Root/Documentation/x86/boot.txt

1             THE LINUX/x86 BOOT PROTOCOL
2             ---------------------------
3
4On the x86 platform, the Linux kernel uses a rather complicated boot
5convention. This has evolved partially due to historical aspects, as
6well as the desire in the early days to have the kernel itself be a
7bootable image, the complicated PC memory model and due to changed
8expectations in the PC industry caused by the effective demise of
9real-mode DOS as a mainstream operating system.
10
11Currently, the following versions of the Linux/x86 boot protocol exist.
12
13Old kernels: zImage/Image support only. Some very early kernels
14        may not even support a command line.
15
16Protocol 2.00: (Kernel 1.3.73) Added bzImage and initrd support, as
17        well as a formalized way to communicate between the
18        boot loader and the kernel. setup.S made relocatable,
19        although the traditional setup area still assumed
20        writable.
21
22Protocol 2.01: (Kernel 1.3.76) Added a heap overrun warning.
23
24Protocol 2.02: (Kernel 2.4.0-test3-pre3) New command line protocol.
25        Lower the conventional memory ceiling. No overwrite
26        of the traditional setup area, thus making booting
27        safe for systems which use the EBDA from SMM or 32-bit
28        BIOS entry points. zImage deprecated but still
29        supported.
30
31Protocol 2.03: (Kernel 2.4.18-pre1) Explicitly makes the highest possible
32        initrd address available to the bootloader.
33
34Protocol 2.04: (Kernel 2.6.14) Extend the syssize field to four bytes.
35
36Protocol 2.05: (Kernel 2.6.20) Make protected mode kernel relocatable.
37        Introduce relocatable_kernel and kernel_alignment fields.
38
39Protocol 2.06: (Kernel 2.6.22) Added a field that contains the size of
40        the boot command line.
41
42Protocol 2.07: (Kernel 2.6.24) Added paravirtualised boot protocol.
43        Introduced hardware_subarch and hardware_subarch_data
44        and KEEP_SEGMENTS flag in load_flags.
45
46Protocol 2.08: (Kernel 2.6.26) Added crc32 checksum and ELF format
47        payload. Introduced payload_offset and payload_length
48        fields to aid in locating the payload.
49
50Protocol 2.09: (Kernel 2.6.26) Added a field of 64-bit physical
51        pointer to single linked list of struct setup_data.
52
53Protocol 2.10: (Kernel 2.6.31) Added a protocol for relaxed alignment
54        beyond the kernel_alignment added, new init_size and
55        pref_address fields. Added extended boot loader IDs.
56
57**** MEMORY LAYOUT
58
59The traditional memory map for the kernel loader, used for Image or
60zImage kernels, typically looks like:
61
62    | |
630A0000 +------------------------+
64    | Reserved for BIOS | Do not use. Reserved for BIOS EBDA.
6509A000 +------------------------+
66    | Command line |
67    | Stack/heap | For use by the kernel real-mode code.
68098000 +------------------------+
69    | Kernel setup | The kernel real-mode code.
70090200 +------------------------+
71    | Kernel boot sector | The kernel legacy boot sector.
72090000 +------------------------+
73    | Protected-mode kernel | The bulk of the kernel image.
74010000 +------------------------+
75    | Boot loader | <- Boot sector entry point 0000:7C00
76001000 +------------------------+
77    | Reserved for MBR/BIOS |
78000800 +------------------------+
79    | Typically used by MBR |
80000600 +------------------------+
81    | BIOS use only |
82000000 +------------------------+
83
84
85When using bzImage, the protected-mode kernel was relocated to
860x100000 ("high memory"), and the kernel real-mode block (boot sector,
87setup, and stack/heap) was made relocatable to any address between
880x10000 and end of low memory. Unfortunately, in protocols 2.00 and
892.01 the 0x90000+ memory range is still used internally by the kernel;
90the 2.02 protocol resolves that problem.
91
92It is desirable to keep the "memory ceiling" -- the highest point in
93low memory touched by the boot loader -- as low as possible, since
94some newer BIOSes have begun to allocate some rather large amounts of
95memory, called the Extended BIOS Data Area, near the top of low
96memory. The boot loader should use the "INT 12h" BIOS call to verify
97how much low memory is available.
98
99Unfortunately, if INT 12h reports that the amount of memory is too
100low, there is usually nothing the boot loader can do but to report an
101error to the user. The boot loader should therefore be designed to
102take up as little space in low memory as it reasonably can. For
103zImage or old bzImage kernels, which need data written into the
1040x90000 segment, the boot loader should make sure not to use memory
105above the 0x9A000 point; too many BIOSes will break above that point.
