blob: e0203662f9e9f8a1a907060437f3fcdf907e2667 [file] [log] [blame]
H. Peter Anvin4039feb2008-05-30 17:16:20 -07001 THE LINUX/x86 BOOT PROTOCOL
2 ---------------------------
Linus Torvalds1da177e2005-04-16 15:20:36 -07003
H. Peter Anvin4039feb2008-05-30 17:16:20 -07004On the x86 platform, the Linux kernel uses a rather complicated boot
Linus Torvalds1da177e2005-04-16 15:20:36 -07005convention. 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
H. Peter Anvin4039feb2008-05-30 17:16:20 -070011Currently, the following versions of the Linux/x86 boot protocol exist.
Linus Torvalds1da177e2005-04-16 15:20:36 -070012
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
H. Peter Anvinf8eeaaf2005-09-06 15:17:24 -070034Protocol 2.04: (Kernel 2.6.14) Extend the syssize field to four bytes.
Bernhard Walle8f9aeca2007-05-02 19:27:10 +020035
Vivek Goyalbe274ee2006-12-07 02:14:04 +010036Protocol 2.05: (Kernel 2.6.20) Make protected mode kernel relocatable.
37 Introduce relocatable_kernel and kernel_alignment fields.
H. Peter Anvinf8eeaaf2005-09-06 15:17:24 -070038
Bernhard Walle8f9aeca2007-05-02 19:27:10 +020039Protocol 2.06: (Kernel 2.6.22) Added a field that contains the size of
Ian Campbell4c0587e2008-04-27 12:21:11 +010040 the boot command line.
Bernhard Walle8f9aeca2007-05-02 19:27:10 +020041
Ian Campbell4c0587e2008-04-27 12:21:11 +010042Protocol 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
Baodong Chen2f6de3a2009-01-03 12:37:06 +080047 payload. Introduced payload_offset and payload_length
Ian Campbell4c0587e2008-04-27 12:21:11 +010048 fields to aid in locating the payload.
49
50Protocol 2.09: (Kernel 2.6.26) Added a field of 64-bit physical
Huang, Yingfb884382008-03-28 10:49:50 +080051 pointer to single linked list of struct setup_data.
Linus Torvalds1da177e2005-04-16 15:20:36 -070052
53**** MEMORY LAYOUT
54
55The traditional memory map for the kernel loader, used for Image or
56zImage kernels, typically looks like:
57
58 | |
590A0000 +------------------------+
60 | Reserved for BIOS | Do not use. Reserved for BIOS EBDA.
6109A000 +------------------------+
H. Peter Anvindec04cf2007-05-17 15:50:47 -070062 | Command line |
63 | Stack/heap | For use by the kernel real-mode code.
Linus Torvalds1da177e2005-04-16 15:20:36 -070064098000 +------------------------+
65 | Kernel setup | The kernel real-mode code.
66090200 +------------------------+
67 | Kernel boot sector | The kernel legacy boot sector.
68090000 +------------------------+
69 | Protected-mode kernel | The bulk of the kernel image.
70010000 +------------------------+
71 | Boot loader | <- Boot sector entry point 0000:7C00
72001000 +------------------------+
73 | Reserved for MBR/BIOS |
74000800 +------------------------+
75 | Typically used by MBR |
76000600 +------------------------+
77 | BIOS use only |
78000000 +------------------------+
79
80
81When using bzImage, the protected-mode kernel was relocated to
820x100000 ("high memory"), and the kernel real-mode block (boot sector,
83setup, and stack/heap) was made relocatable to any address between
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700840x10000 and end of low memory. Unfortunately, in protocols 2.00 and
852.01 the 0x90000+ memory range is still used internally by the kernel;
86the 2.02 protocol resolves that problem.
Linus Torvalds1da177e2005-04-16 15:20:36 -070087
88It is desirable to keep the "memory ceiling" -- the highest point in
89low memory touched by the boot loader -- as low as possible, since
90some newer BIOSes have begun to allocate some rather large amounts of
91memory, called the Extended BIOS Data Area, near the top of low
92memory. The boot loader should use the "INT 12h" BIOS call to verify
93how much low memory is available.
94
95Unfortunately, if INT 12h reports that the amount of memory is too
96low, there is usually nothing the boot loader can do but to report an
97error to the user. The boot loader should therefore be designed to
98take up as little space in low memory as it reasonably can. For
99zImage or old bzImage kernels, which need data written into the
1000x90000 segment, the boot loader should make sure not to use memory
101above the 0x9A000 point; too many BIOSes will break above that point.
102
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700103For a modern bzImage kernel with boot protocol version >= 2.02, a
104memory layout like the following is suggested:
105
106 ~ ~
107 | Protected-mode kernel |
108100000 +------------------------+
109 | I/O memory hole |
1100A0000 +------------------------+
111 | Reserved for BIOS | Leave as much as possible unused
112 ~ ~
113 | Command line | (Can also be below the X+10000 mark)
114X+10000 +------------------------+
115 | Stack/heap | For use by the kernel real-mode code.
116X+08000 +------------------------+
117 | Kernel setup | The kernel real-mode code.
