1 THE LINUX/I386 BOOT PROTOCOL
2 ----------------------------
4 H. Peter Anvin <hpa@zytor.com>
7 On the i386 platform, the Linux kernel uses a rather complicated boot
8 convention. This has evolved partially due to historical aspects, as
9 well as the desire in the early days to have the kernel itself be a
10 bootable image, the complicated PC memory model and due to changed
11 expectations in the PC industry caused by the effective demise of
12 real-mode DOS as a mainstream operating system.
14 Currently, the following versions of the Linux/i386 boot protocol exist.
16 Old kernels: zImage/Image support only. Some very early kernels
17 may not even support a command line.
19 Protocol 2.00: (Kernel 1.3.73) Added bzImage and initrd support, as
20 well as a formalized way to communicate between the
21 boot loader and the kernel. setup.S made relocatable,
22 although the traditional setup area still assumed
25 Protocol 2.01: (Kernel 1.3.76) Added a heap overrun warning.
27 Protocol 2.02: (Kernel 2.4.0-test3-pre3) New command line protocol.
28 Lower the conventional memory ceiling. No overwrite
29 of the traditional setup area, thus making booting
30 safe for systems which use the EBDA from SMM or 32-bit
31 BIOS entry points. zImage deprecated but still
34 Protocol 2.03: (Kernel 2.4.18-pre1) Explicitly makes the highest possible
35 initrd address available to the bootloader.
37 Protocol 2.04: (Kernel 2.6.14) Extend the syssize field to four bytes.
39 Protocol 2.05: (Kernel 2.6.20) Make protected mode kernel relocatable.
40 Introduce relocatable_kernel and kernel_alignment fields.
42 Protocol 2.06: (Kernel 2.6.22) Added a field that contains the size of
48 The traditional memory map for the kernel loader, used for Image or
49 zImage kernels, typically looks like:
52 0A0000 +------------------------+
53 | Reserved for BIOS | Do not use. Reserved for BIOS EBDA.
54 09A000 +------------------------+
56 | Stack/heap | For use by the kernel real-mode code.
57 098000 +------------------------+
58 | Kernel setup | The kernel real-mode code.
59 090200 +------------------------+
60 | Kernel boot sector | The kernel legacy boot sector.
61 090000 +------------------------+
62 | Protected-mode kernel | The bulk of the kernel image.
63 010000 +------------------------+
64 | Boot loader | <- Boot sector entry point 0000:7C00
65 001000 +------------------------+
66 | Reserved for MBR/BIOS |
67 000800 +------------------------+
68 | Typically used by MBR |
69 000600 +------------------------+
71 000000 +------------------------+
74 When using bzImage, the protected-mode kernel was relocated to
75 0x100000 ("high memory"), and the kernel real-mode block (boot sector,
76 setup, and stack/heap) was made relocatable to any address between
77 0x10000 and end of low memory. Unfortunately, in protocols 2.00 and
78 2.01 the 0x90000+ memory range is still used internally by the kernel;
79 the 2.02 protocol resolves that problem.
81 It is desirable to keep the "memory ceiling" -- the highest point in
82 low memory touched by the boot loader -- as low as possible, since
83 some newer BIOSes have begun to allocate some rather large amounts of
84 memory, called the Extended BIOS Data Area, near the top of low
85 memory. The boot loader should use the "INT 12h" BIOS call to verify
86 how much low memory is available.
88 Unfortunately, if INT 12h reports that the amount of memory is too
89 low, there is usually nothing the boot loader can do but to report an
90 error to the user. The boot loader should therefore be designed to
91 take up as little space in low memory as it reasonably can. For
92 zImage or old bzImage kernels, which need data written into the
93 0x90000 segment, the boot loader should make sure not to use memory
94 above the 0x9A000 point; too many BIOSes will break above that point.
96 For a modern bzImage kernel with boot protocol version >= 2.02, a
97 memory layout like the following is suggested:
100 | Protected-mode kernel |
101 100000 +------------------------+
103 0A0000 +------------------------+
104 | Reserved for BIOS | Leave as much as possible unused
106 | Command line | (Can also be below the X+10000 mark)
107 X+10000 +------------------------+
108 | Stack/heap | For use by the kernel real-mode code.
