mirror of
https://github.com/amadvance/snapraid.git
synced 2026-11-22 16:06:30 -05:00
Implement G23 Extended Cauchy construction for AES RAID mode
This commit is contained in:
parent
b5d7474c3b
commit
1b5c926936
22 changed files with 10924 additions and 6047 deletions
24
raid/raid.h
24
raid/raid.h
|
|
@ -5,13 +5,13 @@
|
|||
#define __RAID_H
|
||||
|
||||
/*
|
||||
* GF(2^8) reducing polynomial selection.
|
||||
* RAID operating modes and GF(2^8) field selection.
|
||||
*
|
||||
* RAID parity computation operates in GF(2^8), a finite field where
|
||||
* arithmetic is performed modulo an irreducible polynomial of degree 8.
|
||||
* The choice of operating mode affects both correctness and performance:
|
||||
* all tables, coefficients, and parity data are tied to the selected
|
||||
* polynomial and primitive generator and cannot be mixed between
|
||||
* polynomial and generator matrix and cannot be mixed between
|
||||
* incompatible operating modes.
|
||||
*
|
||||
* Two polynomials are supported:
|
||||
|
|
@ -20,8 +20,9 @@
|
|||
*
|
||||
* The polynomial used by the original RAID-6 specification and the
|
||||
* Linux kernel RAID implementation, used with primitive generator g=2.
|
||||
* Choosing this polynomial ensures on-disk compatibility with existing
|
||||
* RAID arrays and interchange with other RAID implementations.
|
||||
* The P and Q rows preserve the standard RAID5/RAID6 coefficient layout,
|
||||
* providing compatibility with the corresponding Linux RAID parity.
|
||||
* Higher parity rows are SnapRAID's Extended Cauchy extension.
|
||||
*
|
||||
* On CPUs supporting Intel GFNI (Galois Field New Instructions),
|
||||
* generation and recovery are accelerated with vgf2p8affineqb.
|
||||
|
|
@ -31,7 +32,8 @@
|
|||
* 0x1b (x^8 + x^4 + x^3 + x + 1) -- AES polynomial
|
||||
*
|
||||
* The polynomial used by the AES encryption standard (0x11b including
|
||||
* the x^8 term), used with primitive generator g=3.
|
||||
* the x^8 term). Parity uses a 6x251 Extended Cauchy matrix whose Q row
|
||||
* follows the G23 sequence.
|
||||
*
|
||||
* On CPUs supporting Intel GFNI, generation and recovery can use the
|
||||
* vgf2p8mulb instruction, which performs GF(2^8) multiplication directly
|
||||
|
|
@ -76,14 +78,14 @@
|
|||
#define RAID_MODE_VANDERMONDE_RAID 1
|
||||
|
||||
/**
|
||||
* RAID mode supporting up to 6 parities using AES polynomial 0x1b (0x11b) and primitive generator g=3.
|
||||
* RAID mode supporting up to 6 parities and 251 data disks using AES polynomial 0x1b (0x11b).
|
||||
*
|
||||
* It has slightly lower performance than RAID_MODE_CAUCHY_RAID on CPUs without GFNI due
|
||||
* to the additional cost of generator g=3. On CPUs supporting GFNI, it uses native
|
||||
* VGF2P8MULB multiplication and is typically faster than the affine GFNI implementation
|
||||
* required by RAID_MODE_CAUCHY_RAID.
|
||||
* It uses an Extended Cauchy matrix with a G23 Q sequence, allowing optimized
|
||||
* generators to use multiplication by 2 for almost every disk transition.
|
||||
* On CPUs supporting GFNI, it uses native VGF2P8MULB multiplication.
|
||||
*
|
||||
* Parity data is incompatible with RAID_MODE_CAUCHY_RAID.
|
||||
* This parity layout is incompatible with RAID_MODE_CAUCHY_RAID and with the
|
||||
* previous RAID_MODE_CAUCHY_AES layout.
|
||||
*/
|
||||
#define RAID_MODE_CAUCHY_AES 2
|
||||
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue