1.1. Memory Layout¶

The traditional memory map for the kernel loader, used for Image or zImage kernels, typically looks like:

        |                        |
0A0000  +------------------------+
        |  Reserved for BIOS     |      Do not use.  Reserved for BIOS EBDA.
09A000  +------------------------+
        |  Command line          |
        |  Stack/heap            |      For use by the kernel real-mode code.
098000  +------------------------+
        |  Kernel setup          |      The kernel real-mode code.
090200  +------------------------+
        |  Kernel boot sector    |      The kernel legacy boot sector.
090000  +------------------------+
        |  Protected-mode kernel |      The bulk of the kernel image.
010000  +------------------------+
        |  Boot loader           |      <- Boot sector entry point 0000:7C00
001000  +------------------------+
        |  Reserved for MBR/BIOS |
000800  +------------------------+
        |  Typically used by MBR |
000600  +------------------------+
        |  BIOS use only         |
000000  +------------------------+

When using bzImage, the protected-mode kernel was relocated to 0x100000 (“high memory”), and the kernel real-mode block (boot sector, setup, and stack/heap) was made relocatable to any address between 0x10000 and end of low memory. Unfortunately, in protocols 2.00 and 2.01 the 0x90000+ memory range is still used internally by the kernel; the 2.02 protocol resolves that problem.

It is desirable to keep the “memory ceiling” – the highest point in low memory touched by the boot loader – as low as possible, since some newer BIOSes have begun to allocate some rather large amounts of memory, called the Extended BIOS Data Area, near the top of low memory. The boot loader should use the “INT 12h” BIOS call to verify how much low memory is available.

Unfortunately, if INT 12h reports that the amount of memory is too low, there is usually nothing the boot loader can do but to report an error to the user. The boot loader should therefore be designed to take up as little space in low memory as it reasonably can. For zImage or old bzImage kernels, which need data written into the 0x90000 segment, the boot loader should make sure not to use memory above the 0x9A000 point; too many BIOSes will break above that point.

For a modern bzImage kernel with boot protocol version >= 2.02, a memory layout like the following is suggested:

              ~                        ~
              |  Protected-mode kernel |
      100000  +------------------------+
              |  I/O memory hole       |
      0A0000  +------------------------+
              |  Reserved for BIOS     |      Leave as much as possible unused
              ~                        ~
              |  Command line          |      (Can also be below the X+10000 mark)
      X+10000 +------------------------+
              |  Stack/heap            |      For use by the kernel real-mode code.
      X+08000 +------------------------+
              |  Kernel setup          |      The kernel real-mode code.
              |  Kernel boot sector    |      The kernel legacy boot sector.
      X       +------------------------+
              |  Boot loader           |      <- Boot sector entry point 0000:7C00
      001000  +------------------------+
              |  Reserved for MBR/BIOS |
      000800  +------------------------+
              |  Typically used by MBR |
      000600  +------------------------+
              |  BIOS use only         |
      000000  +------------------------+

... where the address X is as low as the design of the boot loader permits.

1.4. The kernel_info¶

The relationships between the headers are analogous to the various data sections:

setup_header = .data boot_params/setup_data = .bss

What is missing from the above list? That’s right:

kernel_info = .rodata

We have been (ab)using .data for things that could go into .rodata or .bss for a long time, for lack of alternatives and – especially early on – inertia. Also, the BIOS stub is responsible for creating boot_params, so it isn’t available to a BIOS-based loader (setup_data is, though).

setup_header is permanently limited to 144 bytes due to the reach of the 2-byte jump field, which doubles as a length field for the structure, combined with the size of the “hole” in struct boot_params that a protected-mode loader or the BIOS stub has to copy it into. It is currently 119 bytes long, which leaves us with 25 very precious bytes. This isn’t something that can be fixed without revising the boot protocol entirely, breaking backwards compatibility.

boot_params proper is limited to 4096 bytes, but can be arbitrarily extended by adding setup_data entries. It cannot be used to communicate properties of the kernel image, because it is .bss and has no image-provided content.

kernel_info solves this by providing an extensible place for information about the kernel image. It is readonly, because the kernel cannot rely on a bootloader copying its contents anywhere, but that is OK; if it becomes necessary it can still contain data items that an enabled bootloader would be expected to copy into a setup_data chunk.

