init/main.c

   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/kernel_stat.h>
  31 #include <linux/start_kernel.h>
  32 #include <linux/security.h>
  33 #include <linux/smp.h>
  34 #include <linux/workqueue.h>
  35 #include <linux/profile.h>
  36 #include <linux/rcupdate.h>
  37 #include <linux/moduleparam.h>
  38 #include <linux/kallsyms.h>
  39 #include <linux/writeback.h>
  40 #include <linux/cpu.h>
  41 #include <linux/cpuset.h>
  42 #include <linux/cgroup.h>
  43 #include <linux/efi.h>
  44 #include <linux/tick.h>
  45 #include <linux/interrupt.h>
  46 #include <linux/taskstats_kern.h>
  47 #include <linux/delayacct.h>
  48 #include <linux/unistd.h>
  49 #include <linux/rmap.h>
  50 #include <linux/mempolicy.h>
  51 #include <linux/key.h>
  52 #include <linux/unwind.h>
  53 #include <linux/buffer_head.h>
  54 #include <linux/debug_locks.h>
  55 #include <linux/debugobjects.h>
  56 #include <linux/lockdep.h>
  57 #include <linux/pid_namespace.h>
  58 #include <linux/device.h>
  59 #include <linux/kthread.h>
  60 #include <linux/sched.h>
  61 #include <linux/signal.h>
  62 #include <linux/idr.h>
  63
  64 #include <asm/io.h>
  65 #include <asm/bugs.h>
  66 #include <asm/setup.h>
  67 #include <asm/sections.h>
  68 #include <asm/cacheflush.h>
  69
  70 #ifdef CONFIG_X86_LOCAL_APIC
  71 #include <asm/smp.h>
  72 #endif
  73
  74 /*
  75  * This is one of the first .c files built. Error out early if we have compiler
  76  * trouble.
  77  */
  78
  79 #if __GNUC__ == 4 && __GNUC_MINOR__ == 1 && __GNUC_PATCHLEVEL__ == 0
  80 #warning gcc-4.1.0 is known to miscompile the kernel.  A different compiler version is recommended.
  81 #endif
  82
  83 static int kernel_init(void *);
  84
  85 extern void init_IRQ(void);
  86 extern void fork_init(unsigned long);
  87 extern void mca_init(void);
  88 extern void sbus_init(void);
  89 extern void prio_tree_init(void);
  90 extern void radix_tree_init(void);
  91 extern void free_initmem(void);
  92 #ifdef  CONFIG_ACPI
  93 extern void acpi_early_init(void);
  94 #else
  95 static inline void acpi_early_init(void) { }
  96 #endif
  97 #ifndef CONFIG_DEBUG_RODATA
  98 static inline void mark_rodata_ro(void) { }
  99 #endif
 100
 101 #ifdef CONFIG_TC
 102 extern void tc_init(void);
 103 #endif
 104
 105 enum system_states system_state;
 106 EXPORT_SYMBOL(system_state);
 107
 108 /*
 109  * Boot command-line arguments
 110  */
 111 #define MAX_INIT_ARGS CONFIG_INIT_ENV_ARG_LIMIT
 112 #define MAX_INIT_ENVS CONFIG_INIT_ENV_ARG_LIMIT
 113
 114 extern void time_init(void);
 115 /* Default late time init is NULL. archs can override this later. */
 116 void (*late_time_init)(void);
 117 extern void softirq_init(void);
 118
 119 /* Untouched command line saved by arch-specific code. */
 120 char __initdata boot_command_line[COMMAND_LINE_SIZE];
 121 /* Untouched saved command line (eg. for /proc) */
 122 char *saved_command_line;
 123 /* Command line for parameter parsing */
 124 static char *static_command_line;
 125
 126 static char *execute_command;
 127 static char *ramdisk_execute_command;
 128
 129 #ifdef CONFIG_SMP
 130 /* Setup configured maximum number of CPUs to activate */
 131 unsigned int __initdata setup_max_cpus = NR_CPUS;
 132
 133 /*
 134  * Setup routine for controlling SMP activation
 135  *
 136  * Command-line option of "nosmp" or "maxcpus=0" will disable SMP
 137  * activation entirely (the MPS table probe still happens, though).
 138  *
 139  * Command-line option of "maxcpus=<NUM>", where <NUM> is an integer
 140  * greater than 0, limits the maximum number of CPUs activated in
 141  * SMP mode to <NUM>.
 142  */
 143 #ifndef CONFIG_X86_IO_APIC
 144 static inline void disable_ioapic_setup(void) {};
 145 #endif
 146
 147 static int __init nosmp(char *str)
 148 {
 149         setup_max_cpus = 0;
 150         disable_ioapic_setup();
 151         return 0;
 152 }
 153
 154 early_param("nosmp", nosmp);
 155
 156 static int __init maxcpus(char *str)
 157 {
 158         get_option(&str, &setup_max_cpus);
 159         if (setup_max_cpus == 0)
 160                 disable_ioapic_setup();
 161
 162         return 0;
 163 }
 164
 165 early_param("maxcpus", maxcpus);
 166 #else
 167 #define setup_max_cpus NR_CPUS
 168 #endif
 169
 170 /*
 171  * If set, this is an indication to the drivers that reset the underlying
 172  * device before going ahead with the initialization otherwise driver might
 173  * rely on the BIOS and skip the reset operation.
 174  *
 175  * This is useful if kernel is booting in an unreliable environment.
 176  * For ex. kdump situaiton where previous kernel has crashed, BIOS has been
 177  * skipped and devices will be in unknown state.
 178  */
 179 unsigned int reset_devices;
 180 EXPORT_SYMBOL(reset_devices);
 181
 182 static int __init set_reset_devices(char *str)
 183 {
 184         reset_devices = 1;
 185         return 1;
 186 }
 187
 188 __setup("reset_devices", set_reset_devices);
 189
 190 static char * argv_init[MAX_INIT_ARGS+2] = { "init", NULL, };
 191 char * envp_init[MAX_INIT_ENVS+2] = { "HOME=/", "TERM=linux", NULL, };
 192 static const char *panic_later, *panic_param;
 193
 194 extern struct obs_kernel_param __setup_start[], __setup_end[];
 195
 196 static int __init obsolete_checksetup(char *line)
 197 {
 198         struct obs_kernel_param *p;
 199         int had_early_param = 0;
 200
 201         p = __setup_start;
 202         do {
 203                 int n = strlen(p->str);
 204                 if (!strncmp(line, p->str, n)) {
 205                         if (p->early) {
 206                                 /* Already done in parse_early_param?
 207                                  * (Needs exact match on param part).
 208                                  * Keep iterating, as we can have early
 209                                  * params and __setups of same names 8( */
 210                                 if (line[n] == '\0' || line[n] == '=')
 211                                         had_early_param = 1;
 212                         } else if (!p->setup_func) {
 213                                 printk(KERN_WARNING "Parameter %s is obsolete,"
 214                                        " ignored\n", p->str);
 215                                 return 1;
 216                         } else if (p->setup_func(line + n))
 217                                 return 1;
 218                 }
 219                 p++;
 220         } while (p < __setup_end);
 221
 222         return had_early_param;
 223 }
 224
 225 /*
 226  * This should be approx 2 Bo*oMips to start (note initial shift), and will
 227  * still work even if initially too large, it will just take slightly longer
 228  */
 229 unsigned long loops_per_jiffy = (1<<12);
 230
 231 EXPORT_SYMBOL(loops_per_jiffy);
 232
 233 static int __init debug_kernel(char *str)
 234 {
 235         console_loglevel = 10;
 236         return 0;
 237 }
 238
 239 static int __init quiet_kernel(char *str)
 240 {
 241         console_loglevel = 4;
 242         return 0;
 243 }
 244
 245 early_param("debug", debug_kernel);
 246 early_param("quiet", quiet_kernel);
 247
 248 static int __init loglevel(char *str)
 249 {
 250         get_option(&str, &console_loglevel);
 251         return 0;
 252 }
 253
 254 early_param("loglevel", loglevel);
 255
 256 /*
 257  * Unknown boot options get handed to init, unless they look like
 258  * failed parameters
 259  */
 260 static int __init unknown_bootoption(char *param, char *val)
 261 {
 262         /* Change NUL term back to "=", to make "param" the whole string. */
 263         if (val) {
 264                 /* param=val or param="val"? */
 265                 if (val == param+strlen(param)+1)
 266                         val[-1] = '=';
 267                 else if (val == param+strlen(param)+2) {
 268                         val[-2] = '=';
 269                         memmove(val-1, val, strlen(val)+1);
 270                         val--;
 271                 } else
 272                         BUG();
 273         }
 274
 275         /* Handle obsolete-style parameters */
 276         if (obsolete_checksetup(param))
 277                 return 0;
 278
 279         /*
 280          * Preemptive maintenance for "why didn't my misspelled command
 281          * line work?"