106
107For a modern bzImage kernel with boot protocol version >= 2.02, a
108memory layout like the following is suggested:
109
110    ~ ~
111        | Protected-mode kernel |
112100000 +------------------------+
113    | I/O memory hole |
1140A0000 +------------------------+
115    | Reserved for BIOS | Leave as much as possible unused
116    ~ ~
117    | Command line | (Can also be below the X+10000 mark)
118X+10000 +------------------------+
119    | Stack/heap | For use by the kernel real-mode code.
120X+08000 +------------------------+
121    | Kernel setup | The kernel real-mode code.
122    | Kernel boot sector | The kernel legacy boot sector.
123X +------------------------+
124    | Boot loader | <- Boot sector entry point 0000:7C00
125001000 +------------------------+
126    | Reserved for MBR/BIOS |
127000800 +------------------------+
128    | Typically used by MBR |
129000600 +------------------------+
130    | BIOS use only |
131000000 +------------------------+
132
133... where the address X is as low as the design of the boot loader
134permits.
135
136
137**** THE REAL-MODE KERNEL HEADER
138
139In the following text, and anywhere in the kernel boot sequence, "a
140sector" refers to 512 bytes. It is independent of the actual sector
141size of the underlying medium.
142
143The first step in loading a Linux kernel should be to load the
144real-mode code (boot sector and setup code) and then examine the
145following header at offset 0x01f1. The real-mode code can total up to
14632K, although the boot loader may choose to load only the first two
147sectors (1K) and then examine the bootup sector size.
148
149The header looks like:
150
151Offset Proto Name Meaning
152/Size
153
15401F1/1 ALL(1 setup_sects The size of the setup in sectors
15501F2/2 ALL root_flags If set, the root is mounted readonly
15601F4/4 2.04+(2 syssize The size of the 32-bit code in 16-byte paras
15701F8/2 ALL ram_size DO NOT USE - for bootsect.S use only
15801FA/2 ALL vid_mode Video mode control
15901FC/2 ALL root_dev Default root device number
16001FE/2 ALL boot_flag 0xAA55 magic number
1610200/2 2.00+ jump Jump instruction
1620202/4 2.00+ header Magic signature "HdrS"
1630206/2 2.00+ version Boot protocol version supported
1640208/4 2.00+ realmode_swtch Boot loader hook (see below)
165020C/2 2.00+ start_sys_seg The load-low segment (0x1000) (obsolete)
166020E/2 2.00+ kernel_version Pointer to kernel version string
1670210/1 2.00+ type_of_loader Boot loader identifier
1680211/1 2.00+ loadflags Boot protocol option flags
1690212/2 2.00+ setup_move_size Move to high memory size (used with hooks)
1700214/4 2.00+ code32_start Boot loader hook (see below)
1710218/4 2.00+ ramdisk_image initrd load address (set by boot loader)
172021C/4 2.00+ ramdisk_size initrd size (set by boot loader)
1730220/4 2.00+ bootsect_kludge DO NOT USE - for bootsect.S use only
1740224/2 2.01+ heap_end_ptr Free memory after setup end
1750226/1 2.02+(3 ext_loader_ver Extended boot loader version
1760227/1 2.02+(3 ext_loader_type Extended boot loader ID
1770228/4 2.02+ cmd_line_ptr 32-bit pointer to the kernel command line
178022C/4 2.03+ ramdisk_max Highest legal initrd address
1790230/4 2.05+ kernel_alignment Physical addr alignment required for kernel
1800234/1 2.05+ relocatable_kernel Whether kernel is relocatable or not
1810235/1 2.10+ min_alignment Minimum alignment, as a power of two
1820236/2 N/A pad3 Unused
1830238/4 2.06+ cmdline_size Maximum size of the kernel command line
184023C/4 2.07+ hardware_subarch Hardware subarchitecture
1850240/8 2.07+ hardware_subarch_data Subarchitecture-specific data
1860248/4 2.08+ payload_offset Offset of kernel payload
187024C/4 2.08+ payload_length Length of kernel payload
1880250/8 2.09+ setup_data 64-bit physical pointer to linked list
189                of struct setup_data
1900258/8 2.10+ pref_address Preferred loading address
1910260/4 2.10+ init_size Linear memory required during initialization
192
193(1) For backwards compatibility, if the setup_sects field contains 0, the
194    real value is 4.
195
196(2) For boot protocol prior to 2.04, the upper two bytes of the syssize
197    field are unusable, which means the size of a bzImage kernel
198    cannot be determined.
199
200(3) Ignored, but safe to set, for boot protocols 2.02-2.09.
201
202If the "HdrS" (0x53726448) magic number is not found at offset 0x202,
203the boot protocol version is "old". Loading an old kernel, the
204following parameters should be assumed:
205
206    Image type = zImage
207    initrd not supported
208    Real-mode kernel must be located at 0x90000.
209
210Otherwise, the "version" field contains the protocol version,
211e.g. protocol version 2.01 will contain 0x0201 in this field. When
212setting fields in the header, you must make sure only to set fields
213supported by the protocol version in use.