All kernel_info data should be part of this structure. Fixed size data have to be put before kernel_info_var_len_data label. Variable size data have to be put after kernel_info_var_len_data label. Each chunk of variable size data has to be prefixed with header/magic and its size, e.g.:

kernel_info:
        .ascii  "LToP"          /* Header, Linux top (structure). */
        .long   kernel_info_var_len_data - kernel_info
        .long   kernel_info_end - kernel_info
        .long   0x01234567      /* Some fixed size data for the bootloaders. */
kernel_info_var_len_data:
example_struct:                 /* Some variable size data for the bootloaders. */
        .ascii  "0123"          /* Header/Magic. */
        .long   example_struct_end - example_struct
        .ascii  "Struct"
        .long   0x89012345
example_struct_end:
example_strings:                /* Some variable size data for the bootloaders. */
        .ascii  "ABCD"          /* Header/Magic. */
        .long   example_strings_end - example_strings
        .asciz  "String_0"
        .asciz  "String_1"
example_strings_end:
kernel_info_end:

This way the kernel_info is self-contained blob.

1.5. Details of the kernel_info Fields¶

Contains the magic number “LToP” (0x506f544c).

This field contains the size of the kernel_info including kernel_info.header. It does not count kernel_info.kernel_info_var_len_data size. This field should be used by the bootloaders to detect supported fixed size fields in the kernel_info and beginning of kernel_info.kernel_info_var_len_data.

This field contains the size of the kernel_info including kernel_info.header and kernel_info.kernel_info_var_len_data.

This field contains maximal allowed type for setup_data and setup_indirect structs.

1.6. The Image Checksum¶

From boot protocol version 2.08 onwards the CRC-32 is calculated over the entire file using the characteristic polynomial 0x04C11DB7 and an initial remainder of 0xffffffff. The checksum is appended to the file; therefore the CRC of the file up to the limit specified in the syssize field of the header is always 0.

1.7. The Kernel Command Line¶

The kernel command line has become an important way for the boot loader to communicate with the kernel. Some of its options are also relevant to the boot loader itself, see “special command line options” below.

The kernel command line is a null-terminated string. The maximum length can be retrieved from the field cmdline_size. Before protocol version 2.06, the maximum was 255 characters. A string that is too long will be automatically truncated by the kernel.

If the boot protocol version is 2.02 or later, the address of the kernel command line is given by the header field cmd_line_ptr (see above.) This address can be anywhere between the end of the setup heap and 0xA0000.

If the protocol version is not 2.02 or higher, the kernel command line is entered using the following protocol:

• At offset 0x0020 (word), “cmd_line_magic”, enter the magic number 0xA33F.

• At offset 0x0022 (word), “cmd_line_offset”, enter the offset of the kernel command line (relative to the start of the real-mode kernel).

• The kernel command line must be within the memory region covered by setup_move_size, so you may need to adjust this field.

1.8. Memory Layout of The Real-Mode Code¶

The real-mode code requires a stack/heap to be set up, as well as memory allocated for the kernel command line. This needs to be done in the real-mode accessible memory in bottom megabyte.

It should be noted that modern machines often have a sizable Extended BIOS Data Area (EBDA). As a result, it is advisable to use as little of the low megabyte as possible.

Unfortunately, under the following circumstances the 0x90000 memory segment has to be used:

• When loading a zImage kernel ((loadflags & 0x01) == 0).

• When loading a 2.01 or earlier boot protocol kernel.

When loading at 0x90000, avoid using memory above 0x9a000.

For boot protocol 2.02 or higher, the command line does not have to be located in the same 64K segment as the real-mode setup code; it is thus permitted to give the stack/heap the full 64K segment and locate the command line above it.