 282          */
 283         if (strchr(param, '.') && (!val || strchr(param, '.') < val)) {
 284                 printk(KERN_ERR "Unknown boot option `%s': ignoring\n", param);
 285                 return 0;
 286         }
 287
 288         if (panic_later)
 289                 return 0;
 290
 291         if (val) {
 292                 /* Environment option */
 293                 unsigned int i;
 294                 for (i = 0; envp_init[i]; i++) {
 295                         if (i == MAX_INIT_ENVS) {
 296                                 panic_later = "Too many boot env vars at `%s'";
 297                                 panic_param = param;
 298                         }
 299                         if (!strncmp(param, envp_init[i], val - param))
 300                                 break;
 301                 }
 302                 envp_init[i] = param;
 303         } else {
 304                 /* Command line option */
 305                 unsigned int i;
 306                 for (i = 0; argv_init[i]; i++) {
 307                         if (i == MAX_INIT_ARGS) {
 308                                 panic_later = "Too many boot init vars at `%s'";
 309                                 panic_param = param;
 310                         }
 311                 }
 312                 argv_init[i] = param;
 313         }
 314         return 0;
 315 }
 316
 317 #ifdef CONFIG_DEBUG_PAGEALLOC
 318 int __read_mostly debug_pagealloc_enabled = 0;
 319 #endif
 320
 321 static int __init init_setup(char *str)
 322 {
 323         unsigned int i;
 324
 325         execute_command = str;
 326         /*
 327          * In case LILO is going to boot us with default command line,
 328          * it prepends "auto" before the whole cmdline which makes
 329          * the shell think it should execute a script with such name.
 330          * So we ignore all arguments entered _before_ init=... [MJ]
 331          */
 332         for (i = 1; i < MAX_INIT_ARGS; i++)
 333                 argv_init[i] = NULL;
 334         return 1;
 335 }
 336 __setup("init=", init_setup);
 337
 338 static int __init rdinit_setup(char *str)
 339 {
 340         unsigned int i;
 341
 342         ramdisk_execute_command = str;
 343         /* See "auto" comment in init_setup */
 344         for (i = 1; i < MAX_INIT_ARGS; i++)
 345                 argv_init[i] = NULL;
 346         return 1;
 347 }
 348 __setup("rdinit=", rdinit_setup);
 349
 350 #ifndef CONFIG_SMP
 351
 352 #ifdef CONFIG_X86_LOCAL_APIC
 353 static void __init smp_init(void)
 354 {
 355         APIC_init_uniprocessor();
 356 }
 357 #else
 358 #define smp_init()      do { } while (0)
 359 #endif
 360
 361 static inline void setup_per_cpu_areas(void) { }
 362 static inline void setup_nr_cpu_ids(void) { }
 363 static inline void smp_prepare_cpus(unsigned int maxcpus) { }
 364
 365 #else
 366
 367 #if NR_CPUS > BITS_PER_LONG
 368 cpumask_t cpu_mask_all __read_mostly = CPU_MASK_ALL;
 369 EXPORT_SYMBOL(cpu_mask_all);
 370 #endif
 371
 372 /* Setup number of possible processor ids */
 373 int nr_cpu_ids __read_mostly = NR_CPUS;
 374 EXPORT_SYMBOL(nr_cpu_ids);
 375
 376 /* An arch may set nr_cpu_ids earlier if needed, so this would be redundant */
 377 static void __init setup_nr_cpu_ids(void)
 378 {
 379         int cpu, highest_cpu = 0;
 380
 381         for_each_possible_cpu(cpu)
 382                 highest_cpu = cpu;
 383
 384         nr_cpu_ids = highest_cpu + 1;
 385 }
 386
 387 #ifndef CONFIG_HAVE_SETUP_PER_CPU_AREA
 388 unsigned long __per_cpu_offset[NR_CPUS] __read_mostly;
 389
 390 EXPORT_SYMBOL(__per_cpu_offset);
 391
 392 static void __init setup_per_cpu_areas(void)
 393 {
 394         unsigned long size, i, old_size;
 395         char *ptr;
 396         unsigned long nr_possible_cpus = num_possible_cpus();
 397         unsigned long align = 1;
 398         unsigned da_size;
 399
 400         /* Copy section for each CPU (we discard the original) */
 401         old_size = PERCPU_ENOUGH_ROOM;
 402         da_size = per_cpu_dyn_array_size(&align);
 403         align = max_t(unsigned long, PAGE_SIZE, align);
 404         size = ALIGN(old_size + da_size, align);
 405         ptr = alloc_bootmem_pages(size * nr_possible_cpus);
 406
 407         for_each_possible_cpu(i) {
 408                 __per_cpu_offset[i] = ptr - __per_cpu_start;
 409                 memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start);
 410                 per_cpu_alloc_dyn_array(i, ptr + old_size);
 411                 ptr += size;
 412         }
 413 }
 414 #endif /* CONFIG_HAVE_SETUP_PER_CPU_AREA */
 415
 416 /* Called by boot processor to activate the rest. */
 417 static void __init smp_init(void)
 418 {
 419         unsigned int cpu;
 420
 421         /*
 422          * Set up the current CPU as possible to migrate to.
