mirror of
https://git.proxmox.com/git/mirror_zfs.git
synced 2024-12-26 03:09:34 +03:00
330c6c0523
The RAIDZ and DRAID code is responsible for reporting checksum errors on their child vdevs. Checksum errors represent events where a disk returned data or parity that should have been correct, but was not. In other words, these are instances of silent data corruption. The checksum errors show up in the vdev stats (and thus `zpool status`'s CKSUM column), and in the event log (`zpool events`). Note, this is in contrast with the more common "noisy" errors where a disk goes offline, in which case ZFS knows that the disk is bad and doesn't try to read it, or the device returns an error on the requested read or write operation. RAIDZ/DRAID generate checksum errors via three code paths: 1. When RAIDZ/DRAID reconstructs a damaged block, checksum errors are reported on any children whose data was not used during the reconstruction. This is handled in `raidz_reconstruct()`. This is the most common type of RAIDZ/DRAID checksum error. 2. When RAIDZ/DRAID is not able to reconstruct a damaged block, that means that the data has been lost. The zio fails and an error is returned to the consumer (e.g. the read(2) system call). This would happen if, for example, three different disks in a RAIDZ2 group are silently damaged. Since the damage is silent, it isn't possible to know which three disks are damaged, so a checksum error is reported against every child that returned data or parity for this read. (For DRAID, typically only one "group" of children is involved in each io.) This case is handled in `vdev_raidz_cksum_finish()`. This is the next most common type of RAIDZ/DRAID checksum error. 3. If RAIDZ/DRAID is not able to reconstruct a damaged block (like in case 2), but there happens to be additional copies of this block due to "ditto blocks" (i.e. multiple DVA's in this blkptr_t), and one of those copies is good, then RAIDZ/DRAID compares each sector of the data or parity that it retrieved with the good data from the other DVA, and if they differ then it reports a checksum error on this child. This differs from case 2 in that the checksum error is reported on only the subset of children that actually have bad data or parity. This case happens very rarely, since normally only metadata has ditto blocks. If the silent damage is extensive, there will be many instances of case 2, and the pool will likely be unrecoverable. The code for handling case 3 is considerably more complicated than the other cases, for two reasons: 1. It needs to run after the main raidz read logic has completed. The data RAIDZ read needs to be preserved until after the alternate DVA has been read, which necessitates refcounts and callbacks managed by the non-raidz-specific zio layer. 2. It's nontrivial to map the sections of data read by RAIDZ to the correct data. For example, the correct data does not include the parity information, so the parity must be recalculated based on the correct data, and then compared to the parity that was read from the RAIDZ children. Due to the complexity of case 3, the rareness of hitting it, and the minimal benefit it provides above case 2, this commit removes the code for case 3. These types of errors will now be handled the same as case 2, i.e. the checksum error will be reported against all children that returned data or parity. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Closes #11735
389 lines
10 KiB
C
389 lines
10 KiB
C
/*
|
|
* CDDL HEADER START
|
|
*
|
|
* The contents of this file are subject to the terms of the
|
|
* Common Development and Distribution License (the "License").
|
|
* You may not use this file except in compliance with the License.
|
|
*
|
|
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
|
|
* or http://www.opensolaris.org/os/licensing.
|
|
* See the License for the specific language governing permissions
|
|
* and limitations under the License.
|
|
*
|
|
* When distributing Covered Code, include this CDDL HEADER in each
|
|
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
|
|
* If applicable, add the following below this CDDL HEADER, with the
|
|
* fields enclosed by brackets "[]" replaced with your own identifying
|
|
* information: Portions Copyright [yyyy] [name of copyright owner]
|
|
*
|
|
* CDDL HEADER END
|
|
*/
|
|
/*
|
|
* Copyright (C) 2016 Gvozden Nešković. All rights reserved.