118 | Kernel boot sector | The kernel legacy boot sector.
119X +------------------------+
120 | Boot loader | <- Boot sector entry point 0000:7C00
121001000 +------------------------+
122 | Reserved for MBR/BIOS |
123000800 +------------------------+
124 | Typically used by MBR |
125000600 +------------------------+
126 | BIOS use only |
127000000 +------------------------+
128
129... where the address X is as low as the design of the boot loader
130permits.
131
Linus Torvalds1da177e2005-04-16 15:20:36 -0700132
133**** THE REAL-MODE KERNEL HEADER
134
135In the following text, and anywhere in the kernel boot sequence, "a
136sector" refers to 512 bytes. It is independent of the actual sector
137size of the underlying medium.
138
139The first step in loading a Linux kernel should be to load the
140real-mode code (boot sector and setup code) and then examine the
141following header at offset 0x01f1. The real-mode code can total up to
14232K, although the boot loader may choose to load only the first two
143sectors (1K) and then examine the bootup sector size.
144
145The header looks like:
146
147Offset Proto Name Meaning
148/Size
149
H. Peter Anvinf8eeaaf2005-09-06 15:17:24 -070015001F1/1 ALL(1 setup_sects The size of the setup in sectors
Linus Torvalds1da177e2005-04-16 15:20:36 -070015101F2/2 ALL root_flags If set, the root is mounted readonly
H. Peter Anvinf8eeaaf2005-09-06 15:17:24 -070015201F4/4 2.04+(2 syssize The size of the 32-bit code in 16-byte paras
Linus Torvalds1da177e2005-04-16 15:20:36 -070015301F8/2 ALL ram_size DO NOT USE - for bootsect.S use only
15401FA/2 ALL vid_mode Video mode control
15501FC/2 ALL root_dev Default root device number
15601FE/2 ALL boot_flag 0xAA55 magic number
1570200/2 2.00+ jump Jump instruction
1580202/4 2.00+ header Magic signature "HdrS"
1590206/2 2.00+ version Boot protocol version supported
1600208/4 2.00+ realmode_swtch Boot loader hook (see below)
Baodong Chene56d0cf2009-01-08 19:24:29 +0800161020C/2 2.00+ start_sys_seg The load-low segment (0x1000) (obsolete)
Linus Torvalds1da177e2005-04-16 15:20:36 -0700162020E/2 2.00+ kernel_version Pointer to kernel version string
1630210/1 2.00+ type_of_loader Boot loader identifier
1640211/1 2.00+ loadflags Boot protocol option flags
1650212/2 2.00+ setup_move_size Move to high memory size (used with hooks)
1660214/4 2.00+ code32_start Boot loader hook (see below)
1670218/4 2.00+ ramdisk_image initrd load address (set by boot loader)
168021C/4 2.00+ ramdisk_size initrd size (set by boot loader)
1690220/4 2.00+ bootsect_kludge DO NOT USE - for bootsect.S use only
1700224/2 2.01+ heap_end_ptr Free memory after setup end
1710226/2 N/A pad1 Unused
1720228/4 2.02+ cmd_line_ptr 32-bit pointer to the kernel command line
Baodong Chene56d0cf2009-01-08 19:24:29 +0800173022C/4 2.03+ ramdisk_max Highest legal initrd address
Vivek Goyald263b212006-12-07 02:14:13 +01001740230/4 2.05+ kernel_alignment Physical addr alignment required for kernel
1750234/1 2.05+ relocatable_kernel Whether kernel is relocatable or not
Baodong Chene56d0cf2009-01-08 19:24:29 +08001760235/1 N/A pad2 Unused
1770236/2 N/A pad3 Unused
Bernhard Walle8f9aeca2007-05-02 19:27:10 +02001780238/4 2.06+ cmdline_size Maximum size of the kernel command line
Rusty Russelle5371ac2007-10-21 16:41:33 -0700179023C/4 2.07+ hardware_subarch Hardware subarchitecture
1800240/8 2.07+ hardware_subarch_data Subarchitecture-specific data
Ian Campbell87253d12008-02-19 11:12:30 +00001810248/4 2.08+ payload_offset Offset of kernel payload
182024C/4 2.08+ payload_length Length of kernel payload
Huang, Yingfb884382008-03-28 10:49:50 +08001830250/8 2.09+ setup_data 64-bit physical pointer to linked list
184 of struct setup_data
Linus Torvalds1da177e2005-04-16 15:20:36 -0700185
H. Peter Anvinf8eeaaf2005-09-06 15:17:24 -0700186(1) For backwards compatibility, if the setup_sects field contains 0, the
187 real value is 4.
188
189(2) For boot protocol prior to 2.04, the upper two bytes of the syssize
190 field are unusable, which means the size of a bzImage kernel
191 cannot be determined.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700192
193If the "HdrS" (0x53726448) magic number is not found at offset 0x202,
194the boot protocol version is "old". Loading an old kernel, the
195following parameters should be assumed:
196
197 Image type = zImage
198 initrd not supported
199 Real-mode kernel must be located at 0x90000.
200
201Otherwise, the "version" field contains the protocol version,
202e.g. protocol version 2.01 will contain 0x0201 in this field. When
203setting fields in the header, you must make sure only to set fields
204supported by the protocol version in use.