 423          * The other ones will be done by cpu_up/cpu_down()
 424          */
 425         cpu = smp_processor_id();
 426         cpu_set(cpu, cpu_active_map);
 427
 428         /* FIXME: This should be done in userspace --RR */
 429         for_each_present_cpu(cpu) {
 430                 if (num_online_cpus() >= setup_max_cpus)
 431                         break;
 432                 if (!cpu_online(cpu))
 433                         cpu_up(cpu);
 434         }
 435
 436         /* Any cleanup work */
 437         printk(KERN_INFO "Brought up %ld CPUs\n", (long)num_online_cpus());
 438         smp_cpus_done(setup_max_cpus);
 439 }
 440
 441 #endif
 442
 443 /*
 444  * We need to store the untouched command line for future reference.
 445  * We also need to store the touched command line since the parameter
 446  * parsing is performed in place, and we should allow a component to
 447  * store reference of name/value for future reference.
 448  */
 449 static void __init setup_command_line(char *command_line)
 450 {
 451         saved_command_line = alloc_bootmem(strlen (boot_command_line)+1);
 452         static_command_line = alloc_bootmem(strlen (command_line)+1);
 453         strcpy (saved_command_line, boot_command_line);
 454         strcpy (static_command_line, command_line);
 455 }
 456
 457 /*
 458  * We need to finalize in a non-__init function or else race conditions
 459  * between the root thread and the init thread may cause start_kernel to
 460  * be reaped by free_initmem before the root thread has proceeded to
 461  * cpu_idle.
 462  *
 463  * gcc-3.4 accidentally inlines this function, so use noinline.
 464  */
 465
 466 static void noinline __init_refok rest_init(void)
 467         __releases(kernel_lock)
 468 {
 469         int pid;
 470
 471         kernel_thread(kernel_init, NULL, CLONE_FS | CLONE_SIGHAND);
 472         numa_default_policy();
 473         pid = kernel_thread(kthreadd, NULL, CLONE_FS | CLONE_FILES);
 474         kthreadd_task = find_task_by_pid_ns(pid, &init_pid_ns);
 475         unlock_kernel();
 476
 477         /*
 478          * The boot idle thread must execute schedule()
 479          * at least once to get things moving:
 480          */
 481         init_idle_bootup_task(current);
 482         preempt_enable_no_resched();
 483         schedule();
 484         preempt_disable();
 485
 486         /* Call into cpu_idle with preempt disabled */
 487         cpu_idle();
 488 }
 489
 490 /* Check for early params. */
 491 static int __init do_early_param(char *param, char *val)
 492 {
 493         struct obs_kernel_param *p;
 494
 495         for (p = __setup_start; p < __setup_end; p++) {
 496                 if ((p->early && strcmp(param, p->str) == 0) ||
 497                     (strcmp(param, "console") == 0 &&
 498                      strcmp(p->str, "earlycon") == 0)
 499                 ) {
 500                         if (p->setup_func(val) != 0)
 501                                 printk(KERN_WARNING
 502                                        "Malformed early option '%s'\n", param);
 503                 }
 504         }
 505         /* We accept everything at this stage. */
 506         return 0;
 507 }
 508
 509 /* Arch code calls this early on, or if not, just before other parsing. */
 510 void __init parse_early_param(void)
 511 {
 512         static __initdata int done = 0;
 513         static __initdata char tmp_cmdline[COMMAND_LINE_SIZE];
 514
 515         if (done)
 516                 return;
 517
 518         /* All fall through to do_early_param. */
 519         strlcpy(tmp_cmdline, boot_command_line, COMMAND_LINE_SIZE);
 520         parse_args("early options", tmp_cmdline, NULL, 0, do_early_param);
 521         done = 1;
 522 }
 523
 524 /*
 525  *      Activate the first processor.