214
215
216**** DETAILS OF HEADER FIELDS
217
218For each field, some are information from the kernel to the bootloader
219("read"), some are expected to be filled out by the bootloader
220("write"), and some are expected to be read and modified by the
221bootloader ("modify").
222
223All general purpose boot loaders should write the fields marked
224(obligatory). Boot loaders who want to load the kernel at a
225nonstandard address should fill in the fields marked (reloc); other
226boot loaders can ignore those fields.
227
228The byte order of all fields is littleendian (this is x86, after all.)
229
230Field name: setup_sects
231Type: read
232Offset/size: 0x1f1/1
233Protocol: ALL
234
235  The size of the setup code in 512-byte sectors. If this field is
236  0, the real value is 4. The real-mode code consists of the boot
237  sector (always one 512-byte sector) plus the setup code.
238
239Field name: root_flags
240Type: modify (optional)
241Offset/size: 0x1f2/2
242Protocol: ALL
243
244  If this field is nonzero, the root defaults to readonly. The use of
245  this field is deprecated; use the "ro" or "rw" options on the
246  command line instead.
247
248Field name: syssize
249Type: read
250Offset/size: 0x1f4/4 (protocol 2.04+) 0x1f4/2 (protocol ALL)
251Protocol: 2.04+
252
253  The size of the protected-mode code in units of 16-byte paragraphs.
254  For protocol versions older than 2.04 this field is only two bytes
255  wide, and therefore cannot be trusted for the size of a kernel if
256  the LOAD_HIGH flag is set.
257
258Field name: ram_size
259Type: kernel internal
260Offset/size: 0x1f8/2
261Protocol: ALL
262
263  This field is obsolete.
264
265Field name: vid_mode
266Type: modify (obligatory)
267Offset/size: 0x1fa/2
268
269  Please see the section on SPECIAL COMMAND LINE OPTIONS.
270
271Field name: root_dev
272Type: modify (optional)
273Offset/size: 0x1fc/2
274Protocol: ALL
275
276  The default root device device number. The use of this field is
277  deprecated, use the "root=" option on the command line instead.
278
279Field name: boot_flag
280Type: read
281Offset/size: 0x1fe/2
282Protocol: ALL
283
284  Contains 0xAA55. This is the closest thing old Linux kernels have
285  to a magic number.
286
287Field name: jump
288Type: read
289Offset/size: 0x200/2
290Protocol: 2.00+
291
292  Contains an x86 jump instruction, 0xEB followed by a signed offset
293  relative to byte 0x202. This can be used to determine the size of
294  the header.
295
296Field name: header
297Type: read
298Offset/size: 0x202/4
299Protocol: 2.00+
300
301  Contains the magic number "HdrS" (0x53726448).
302
303Field name: version
304Type: read
305Offset/size: 0x206/2
306Protocol: 2.00+
307
308  Contains the boot protocol version, in (major << 8)+minor format,
309  e.g. 0x0204 for version 2.04, and 0x0a11 for a hypothetical version
310  10.17.
311
312Field name: realmode_swtch
313Type: modify (optional)
314Offset/size: 0x208/4
315Protocol: 2.00+
316
317  Boot loader hook (see ADVANCED BOOT LOADER HOOKS below.)
318
319Field name: start_sys_seg
320Type: read
321Offset/size: 0x20c/2
322Protocol: 2.00+
323
324  The load low segment (0x1000). Obsolete.
325
326Field name: kernel_version
327Type: read
328Offset/size: 0x20e/2
329Protocol: 2.00+
330
331  If set to a nonzero value, contains a pointer to a NUL-terminated
332  human-readable kernel version number string, less 0x200. This can
333  be used to display the kernel version to the user. This value
334  should be less than (0x200*setup_sects).
335
336  For example, if this value is set to 0x1c00, the kernel version
337  number string can be found at offset 0x1e00 in the kernel file.
338  This is a valid value if and only if the "setup_sects" field
339  contains the value 15 or higher, as:
340
341    0x1c00 < 15*0x200 (= 0x1e00) but
342    0x1c00 >= 14*0x200 (= 0x1c00)
343
344    0x1c00 >> 9 = 14, so the minimum value for setup_secs is 15.
345
346Field name: type_of_loader
347Type: write (obligatory)
348Offset/size: 0x210/1
349Protocol: 2.00+
350
351  If your boot loader has an assigned id (see table below), enter
352  0xTV here, where T is an identifier for the boot loader and V is
353  a version number. Otherwise, enter 0xFF here.
354
355  For boot loader IDs above T = 0xD, write T = 0xE to this field and
356  write the extended ID minus 0x10 to the ext_loader_type field.
357  Similarly, the ext_loader_ver field can be used to provide more than
358  four bits for the bootloader version.