205
Linus Torvalds1da177e2005-04-16 15:20:36 -0700206
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700207**** DETAILS OF HEADER FIELDS
Linus Torvalds1da177e2005-04-16 15:20:36 -0700208
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700209For each field, some are information from the kernel to the bootloader
210("read"), some are expected to be filled out by the bootloader
211("write"), and some are expected to be read and modified by the
212bootloader ("modify").
Linus Torvalds1da177e2005-04-16 15:20:36 -0700213
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700214All general purpose boot loaders should write the fields marked
215(obligatory). Boot loaders who want to load the kernel at a
216nonstandard address should fill in the fields marked (reloc); other
217boot loaders can ignore those fields.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700218
H. Peter Anvindb2668f2007-05-23 16:59:27 -0700219The byte order of all fields is littleendian (this is x86, after all.)
220
Rusty Russelle5371ac2007-10-21 16:41:33 -0700221Field name: setup_sects
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700222Type: read
223Offset/size: 0x1f1/1
224Protocol: ALL
225
226 The size of the setup code in 512-byte sectors. If this field is
227 0, the real value is 4. The real-mode code consists of the boot
228 sector (always one 512-byte sector) plus the setup code.
229
230Field name: root_flags
231Type: modify (optional)
232Offset/size: 0x1f2/2
233Protocol: ALL
234
235 If this field is nonzero, the root defaults to readonly. The use of
236 this field is deprecated; use the "ro" or "rw" options on the
237 command line instead.
238
239Field name: syssize
240Type: read
241Offset/size: 0x1f4/4 (protocol 2.04+) 0x1f4/2 (protocol ALL)
242Protocol: 2.04+
243
244 The size of the protected-mode code in units of 16-byte paragraphs.
245 For protocol versions older than 2.04 this field is only two bytes
246 wide, and therefore cannot be trusted for the size of a kernel if
247 the LOAD_HIGH flag is set.
248
249Field name: ram_size
250Type: kernel internal
251Offset/size: 0x1f8/2
252Protocol: ALL
253
254 This field is obsolete.
255
256Field name: vid_mode
257Type: modify (obligatory)
258Offset/size: 0x1fa/2
259
260 Please see the section on SPECIAL COMMAND LINE OPTIONS.
261
262Field name: root_dev
263Type: modify (optional)
264Offset/size: 0x1fc/2
265Protocol: ALL
266
267 The default root device device number. The use of this field is
268 deprecated, use the "root=" option on the command line instead.
269
270Field name: boot_flag
271Type: read
272Offset/size: 0x1fe/2
273Protocol: ALL
274
275 Contains 0xAA55. This is the closest thing old Linux kernels have
276 to a magic number.
277
278Field name: jump
279Type: read
280Offset/size: 0x200/2
281Protocol: 2.00+
282
283 Contains an x86 jump instruction, 0xEB followed by a signed offset
284 relative to byte 0x202. This can be used to determine the size of
285 the header.
286
287Field name: header
288Type: read
289Offset/size: 0x202/4
290Protocol: 2.00+
291
292 Contains the magic number "HdrS" (0x53726448).
293
294Field name: version
295Type: read
296Offset/size: 0x206/2
297Protocol: 2.00+
298
H. Peter Anvindb2668f2007-05-23 16:59:27 -0700299 Contains the boot protocol version, in (major << 8)+minor format,
300 e.g. 0x0204 for version 2.04, and 0x0a11 for a hypothetical version
301 10.17.
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700302
Baodong Chene56d0cf2009-01-08 19:24:29 +0800303Field name: realmode_swtch
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700304Type: modify (optional)
305Offset/size: 0x208/4
306Protocol: 2.00+
307
H. Peter Anvindb2668f2007-05-23 16:59:27 -0700308 Boot loader hook (see ADVANCED BOOT LOADER HOOKS below.)
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700309
Baodong Chene56d0cf2009-01-08 19:24:29 +0800310Field name: start_sys_seg
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700311Type: read
H. Peter Anvina021e512008-07-22 15:33:57 -0400312Offset/size: 0x20c/2
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700313Protocol: 2.00+
314
315 The load low segment (0x1000). Obsolete.
316
317Field name: kernel_version
318Type: read
319Offset/size: 0x20e/2
320Protocol: 2.00+
321
322 If set to a nonzero value, contains a pointer to a NUL-terminated
323 human-readable kernel version number string, less 0x200. This can
324 be used to display the kernel version to the user. This value
H. Peter Anvindb2668f2007-05-23 16:59:27 -0700325 should be less than (0x200*setup_sects).
326
327 For example, if this value is set to 0x1c00, the kernel version
328 number string can be found at offset 0x1e00 in the kernel file.
329 This is a valid value if and only if the "setup_sects" field
330 contains the value 15 or higher, as:
331
332 0x1c00 < 15*0x200 (= 0x1e00) but
333 0x1c00 >= 14*0x200 (= 0x1c00)
334
335 0x1c00 >> 9 = 14, so the minimum value for setup_secs is 15.