 526  */
 527
 528 static void __init boot_cpu_init(void)
 529 {
 530         int cpu = smp_processor_id();
 531         /* Mark the boot cpu "present", "online" etc for SMP and UP case */
 532         cpu_set(cpu, cpu_online_map);
 533         cpu_set(cpu, cpu_present_map);
 534         cpu_set(cpu, cpu_possible_map);
 535 }
 536
 537 void __init __weak smp_setup_processor_id(void)
 538 {
 539 }
 540
 541 void __init __weak thread_info_cache_init(void)
 542 {
 543 }
 544
 545 void pre_alloc_dyn_array(void)
 546 {
 547 #ifdef CONFIG_HAVE_DYN_ARRAY
 548         unsigned long total_size = 0, size, phys;
 549         unsigned long max_align = 1;
 550         struct dyn_array **daa;
 551         char *ptr;
 552
 553         /* get the total size at first */
 554         for (daa = __dyn_array_start ; daa < __dyn_array_end; daa++) {
 555                 struct dyn_array *da = *daa;
 556
 557                 size = da->size * (*da->nr);
 558                 print_fn_descriptor_symbol("dyn_array %s ", da->name);
 559                 printk(KERN_CONT "size:%#lx nr:%d align:%#lx\n",
 560                         da->size, *da->nr, da->align);
 561                 total_size += roundup(size, da->align);
 562                 if (da->align > max_align)
 563                         max_align = da->align;
 564         }
 565         if (total_size)
 566                 printk(KERN_DEBUG "dyn_array total_size: %#lx\n",
 567                          total_size);
 568         else
 569                 return;
 570
 571         /* allocate them all together */
 572         max_align = max_t(unsigned long, max_align, PAGE_SIZE);
 573         ptr = __alloc_bootmem_nopanic(total_size, max_align, 0);
 574         if (!ptr)
 575                 panic("Can not alloc dyn_alloc\n");
 576
 577         phys = virt_to_phys(ptr);
 578         for (daa = __dyn_array_start ; daa < __dyn_array_end; daa++) {
 579                 struct dyn_array *da = *daa;
 580
 581                 size = da->size * (*da->nr);
 582                 print_fn_descriptor_symbol("dyn_array %s ", da->name);
 583
 584                 phys = roundup(phys, da->align);
 585                 *da->name = phys_to_virt(phys);
 586                 printk(KERN_CONT " ==> [%#lx - %#lx]\n", phys, phys + size);
 587
 588                 phys += size;
 589
 590                 if (da->init_work)
 591                         da->init_work(da);
 592         }
 593 #endif
 594 }
 595
 596 unsigned long per_cpu_dyn_array_size(unsigned long *align)
 597 {
 598         unsigned long total_size = 0;
 599 #ifdef CONFIG_HAVE_DYN_ARRAY
 600         unsigned long size;
 601         struct dyn_array **daa;
 602         unsigned max_align = 1;
 603
 604         for (daa = __per_cpu_dyn_array_start ; daa < __per_cpu_dyn_array_end; daa++) {
 605                 struct dyn_array *da = *daa;
 606
 607                 size = da->size * (*da->nr);
 608                 print_fn_descriptor_symbol("per_cpu_dyn_array %s ", da->name);
 609                 printk(KERN_CONT "size:%#lx nr:%d align:%#lx\n",
 610                         da->size, *da->nr, da->align);
 611                 total_size += roundup(size, da->align);
 612                 if (da->align > max_align)
 613                         max_align = da->align;
 614         }
 615         if (total_size) {
 616                 printk(KERN_DEBUG "per_cpu_dyn_array total_size: %#lx\n",
 617                          total_size);
 618                 *align = max_align;
 619         }
 620 #endif
 621         return total_size;
 622 }
 623
 624 void per_cpu_alloc_dyn_array(int cpu, char *ptr)
 625 {
 626 #ifdef CONFIG_HAVE_DYN_ARRAY
 627         unsigned long size, phys;
 628         struct dyn_array **daa;
 629         unsigned long addr;
 630         void **array;
 631
 632         phys = virt_to_phys(ptr);
 633         for (daa = __per_cpu_dyn_array_start ; daa < __per_cpu_dyn_array_end; daa++) {