109 X+08000 +------------------------+
110 | Kernel setup | The kernel real-mode code.
111 | Kernel boot sector | The kernel legacy boot sector.
112 X +------------------------+
113 | Boot loader | <- Boot sector entry point 0000:7C00
114 001000 +------------------------+
115 | Reserved for MBR/BIOS |
116 000800 +------------------------+
117 | Typically used by MBR |
118 000600 +------------------------+
120 000000 +------------------------+
122 ... where the address X is as low as the design of the boot loader
126 **** THE REAL-MODE KERNEL HEADER
128 In the following text, and anywhere in the kernel boot sequence, "a
129 sector" refers to 512 bytes. It is independent of the actual sector
130 size of the underlying medium.
132 The first step in loading a Linux kernel should be to load the
133 real-mode code (boot sector and setup code) and then examine the
134 following header at offset 0x01f1. The real-mode code can total up to
135 32K, although the boot loader may choose to load only the first two
136 sectors (1K) and then examine the bootup sector size.
138 The header looks like:
140 Offset Proto Name Meaning
143 01F1/1 ALL(1 setup_sects The size of the setup in sectors
144 01F2/2 ALL root_flags If set, the root is mounted readonly
145 01F4/4 2.04+(2 syssize The size of the 32-bit code in 16-byte paras
146 01F8/2 ALL ram_size DO NOT USE - for bootsect.S use only
147 01FA/2 ALL vid_mode Video mode control
148 01FC/2 ALL root_dev Default root device number
149 01FE/2 ALL boot_flag 0xAA55 magic number
150 0200/2 2.00+ jump Jump instruction
151 0202/4 2.00+ header Magic signature "HdrS"
152 0206/2 2.00+ version Boot protocol version supported
153 0208/4 2.00+ realmode_swtch Boot loader hook (see below)
154 020C/2 2.00+ start_sys The load-low segment (0x1000) (obsolete)
155 020E/2 2.00+ kernel_version Pointer to kernel version string
156 0210/1 2.00+ type_of_loader Boot loader identifier
157 0211/1 2.00+ loadflags Boot protocol option flags
158 0212/2 2.00+ setup_move_size Move to high memory size (used with hooks)
159 0214/4 2.00+ code32_start Boot loader hook (see below)
160 0218/4 2.00+ ramdisk_image initrd load address (set by boot loader)
161 021C/4 2.00+ ramdisk_size initrd size (set by boot loader)
162 0220/4 2.00+ bootsect_kludge DO NOT USE - for bootsect.S use only
163 0224/2 2.01+ heap_end_ptr Free memory after setup end
164 0226/2 N/A pad1 Unused
165 0228/4 2.02+ cmd_line_ptr 32-bit pointer to the kernel command line
166 022C/4 2.03+ initrd_addr_max Highest legal initrd address
167 0230/4 2.05+ kernel_alignment Physical addr alignment required for kernel
168 0234/1 2.05+ relocatable_kernel Whether kernel is relocatable or not
169 0235/3 N/A pad2 Unused
170 0238/4 2.06+ cmdline_size Maximum size of the kernel command line
172 (1) For backwards compatibility, if the setup_sects field contains 0, the
175 (2) For boot protocol prior to 2.04, the upper two bytes of the syssize
176 field are unusable, which means the size of a bzImage kernel
177 cannot be determined.
179 If the "HdrS" (0x53726448) magic number is not found at offset 0x202,
180 the boot protocol version is "old". Loading an old kernel, the
181 following parameters should be assumed:
185 Real-mode kernel must be located at 0x90000.
187 Otherwise, the "version" field contains the protocol version,
188 e.g. protocol version 2.01 will contain 0x0201 in this field. When
189 setting fields in the header, you must make sure only to set fields
190 supported by the protocol version in use.
193 **** DETAILS OF HEADER FIELDS
195 For each field, some are information from the kernel to the bootloader
196 ("read"), some are expected to be filled out by the bootloader
197 ("write"), and some are expected to be read and modified by the
198 bootloader ("modify").
