Documentation/x86/x86_64/mm.txt

   1
   2 Virtual memory map with 4 level page tables:
   3
   4 0000000000000000 - 00007fffffffffff (=47 bits) user space, different per mm
   5 hole caused by [47:63] sign extension
   6 ffff800000000000 - ffff87ffffffffff (=43 bits) guard hole, reserved for hypervisor
   7 ffff880000000000 - ffffc7ffffffffff (=64 TB) direct mapping of all phys. memory
   8 ffffc80000000000 - ffffc8ffffffffff (=40 bits) hole
   9 ffffc90000000000 - ffffe8ffffffffff (=45 bits) vmalloc/ioremap space
  10 ffffe90000000000 - ffffe9ffffffffff (=40 bits) hole
  11 ffffea0000000000 - ffffeaffffffffff (=40 bits) virtual memory map (1TB)
  12 ... unused hole ...
  13 ffffec0000000000 - fffffbffffffffff (=44 bits) kasan shadow memory (16TB)
  14 ... unused hole ...
  15 ffffff0000000000 - ffffff7fffffffff (=39 bits) %esp fixup stacks
  16 ... unused hole ...
  17 ffffffef00000000 - fffffffeffffffff (=64 GB) EFI region mapping space
  18 ... unused hole ...
  19 ffffffff80000000 - ffffffff9fffffff (=512 MB)  kernel text mapping, from phys 0
  20 ffffffffa0000000 - [fixmap start]   (~1526 MB) module mapping space (variable)
  21 [fixmap start]   - ffffffffff5fffff kernel-internal fixmap range
  22 ffffffffff600000 - ffffffffff600fff (=4 kB) legacy vsyscall ABI
  23 ffffffffffe00000 - ffffffffffffffff (=2 MB) unused hole
  24
  25 Virtual memory map with 5 level page tables:
  26
  27 0000000000000000 - 00ffffffffffffff (=56 bits) user space, different per mm
  28 hole caused by [56:63] sign extension
  29 ff00000000000000 - ff0fffffffffffff (=52 bits) guard hole, reserved for hypervisor
  30 ff10000000000000 - ff8fffffffffffff (=55 bits) direct mapping of all phys. memory
  31 ff90000000000000 - ff91ffffffffffff (=49 bits) hole
  32 ff92000000000000 - ffd1ffffffffffff (=54 bits) vmalloc/ioremap space
  33 ffd2000000000000 - ffd3ffffffffffff (=49 bits) hole
  34 ffd4000000000000 - ffd5ffffffffffff (=49 bits) virtual memory map (512TB)
  35 ... unused hole ...
  36 ffdf000000000000 - fffffc0000000000 (=53 bits) kasan shadow memory (8PB)
  37 ... unused hole ...
  38 ffffff0000000000 - ffffff7fffffffff (=39 bits) %esp fixup stacks
  39 ... unused hole ...
  40 ffffffef00000000 - fffffffeffffffff (=64 GB) EFI region mapping space
  41 ... unused hole ...
  42 ffffffff80000000 - ffffffff9fffffff (=512 MB)  kernel text mapping, from phys 0
  43 ffffffffa0000000 - [fixmap start]   (~1526 MB) module mapping space
  44 [fixmap start]   - ffffffffff5fffff kernel-internal fixmap range
  45 ffffffffff600000 - ffffffffff600fff (=4 kB) legacy vsyscall ABI
  46 ffffffffffe00000 - ffffffffffffffff (=2 MB) unused hole
  47
  48 Architecture defines a 64-bit virtual address. Implementations can support
  49 less. Currently supported are 48- and 57-bit virtual addresses. Bits 63
  50 through to the most-significant implemented bit are sign extended.
  51 This causes hole between user space and kernel addresses if you interpret them
  52 as unsigned.
  53
  54 The direct mapping covers all memory in the system up to the highest
  55 memory address (this means in some cases it can also include PCI memory
  56 holes).
  57
  58 vmalloc space is lazily synchronized into the different PML4/PML5 pages of
  59 the processes using the page fault handler, with init_top_pgt as
  60 reference.
  61
  62 We map EFI runtime services in the 'efi_pgd' PGD in a 64Gb large virtual
  63 memory window (this size is arbitrary, it can be raised later if needed).
  64 The mappings are not part of any other kernel PGD and are only available
  65 during EFI runtime calls.
  66
  67 The module mapping space size changes based on the CONFIG requirements for the
  68 following fixmap section.
  69
  70 Note that if CONFIG_RANDOMIZE_MEMORY is enabled, the direct mapping of all
  71 physical memory, vmalloc/ioremap space and virtual memory map are randomized.
  72 Their order is preserved but their base will be offset early at boot time.