843412cc79
Allocating pre-erased blocks gets quite complicated due to our
restricted flash model, but at least the actual pre-erasing is
relatively straightforward:
- We keep track of known preerased state in lfs3->gbmap.preeraser.
- If LFS3_GC_PREERASE is provided during gc work, we increment the
preeraser's known window by scanning the gbmap.
- Any BMFREE ranges we find, we erase a block at a time, and store the
resulting ecksum in a BMERASED range in the gbmap.
- We keep track of how many blocks we erased, and stop early if this
exceeds cfg.gc_preerase_count. This just lets users tune how many
blocks to preerase in case something (?) prevents preerased blocks
from being used.
Some notes:
- We don't really do anything with ranges in lfs3_alloc_preerase. In
theory we could bulk in erase to minimize the number of commits to the
gbmap, but we expect erase to dominate, so this probably isn't worth
it.
And if erase doesn't dominate, why would you bother pre-erasing
blocks?
- Preerasing isn't really a traversal operation, and is managed by a
sort of secondary state machine in lfs3_fs_gc_.
This also means lfs3_trv_read with LFS3_T_PREERASE does nothing, but I
guess that is ok? It's tempting to try to make lfs3_trv_read also
preerase, but it's unclear what block it should return -- it's
probably the wrong API.
- Introducing ecksums actually went quite a bit smoother than I
expected. Though it helps ecksums are the only optional payload, no
type punning or anything.
Ecksums do muddy the gbmap's design a bit, unfortunately. The main
issue being that we can only merge BMERASED ranges with equal ecksums.
This makes BMERASED ranges less compressable than the others, and may
be one reason to limit cfg.gc_preerase_count.
However:
1. This is where I think it's useful to emphasize that the gbmap's
responsibility is to track _free_ blocks, in-use blocks are
secondary.
When allocating, we're going to stop at the first BMFREE/BMERASED,
but may need to skip over an unbounded number of BMINUSE/BMBAD
blocks. So the compressability of BMFREE/BMERASED ranges should
have less of an impact on block allocation.
2. In practice, most flash uses consistent erase values, so the
resulting ecksums will probably be compressable. The exceptions are
noop-erases (SD/eMMC, RAM, NVRAM, etc), and encryption with block
address permutation?
Though noop-erases are a pretty big exception.
Code changes:
code stack ctx
before: 35116 2136 660
after: 35116 (+0.0%) 2136 (+0.0%) 660 (+0.0%)
code stack ctx
gbmap+np before: 38040 2136 776
gbmap+np after: 38188 (+0.4%) 2144 (+0.4%) 776 (+0.0%)
code stack ctx
gbmap+yp before: 38040 2136 776
gbmap+yp after: 38608 (+1.5%) 2144 (+0.4%) 796 (+2.6%)
192 lines
6.7 KiB
C
192 lines
6.7 KiB
C
/*
|
|
* Runner for littlefs benchmarks
|
|
*
|
|
* Copyright (c) 2022, The littlefs authors.
|
|
* SPDX-License-Identifier: BSD-3-Clause
|
|
*/
|
|
#ifndef BENCH_RUNNER_H
|
|
#define BENCH_RUNNER_H
|
|
|
|
|
|
// override LFS3_TRACE
|
|
void bench_trace(const char *fmt, ...);
|
|
|
|
#define LFS3_TRACE_(fmt, ...) \
|
|
bench_trace("%s:%d:trace: " fmt "%s\n", \
|
|
__FILE__, \
|
|
__LINE__, \
|
|
__VA_ARGS__)
|
|
#define LFS3_TRACE(...) LFS3_TRACE_(__VA_ARGS__, "")
|
|
#define LFS3_EMUBD_TRACE(...) LFS3_TRACE_(__VA_ARGS__, "")
|
|
|
|
// BENCH_START/BENCH_STOP macros measure readed/proged/erased bytes
|
|
// through emubd
|
|
void bench_start(const char *m, uintmax_t n);
|
|
void bench_stop(const char *m);
|
|
|
|
#define BENCH_START(m, n) bench_start(m, n)
|
|
#define BENCH_STOP(m) bench_stop(m)
|
|
