Merge 3.9-rc5 into staging-next
[deliverable/linux.git] / drivers / staging / zcache / Kconfig
1 config ZCACHE
2 bool "Dynamic compression of swap pages and clean pagecache pages"
3 depends on CRYPTO=y && SWAP=y && CLEANCACHE && FRONTSWAP
4 select CRYPTO_LZO
5 default n
6 help
7 Zcache doubles RAM efficiency while providing a significant
8 performance boosts on many workloads. Zcache uses
9 compression and an in-kernel implementation of transcendent
10 memory to store clean page cache pages and swap in RAM,
11 providing a noticeable reduction in disk I/O.
12
13 config ZCACHE_DEBUG
14 bool "Enable debug statistics"
15 depends on DEBUG_FS && ZCACHE
16 default n
17 help
18 This is used to provide an debugfs directory with counters of
19 how zcache is doing. You probably want to set this to 'N'.
20
21 config RAMSTER
22 bool "Cross-machine RAM capacity sharing, aka peer-to-peer tmem"
23 depends on CONFIGFS_FS=y && SYSFS=y && !HIGHMEM && ZCACHE=y
24 depends on NET
25 # must ensure struct page is 8-byte aligned
26 select HAVE_ALIGNED_STRUCT_PAGE if !64BIT
27 default n
28 help
29 RAMster allows RAM on other machines in a cluster to be utilized
30 dynamically and symmetrically instead of swapping to a local swap
31 disk, thus improving performance on memory-constrained workloads
32 while minimizing total RAM across the cluster. RAMster, like
33 zcache2, compresses swap pages into local RAM, but then remotifies
34 the compressed pages to another node in the RAMster cluster.
35
36 # Depends on not-yet-upstreamed mm patches to export end_swap_bio_write and
37 # __add_to_swap_cache, and implement __swap_writepage (which is swap_writepage
38 # without the frontswap call. When these are in-tree, the dependency on
39 # BROKEN can be removed
40 config ZCACHE_WRITEBACK
41 bool "Allow compressed swap pages to be writtenback to swap disk"
42 depends on ZCACHE=y && BROKEN
43 default n
44 help
45 Zcache caches compressed swap pages (and other data) in RAM which
46 often improves performance by avoiding I/O's due to swapping.
47 In some workloads with very long-lived large processes, it can
48 instead reduce performance. Writeback decompresses zcache-compressed
49 pages (in LRU order) when under memory pressure and writes them to
50 the backing swap disk to ameliorate this problem. Policy driving
51 writeback is still under development.
This page took 0.068138 seconds and 5 git commands to generate.