 634                 struct dyn_array *da = *daa;
 635
 636                 size = da->size * (*da->nr);
 637                 print_fn_descriptor_symbol("per_cpu_dyn_array %s ", da->name);
 638
 639                 phys = roundup(phys, da->align);
 640                 addr = (unsigned long)da->name;
 641                 addr += per_cpu_offset(cpu);
 642                 array = (void **)addr;
 643                 *array = phys_to_virt(phys);
 644                 *da->name = *array; /* so init_work could use it directly */
 645                 printk(KERN_CONT " ==> [%#lx - %#lx]\n", phys, phys + size);
 646
 647                 phys += size;
 648
 649                 if (da->init_work) {
 650                         da->init_work(da);
 651                 }
 652         }
 653 #endif
 654 }
 655
 656 asmlinkage void __init start_kernel(void)
 657 {
 658         char * command_line;
 659         extern struct kernel_param __start___param[], __stop___param[];
 660
 661         smp_setup_processor_id();
 662
 663         /*
 664          * Need to run as early as possible, to initialize the
 665          * lockdep hash:
 666          */
 667         unwind_init();
 668         lockdep_init();
 669         debug_objects_early_init();
 670         cgroup_init_early();
 671
 672         local_irq_disable();
 673         early_boot_irqs_off();
 674         early_init_irq_lock_class();
 675
 676 /*
 677  * Interrupts are still disabled. Do necessary setups, then
 678  * enable them
 679  */
 680         lock_kernel();
 681         tick_init();
 682         boot_cpu_init();
 683         page_address_init();
 684         printk(KERN_NOTICE);
 685         printk(linux_banner);
 686         setup_arch(&command_line);
 687         pre_alloc_dyn_array();
 688         mm_init_owner(&init_mm, &init_task);
 689         setup_command_line(command_line);
 690         unwind_setup();
 691         setup_per_cpu_areas();
 692         setup_nr_cpu_ids();
 693         smp_prepare_boot_cpu(); /* arch-specific boot-cpu hooks */
 694
 695         /*
 696          * Set up the scheduler prior starting any interrupts (such as the
 697          * timer interrupt). Full topology setup happens at smp_init()
 698          * time - but meanwhile we still have a functioning scheduler.
 699          */
 700         sched_init();
 701         /*
 702          * Disable preemption - early bootup scheduling is extremely
 703          * fragile until we cpu_idle() for the first time.
 704          */
 705         preempt_disable();
 706         build_all_zonelists();
 707         page_alloc_init();
 708         printk(KERN_NOTICE "Kernel command line: %s\n", boot_command_line);
 709         parse_early_param();
 710         parse_args("Booting kernel", static_command_line, __start___param,
 711                    __stop___param - __start___param,
 712                    &unknown_bootoption);
 713         if (!irqs_disabled()) {
 714                 printk(KERN_WARNING "start_kernel(): bug: interrupts were "
 715                                 "enabled *very* early, fixing it\n");
 716                 local_irq_disable();
 717         }
 718         sort_main_extable();
 719         trap_init();
 720         rcu_init();
 721         init_IRQ();
 722         pidhash_init();
 723         init_timers();
 724         hrtimers_init();
 725         softirq_init();
 726         timekeeping_init();
 727         time_init();
 728         sched_clock_init();
 729         profile_init();
 730         if (!irqs_disabled())
 731                 printk("start_kernel(): bug: interrupts were enabled early\n");
 732         early_boot_irqs_on();
 733         local_irq_enable();
 734
 735         /*
 736          * HACK ALERT! This is early. We're enabling the console before
 737          * we've done PCI setups etc, and console_init() must be aware of
 738          * this. But we do want output early, in case something goes wrong.