200 All general purpose boot loaders should write the fields marked
201 (obligatory). Boot loaders who want to load the kernel at a
202 nonstandard address should fill in the fields marked (reloc); other
203 boot loaders can ignore those fields.
205 The byte order of all fields is littleendian (this is x86, after all.)
207 Field name: setup_secs
212 The size of the setup code in 512-byte sectors. If this field is
213 0, the real value is 4. The real-mode code consists of the boot
214 sector (always one 512-byte sector) plus the setup code.
216 Field name: root_flags
217 Type: modify (optional)
221 If this field is nonzero, the root defaults to readonly. The use of
222 this field is deprecated; use the "ro" or "rw" options on the
223 command line instead.
227 Offset/size: 0x1f4/4 (protocol 2.04+) 0x1f4/2 (protocol ALL)
230 The size of the protected-mode code in units of 16-byte paragraphs.
231 For protocol versions older than 2.04 this field is only two bytes
232 wide, and therefore cannot be trusted for the size of a kernel if
233 the LOAD_HIGH flag is set.
236 Type: kernel internal
240 This field is obsolete.
243 Type: modify (obligatory)
246 Please see the section on SPECIAL COMMAND LINE OPTIONS.
249 Type: modify (optional)
253 The default root device device number. The use of this field is
254 deprecated, use the "root=" option on the command line instead.
256 Field name: boot_flag
261 Contains 0xAA55. This is the closest thing old Linux kernels have
269 Contains an x86 jump instruction, 0xEB followed by a signed offset
270 relative to byte 0x202. This can be used to determine the size of
278 Contains the magic number "HdrS" (0x53726448).
285 Contains the boot protocol version, in (major << 8)+minor format,
286 e.g. 0x0204 for version 2.04, and 0x0a11 for a hypothetical version
289 Field name: readmode_swtch
290 Type: modify (optional)
294 Boot loader hook (see ADVANCED BOOT LOADER HOOKS below.)
296 Field name: start_sys
301 The load low segment (0x1000). Obsolete.
303 Field name: kernel_version
308 If set to a nonzero value, contains a pointer to a NUL-terminated
309 human-readable kernel version number string, less 0x200. This can
310 be used to display the kernel version to the user. This value
311 should be less than (0x200*setup_sects).
313 For example, if this value is set to 0x1c00, the kernel version
314 number string can be found at offset 0x1e00 in the kernel file.
315 This is a valid value if and only if the "setup_sects" field
316 contains the value 15 or higher, as:
318 0x1c00 < 15*0x200 (= 0x1e00) but
319 0x1c00 >= 14*0x200 (= 0x1c00)
321 0x1c00 >> 9 = 14, so the minimum value for setup_secs is 15.
323 Field name: type_of_loader
324 Type: write (obligatory)
328 If your boot loader has an assigned id (see table below), enter
329 0xTV here, where T is an identifier for the boot loader and V is
330 a version number. Otherwise, enter 0xFF here.
332 Assigned boot loader ids:
333 0 LILO (0x00 reserved for pre-2.00 bootloader)
335 2 bootsect-loader (0x20, all other values reserved)
345 Please contact <hpa@zytor.com> if you need a bootloader ID
348 Field name: loadflags
349 Type: modify (obligatory)
353 This field is a bitmask.
355 Bit 0 (read): LOADED_HIGH
356 - If 0, the protected-mode code is loaded at 0x10000.
357 - If 1, the protected-mode code is loaded at 0x100000.
359 Bit 7 (write): CAN_USE_HEAP
360 Set this bit to 1 to indicate that the value entered in the
361 heap_end_ptr is valid. If this field is clear, some setup code
362 functionality will be disabled.
364 Field name: setup_move_size
365 Type: modify (obligatory)
369 When using protocol 2.00 or 2.01, if the real mode kernel is not
370 loaded at 0x90000, it gets moved there later in the loading
371 sequence. Fill in this field if you want additional data (such as
372 the kernel command line) moved in addition to the real-mode kernel
375 The unit is bytes starting with the beginning of the boot sector.
377 This field is can be ignored when the protocol is 2.02 or higher, or
378 if the real-mode code is loaded at 0x90000.