359
360  For example, for T = 0x15, V = 0x234, write:
361
362  type_of_loader <- 0xE4
363  ext_loader_type <- 0x05
364  ext_loader_ver <- 0x23
365
366  Assigned boot loader ids:
367    0 LILO (0x00 reserved for pre-2.00 bootloader)
368    1 Loadlin
369    2 bootsect-loader (0x20, all other values reserved)
370    3 Syslinux
371    4 Etherboot/gPXE
372    5 ELILO
373    7 GRUB
374    8 U-Boot
375    9 Xen
376    A Gujin
377    B Qemu
378    C Arcturus Networks uCbootloader
379    E Extended (see ext_loader_type)
380    F Special (0xFF = undefined)
381
382  Please contact <hpa@zytor.com> if you need a bootloader ID
383  value assigned.
384
385Field name: loadflags
386Type: modify (obligatory)
387Offset/size: 0x211/1
388Protocol: 2.00+
389
390  This field is a bitmask.
391
392  Bit 0 (read): LOADED_HIGH
393    - If 0, the protected-mode code is loaded at 0x10000.
394    - If 1, the protected-mode code is loaded at 0x100000.
395
396  Bit 5 (write): QUIET_FLAG
397    - If 0, print early messages.
398    - If 1, suppress early messages.
399        This requests to the kernel (decompressor and early
400        kernel) to not write early messages that require
401        accessing the display hardware directly.
402
403  Bit 6 (write): KEEP_SEGMENTS
404    Protocol: 2.07+
405    - If 0, reload the segment registers in the 32bit entry point.
406    - If 1, do not reload the segment registers in the 32bit entry point.
407        Assume that %cs %ds %ss %es are all set to flat segments with
408        a base of 0 (or the equivalent for their environment).
409
410  Bit 7 (write): CAN_USE_HEAP
411    Set this bit to 1 to indicate that the value entered in the
412    heap_end_ptr is valid. If this field is clear, some setup code
413    functionality will be disabled.
414
415Field name: setup_move_size
416Type: modify (obligatory)
417Offset/size: 0x212/2
418Protocol: 2.00-2.01
419
420  When using protocol 2.00 or 2.01, if the real mode kernel is not
421  loaded at 0x90000, it gets moved there later in the loading
422  sequence. Fill in this field if you want additional data (such as
423  the kernel command line) moved in addition to the real-mode kernel
424  itself.
425
426  The unit is bytes starting with the beginning of the boot sector.
427  
428  This field is can be ignored when the protocol is 2.02 or higher, or
429  if the real-mode code is loaded at 0x90000.
430
431Field name: code32_start
432Type: modify (optional, reloc)
433Offset/size: 0x214/4
434Protocol: 2.00+
435
436  The address to jump to in protected mode. This defaults to the load
437  address of the kernel, and can be used by the boot loader to
438  determine the proper load address.
439
440  This field can be modified for two purposes:
441
442  1. as a boot loader hook (see ADVANCED BOOT LOADER HOOKS below.)
443
444  2. if a bootloader which does not install a hook loads a
445     relocatable kernel at a nonstandard address it will have to modify
446     this field to point to the load address.
447
448Field name: ramdisk_image
449Type: write (obligatory)
450Offset/size: 0x218/4
451Protocol: 2.00+
452
453  The 32-bit linear address of the initial ramdisk or ramfs. Leave at
454  zero if there is no initial ramdisk/ramfs.
455
456Field name: ramdisk_size
457Type: write (obligatory)
458Offset/size: 0x21c/4
459Protocol: 2.00+
460
461  Size of the initial ramdisk or ramfs. Leave at zero if there is no
462  initial ramdisk/ramfs.
463
464Field name: bootsect_kludge
465Type: kernel internal
466Offset/size: 0x220/4
467Protocol: 2.00+
468
469  This field is obsolete.
470
471Field name: heap_end_ptr
472Type: write (obligatory)
473Offset/size: 0x224/2
474Protocol: 2.01+
475
476  Set this field to the offset (from the beginning of the real-mode
477  code) of the end of the setup stack/heap, minus 0x0200.
478
479Field name: ext_loader_ver
480Type: write (optional)
481Offset/size: 0x226/1
482Protocol: 2.02+
483
484  This field is used as an extension of the version number in the
485  type_of_loader field. The total version number is considered to be
486  (type_of_loader & 0x0f) + (ext_loader_ver << 4).
487
488  The use of this field is boot loader specific. If not written, it
489  is zero.
490
491  Kernels prior to 2.6.31 did not recognize this field, but it is safe
492  to write for protocol version 2.02 or higher.
