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 fffffe8000000000 - fffffeffffffffff (=39 bits) cpu_entry_area mapping
  16 ffffff0000000000 - ffffff7fffffffff (=39 bits) %esp fixup stacks
  17 ... unused hole ...
  18 ffffffef00000000 - fffffffeffffffff (=64 GB) EFI region mapping space
  19 ... unused hole ...
  20 ffffffff80000000 - ffffffff9fffffff (=512 MB)  kernel text mapping, from phys 0
  21 ffffffffa0000000 - [fixmap start]   (~1526 MB) module mapping space (variable)
  22 [fixmap start]   - ffffffffff5fffff kernel-internal fixmap range
  23 ffffffffff600000 - ffffffffff600fff (=4 kB) legacy vsyscall ABI
  24 ffffffffffe00000 - ffffffffffffffff (=2 MB) unused hole
  25
  26 Virtual memory map with 5 level page tables:
  27
  28 0000000000000000 - 00ffffffffffffff (=56 bits) user space, different per mm
  29 hole caused by [56:63] sign extension
  30 ff00000000000000 - ff0fffffffffffff (=52 bits) guard hole, reserved for hypervisor
  31 ff10000000000000 - ff8fffffffffffff (=55 bits) direct mapping of all phys. memory
  32 ff90000000000000 - ff9fffffffffffff (=52 bits) hole
  33 ffa0000000000000 - ffd1ffffffffffff (=54 bits) vmalloc/ioremap space (12800 TB)
  34 ffd2000000000000 - ffd3ffffffffffff (=49 bits) hole
  35 ffd4000000000000 - ffd5ffffffffffff (=49 bits) virtual memory map (512TB)
  36 ... unused hole ...
  37 ffdf000000000000 - fffffc0000000000 (=53 bits) kasan shadow memory (8PB)
  38 ... unused hole ...
  39 fffffe8000000000 - fffffeffffffffff (=39 bits) cpu_entry_area mapping
  40 ffffff0000000000 - ffffff7fffffffff (=39 bits) %esp fixup stacks
  41 ... unused hole ...
  42 ffffffef00000000 - fffffffeffffffff (=64 GB) EFI region mapping space
  43 ... unused hole ...
  44 ffffffff80000000 - ffffffff9fffffff (=512 MB)  kernel text mapping, from phys 0
  45 ffffffffa0000000 - [fixmap start]   (~1526 MB) module mapping space
  46 [fixmap start]   - ffffffffff5fffff kernel-internal fixmap range
  47 ffffffffff600000 - ffffffffff600fff (=4 kB) legacy vsyscall ABI
  48 ffffffffffe00000 - ffffffffffffffff (=2 MB) unused hole
  49
  50 Architecture defines a 64-bit virtual address. Implementations can support
  51 less. Currently supported are 48- and 57-bit virtual addresses. Bits 63
  52 through to the most-significant implemented bit are sign extended.
  53 This causes hole between user space and kernel addresses if you interpret them
  54 as unsigned.
  55
  56 The direct mapping covers all memory in the system up to the highest
  57 memory address (this means in some cases it can also include PCI memory
  58 holes).
  59
  60 vmalloc space is lazily synchronized into the different PML4/PML5 pages of
  61 the processes using the page fault handler, with init_top_pgt as
  62 reference.
  63
  64 We map EFI runtime services in the 'efi_pgd' PGD in a 64Gb large virtual
  65 memory window (this size is arbitrary, it can be raised later if needed).
  66 The mappings are not part of any other kernel PGD and are only available
  67 during EFI runtime calls.
  68
  69 The module mapping space size changes based on the CONFIG requirements for the
  70 following fixmap section.
  71
  72 Note that if CONFIG_RANDOMIZE_MEMORY is enabled, the direct mapping of all
  73 physical memory, vmalloc/ioremap space and virtual memory map are randomized.
  74 Their order is preserved but their base will be offset early at boot time.