xref: /openbmc/linux/Documentation/ABI/testing/sysfs-memory-page-offline (revision facb6011f3993947283fa15d039dacb4ad140230)
1*facb6011SAndi KleenWhat:		/sys/devices/system/memory/soft_offline_page
2*facb6011SAndi KleenDate:		Sep 2009
3*facb6011SAndi KleenKernelVersion:	2.6.33
4*facb6011SAndi KleenContact:	andi@firstfloor.org
5*facb6011SAndi KleenDescription:
6*facb6011SAndi Kleen		Soft-offline the memory page containing the physical address
7*facb6011SAndi Kleen		written into this file. Input is a hex number specifying the
8*facb6011SAndi Kleen		physical address of the page. The kernel will then attempt
9*facb6011SAndi Kleen		to soft-offline it, by moving the contents elsewhere or
10*facb6011SAndi Kleen		dropping it if possible. The kernel will then be placed
11*facb6011SAndi Kleen		on the bad page list and never be reused.
12*facb6011SAndi Kleen
13*facb6011SAndi Kleen		The offlining is done in kernel specific granuality.
14*facb6011SAndi Kleen		Normally it's the base page size of the kernel, but
15*facb6011SAndi Kleen		this might change.
16*facb6011SAndi Kleen
17*facb6011SAndi Kleen		The page must be still accessible, not poisoned. The
18*facb6011SAndi Kleen		kernel will never kill anything for this, but rather
19*facb6011SAndi Kleen		fail the offline.  Return value is the size of the
20*facb6011SAndi Kleen		number, or a error when the offlining failed.  Reading
21*facb6011SAndi Kleen		the file is not allowed.
22*facb6011SAndi Kleen
23*facb6011SAndi KleenWhat:		/sys/devices/system/memory/hard_offline_page
24*facb6011SAndi KleenDate:		Sep 2009
25*facb6011SAndi KleenKernelVersion:	2.6.33
26*facb6011SAndi KleenContact:	andi@firstfloor.org
27*facb6011SAndi KleenDescription:
28*facb6011SAndi Kleen		Hard-offline the memory page containing the physical
29*facb6011SAndi Kleen		address written into this file. Input is a hex number
30*facb6011SAndi Kleen		specifying the physical address of the page. The
31*facb6011SAndi Kleen		kernel will then attempt to hard-offline the page, by
32*facb6011SAndi Kleen		trying to drop the page or killing any owner or
33*facb6011SAndi Kleen		triggering IO errors if needed.  Note this may kill
34*facb6011SAndi Kleen		any processes owning the page. The kernel will avoid
35*facb6011SAndi Kleen		to access this page assuming it's poisoned by the
36*facb6011SAndi Kleen		hardware.
37*facb6011SAndi Kleen
38*facb6011SAndi Kleen		The offlining is done in kernel specific granuality.
39*facb6011SAndi Kleen		Normally it's the base page size of the kernel, but
40*facb6011SAndi Kleen		this might change.
41*facb6011SAndi Kleen
42*facb6011SAndi Kleen		Return value is the size of the number, or a error when
43*facb6011SAndi Kleen		the offlining failed.
44*facb6011SAndi Kleen		Reading the file is not allowed.
45