493
494Field name: ext_loader_type
495Type: write (obligatory if (type_of_loader & 0xf0) == 0xe0)
496Offset/size: 0x227/1
497Protocol: 2.02+
498
499  This field is used as an extension of the type number in
500  type_of_loader field. If the type in type_of_loader is 0xE, then
501  the actual type is (ext_loader_type + 0x10).
502
503  This field is ignored if the type in type_of_loader is not 0xE.
504
505  Kernels prior to 2.6.31 did not recognize this field, but it is safe
506  to write for protocol version 2.02 or higher.
507
508Field name: cmd_line_ptr
509Type: write (obligatory)
510Offset/size: 0x228/4
511Protocol: 2.02+
512
513  Set this field to the linear address of the kernel command line.
514  The kernel command line can be located anywhere between the end of
515  the setup heap and 0xA0000; it does not have to be located in the
516  same 64K segment as the real-mode code itself.
517
518  Fill in this field even if your boot loader does not support a
519  command line, in which case you can point this to an empty string
520  (or better yet, to the string "auto".) If this field is left at
521  zero, the kernel will assume that your boot loader does not support
522  the 2.02+ protocol.
523
524Field name: ramdisk_max
525Type: read
526Offset/size: 0x22c/4
527Protocol: 2.03+
528
529  The maximum address that may be occupied by the initial
530  ramdisk/ramfs contents. For boot protocols 2.02 or earlier, this
531  field is not present, and the maximum address is 0x37FFFFFF. (This
532  address is defined as the address of the highest safe byte, so if
533  your ramdisk is exactly 131072 bytes long and this field is
534  0x37FFFFFF, you can start your ramdisk at 0x37FE0000.)
535
536Field name: kernel_alignment
537Type: read/modify (reloc)
538Offset/size: 0x230/4
539Protocol: 2.05+ (read), 2.10+ (modify)
540
541  Alignment unit required by the kernel (if relocatable_kernel is
542  true.) A relocatable kernel that is loaded at an alignment
543  incompatible with the value in this field will be realigned during
544  kernel initialization.
545
546  Starting with protocol version 2.10, this reflects the kernel
547  alignment preferred for optimal performance; it is possible for the
548  loader to modify this field to permit a lesser alignment. See the
549  min_alignment and pref_address field below.
550
551Field name: relocatable_kernel
552Type: read (reloc)
553Offset/size: 0x234/1
554Protocol: 2.05+
555
556  If this field is nonzero, the protected-mode part of the kernel can
557  be loaded at any address that satisfies the kernel_alignment field.
558  After loading, the boot loader must set the code32_start field to
559  point to the loaded code, or to a boot loader hook.
560
561Field name: min_alignment
562Type: read (reloc)
563Offset/size: 0x235/1
564Protocol: 2.10+
565
566  This field, if nonzero, indicates as a power of two the minimum
567  alignment required, as opposed to preferred, by the kernel to boot.
568  If a boot loader makes use of this field, it should update the
569  kernel_alignment field with the alignment unit desired; typically:
570
571    kernel_alignment = 1 << min_alignment
572
573  There may be a considerable performance cost with an excessively
574  misaligned kernel. Therefore, a loader should typically try each
575  power-of-two alignment from kernel_alignment down to this alignment.
576
577Field name: cmdline_size
578Type: read
579Offset/size: 0x238/4
580Protocol: 2.06+
581
582  The maximum size of the command line without the terminating
583  zero. This means that the command line can contain at most
584  cmdline_size characters. With protocol version 2.05 and earlier, the
585  maximum size was 255.
586
587Field name: hardware_subarch
588Type: write (optional, defaults to x86/PC)
589Offset/size: 0x23c/4
590Protocol: 2.07+
591
592  In a paravirtualized environment the hardware low level architectural
593  pieces such as interrupt handling, page table handling, and
594  accessing process control registers needs to be done differently.
595
596  This field allows the bootloader to inform the kernel we are in one
597  one of those environments.
598
599  0x00000000 The default x86/PC environment
600  0x00000001 lguest
601  0x00000002 Xen
602  0x00000003 Moorestown MID
603
604Field name: hardware_subarch_data
605Type: write (subarch-dependent)
606Offset/size: 0x240/8
607Protocol: 2.07+
608
609  A pointer to data that is specific to hardware subarch
610  This field is currently unused for the default x86/PC environment,
611  do not modify.
612
613Field name: payload_offset
614Type: read
615Offset/size: 0x248/4
616Protocol: 2.08+
617
618  If non-zero then this field contains the offset from the beginning
619  of the protected-mode code to the payload.
620
621  The payload may be compressed. The format of both the compressed and
622  uncompressed data should be determined using the standard magic
623  numbers. The currently supported compression formats are gzip
624  (magic numbers 1F 8B or 1F 9E), bzip2 (magic number 42 5A) and LZMA
625  (magic number 5D 00). The uncompressed payload is currently always ELF
626  (magic number 7F 45 4C 46).