|
|
*/
|
|
|
|
#ifndef _VDEV_RAIDZ_H
|
|
#define _VDEV_RAIDZ_H
|
|
|
|
#include <sys/types.h>
|
|
#include <sys/debug.h>
|
|
#include <sys/kstat.h>
|
|
#include <sys/abd.h>
|
|
#include <sys/vdev_impl.h>
|
|
|
|
#ifdef __cplusplus
|
|
extern "C" {
|
|
#endif
|
|
|
|
#define CODE_P (0U)
|
|
#define CODE_Q (1U)
|
|
#define CODE_R (2U)
|
|
|
|
#define PARITY_P (1U)
|
|
#define PARITY_PQ (2U)
|
|
#define PARITY_PQR (3U)
|
|
|
|
#define TARGET_X (0U)
|
|
#define TARGET_Y (1U)
|
|
#define TARGET_Z (2U)
|
|
|
|
/*
|
|
* Parity generation methods indexes
|
|
*/
|
|
enum raidz_math_gen_op {
|
|
RAIDZ_GEN_P = 0,
|
|
RAIDZ_GEN_PQ,
|
|
RAIDZ_GEN_PQR,
|
|
RAIDZ_GEN_NUM = 3
|
|
};
|
|
/*
|
|
* Data reconstruction methods indexes
|
|
*/
|
|
enum raidz_rec_op {
|
|
RAIDZ_REC_P = 0,
|
|
RAIDZ_REC_Q,
|
|
RAIDZ_REC_R,
|
|
RAIDZ_REC_PQ,
|
|
RAIDZ_REC_PR,
|
|
RAIDZ_REC_QR,
|
|
RAIDZ_REC_PQR,
|
|
RAIDZ_REC_NUM = 7
|
|
};
|
|
|
|
extern const char *raidz_gen_name[RAIDZ_GEN_NUM];
|
|
extern const char *raidz_rec_name[RAIDZ_REC_NUM];
|
|
|
|
/*
|
|
* Methods used to define raidz implementation
|
|
*
|
|
* @raidz_gen_f Parity generation function
|
|
* @par1 pointer to raidz_map
|
|
* @raidz_rec_f Data reconstruction function
|
|
* @par1 pointer to raidz_map
|
|
* @par2 array of reconstruction targets
|
|
* @will_work_f Function returns TRUE if impl. is supported on the system
|
|
* @init_impl_f Function is called once on init
|
|
* @fini_impl_f Function is called once on fini
|
|
*/
|
|
typedef void (*raidz_gen_f)(void *);
|
|
typedef int (*raidz_rec_f)(void *, const int *);
|
|
typedef boolean_t (*will_work_f)(void);
|
|
typedef void (*init_impl_f)(void);
|
|
typedef void (*fini_impl_f)(void);
|
|
|
|
#define RAIDZ_IMPL_NAME_MAX (20)
|
|
|
|
typedef struct raidz_impl_ops {
|
|
init_impl_f init;
|
|
fini_impl_f fini;
|
|
raidz_gen_f gen[RAIDZ_GEN_NUM]; /* Parity generate functions */
|
|
raidz_rec_f rec[RAIDZ_REC_NUM]; /* Data reconstruction functions */
|
|
will_work_f is_supported; /* Support check function */
|
|
char name[RAIDZ_IMPL_NAME_MAX]; /* Name of the implementation */
|
|
} raidz_impl_ops_t;
|
|
|
|
typedef struct raidz_col {
|
|
uint64_t rc_devidx; /* child device index for I/O */
|
|
uint64_t rc_offset; /* device offset */
|
|
uint64_t rc_size; /* I/O size */
|
|
abd_t rc_abdstruct; /* rc_abd probably points here */
|
|
abd_t *rc_abd; /* I/O data */
|
|
abd_t *rc_orig_data; /* pre-reconstruction */
|
|
int rc_error; /* I/O error for this device */
|
|
uint8_t rc_tried; /* Did we attempt this I/O column? */
|
|
uint8_t rc_skipped; /* Did we skip this I/O column? */
|
|
uint8_t rc_need_orig_restore; /* need to restore from orig_data? */