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700336
337Field name: type_of_loader
338Type: write (obligatory)
339Offset/size: 0x210/1
340Protocol: 2.00+
341
342 If your boot loader has an assigned id (see table below), enter
343 0xTV here, where T is an identifier for the boot loader and V is
344 a version number. Otherwise, enter 0xFF here.
345
346 Assigned boot loader ids:
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700347 0 LILO (0x00 reserved for pre-2.00 bootloader)
Linus Torvalds1da177e2005-04-16 15:20:36 -0700348 1 Loadlin
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700349 2 bootsect-loader (0x20, all other values reserved)
Linus Torvalds1da177e2005-04-16 15:20:36 -0700350 3 SYSLINUX
351 4 EtherBoot
352 5 ELILO
KOSAKI Motohiro9ee670f2008-11-11 11:49:01 +0900353 7 GRUB
Linus Torvalds1da177e2005-04-16 15:20:36 -0700354 8 U-BOOT
Jeremy Fitzhardinge354332e2006-09-12 20:35:57 -0700355 9 Xen
H. Peter Anvinc229ec52007-01-26 09:15:47 -0800356 A Gujin
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700357 B Qemu
Linus Torvalds1da177e2005-04-16 15:20:36 -0700358
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700359 Please contact <hpa@zytor.com> if you need a bootloader ID
360 value assigned.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700361
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700362Field name: loadflags
363Type: modify (obligatory)
364Offset/size: 0x211/1
365Protocol: 2.00+
Linus Torvalds1da177e2005-04-16 15:20:36 -0700366
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700367 This field is a bitmask.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700368
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700369 Bit 0 (read): LOADED_HIGH
370 - If 0, the protected-mode code is loaded at 0x10000.
371 - If 1, the protected-mode code is loaded at 0x100000.
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700372
H. Peter Anvin4039feb2008-05-30 17:16:20 -0700373 Bit 5 (write): QUIET_FLAG
374 - If 0, print early messages.
375 - If 1, suppress early messages.
376 This requests to the kernel (decompressor and early
377 kernel) to not write early messages that require
378 accessing the display hardware directly.
379
Rusty Russelle5371ac2007-10-21 16:41:33 -0700380 Bit 6 (write): KEEP_SEGMENTS
381 Protocol: 2.07+
H. Peter Anvin4039feb2008-05-30 17:16:20 -0700382 - If 0, reload the segment registers in the 32bit entry point.
383 - If 1, do not reload the segment registers in the 32bit entry point.
Rusty Russelle5371ac2007-10-21 16:41:33 -0700384 Assume that %cs %ds %ss %es are all set to flat segments with
385 a base of 0 (or the equivalent for their environment).
386
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700387 Bit 7 (write): CAN_USE_HEAP
388 Set this bit to 1 to indicate that the value entered in the
389 heap_end_ptr is valid. If this field is clear, some setup code
390 functionality will be disabled.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700391
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700392Field name: setup_move_size
393Type: modify (obligatory)
394Offset/size: 0x212/2
395Protocol: 2.00-2.01
Linus Torvalds1da177e2005-04-16 15:20:36 -0700396
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700397 When using protocol 2.00 or 2.01, if the real mode kernel is not
398 loaded at 0x90000, it gets moved there later in the loading
399 sequence. Fill in this field if you want additional data (such as
400 the kernel command line) moved in addition to the real-mode kernel
401 itself.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700402
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700403 The unit is bytes starting with the beginning of the boot sector.
404
405 This field is can be ignored when the protocol is 2.02 or higher, or
406 if the real-mode code is loaded at 0x90000.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700407
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700408Field name: code32_start
409Type: modify (optional, reloc)
410Offset/size: 0x214/4
411Protocol: 2.00+
412
413 The address to jump to in protected mode. This defaults to the load
414 address of the kernel, and can be used by the boot loader to
415 determine the proper load address.
416
417 This field can be modified for two purposes:
418
H. Peter Anvindb2668f2007-05-23 16:59:27 -0700419 1. as a boot loader hook (see ADVANCED BOOT LOADER HOOKS below.)
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700420
421 2. if a bootloader which does not install a hook loads a
422 relocatable kernel at a nonstandard address it will have to modify
423 this field to point to the load address.
424
425Field name: ramdisk_image
426Type: write (obligatory)
427Offset/size: 0x218/4
428Protocol: 2.00+
429
430 The 32-bit linear address of the initial ramdisk or ramfs. Leave at
431 zero if there is no initial ramdisk/ramfs.
432
433Field name: ramdisk_size
434Type: write (obligatory)
435Offset/size: 0x21c/4
436Protocol: 2.00+
437
438 Size of the initial ramdisk or ramfs. Leave at zero if there is no
439 initial ramdisk/ramfs.