627  
628Field name: payload_length
629Type: read
630Offset/size: 0x24c/4
631Protocol: 2.08+
632
633  The length of the payload.
634
635Field name: setup_data
636Type: write (special)
637Offset/size: 0x250/8
638Protocol: 2.09+
639
640  The 64-bit physical pointer to NULL terminated single linked list of
641  struct setup_data. This is used to define a more extensible boot
642  parameters passing mechanism. The definition of struct setup_data is
643  as follow:
644
645  struct setup_data {
646      u64 next;
647      u32 type;
648      u32 len;
649      u8 data[0];
650  };
651
652  Where, the next is a 64-bit physical pointer to the next node of
653  linked list, the next field of the last node is 0; the type is used
654  to identify the contents of data; the len is the length of data
655  field; the data holds the real payload.
656
657  This list may be modified at a number of points during the bootup
658  process. Therefore, when modifying this list one should always make
659  sure to consider the case where the linked list already contains
660  entries.
661
662Field name: pref_address
663Type: read (reloc)
664Offset/size: 0x258/8
665Protocol: 2.10+
666
667  This field, if nonzero, represents a preferred load address for the
668  kernel. A relocating bootloader should attempt to load at this
669  address if possible.
670
671  A non-relocatable kernel will unconditionally move itself and to run
672  at this address.
673
674Field name: init_size
675Type: read
676Offset/size: 0x25c/4
677
678  This field indicates the amount of linear contiguous memory starting
679  at the kernel runtime start address that the kernel needs before it
680  is capable of examining its memory map. This is not the same thing
681  as the total amount of memory the kernel needs to boot, but it can
682  be used by a relocating boot loader to help select a safe load
683  address for the kernel.
684
685  The kernel runtime start address is determined by the following algorithm:
686
687  if (relocatable_kernel)
688    runtime_start = align_up(load_address, kernel_alignment)
689  else
690    runtime_start = pref_address
691
692
693**** THE IMAGE CHECKSUM
694
695From boot protocol version 2.08 onwards the CRC-32 is calculated over
696the entire file using the characteristic polynomial 0x04C11DB7 and an
697initial remainder of 0xffffffff. The checksum is appended to the
698file; therefore the CRC of the file up to the limit specified in the
699syssize field of the header is always 0.
700
701
702**** THE KERNEL COMMAND LINE
703
704The kernel command line has become an important way for the boot
705loader to communicate with the kernel. Some of its options are also
706relevant to the boot loader itself, see "special command line options"
707below.
708
709The kernel command line is a null-terminated string. The maximum
710length can be retrieved from the field cmdline_size. Before protocol
711version 2.06, the maximum was 255 characters. A string that is too
712long will be automatically truncated by the kernel.
713
714If the boot protocol version is 2.02 or later, the address of the
715kernel command line is given by the header field cmd_line_ptr (see
716above.) This address can be anywhere between the end of the setup
717heap and 0xA0000.
718
719If the protocol version is *not* 2.02 or higher, the kernel
720command line is entered using the following protocol:
721
722    At offset 0x0020 (word), "cmd_line_magic", enter the magic
723    number 0xA33F.
724
725    At offset 0x0022 (word), "cmd_line_offset", enter the offset
726    of the kernel command line (relative to the start of the
727    real-mode kernel).
728    
729    The kernel command line *must* be within the memory region
730    covered by setup_move_size, so you may need to adjust this
731    field.
732
733
734**** MEMORY LAYOUT OF THE REAL-MODE CODE
735
736The real-mode code requires a stack/heap to be set up, as well as
737memory allocated for the kernel command line. This needs to be done
738in the real-mode accessible memory in bottom megabyte.
739
740It should be noted that modern machines often have a sizable Extended
741BIOS Data Area (EBDA). As a result, it is advisable to use as little
742of the low megabyte as possible.
743
744Unfortunately, under the following circumstances the 0x90000 memory
745segment has to be used:
746
747    - When loading a zImage kernel ((loadflags & 0x01) == 0).
748    - When loading a 2.01 or earlier boot protocol kernel.
749
750      -> For the 2.00 and 2.01 boot protocols, the real-mode code
751         can be loaded at another address, but it is internally
752         relocated to 0x90000. For the "old" protocol, the
753         real-mode code must be loaded at 0x90000.
754
755When loading at 0x90000, avoid using memory above 0x9a000.
756
757For boot protocol 2.02 or higher, the command line does not have to be
758located in the same 64K segment as the real-mode setup code; it is
759thus permitted to give the stack/heap the full 64K segment and locate
760the command line above it.
761
762The kernel command line should not be located below the real-mode
763code, nor should it be located in high memory.