|
|
uint8_t rc_repair; /* Write good data to this column */
|
|
} raidz_col_t;
|
|
|
|
typedef struct raidz_row {
|
|
uint64_t rr_cols; /* Regular column count */
|
|
uint64_t rr_scols; /* Count including skipped columns */
|
|
uint64_t rr_bigcols; /* Remainder data column count */
|
|
uint64_t rr_missingdata; /* Count of missing data devices */
|
|
uint64_t rr_missingparity; /* Count of missing parity devices */
|
|
uint64_t rr_firstdatacol; /* First data column/parity count */
|
|
abd_t *rr_abd_empty; /* dRAID empty sector buffer */
|
|
int rr_nempty; /* empty sectors included in parity */
|
|
#ifdef ZFS_DEBUG
|
|
uint64_t rr_offset; /* Logical offset for *_io_verify() */
|
|
uint64_t rr_size; /* Physical size for *_io_verify() */
|
|
#endif
|
|
raidz_col_t rr_col[0]; /* Flexible array of I/O columns */
|
|
} raidz_row_t;
|
|
|
|
typedef struct raidz_map {
|
|
boolean_t rm_ecksuminjected; /* checksum error was injected */
|
|
int rm_nrows; /* Regular row count */
|
|
int rm_nskip; /* RAIDZ sectors skipped for padding */
|
|
int rm_skipstart; /* Column index of padding start */
|
|
const raidz_impl_ops_t *rm_ops; /* RAIDZ math operations */
|
|
raidz_row_t *rm_row[0]; /* flexible array of rows */
|
|
} raidz_map_t;
|
|
|
|
|
|
#define RAIDZ_ORIGINAL_IMPL (INT_MAX)
|
|
|
|
extern const raidz_impl_ops_t vdev_raidz_scalar_impl;
|
|
extern boolean_t raidz_will_scalar_work(void);
|
|
|
|
#if defined(__x86_64) && defined(HAVE_SSE2) /* only x86_64 for now */
|
|
extern const raidz_impl_ops_t vdev_raidz_sse2_impl;
|
|
#endif
|
|
#if defined(__x86_64) && defined(HAVE_SSSE3) /* only x86_64 for now */
|
|
extern const raidz_impl_ops_t vdev_raidz_ssse3_impl;
|
|
#endif
|
|
#if defined(__x86_64) && defined(HAVE_AVX2) /* only x86_64 for now */
|
|
extern const raidz_impl_ops_t vdev_raidz_avx2_impl;
|
|
#endif
|
|
#if defined(__x86_64) && defined(HAVE_AVX512F) /* only x86_64 for now */
|
|
extern const raidz_impl_ops_t vdev_raidz_avx512f_impl;
|
|
#endif
|
|
#if defined(__x86_64) && defined(HAVE_AVX512BW) /* only x86_64 for now */
|
|
extern const raidz_impl_ops_t vdev_raidz_avx512bw_impl;
|
|
#endif
|
|
#if defined(__aarch64__)
|
|
extern const raidz_impl_ops_t vdev_raidz_aarch64_neon_impl;
|
|
extern const raidz_impl_ops_t vdev_raidz_aarch64_neonx2_impl;
|
|
#endif
|
|
#if defined(__powerpc__)
|
|
extern const raidz_impl_ops_t vdev_raidz_powerpc_altivec_impl;
|
|
#endif
|
|
|
|
/*
|
|
* Commonly used raidz_map helpers
|
|
*
|
|
* raidz_parity Returns parity of the RAIDZ block
|
|
* raidz_ncols Returns number of columns the block spans
|
|
* Note, all rows have the same number of columns.