 739          */
 740         console_init();
 741         if (panic_later)
 742                 panic(panic_later, panic_param);
 743
 744         lockdep_info();
 745
 746         /*
 747          * Need to run this when irqs are enabled, because it wants
 748          * to self-test [hard/soft]-irqs on/off lock inversion bugs
 749          * too:
 750          */
 751         locking_selftest();
 752
 753 #ifdef CONFIG_BLK_DEV_INITRD
 754         if (initrd_start && !initrd_below_start_ok &&
 755             page_to_pfn(virt_to_page((void *)initrd_start)) < min_low_pfn) {
 756                 printk(KERN_CRIT "initrd overwritten (0x%08lx < 0x%08lx) - "
 757                     "disabling it.\n",
 758                     page_to_pfn(virt_to_page((void *)initrd_start)),
 759                     min_low_pfn);
 760                 initrd_start = 0;
 761         }
 762 #endif
 763         vfs_caches_init_early();
 764         cpuset_init_early();
 765         mem_init();
 766         enable_debug_pagealloc();
 767         cpu_hotplug_init();
 768         kmem_cache_init();
 769         debug_objects_mem_init();
 770         idr_init_cache();
 771         setup_per_cpu_pageset();
 772         numa_policy_init();
 773         if (late_time_init)
 774                 late_time_init();
 775         calibrate_delay();
 776         pidmap_init();
 777         pgtable_cache_init();
 778         prio_tree_init();
 779         anon_vma_init();
 780 #ifdef CONFIG_X86
 781         if (efi_enabled)
 782                 efi_enter_virtual_mode();
 783 #endif
 784         thread_info_cache_init();
 785         fork_init(num_physpages);
 786         proc_caches_init();
 787         buffer_init();
 788         unnamed_dev_init();
 789         key_init();
 790         security_init();
 791         vfs_caches_init(num_physpages);
 792         radix_tree_init();
 793         signals_init();
 794         /* rootfs populating might need page-writeback */
 795         page_writeback_init();
 796 #ifdef CONFIG_PROC_FS
 797         proc_root_init();
 798 #endif
 799         cgroup_init();
 800         cpuset_init();
 801         taskstats_init_early();
 802         delayacct_init();
 803
 804         check_bugs();
 805
 806         acpi_early_init(); /* before LAPIC and SMP init */
 807
 808         /* Do the rest non-__init'ed, we're now alive */
 809         rest_init();
 810 }
 811
 812 static int initcall_debug;
 813
 814 static int __init initcall_debug_setup(char *str)
 815 {
 816         initcall_debug = 1;
 817         return 1;
 818 }
 819 __setup("initcall_debug", initcall_debug_setup);
 820
 821 int do_one_initcall(initcall_t fn)
 822 {
 823         int count = preempt_count();
 824         ktime_t t0, t1, delta;
 825         char msgbuf[64];
 826         int result;
 827
 828         if (initcall_debug) {
 829                 printk("calling  %pF @ %i\n", fn, task_pid_nr(current));
 830                 t0 = ktime_get();
 831         }
 832
 833         result = fn();
 834
 835         if (initcall_debug) {
 836                 t1 = ktime_get();
 837                 delta = ktime_sub(t1, t0);
 838
 839                 printk("initcall %pF returned %d after %Ld msecs\n",
 840                         fn, result,
 841                         (unsigned long long) delta.tv64 >> 20);
 842         }
 843
 844         msgbuf[0] = 0;
 845
 846         if (result && result != -ENODEV && initcall_debug)
 847                 sprintf(msgbuf, "error code %d ", result);
 848
 849         if (preempt_count() != count) {
 850                 strlcat(msgbuf, "preemption imbalance ", sizeof(msgbuf));
 851                 preempt_count() = count;
 852         }
 853         if (irqs_disabled()) {
 854                 strlcat(msgbuf, "disabled interrupts ", sizeof(msgbuf));
 855                 local_irq_enable();
 856         }
 857         if (msgbuf[0]) {
 858                 printk("initcall %pF returned with %s\n", fn, msgbuf);
 859         }
 860
 861         return result;
 862 }
 863
 864
 865 extern initcall_t __initcall_start[], __initcall_end[], __early_initcall_end[];
 866
 867 static void __init do_initcalls(void)
 868 {
 869         initcall_t *call;
 870
 871         for (call = __early_initcall_end; call < __initcall_end; call++)
 872                 do_one_initcall(*call);
 873
 874         /* Make sure there is no pending stuff from the initcall sequence */
 875         flush_scheduled_work();
 876 }
 877
 878 /*
 879  * Ok, the machine is now initialized. None of the devices
 880  * have been touched yet, but the CPU subsystem is up and
 881  * running, and memory and process management works.