764
765
766**** SAMPLE BOOT CONFIGURATION
767
768As a sample configuration, assume the following layout of the real
769mode segment:
770
771    When loading below 0x90000, use the entire segment:
772
773    0x0000-0x7fff Real mode kernel
774    0x8000-0xdfff Stack and heap
775    0xe000-0xffff Kernel command line
776
777    When loading at 0x90000 OR the protocol version is 2.01 or earlier:
778
779    0x0000-0x7fff Real mode kernel
780    0x8000-0x97ff Stack and heap
781    0x9800-0x9fff Kernel command line
782
783Such a boot loader should enter the following fields in the header:
784
785    unsigned long base_ptr; /* base address for real-mode segment */
786
787    if ( setup_sects == 0 ) {
788        setup_sects = 4;
789    }
790
791    if ( protocol >= 0x0200 ) {
792        type_of_loader = <type code>;
793        if ( loading_initrd ) {
794            ramdisk_image = <initrd_address>;
795            ramdisk_size = <initrd_size>;
796        }
797
798        if ( protocol >= 0x0202 && loadflags & 0x01 )
799            heap_end = 0xe000;
800        else
801            heap_end = 0x9800;
802
803        if ( protocol >= 0x0201 ) {
804            heap_end_ptr = heap_end - 0x200;
805            loadflags |= 0x80; /* CAN_USE_HEAP */
806        }
807
808        if ( protocol >= 0x0202 ) {
809            cmd_line_ptr = base_ptr + heap_end;
810            strcpy(cmd_line_ptr, cmdline);
811        } else {
812            cmd_line_magic = 0xA33F;
813            cmd_line_offset = heap_end;
814            setup_move_size = heap_end + strlen(cmdline)+1;
815            strcpy(base_ptr+cmd_line_offset, cmdline);
816        }
817    } else {
818        /* Very old kernel */
819
820        heap_end = 0x9800;
821
822        cmd_line_magic = 0xA33F;
823        cmd_line_offset = heap_end;
824
825        /* A very old kernel MUST have its real-mode code
826           loaded at 0x90000 */
827
828        if ( base_ptr != 0x90000 ) {
829            /* Copy the real-mode kernel */
830            memcpy(0x90000, base_ptr, (setup_sects+1)*512);
831            base_ptr = 0x90000; /* Relocated */
832        }
833
834        strcpy(0x90000+cmd_line_offset, cmdline);
835
836        /* It is recommended to clear memory up to the 32K mark */
837        memset(0x90000 + (setup_sects+1)*512, 0,
838               (64-(setup_sects+1))*512);
839    }
840
841
842**** LOADING THE REST OF THE KERNEL
843
844The 32-bit (non-real-mode) kernel starts at offset (setup_sects+1)*512
845in the kernel file (again, if setup_sects == 0 the real value is 4.)
846It should be loaded at address 0x10000 for Image/zImage kernels and
8470x100000 for bzImage kernels.
848
849The kernel is a bzImage kernel if the protocol >= 2.00 and the 0x01
850bit (LOAD_HIGH) in the loadflags field is set:
851
852    is_bzImage = (protocol >= 0x0200) && (loadflags & 0x01);
853    load_address = is_bzImage ? 0x100000 : 0x10000;
854
855Note that Image/zImage kernels can be up to 512K in size, and thus use
856the entire 0x10000-0x90000 range of memory. This means it is pretty
857much a requirement for these kernels to load the real-mode part at
8580x90000. bzImage kernels allow much more flexibility.
859
860
861**** SPECIAL COMMAND LINE OPTIONS
862
863If the command line provided by the boot loader is entered by the
864user, the user may expect the following command line options to work.
865They should normally not be deleted from the kernel command line even
866though not all of them are actually meaningful to the kernel. Boot
867loader authors who need additional command line options for the boot
868loader itself should get them registered in
869Documentation/kernel-parameters.txt to make sure they will not
870conflict with actual kernel options now or in the future.
871
872  vga=<mode>
873    <mode> here is either an integer (in C notation, either
874    decimal, octal, or hexadecimal) or one of the strings
875    "normal" (meaning 0xFFFF), "ext" (meaning 0xFFFE) or "ask"
876    (meaning 0xFFFD). This value should be entered into the
877    vid_mode field, as it is used by the kernel before the command
878    line is parsed.
879
880  mem=<size>
881    <size> is an integer in C notation optionally followed by
882    (case insensitive) K, M, G, T, P or E (meaning << 10, << 20,
883    << 30, << 40, << 50 or << 60). This specifies the end of
884    memory to the kernel. This affects the possible placement of
885    an initrd, since an initrd should be placed near end of
886    memory. Note that this is an option to *both* the kernel and
887    the bootloader!
888
889  initrd=<file>
890    An initrd should be loaded. The meaning of <file> is
891    obviously bootloader-dependent, and some boot loaders
892    (e.g. LILO) do not have such a command.