440
441Field name: bootsect_kludge
442Type: kernel internal
443Offset/size: 0x220/4
444Protocol: 2.00+
445
446 This field is obsolete.
447
448Field name: heap_end_ptr
449Type: write (obligatory)
450Offset/size: 0x224/2
451Protocol: 2.01+
452
453 Set this field to the offset (from the beginning of the real-mode
454 code) of the end of the setup stack/heap, minus 0x0200.
455
456Field name: cmd_line_ptr
457Type: write (obligatory)
458Offset/size: 0x228/4
459Protocol: 2.02+
460
461 Set this field to the linear address of the kernel command line.
462 The kernel command line can be located anywhere between the end of
463 the setup heap and 0xA0000; it does not have to be located in the
464 same 64K segment as the real-mode code itself.
465
466 Fill in this field even if your boot loader does not support a
467 command line, in which case you can point this to an empty string
468 (or better yet, to the string "auto".) If this field is left at
469 zero, the kernel will assume that your boot loader does not support
470 the 2.02+ protocol.
471
Baodong Chene56d0cf2009-01-08 19:24:29 +0800472Field name: ramdisk_max
H. Peter Anvindec04cf2007-05-17 15:50:47 -0700473Type: read
474Offset/size: 0x22c/4
475Protocol: 2.03+
476
477 The maximum address that may be occupied by the initial
478 ramdisk/ramfs contents. For boot protocols 2.02 or earlier, this
479 field is not present, and the maximum address is 0x37FFFFFF. (This
480 address is defined as the address of the highest safe byte, so if
481 your ramdisk is exactly 131072 bytes long and this field is
482 0x37FFFFFF, you can start your ramdisk at 0x37FE0000.)
483
484Field name: kernel_alignment
485Type: read (reloc)
486Offset/size: 0x230/4
487Protocol: 2.05+
488
489 Alignment unit required by the kernel (if relocatable_kernel is true.)
490
491Field name: relocatable_kernel
492Type: read (reloc)
493Offset/size: 0x234/1
494Protocol: 2.05+
495
496 If this field is nonzero, the protected-mode part of the kernel can
497 be loaded at any address that satisfies the kernel_alignment field.
498 After loading, the boot loader must set the code32_start field to
499 point to the loaded code, or to a boot loader hook.
500
501Field name: cmdline_size
502Type: read
503Offset/size: 0x238/4
504Protocol: 2.06+
505
506 The maximum size of the command line without the terminating
507 zero. This means that the command line can contain at most
508 cmdline_size characters. With protocol version 2.05 and earlier, the
509 maximum size was 255.
Bernhard Walle8f9aeca2007-05-02 19:27:10 +0200510
Rusty Russelle5371ac2007-10-21 16:41:33 -0700511Field name: hardware_subarch
H. Peter Anvin4039feb2008-05-30 17:16:20 -0700512Type: write (optional, defaults to x86/PC)
Rusty Russelle5371ac2007-10-21 16:41:33 -0700513Offset/size: 0x23c/4
514Protocol: 2.07+
515
516 In a paravirtualized environment the hardware low level architectural
517 pieces such as interrupt handling, page table handling, and
518 accessing process control registers needs to be done differently.
519
520 This field allows the bootloader to inform the kernel we are in one
521 one of those environments.
522
523 0x00000000 The default x86/PC environment
524 0x00000001 lguest
525 0x00000002 Xen
526
527Field name: hardware_subarch_data
H. Peter Anvin4039feb2008-05-30 17:16:20 -0700528Type: write (subarch-dependent)
Rusty Russelle5371ac2007-10-21 16:41:33 -0700529Offset/size: 0x240/8
530Protocol: 2.07+
531
532 A pointer to data that is specific to hardware subarch
H. Peter Anvin4039feb2008-05-30 17:16:20 -0700533 This field is currently unused for the default x86/PC environment,
534 do not modify.
Rusty Russelle5371ac2007-10-21 16:41:33 -0700535
Ian Campbell87253d12008-02-19 11:12:30 +0000536Field name: payload_offset
Ian Campbell099e1372008-02-13 20:54:58 +0000537Type: read
538Offset/size: 0x248/4
539Protocol: 2.08+
540
Philipp Kohlbechere1286f22008-12-14 18:50:01 +0100541 If non-zero then this field contains the offset from the beginning
542 of the protected-mode code to the payload.
Ian Campbell87253d12008-02-19 11:12:30 +0000543
544 The payload may be compressed. The format of both the compressed and
545 uncompressed data should be determined using the standard magic
H. Peter Anvinee287582009-01-14 16:07:38 -0800546 numbers. The currently supported compression formats are gzip
547 (magic numbers 1F 8B or 1F 9E), bzip2 (magic number 42 5A) and LZMA
548 (magic number 5D 00). The uncompressed payload is currently always ELF
549 (magic number 7F 45 4C 46).