 882  *
 883  * Now we can finally start doing some real work..
 884  */
 885 static void __init do_basic_setup(void)
 886 {
 887         rcu_init_sched(); /* needed by module_init stage. */
 888         /* drivers will send hotplug events */
 889         init_workqueues();
 890         usermodehelper_init();
 891         driver_init();
 892         init_irq_proc();
 893         do_initcalls();
 894 }
 895
 896 static void __init do_pre_smp_initcalls(void)
 897 {
 898         initcall_t *call;
 899
 900         for (call = __initcall_start; call < __early_initcall_end; call++)
 901                 do_one_initcall(*call);
 902 }
 903
 904 static void run_init_process(char *init_filename)
 905 {
 906         argv_init[0] = init_filename;
 907         kernel_execve(init_filename, argv_init, envp_init);
 908 }
 909
 910 /* This is a non __init function. Force it to be noinline otherwise gcc
 911  * makes it inline to init() and it becomes part of init.text section
 912  */
 913 static int noinline init_post(void)
 914 {
 915         free_initmem();
 916         unlock_kernel();
 917         mark_rodata_ro();
 918         system_state = SYSTEM_RUNNING;
 919         numa_default_policy();
 920
 921         if (sys_open((const char __user *) "/dev/console", O_RDWR, 0) < 0)
 922                 printk(KERN_WARNING "Warning: unable to open an initial console.\n");
 923
 924         (void) sys_dup(0);
 925         (void) sys_dup(0);
 926
 927         current->signal->flags |= SIGNAL_UNKILLABLE;
 928
 929         if (ramdisk_execute_command) {
 930                 run_init_process(ramdisk_execute_command);
 931                 printk(KERN_WARNING "Failed to execute %s\n",
 932                                 ramdisk_execute_command);
 933         }
 934
 935         /*
 936          * We try each of these until one succeeds.
 937          *
 938          * The Bourne shell can be used instead of init if we are
 939          * trying to recover a really broken machine.
 940          */
 941         if (execute_command) {
 942                 run_init_process(execute_command);
 943                 printk(KERN_WARNING "Failed to execute %s.  Attempting "
 944                                         "defaults...\n", execute_command);
 945         }
 946         run_init_process("/sbin/init");
 947         run_init_process("/etc/init");
 948         run_init_process("/bin/init");
 949         run_init_process("/bin/sh");
 950
 951         panic("No init found.  Try passing init= option to kernel.");
 952 }
 953
 954 static int __init kernel_init(void * unused)
 955 {
 956         lock_kernel();
 957         /*
 958          * init can run on any cpu.
 959          */
 960         set_cpus_allowed_ptr(current, CPU_MASK_ALL_PTR);
 961         /*
 962          * Tell the world that we're going to be the grim
 963          * reaper of innocent orphaned children.
 964          *
 965          * We don't want people to have to make incorrect
 966          * assumptions about where in the task array this
 967          * can be found.
 968          */
 969         init_pid_ns.child_reaper = current;
 970
 971         cad_pid = task_pid(current);
 972
 973         smp_prepare_cpus(setup_max_cpus);
 974
 975         do_pre_smp_initcalls();
 976
 977         smp_init();
 978         sched_init_smp();
 979
 980         cpuset_init_smp();
 981
 982         do_basic_setup();
 983
 984         /*
 985          * check if there is an early userspace init.  If yes, let it do all
 986          * the work
 987          */
 988
 989         if (!ramdisk_execute_command)
 990                 ramdisk_execute_command = "/init";
 991
 992         if (sys_access((const char __user *) ramdisk_execute_command, 0) != 0) {
 993                 ramdisk_execute_command = NULL;
 994                 prepare_namespace();
 995         }
 996
 997         /*
 998          * Ok, we have completed the initial bootup, and
 999          * we're essentially up and running. Get rid of the
1000          * initmem segments and start the user-mode stuff..
1001          */
1002         init_post();
1003         return 0;
1004 }