893
894In addition, some boot loaders add the following options to the
895user-specified command line:
896
897  BOOT_IMAGE=<file>
898    The boot image which was loaded. Again, the meaning of <file>
899    is obviously bootloader-dependent.
900
901  auto
902    The kernel was booted without explicit user intervention.
903
904If these options are added by the boot loader, it is highly
905recommended that they are located *first*, before the user-specified
906or configuration-specified command line. Otherwise, "init=/bin/sh"
907gets confused by the "auto" option.
908
909
910**** RUNNING THE KERNEL
911
912The kernel is started by jumping to the kernel entry point, which is
913located at *segment* offset 0x20 from the start of the real mode
914kernel. This means that if you loaded your real-mode kernel code at
9150x90000, the kernel entry point is 9020:0000.
916
917At entry, ds = es = ss should point to the start of the real-mode
918kernel code (0x9000 if the code is loaded at 0x90000), sp should be
919set up properly, normally pointing to the top of the heap, and
920interrupts should be disabled. Furthermore, to guard against bugs in
921the kernel, it is recommended that the boot loader sets fs = gs = ds =
922es = ss.
923
924In our example from above, we would do:
925
926    /* Note: in the case of the "old" kernel protocol, base_ptr must
927       be == 0x90000 at this point; see the previous sample code */
928
929    seg = base_ptr >> 4;
930
931    cli(); /* Enter with interrupts disabled! */
932
933    /* Set up the real-mode kernel stack */
934    _SS = seg;
935    _SP = heap_end;
936
937    _DS = _ES = _FS = _GS = seg;
938    jmp_far(seg+0x20, 0); /* Run the kernel */
939
940If your boot sector accesses a floppy drive, it is recommended to
941switch off the floppy motor before running the kernel, since the
942kernel boot leaves interrupts off and thus the motor will not be
943switched off, especially if the loaded kernel has the floppy driver as
944a demand-loaded module!
945
946
947**** ADVANCED BOOT LOADER HOOKS
948
949If the boot loader runs in a particularly hostile environment (such as
950LOADLIN, which runs under DOS) it may be impossible to follow the
951standard memory location requirements. Such a boot loader may use the
952following hooks that, if set, are invoked by the kernel at the
953appropriate time. The use of these hooks should probably be
954considered an absolutely last resort!
955
956IMPORTANT: All the hooks are required to preserve %esp, %ebp, %esi and
957%edi across invocation.
958
959  realmode_swtch:
960    A 16-bit real mode far subroutine invoked immediately before
961    entering protected mode. The default routine disables NMI, so
962    your routine should probably do so, too.
963
964  code32_start:
965    A 32-bit flat-mode routine *jumped* to immediately after the
966    transition to protected mode, but before the kernel is
967    uncompressed. No segments, except CS, are guaranteed to be
968    set up (current kernels do, but older ones do not); you should
969    set them up to BOOT_DS (0x18) yourself.
970
971    After completing your hook, you should jump to the address
972    that was in this field before your boot loader overwrote it
973    (relocated, if appropriate.)
974
975
976**** 32-bit BOOT PROTOCOL
977
978For machine with some new BIOS other than legacy BIOS, such as EFI,
979LinuxBIOS, etc, and kexec, the 16-bit real mode setup code in kernel
980based on legacy BIOS can not be used, so a 32-bit boot protocol needs
981to be defined.
982
983In 32-bit boot protocol, the first step in loading a Linux kernel
984should be to setup the boot parameters (struct boot_params,
985traditionally known as "zero page"). The memory for struct boot_params
986should be allocated and initialized to all zero. Then the setup header
987from offset 0x01f1 of kernel image on should be loaded into struct
988boot_params and examined. The end of setup header can be calculated as
989follow:
990
991    0x0202 + byte value at offset 0x0201
992
993In addition to read/modify/write the setup header of the struct
994boot_params as that of 16-bit boot protocol, the boot loader should
995also fill the additional fields of the struct boot_params as that
996described in zero-page.txt.
997
998After setupping the struct boot_params, the boot loader can load the
99932/64-bit kernel in the same way as that of 16-bit boot protocol.
1000
1001In 32-bit boot protocol, the kernel is started by jumping to the
100232-bit kernel entry point, which is the start address of loaded
100332/64-bit kernel.
1004
1005At entry, the CPU must be in 32-bit protected mode with paging
1006disabled; a GDT must be loaded with the descriptors for selectors
1007__BOOT_CS(0x10) and __BOOT_DS(0x18); both descriptors must be 4G flat
1008segment; __BOOS_CS must have execute/read permission, and __BOOT_DS
1009must have read/write permission; CS must be __BOOT_CS and DS, ES, SS
1010must be __BOOT_DS; interrupt must be disabled; %esi must hold the base
1011address of the struct boot_params; %ebp, %edi and %ebx must be zero.
1012

Archive Download this file



interactive