Ian Campbell099e1372008-02-13 20:54:58 +0000550
Ian Campbell87253d12008-02-19 11:12:30 +0000551Field name: payload_length
Ian Campbell099e1372008-02-13 20:54:58 +0000552Type: read
553Offset/size: 0x24c/4
554Protocol: 2.08+
555
Ian Campbell87253d12008-02-19 11:12:30 +0000556 The length of the payload.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700557
H. Peter Anvin4039feb2008-05-30 17:16:20 -0700558Field name: setup_data
559Type: write (special)
560Offset/size: 0x250/8
561Protocol: 2.09+
562
563 The 64-bit physical pointer to NULL terminated single linked list of
564 struct setup_data. This is used to define a more extensible boot
565 parameters passing mechanism. The definition of struct setup_data is
566 as follow:
567
568 struct setup_data {
569 u64 next;
570 u32 type;
571 u32 len;
572 u8 data[0];
573 };
574
575 Where, the next is a 64-bit physical pointer to the next node of
576 linked list, the next field of the last node is 0; the type is used
577 to identify the contents of data; the len is the length of data
578 field; the data holds the real payload.
579
580 This list may be modified at a number of points during the bootup
581 process. Therefore, when modifying this list one should always make
582 sure to consider the case where the linked list already contains
583 entries.
584
585
Ian Campbell7d6e7372008-02-17 20:06:35 +0100586**** THE IMAGE CHECKSUM
587
588From boot protocol version 2.08 onwards the CRC-32 is calculated over
589the entire file using the characteristic polynomial 0x04C11DB7 and an
590initial remainder of 0xffffffff. The checksum is appended to the
591file; therefore the CRC of the file up to the limit specified in the
592syssize field of the header is always 0.
593
H. Peter Anvin4039feb2008-05-30 17:16:20 -0700594
Linus Torvalds1da177e2005-04-16 15:20:36 -0700595**** THE KERNEL COMMAND LINE
596
597The kernel command line has become an important way for the boot
598loader to communicate with the kernel. Some of its options are also
599relevant to the boot loader itself, see "special command line options"
600below.
601
Bernhard Walle8f9aeca2007-05-02 19:27:10 +0200602The kernel command line is a null-terminated string. The maximum
603length can be retrieved from the field cmdline_size. Before protocol
604version 2.06, the maximum was 255 characters. A string that is too
605long will be automatically truncated by the kernel.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700606
607If the boot protocol version is 2.02 or later, the address of the
608kernel command line is given by the header field cmd_line_ptr (see
H. Peter Anvinf8eeaaf2005-09-06 15:17:24 -0700609above.) This address can be anywhere between the end of the setup
610heap and 0xA0000.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700611
612If the protocol version is *not* 2.02 or higher, the kernel
613command line is entered using the following protocol:
614
615 At offset 0x0020 (word), "cmd_line_magic", enter the magic
616 number 0xA33F.
617
618 At offset 0x0022 (word), "cmd_line_offset", enter the offset
619 of the kernel command line (relative to the start of the
620 real-mode kernel).
621
622 The kernel command line *must* be within the memory region
623 covered by setup_move_size, so you may need to adjust this
624 field.
625
626
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700627**** MEMORY LAYOUT OF THE REAL-MODE CODE
628
629The real-mode code requires a stack/heap to be set up, as well as
630memory allocated for the kernel command line. This needs to be done
631in the real-mode accessible memory in bottom megabyte.
632
633It should be noted that modern machines often have a sizable Extended
634BIOS Data Area (EBDA). As a result, it is advisable to use as little
635of the low megabyte as possible.
636
637Unfortunately, under the following circumstances the 0x90000 memory
638segment has to be used:
639
640 - When loading a zImage kernel ((loadflags & 0x01) == 0).
641 - When loading a 2.01 or earlier boot protocol kernel.
642
643 -> For the 2.00 and 2.01 boot protocols, the real-mode code
644 can be loaded at another address, but it is internally
645 relocated to 0x90000. For the "old" protocol, the
646 real-mode code must be loaded at 0x90000.
647
648When loading at 0x90000, avoid using memory above 0x9a000.
649
650For boot protocol 2.02 or higher, the command line does not have to be
651located in the same 64K segment as the real-mode setup code; it is
652thus permitted to give the stack/heap the full 64K segment and locate
653the command line above it.
654
655The kernel command line should not be located below the real-mode
656code, nor should it be located in high memory.