|
|
* raidz_nbigcols Returns number of big columns
|
|
* raidz_col_p Returns pointer to a column
|
|
* raidz_col_size Returns size of a column
|
|
* raidz_big_size Returns size of big columns
|
|
* raidz_short_size Returns size of short columns
|
|
*/
|
|
#define raidz_parity(rm) ((rm)->rm_row[0]->rr_firstdatacol)
|
|
#define raidz_ncols(rm) ((rm)->rm_row[0]->rr_cols)
|
|
#define raidz_nbigcols(rm) ((rm)->rm_bigcols)
|
|
#define raidz_col_p(rm, c) ((rm)->rm_col + (c))
|
|
#define raidz_col_size(rm, c) ((rm)->rm_col[c].rc_size)
|
|
#define raidz_big_size(rm) (raidz_col_size(rm, CODE_P))
|
|
#define raidz_short_size(rm) (raidz_col_size(rm, raidz_ncols(rm)-1))
|
|
|
|
/*
|
|
* Macro defines an RAIDZ parity generation method
|
|
*
|
|
* @code parity the function produce
|
|
* @impl name of the implementation
|
|
*/
|
|
#define _RAIDZ_GEN_WRAP(code, impl) \
|
|
static void \
|
|
impl ## _gen_ ## code(void *rrp) \
|
|
{ \
|
|
raidz_row_t *rr = (raidz_row_t *)rrp; \
|
|
raidz_generate_## code ## _impl(rr); \
|
|
}
|
|
|
|
/*
|
|
* Macro defines an RAIDZ data reconstruction method
|
|
*
|
|
* @code parity the function produce
|
|
* @impl name of the implementation
|
|
*/
|
|
#define _RAIDZ_REC_WRAP(code, impl) \
|
|
static int \
|
|
impl ## _rec_ ## code(void *rrp, const int *tgtidx) \
|
|
{ \
|
|
raidz_row_t *rr = (raidz_row_t *)rrp; \
|
|
return (raidz_reconstruct_## code ## _impl(rr, tgtidx)); \
|
|
}
|
|
|
|
/*
|
|
* Define all gen methods for an implementation
|
|
*
|
|
* @impl name of the implementation
|
|
*/
|
|
#define DEFINE_GEN_METHODS(impl) \
|
|
_RAIDZ_GEN_WRAP(p, impl); \
|
|
_RAIDZ_GEN_WRAP(pq, impl); \
|
|
_RAIDZ_GEN_WRAP(pqr, impl)
|
|
|
|
/*
|
|
* Define all rec functions for an implementation
|
|
*
|
|
* @impl name of the implementation
|
|
*/
|
|
#define DEFINE_REC_METHODS(impl) \
|
|
_RAIDZ_REC_WRAP(p, impl); \
|
|
_RAIDZ_REC_WRAP(q, impl); \
|
|
_RAIDZ_REC_WRAP(r, impl); \
|
|
_RAIDZ_REC_WRAP(pq, impl); \
|
|
_RAIDZ_REC_WRAP(pr, impl); \
|
|
_RAIDZ_REC_WRAP(qr, impl); \
|
|
_RAIDZ_REC_WRAP(pqr, impl)
|
|
|
|
#define RAIDZ_GEN_METHODS(impl) \
|
|
{ \
|
|
[RAIDZ_GEN_P] = & impl ## _gen_p, \
|
|
[RAIDZ_GEN_PQ] = & impl ## _gen_pq, \
|
|
[RAIDZ_GEN_PQR] = & impl ## _gen_pqr \
|
|
}
|
|
|
|
#define RAIDZ_REC_METHODS(impl) \
|
|
{ \
|
|
[RAIDZ_REC_P] = & impl ## _rec_p, \
|
|
[RAIDZ_REC_Q] = & impl ## _rec_q, \
|
|
[RAIDZ_REC_R] = & impl ## _rec_r, \
|
|
[RAIDZ_REC_PQ] = & impl ## _rec_pq, \
|
|
[RAIDZ_REC_PR] = & impl ## _rec_pr, \
|
|
[RAIDZ_REC_QR] = & impl ## _rec_qr, \
|
|
[RAIDZ_REC_PQR] = & impl ## _rec_pqr \
|
|
}
|
|
|
|
|
|
typedef struct raidz_impl_kstat {
|
|
uint64_t gen[RAIDZ_GEN_NUM]; /* gen method speed B/s */
|
|
uint64_t rec[RAIDZ_REC_NUM]; /* rec method speed B/s */
|
|
} raidz_impl_kstat_t;
|
|
|
|
/*
|
|
* Enumerate various multiplication constants
|
|
* used in reconstruction methods
|
|
*/
|
|
typedef enum raidz_mul_info {
|
|
/* Reconstruct Q */
|
|
MUL_Q_X = 0,
|
|
/* Reconstruct R */
|
|
MUL_R_X = 0,
|
|
/* Reconstruct PQ */
|
|
MUL_PQ_X = 0,
|
|
MUL_PQ_Y = 1,
|
|
/* Reconstruct PR */
|
|
MUL_PR_X = 0,
|
|
MUL_PR_Y = 1,
|
|
/* Reconstruct QR */
|
|
MUL_QR_XQ = 0,
|
|
MUL_QR_X = 1,
|
|
MUL_QR_YQ = 2,
|
|
MUL_QR_Y = 3,
|
|
/* Reconstruct PQR */
|
|
MUL_PQR_XP = 0,
|
|
MUL_PQR_XQ = 1,
|
|
MUL_PQR_XR = 2,
|
|
MUL_PQR_YU = 3,
|
|
MUL_PQR_YP = 4,
|
|
MUL_PQR_YQ = 5,
|
|
|
|
MUL_CNT = 6
|
|
} raidz_mul_info_t;
|
|
|
|
/*
|
|
* Powers of 2 in the Galois field.