380 Field name: code32_start
381 Type: modify (optional, reloc)
385 The address to jump to in protected mode. This defaults to the load
386 address of the kernel, and can be used by the boot loader to
387 determine the proper load address.
389 This field can be modified for two purposes:
391 1. as a boot loader hook (see ADVANCED BOOT LOADER HOOKS below.)
393 2. if a bootloader which does not install a hook loads a
394 relocatable kernel at a nonstandard address it will have to modify
395 this field to point to the load address.
397 Field name: ramdisk_image
398 Type: write (obligatory)
402 The 32-bit linear address of the initial ramdisk or ramfs. Leave at
403 zero if there is no initial ramdisk/ramfs.
405 Field name: ramdisk_size
406 Type: write (obligatory)
410 Size of the initial ramdisk or ramfs. Leave at zero if there is no
411 initial ramdisk/ramfs.
413 Field name: bootsect_kludge
414 Type: kernel internal
418 This field is obsolete.
420 Field name: heap_end_ptr
421 Type: write (obligatory)
425 Set this field to the offset (from the beginning of the real-mode
426 code) of the end of the setup stack/heap, minus 0x0200.
428 Field name: cmd_line_ptr
429 Type: write (obligatory)
433 Set this field to the linear address of the kernel command line.
434 The kernel command line can be located anywhere between the end of
435 the setup heap and 0xA0000; it does not have to be located in the
436 same 64K segment as the real-mode code itself.
438 Fill in this field even if your boot loader does not support a
439 command line, in which case you can point this to an empty string
440 (or better yet, to the string "auto".) If this field is left at
441 zero, the kernel will assume that your boot loader does not support
444 Field name: initrd_addr_max
449 The maximum address that may be occupied by the initial
450 ramdisk/ramfs contents. For boot protocols 2.02 or earlier, this
451 field is not present, and the maximum address is 0x37FFFFFF. (This
452 address is defined as the address of the highest safe byte, so if
453 your ramdisk is exactly 131072 bytes long and this field is
454 0x37FFFFFF, you can start your ramdisk at 0x37FE0000.)
456 Field name: kernel_alignment
461 Alignment unit required by the kernel (if relocatable_kernel is true.)
463 Field name: relocatable_kernel
468 If this field is nonzero, the protected-mode part of the kernel can
469 be loaded at any address that satisfies the kernel_alignment field.
470 After loading, the boot loader must set the code32_start field to
471 point to the loaded code, or to a boot loader hook.
473 Field name: cmdline_size
478 The maximum size of the command line without the terminating
479 zero. This means that the command line can contain at most
480 cmdline_size characters. With protocol version 2.05 and earlier, the
481 maximum size was 255.
484 **** THE KERNEL COMMAND LINE
486 The kernel command line has become an important way for the boot
487 loader to communicate with the kernel. Some of its options are also
488 relevant to the boot loader itself, see "special command line options"
491 The kernel command line is a null-terminated string. The maximum
492 length can be retrieved from the field cmdline_size. Before protocol
493 version 2.06, the maximum was 255 characters. A string that is too
494 long will be automatically truncated by the kernel.
496 If the boot protocol version is 2.02 or later, the address of the
497 kernel command line is given by the header field cmd_line_ptr (see
498 above.) This address can be anywhere between the end of the setup
501 If the protocol version is *not* 2.02 or higher, the kernel
502 command line is entered using the following protocol:
504 At offset 0x0020 (word), "cmd_line_magic", enter the magic
507 At offset 0x0022 (word), "cmd_line_offset", enter the offset
508 of the kernel command line (relative to the start of the
511 The kernel command line *must* be within the memory region
512 covered by setup_move_size, so you may need to adjust this
516 **** MEMORY LAYOUT OF THE REAL-MODE CODE
518 The real-mode code requires a stack/heap to be set up, as well as
519 memory allocated for the kernel command line. This needs to be done
520 in the real-mode accessible memory in bottom megabyte.