657
658
Linus Torvalds1da177e2005-04-16 15:20:36 -0700659**** SAMPLE BOOT CONFIGURATION
660
661As a sample configuration, assume the following layout of the real
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700662mode segment:
Linus Torvalds1da177e2005-04-16 15:20:36 -0700663
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700664 When loading below 0x90000, use the entire segment:
665
666 0x0000-0x7fff Real mode kernel
667 0x8000-0xdfff Stack and heap
668 0xe000-0xffff Kernel command line
669
670 When loading at 0x90000 OR the protocol version is 2.01 or earlier:
671
672 0x0000-0x7fff Real mode kernel
673 0x8000-0x97ff Stack and heap
674 0x9800-0x9fff Kernel command line
Linus Torvalds1da177e2005-04-16 15:20:36 -0700675
676Such a boot loader should enter the following fields in the header:
677
678 unsigned long base_ptr; /* base address for real-mode segment */
679
680 if ( setup_sects == 0 ) {
681 setup_sects = 4;
682 }
683
684 if ( protocol >= 0x0200 ) {
685 type_of_loader = <type code>;
686 if ( loading_initrd ) {
687 ramdisk_image = <initrd_address>;
688 ramdisk_size = <initrd_size>;
689 }
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700690
691 if ( protocol >= 0x0202 && loadflags & 0x01 )
692 heap_end = 0xe000;
693 else
694 heap_end = 0x9800;
695
Linus Torvalds1da177e2005-04-16 15:20:36 -0700696 if ( protocol >= 0x0201 ) {
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700697 heap_end_ptr = heap_end - 0x200;
Linus Torvalds1da177e2005-04-16 15:20:36 -0700698 loadflags |= 0x80; /* CAN_USE_HEAP */
699 }
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700700
Linus Torvalds1da177e2005-04-16 15:20:36 -0700701 if ( protocol >= 0x0202 ) {
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700702 cmd_line_ptr = base_ptr + heap_end;
703 strcpy(cmd_line_ptr, cmdline);
Linus Torvalds1da177e2005-04-16 15:20:36 -0700704 } else {
705 cmd_line_magic = 0xA33F;
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700706 cmd_line_offset = heap_end;
707 setup_move_size = heap_end + strlen(cmdline)+1;
708 strcpy(base_ptr+cmd_line_offset, cmdline);
Linus Torvalds1da177e2005-04-16 15:20:36 -0700709 }
710 } else {
711 /* Very old kernel */
712
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700713 heap_end = 0x9800;
714
Linus Torvalds1da177e2005-04-16 15:20:36 -0700715 cmd_line_magic = 0xA33F;
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700716 cmd_line_offset = heap_end;
Linus Torvalds1da177e2005-04-16 15:20:36 -0700717
718 /* A very old kernel MUST have its real-mode code
719 loaded at 0x90000 */
720
721 if ( base_ptr != 0x90000 ) {
722 /* Copy the real-mode kernel */
723 memcpy(0x90000, base_ptr, (setup_sects+1)*512);
Linus Torvalds1da177e2005-04-16 15:20:36 -0700724 base_ptr = 0x90000; /* Relocated */
725 }
726
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700727 strcpy(0x90000+cmd_line_offset, cmdline);
728
Linus Torvalds1da177e2005-04-16 15:20:36 -0700729 /* It is recommended to clear memory up to the 32K mark */
730 memset(0x90000 + (setup_sects+1)*512, 0,
731 (64-(setup_sects+1))*512);
732 }
733
734
735**** LOADING THE REST OF THE KERNEL
736
H. Peter Anvinf8eeaaf2005-09-06 15:17:24 -0700737The 32-bit (non-real-mode) kernel starts at offset (setup_sects+1)*512
738in the kernel file (again, if setup_sects == 0 the real value is 4.)
739It should be loaded at address 0x10000 for Image/zImage kernels and
Linus Torvalds1da177e2005-04-16 15:20:36 -07007400x100000 for bzImage kernels.
741
742The kernel is a bzImage kernel if the protocol >= 2.00 and the 0x01
743bit (LOAD_HIGH) in the loadflags field is set:
744
745 is_bzImage = (protocol >= 0x0200) && (loadflags & 0x01);
746 load_address = is_bzImage ? 0x100000 : 0x10000;
747
748Note that Image/zImage kernels can be up to 512K in size, and thus use
749the entire 0x10000-0x90000 range of memory. This means it is pretty
750much a requirement for these kernels to load the real-mode part at
7510x90000. bzImage kernels allow much more flexibility.
752
753
754**** SPECIAL COMMAND LINE OPTIONS
755
756If the command line provided by the boot loader is entered by the
757user, the user may expect the following command line options to work.
758They should normally not be deleted from the kernel command line even
759though not all of them are actually meaningful to the kernel. Boot
760loader authors who need additional command line options for the boot
761loader itself should get them registered in
762Documentation/kernel-parameters.txt to make sure they will not
763conflict with actual kernel options now or in the future.
764
765 vga=<mode>
766 <mode> here is either an integer (in C notation, either
767 decimal, octal, or hexadecimal) or one of the strings
768 "normal" (meaning 0xFFFF), "ext" (meaning 0xFFFE) or "ask"
769 (meaning 0xFFFD). This value should be entered into the
770 vid_mode field, as it is used by the kernel before the command
771 line is parsed.
772
773 mem=<size>
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700774 <size> is an integer in C notation optionally followed by
775 (case insensitive) K, M, G, T, P or E (meaning << 10, << 20,
776 << 30, << 40, << 50 or << 60). This specifies the end of
777 memory to the kernel. This affects the possible placement of
778 an initrd, since an initrd should be placed near end of
Linus Torvalds1da177e2005-04-16 15:20:36 -0700779 memory. Note that this is an option to *both* the kernel and
780 the bootloader!
781
782 initrd=<file>
783 An initrd should be loaded. The meaning of <file> is
784 obviously bootloader-dependent, and some boot loaders
785 (e.g. LILO) do not have such a command.
786
787In addition, some boot loaders add the following options to the
788user-specified command line:
789
790 BOOT_IMAGE=<file>
791 The boot image which was loaded. Again, the meaning of <file>
792 is obviously bootloader-dependent.
