 0) {
876 		ramdisk_execute_command = NULL;
877 		prepare_namespace();
878 	}
879 
880 	/*
881 	 * Ok, we have completed the initial bootup, and
882 	 * we're essentially up and running. Get rid of the
883 	 * initmem segments and start the user-mode stuff..
884 	 */
885 	stop_boot_trace();
886 	init_post();
887 	return 0;
888 }
889