The kernel command line should not be located below the real-mode code, nor should it be located in high memory.

1.9. Sample Boot Configuartion¶

As a sample configuration, assume the following layout of the real mode segment.

When loading below 0x90000, use the entire segment:

0x0000-0x7fff	Real mode kernel
0x8000-0xdfff	Stack and heap
0xe000-0xffff	Kernel command line

When loading at 0x90000 OR the protocol version is 2.01 or earlier:

0x0000-0x7fff	Real mode kernel
0x8000-0x97ff	Stack and heap
0x9800-0x9fff	Kernel command line

Such a boot loader should enter the following fields in the header:

unsigned long base_ptr; /* base address for real-mode segment */

if ( setup_sects == 0 ) {
        setup_sects = 4;
}

if ( protocol >= 0x0200 ) {
        type_of_loader = <type code>;
        if ( loading_initrd ) {
                ramdisk_image = <initrd_address>;
                ramdisk_size = <initrd_size>;
        }

        if ( protocol >= 0x0202 && loadflags & 0x01 )
                heap_end = 0xe000;
        else
                heap_end = 0x9800;

        if ( protocol >= 0x0201 ) {
                heap_end_ptr = heap_end - 0x200;
                loadflags |= 0x80; /* CAN_USE_HEAP */
        }

        if ( protocol >= 0x0202 ) {
                cmd_line_ptr = base_ptr + heap_end;
                strcpy(cmd_line_ptr, cmdline);
        } else {
                cmd_line_magic  = 0xA33F;
                cmd_line_offset = heap_end;
                setup_move_size = heap_end + strlen(cmdline)+1;
                strcpy(base_ptr+cmd_line_offset, cmdline);
        }
} else {
        /* Very old kernel */

        heap_end = 0x9800;

        cmd_line_magic  = 0xA33F;
        cmd_line_offset = heap_end;

        /* A very old kernel MUST have its real-mode code
           loaded at 0x90000 */

        if ( base_ptr != 0x90000 ) {
                /* Copy the real-mode kernel */
                memcpy(0x90000, base_ptr, (setup_sects+1)*512);
                base_ptr = 0x90000;              /* Relocated */
        }

        strcpy(0x90000+cmd_line_offset, cmdline);

        /* It is recommended to clear memory up to the 32K mark */
        memset(0x90000 + (setup_sects+1)*512, 0,
               (64-(setup_sects+1))*512);
}

1.10. Loading The Rest of The Kernel¶

The 32-bit (non-real-mode) kernel starts at offset (setup_sects+1)*512 in the kernel file (again, if setup_sects == 0 the real value is 4.) It should be loaded at address 0x10000 for Image/zImage kernels and 0x100000 for bzImage kernels.

The kernel is a bzImage kernel if the protocol >= 2.00 and the 0x01 bit (LOAD_HIGH) in the loadflags field is set:

is_bzImage = (protocol >= 0x0200) && (loadflags & 0x01);
load_address = is_bzImage ? 0x100000 : 0x10000;

Note that Image/zImage kernels can be up to 512K in size, and thus use the entire 0x10000-0x90000 range of memory. This means it is pretty much a requirement for these kernels to load the real-mode part at 0x90000. bzImage kernels allow much more flexibility.

1.11. Special Command Line Options¶

If the command line provided by the boot loader is entered by the user, the user may expect the following command line options to work. They should normally not be deleted from the kernel command line even though not all of them are actually meaningful to the kernel. Boot loader authors who need additional command line options for the boot loader itself should get them registered in The kernel’s command-line parameters to make sure they will not conflict with actual kernel options now or in the future.

vga=<mode>

<mode> here is either an integer (in C notation, either decimal, octal, or hexadecimal) or one of the strings “normal” (meaning 0xFFFF), “ext” (meaning 0xFFFE) or “ask” (meaning 0xFFFD). This value should be entered into the vid_mode field, as it is used by the kernel before the command line is parsed.

mem=<size>

<size> is an integer in C notation optionally followed by (case insensitive) K, M, G, T, P or E (meaning << 10, << 20, << 30, << 40, << 50 or << 60). This specifies the end of memory to the kernel. This affects the possible placement of an initrd, since an initrd should be placed near end of memory. Note that this is an option to both the kernel and the bootloader!

initrd=<file>

An initrd should be loaded. The meaning of <file> is obviously bootloader-dependent, and some boot loaders (e.g. LILO) do not have such a command.