793
794 auto
795 The kernel was booted without explicit user intervention.
796
797If these options are added by the boot loader, it is highly
798recommended that they are located *first*, before the user-specified
799or configuration-specified command line. Otherwise, "init=/bin/sh"
800gets confused by the "auto" option.
801
802
803**** RUNNING THE KERNEL
804
805The kernel is started by jumping to the kernel entry point, which is
806located at *segment* offset 0x20 from the start of the real mode
807kernel. This means that if you loaded your real-mode kernel code at
8080x90000, the kernel entry point is 9020:0000.
809
810At entry, ds = es = ss should point to the start of the real-mode
811kernel code (0x9000 if the code is loaded at 0x90000), sp should be
812set up properly, normally pointing to the top of the heap, and
813interrupts should be disabled. Furthermore, to guard against bugs in
814the kernel, it is recommended that the boot loader sets fs = gs = ds =
815es = ss.
816
817In our example from above, we would do:
818
819 /* Note: in the case of the "old" kernel protocol, base_ptr must
820 be == 0x90000 at this point; see the previous sample code */
821
822 seg = base_ptr >> 4;
823
824 cli(); /* Enter with interrupts disabled! */
825
826 /* Set up the real-mode kernel stack */
827 _SS = seg;
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700828 _SP = heap_end;
Linus Torvalds1da177e2005-04-16 15:20:36 -0700829
830 _DS = _ES = _FS = _GS = seg;
831 jmp_far(seg+0x20, 0); /* Run the kernel */
832
833If your boot sector accesses a floppy drive, it is recommended to
834switch off the floppy motor before running the kernel, since the
835kernel boot leaves interrupts off and thus the motor will not be
836switched off, especially if the loaded kernel has the floppy driver as
837a demand-loaded module!
838
839
H. Peter Anvindb2668f2007-05-23 16:59:27 -0700840**** ADVANCED BOOT LOADER HOOKS
Linus Torvalds1da177e2005-04-16 15:20:36 -0700841
842If the boot loader runs in a particularly hostile environment (such as
843LOADLIN, which runs under DOS) it may be impossible to follow the
844standard memory location requirements. Such a boot loader may use the
845following hooks that, if set, are invoked by the kernel at the
846appropriate time. The use of these hooks should probably be
847considered an absolutely last resort!
848
849IMPORTANT: All the hooks are required to preserve %esp, %ebp, %esi and
850%edi across invocation.
851
852 realmode_swtch:
853 A 16-bit real mode far subroutine invoked immediately before
854 entering protected mode. The default routine disables NMI, so
855 your routine should probably do so, too.
856
857 code32_start:
858 A 32-bit flat-mode routine *jumped* to immediately after the
859 transition to protected mode, but before the kernel is
H. Peter Anvinde372ec2007-05-08 20:37:02 -0700860 uncompressed. No segments, except CS, are guaranteed to be
861 set up (current kernels do, but older ones do not); you should
862 set them up to BOOT_DS (0x18) yourself.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700863
864 After completing your hook, you should jump to the address
H. Peter Anvindb2668f2007-05-23 16:59:27 -0700865 that was in this field before your boot loader overwrote it
866 (relocated, if appropriate.)
Huang, Yingaa694322007-10-24 10:18:49 +0800867
868
869**** 32-bit BOOT PROTOCOL
870
871For machine with some new BIOS other than legacy BIOS, such as EFI,
872LinuxBIOS, etc, and kexec, the 16-bit real mode setup code in kernel
873based on legacy BIOS can not be used, so a 32-bit boot protocol needs
874to be defined.
875
876In 32-bit boot protocol, the first step in loading a Linux kernel
877should be to setup the boot parameters (struct boot_params,
878traditionally known as "zero page"). The memory for struct boot_params
879should be allocated and initialized to all zero. Then the setup header
880from offset 0x01f1 of kernel image on should be loaded into struct
881boot_params and examined. The end of setup header can be calculated as
882follow:
883
884 0x0202 + byte value at offset 0x0201
885
886In addition to read/modify/write the setup header of the struct
887boot_params as that of 16-bit boot protocol, the boot loader should
888also fill the additional fields of the struct boot_params as that
889described in zero-page.txt.
890
891After setupping the struct boot_params, the boot loader can load the
89232/64-bit kernel in the same way as that of 16-bit boot protocol.
893
894In 32-bit boot protocol, the kernel is started by jumping to the
89532-bit kernel entry point, which is the start address of loaded
89632/64-bit kernel.
897
898At entry, the CPU must be in 32-bit protected mode with paging
899disabled; a GDT must be loaded with the descriptors for selectors
900__BOOT_CS(0x10) and __BOOT_DS(0x18); both descriptors must be 4G flat
901segment; __BOOS_CS must have execute/read permission, and __BOOT_DS
902must have read/write permission; CS must be __BOOT_CS and DS, ES, SS
903must be __BOOT_DS; interrupt must be disabled; %esi must hold the base
904address of the struct boot_params; %ebp, %edi and %ebx must be zero.