522 It should be noted that modern machines often have a sizable Extended
523 BIOS Data Area (EBDA). As a result, it is advisable to use as little
524 of the low megabyte as possible.
526 Unfortunately, under the following circumstances the 0x90000 memory
527 segment has to be used:
529 - When loading a zImage kernel ((loadflags & 0x01) == 0).
530 - When loading a 2.01 or earlier boot protocol kernel.
532 -> For the 2.00 and 2.01 boot protocols, the real-mode code
533 can be loaded at another address, but it is internally
534 relocated to 0x90000. For the "old" protocol, the
535 real-mode code must be loaded at 0x90000.
537 When loading at 0x90000, avoid using memory above 0x9a000.
539 For boot protocol 2.02 or higher, the command line does not have to be
540 located in the same 64K segment as the real-mode setup code; it is
541 thus permitted to give the stack/heap the full 64K segment and locate
542 the command line above it.
544 The kernel command line should not be located below the real-mode
545 code, nor should it be located in high memory.
548 **** SAMPLE BOOT CONFIGURATION
550 As a sample configuration, assume the following layout of the real
553 When loading below 0x90000, use the entire segment:
555 0x0000-0x7fff Real mode kernel
556 0x8000-0xdfff Stack and heap
557 0xe000-0xffff Kernel command line
559 When loading at 0x90000 OR the protocol version is 2.01 or earlier:
561 0x0000-0x7fff Real mode kernel
562 0x8000-0x97ff Stack and heap
563 0x9800-0x9fff Kernel command line
565 Such a boot loader should enter the following fields in the header:
567 unsigned long base_ptr; /* base address for real-mode segment */
569 if ( setup_sects == 0 ) {
573 if ( protocol >= 0x0200 ) {
574 type_of_loader = <type code>;
575 if ( loading_initrd ) {
576 ramdisk_image = <initrd_address>;
577 ramdisk_size = <initrd_size>;
580 if ( protocol >= 0x0202 && loadflags & 0x01 )
585 if ( protocol >= 0x0201 ) {
586 heap_end_ptr = heap_end - 0x200;
587 loadflags |= 0x80; /* CAN_USE_HEAP */
590 if ( protocol >= 0x0202 ) {
591 cmd_line_ptr = base_ptr + heap_end;
592 strcpy(cmd_line_ptr, cmdline);
594 cmd_line_magic = 0xA33F;
595 cmd_line_offset = heap_end;
596 setup_move_size = heap_end + strlen(cmdline)+1;
597 strcpy(base_ptr+cmd_line_offset, cmdline);
600 /* Very old kernel */
604 cmd_line_magic = 0xA33F;
605 cmd_line_offset = heap_end;
607 /* A very old kernel MUST have its real-mode code
610 if ( base_ptr != 0x90000 ) {
611 /* Copy the real-mode kernel */
612 memcpy(0x90000, base_ptr, (setup_sects+1)*512);
613 base_ptr = 0x90000; /* Relocated */
616 strcpy(0x90000+cmd_line_offset, cmdline);
618 /* It is recommended to clear memory up to the 32K mark */
619 memset(0x90000 + (setup_sects+1)*512, 0,
620 (64-(setup_sects+1))*512);
624 **** LOADING THE REST OF THE KERNEL
626 The 32-bit (non-real-mode) kernel starts at offset (setup_sects+1)*512
627 in the kernel file (again, if setup_sects == 0 the real value is 4.)
628 It should be loaded at address 0x10000 for Image/zImage kernels and
629 0x100000 for bzImage kernels.
631 The kernel is a bzImage kernel if the protocol >= 2.00 and the 0x01
632 bit (LOAD_HIGH) in the loadflags field is set:
634 is_bzImage = (protocol >= 0x0200) && (loadflags & 0x01);
635 load_address = is_bzImage ? 0x100000 : 0x10000;
637 Note that Image/zImage kernels can be up to 512K in size, and thus use
638 the entire 0x10000-0x90000 range of memory. This means it is pretty
639 much a requirement for these kernels to load the real-mode part at
640 0x90000. bzImage kernels allow much more flexibility.