In addition, some boot loaders add the following options to the user-specified command line:

BOOT_IMAGE=<file>

The boot image which was loaded. Again, the meaning of <file> is obviously bootloader-dependent.

auto

The kernel was booted without explicit user intervention.

If these options are added by the boot loader, it is highly recommended that they are located first, before the user-specified or configuration-specified command line. Otherwise, “init=/bin/sh” gets confused by the “auto” option.

1.12. Running the Kernel¶

The kernel is started by jumping to the kernel entry point, which is located at segment offset 0x20 from the start of the real mode kernel. This means that if you loaded your real-mode kernel code at 0x90000, the kernel entry point is 9020:0000.

At entry, ds = es = ss should point to the start of the real-mode kernel code (0x9000 if the code is loaded at 0x90000), sp should be set up properly, normally pointing to the top of the heap, and interrupts should be disabled. Furthermore, to guard against bugs in the kernel, it is recommended that the boot loader sets fs = gs = ds = es = ss.

In our example from above, we would do:

/* Note: in the case of the "old" kernel protocol, base_ptr must
   be == 0x90000 at this point; see the previous sample code */

seg = base_ptr >> 4;

cli();  /* Enter with interrupts disabled! */

/* Set up the real-mode kernel stack */
_SS = seg;
_SP = heap_end;

_DS = _ES = _FS = _GS = seg;
jmp_far(seg+0x20, 0);   /* Run the kernel */

If your boot sector accesses a floppy drive, it is recommended to switch off the floppy motor before running the kernel, since the kernel boot leaves interrupts off and thus the motor will not be switched off, especially if the loaded kernel has the floppy driver as a demand-loaded module!

1.13. Advanced Boot Loader Hooks¶

If the boot loader runs in a particularly hostile environment (such as LOADLIN, which runs under DOS) it may be impossible to follow the standard memory location requirements. Such a boot loader may use the following hooks that, if set, are invoked by the kernel at the appropriate time. The use of these hooks should probably be considered an absolutely last resort!

IMPORTANT: All the hooks are required to preserve %esp, %ebp, %esi and %edi across invocation.

realmode_swtch:

A 16-bit real mode far subroutine invoked immediately before entering protected mode. The default routine disables NMI, so your routine should probably do so, too.

code32_start:

A 32-bit flat-mode routine jumped to immediately after the transition to protected mode, but before the kernel is uncompressed. No segments, except CS, are guaranteed to be set up (current kernels do, but older ones do not); you should set them up to BOOT_DS (0x18) yourself.

After completing your hook, you should jump to the address that was in this field before your boot loader overwrote it (relocated, if appropriate.)

1.14. 32-bit Boot Protocol¶

For machine with some new BIOS other than legacy BIOS, such as EFI, LinuxBIOS, etc, and kexec, the 16-bit real mode setup code in kernel based on legacy BIOS can not be used, so a 32-bit boot protocol needs to be defined.

In 32-bit boot protocol, the first step in loading a Linux kernel should be to setup the boot parameters (struct boot_params, traditionally known as “zero page”). The memory for struct boot_params should be allocated and initialized to all zero. Then the setup header from offset 0x01f1 of kernel image on should be loaded into struct boot_params and examined. The end of setup header can be calculated as follow:

0x0202 + byte value at offset 0x0201

In addition to read/modify/write the setup header of the struct boot_params as that of 16-bit boot protocol, the boot loader should also fill the additional fields of the struct boot_params as described in chapter Zero Page.

After setting up the struct boot_params, the boot loader can load the 32/64-bit kernel in the same way as that of 16-bit boot protocol.

