| 1 |
=============== |
| 2 |
Soft-Dirty PTEs |
| 3 |
=============== |
| 4 |
|
| 5 |
The soft-dirty is a bit on a PTE which helps to track which pages a task |
| 6 |
writes to. In order to do this tracking one should |
| 7 |
|
| 8 |
1. Clear soft-dirty bits from the task's PTEs. |
| 9 |
|
| 10 |
This is done by writing "4" into the ``/proc/PID/clear_refs`` file of the |
| 11 |
task in question. |
| 12 |
|
| 13 |
2. Wait some time. |
| 14 |
|
| 15 |
3. Read soft-dirty bits from the PTEs. |
| 16 |
|
| 17 |
This is done by reading from the ``/proc/PID/pagemap``. The bit 55 of the |
| 18 |
64-bit qword is the soft-dirty one. If set, the respective PTE was |
| 19 |
written to since step 1. |
| 20 |
|
| 21 |
|
| 22 |
Internally, to do this tracking, the writable bit is cleared from PTEs |
| 23 |
when the soft-dirty bit is cleared. So, after this, when the task tries to |
| 24 |
modify a page at some virtual address the #PF occurs and the kernel sets |
| 25 |
the soft-dirty bit on the respective PTE. |
| 26 |
|
| 27 |
Note, that although all the task's address space is marked as r/o after the |
| 28 |
soft-dirty bits clear, the #PF-s that occur after that are processed fast. |
| 29 |
This is so, since the pages are still mapped to physical memory, and thus all |
| 30 |
the kernel does is finds this fact out and puts both writable and soft-dirty |
| 31 |
bits on the PTE. |
| 32 |
|
| 33 |
While in most cases tracking memory changes by #PF-s is more than enough |
| 34 |
there is still a scenario when we can lose soft dirty bits -- a task |
| 35 |
unmaps a previously mapped memory region and then maps a new one at exactly |
| 36 |
the same place. When unmap is called, the kernel internally clears PTE values |
| 37 |
including soft dirty bits. To notify user space application about such |
| 38 |
memory region renewal the kernel always marks new memory regions (and |
| 39 |
expanded regions) as soft dirty. |
| 40 |
|
| 41 |
This feature is actively used by the checkpoint-restore project. You |
| 42 |
can find more details about it on http://criu.org |
| 43 |
|
| 44 |
|
| 45 |
-- Pavel Emelyanov, Apr 9, 2013 |