|
|
*/
|
|
extern const uint8_t vdev_raidz_pow2[256] __attribute__((aligned(256)));
|
|
/* Logs of 2 in the Galois field defined above. */
|
|
extern const uint8_t vdev_raidz_log2[256] __attribute__((aligned(256)));
|
|
|
|
/*
|
|
* Multiply a given number by 2 raised to the given power.
|
|
*/
|
|
static inline uint8_t
|
|
vdev_raidz_exp2(const uint8_t a, const unsigned exp)
|
|
{
|
|
if (a == 0)
|
|
return (0);
|
|
|
|
return (vdev_raidz_pow2[(exp + (unsigned)vdev_raidz_log2[a]) % 255]);
|
|
}
|
|
|
|
/*
|
|
* Galois Field operations.
|
|
*
|
|
* gf_exp2 - computes 2 raised to the given power
|
|
* gf_exp2 - computes 4 raised to the given power
|
|
* gf_mul - multiplication
|
|
* gf_div - division
|
|
* gf_inv - multiplicative inverse
|
|
*/
|
|
typedef unsigned gf_t;
|
|
typedef unsigned gf_log_t;
|
|
|
|
static inline gf_t
|
|
gf_mul(const gf_t a, const gf_t b)
|
|
{
|
|
gf_log_t logsum;
|
|
|
|
if (a == 0 || b == 0)
|
|
return (0);
|
|
|
|
logsum = (gf_log_t)vdev_raidz_log2[a] + (gf_log_t)vdev_raidz_log2[b];
|
|
|
|
return ((gf_t)vdev_raidz_pow2[logsum % 255]);
|
|
}
|
|
|
|
static inline gf_t
|
|
gf_div(const gf_t a, const gf_t b)
|
|
{
|
|
gf_log_t logsum;
|
|
|
|
ASSERT3U(b, >, 0);
|
|
if (a == 0)
|
|
return (0);
|
|
|
|
logsum = (gf_log_t)255 + (gf_log_t)vdev_raidz_log2[a] -
|
|
(gf_log_t)vdev_raidz_log2[b];
|
|
|
|
return ((gf_t)vdev_raidz_pow2[logsum % 255]);
|
|
}
|
|
|
|
static inline gf_t
|
|
gf_inv(const gf_t a)
|
|
{
|
|
gf_log_t logsum;
|
|
|
|
ASSERT3U(a, >, 0);
|
|
|
|
logsum = (gf_log_t)255 - (gf_log_t)vdev_raidz_log2[a];
|
|
|
|
return ((gf_t)vdev_raidz_pow2[logsum]);
|
|
}
|
|
|
|
static inline gf_t
|
|
gf_exp2(gf_log_t exp)
|
|
{
|
|
return (vdev_raidz_pow2[exp % 255]);
|
|
}
|
|
|
|
static inline gf_t
|
|
gf_exp4(gf_log_t exp)
|
|
{
|
|
ASSERT3U(exp, <=, 255);
|
|
return ((gf_t)vdev_raidz_pow2[(2 * exp) % 255]);
|
|
}
|
|
|
|
#ifdef __cplusplus
|
|
}
|
|
#endif
|
|
|
|
#endif /* _VDEV_RAIDZ_H */
|