In 32-bit boot protocol, the kernel is started by jumping to the 32-bit kernel entry point, which is the start address of loaded 32/64-bit kernel.

At entry, the CPU must be in 32-bit protected mode with paging disabled; a GDT must be loaded with the descriptors for selectors __BOOT_CS(0x10) and __BOOT_DS(0x18); both descriptors must be 4G flat segment; __BOOT_CS must have execute/read permission, and __BOOT_DS must have read/write permission; CS must be __BOOT_CS and DS, ES, SS must be __BOOT_DS; interrupt must be disabled; %esi must hold the base address of the struct boot_params; %ebp, %edi and %ebx must be zero.

1.15. 64-bit Boot Protocol¶

For machine with 64bit cpus and 64bit kernel, we could use 64bit bootloader and we need a 64-bit boot protocol.

In 64-bit boot protocol, the first step in loading a Linux kernel should be to setup the boot parameters (struct boot_params, traditionally known as “zero page”). The memory for struct boot_params could be allocated anywhere (even above 4G) and initialized to all zero. Then, the setup header at offset 0x01f1 of kernel image on should be loaded into struct boot_params and examined. The end of setup header can be calculated as follows:

0x0202 + byte value at offset 0x0201

In addition to read/modify/write the setup header of the struct boot_params as that of 16-bit boot protocol, the boot loader should also fill the additional fields of the struct boot_params as described in chapter Zero Page.

After setting up the struct boot_params, the boot loader can load 64-bit kernel in the same way as that of 16-bit boot protocol, but kernel could be loaded above 4G.

In 64-bit boot protocol, the kernel is started by jumping to the 64-bit kernel entry point, which is the start address of loaded 64-bit kernel plus 0x200.

At entry, the CPU must be in 64-bit mode with paging enabled. The range with setup_header.init_size from start address of loaded kernel and zero page and command line buffer get ident mapping; a GDT must be loaded with the descriptors for selectors __BOOT_CS(0x10) and __BOOT_DS(0x18); both descriptors must be 4G flat segment; __BOOT_CS must have execute/read permission, and __BOOT_DS must have read/write permission; CS must be __BOOT_CS and DS, ES, SS must be __BOOT_DS; interrupt must be disabled; %rsi must hold the base address of the struct boot_params.

1.16. EFI Handover Protocol (deprecated)¶

This protocol allows boot loaders to defer initialisation to the EFI boot stub. The boot loader is required to load the kernel/initrd(s) from the boot media and jump to the EFI handover protocol entry point which is hdr->handover_offset bytes from the beginning of startup_{32,64}.

The boot loader MUST respect the kernel’s PE/COFF metadata when it comes to section alignment, the memory footprint of the executable image beyond the size of the file itself, and any other aspect of the PE/COFF header that may affect correct operation of the image as a PE/COFF binary in the execution context provided by the EFI firmware.

The function prototype for the handover entry point looks like this:

efi_main(void *handle, efi_system_table_t *table, struct boot_params *bp)

‘handle’ is the EFI image handle passed to the boot loader by the EFI firmware, ‘table’ is the EFI system table - these are the first two arguments of the “handoff state” as described in section 2.3 of the UEFI specification. ‘bp’ is the boot loader-allocated boot params.

The boot loader must fill out the following fields in bp:

- hdr.cmd_line_ptr
- hdr.ramdisk_image (if applicable)
- hdr.ramdisk_size  (if applicable)

All other fields should be zero.

NOTE: The EFI Handover Protocol is deprecated in favour of the ordinary PE/COFF

entry point, combined with the LINUX_EFI_INITRD_MEDIA_GUID based initrd loading protocol (refer to [0] for an example of the bootloader side of this), which removes the need for any knowledge on the part of the EFI bootloader regarding the internal representation of boot_params or any requirements/limitations regarding the placement of the command line and ramdisk in memory, or the placement of the kernel image itself.

[0] https://github.com/u-boot/u-boot/commit/ec80b4735a593961fe701cc3a5d717d4739b0fd0