643 **** SPECIAL COMMAND LINE OPTIONS
645 If the command line provided by the boot loader is entered by the
646 user, the user may expect the following command line options to work.
647 They should normally not be deleted from the kernel command line even
648 though not all of them are actually meaningful to the kernel. Boot
649 loader authors who need additional command line options for the boot
650 loader itself should get them registered in
651 Documentation/kernel-parameters.txt to make sure they will not
652 conflict with actual kernel options now or in the future.
655 <mode> here is either an integer (in C notation, either
656 decimal, octal, or hexadecimal) or one of the strings
657 "normal" (meaning 0xFFFF), "ext" (meaning 0xFFFE) or "ask"
658 (meaning 0xFFFD). This value should be entered into the
659 vid_mode field, as it is used by the kernel before the command
663 <size> is an integer in C notation optionally followed by
664 (case insensitive) K, M, G, T, P or E (meaning << 10, << 20,
665 << 30, << 40, << 50 or << 60). This specifies the end of
666 memory to the kernel. This affects the possible placement of
667 an initrd, since an initrd should be placed near end of
668 memory. Note that this is an option to *both* the kernel and
672 An initrd should be loaded. The meaning of <file> is
673 obviously bootloader-dependent, and some boot loaders
674 (e.g. LILO) do not have such a command.
676 In addition, some boot loaders add the following options to the
677 user-specified command line:
680 The boot image which was loaded. Again, the meaning of <file>
681 is obviously bootloader-dependent.
684 The kernel was booted without explicit user intervention.
686 If these options are added by the boot loader, it is highly
687 recommended that they are located *first*, before the user-specified
688 or configuration-specified command line. Otherwise, "init=/bin/sh"
689 gets confused by the "auto" option.
692 **** RUNNING THE KERNEL
694 The kernel is started by jumping to the kernel entry point, which is
695 located at *segment* offset 0x20 from the start of the real mode
696 kernel. This means that if you loaded your real-mode kernel code at
697 0x90000, the kernel entry point is 9020:0000.
699 At entry, ds = es = ss should point to the start of the real-mode
700 kernel code (0x9000 if the code is loaded at 0x90000), sp should be
701 set up properly, normally pointing to the top of the heap, and
702 interrupts should be disabled. Furthermore, to guard against bugs in
703 the kernel, it is recommended that the boot loader sets fs = gs = ds =
706 In our example from above, we would do:
708 /* Note: in the case of the "old" kernel protocol, base_ptr must
709 be == 0x90000 at this point; see the previous sample code */
713 cli(); /* Enter with interrupts disabled! */
715 /* Set up the real-mode kernel stack */
719 _DS = _ES = _FS = _GS = seg;
720 jmp_far(seg+0x20, 0); /* Run the kernel */
722 If your boot sector accesses a floppy drive, it is recommended to
723 switch off the floppy motor before running the kernel, since the
724 kernel boot leaves interrupts off and thus the motor will not be
725 switched off, especially if the loaded kernel has the floppy driver as
726 a demand-loaded module!
729 **** ADVANCED BOOT LOADER HOOKS
731 If the boot loader runs in a particularly hostile environment (such as
732 LOADLIN, which runs under DOS) it may be impossible to follow the
733 standard memory location requirements. Such a boot loader may use the
734 following hooks that, if set, are invoked by the kernel at the
735 appropriate time. The use of these hooks should probably be
736 considered an absolutely last resort!
738 IMPORTANT: All the hooks are required to preserve %esp, %ebp, %esi and
739 %edi across invocation.
742 A 16-bit real mode far subroutine invoked immediately before
743 entering protected mode. The default routine disables NMI, so
744 your routine should probably do so, too.
747 A 32-bit flat-mode routine *jumped* to immediately after the
748 transition to protected mode, but before the kernel is
749 uncompressed. No segments, except CS, are guaranteed to be
750 set up (current kernels do, but older ones do not); you should
751 set them up to BOOT_DS (0x18) yourself.
753 After completing your hook, you should jump to the address
754 that was in this field before your boot loader overwrote it
755 (relocated, if appropriate.)