|
|
// BENCH_RESULT/BENCH_FRESULT allow for explicit non-io measurements
|
|
void bench_result(const char *m, uintmax_t n, uintmax_t result);
|
|
void bench_fresult(const char *m, uintmax_t n, double result);
|
|
|
|
#define BENCH_RESULT(m, n, result) bench_result(m, n, result)
|
|
#define BENCH_FRESULT(m, n, result) bench_fresult(m, n, result)
|
|
|
|
|
|
// note these are indirectly included in any generated files
|
|
#include "bd/lfs3_emubd.h"
|
|
#include <stdio.h>
|
|
|
|
// give source a chance to define feature macros
|
|
#undef _FEATURES_H
|
|
#undef _STDIO_H
|
|
|
|
|
|
// generated bench configurations
|
|
struct lfs3_cfg;
|
|
|
|
enum bench_flags {
|
|
BENCH_INTERNAL = 0x1,
|
|
};
|
|
typedef uint8_t bench_flags_t;
|
|
|
|
typedef struct bench_define {
|
|
const char *name;
|
|
intmax_t *define;
|
|
intmax_t (*cb)(void *data, size_t i);
|
|
void *data;
|
|
size_t permutations;
|
|
} bench_define_t;
|
|
|
|
struct bench_case {
|
|
const char *name;
|
|
const char *path;
|
|
bench_flags_t flags;
|
|
|
|
const bench_define_t *defines;
|
|
size_t permutations;
|
|
|
|
bool (*if_)(void);
|
|
void (*run)(struct lfs3_cfg *cfg);
|
|
};
|
|
|
|
struct bench_suite {
|
|
const char *name;
|
|
const char *path;
|
|
bench_flags_t flags;
|
|
|
|
const bench_define_t *defines;
|
|
size_t define_count;
|
|
|
|
const struct bench_case *cases;
|
|
size_t case_count;
|
|
};
|
|
|
|
extern const struct bench_suite *const bench_suites[];
|
|
extern const size_t bench_suite_count;
|
|
|
|
|
|
// deterministic prng for pseudo-randomness in benches
|
|
uint32_t bench_prng(uint32_t *state);
|
|
|
|
#define BENCH_PRNG(state) bench_prng(state)
|
|
|
|
// generation of specific permutations of an array for exhaustive benching
|
|
size_t bench_factorial(size_t x);
|
|
void bench_permutation(size_t i, uint32_t *buffer, size_t size);
|
|
|
|
#define BENCH_FACTORIAL(x) bench_factorial(x)
|
|
#define BENCH_PERMUTATION(i, buffer, size) bench_permutation(i, buffer, size)
|
|
|
|
|
|
// a few preconfigured defines that control how benches run
|
|
#define BENCH_IMPLICIT_DEFINES \
|
|
/* name value (overridable) */ \
|
|
BENCH_DEFINE(READ_SIZE, 1 ) \
|
|
BENCH_DEFINE(PROG_SIZE, 1 ) \
|
|
BENCH_DEFINE(BLOCK_SIZE, 4096 ) \
|
|
BENCH_DEFINE(BLOCK_COUNT, DISK_SIZE/BLOCK_SIZE ) \
|
|
BENCH_DEFINE(DISK_SIZE, 1024*1024 ) \
|
|
BENCH_DEFINE(BLOCK_RECYCLES, -1 ) \
|
|
BENCH_DEFINE(RCACHE_SIZE, LFS3_MAX(16, READ_SIZE) ) \
|
|
BENCH_DEFINE(PCACHE_SIZE, LFS3_MAX(16, PROG_SIZE) ) \
|
|
BENCH_DEFINE(FCACHE_SIZE, 16 ) \
|
|
BENCH_DEFINE(LOOKAHEAD_SIZE, 16 ) \
|
|
BENCH_DEFINE(GC_FLAGS, LFS3_GC_GC ) \
|
|
BENCH_DEFINE(GC_STEPS, 0 ) \
|
|
BENCH_DEFINE(GC_LOOKAHEAD_THRESH, -1 ) \
|
|
BENCH_DEFINE(GC_LOOKGBMAP_THRESH, -1 ) \
|
|
BENCH_DEFINE(GC_PREERASE_COUNT, -1 ) \
|
|
BENCH_DEFINE(GC_COMPACT_THRESH, 0 ) \
|
|
BENCH_DEFINE(SHRUB_SIZE, BLOCK_SIZE/4 ) \
|
|
BENCH_DEFINE(FRAGMENT_SIZE, LFS3_MIN(BLOCK_SIZE/8, 512) ) \
|
|
BENCH_DEFINE(CRYSTAL_THRESH, BLOCK_SIZE/8 ) \
|
|
BENCH_DEFINE(LOOKGBMAP_THRESH, BLOCK_COUNT/4 ) \
|
|
BENCH_DEFINE(ERASE_VALUE, 0xff ) \
|
|
BENCH_DEFINE(ERASE_CYCLES, 0 ) \
|
|
BENCH_DEFINE(BADBLOCK_BEHAVIOR, LFS3_EMUBD_BADBLOCK_PROGERROR ) \
|
|
BENCH_DEFINE(POWERLOSS_BEHAVIOR, LFS3_EMUBD_POWERLOSS_ATOMIC ) \
|
|
BENCH_DEFINE(EMUBD_SEED, 0 )
|
|
|
|
// declare defines as global intmax_ts
|
|
#define BENCH_DEFINE(k, v) \
|
|
extern intmax_t k;
|
|
|
|
BENCH_IMPLICIT_DEFINES
|
|
#undef BENCH_DEFINE
|
|
|
|
// map defines to cfg struct fields
|
|
#define BENCH_CFG \
|
|
.read_size = READ_SIZE, \
|
|
.prog_size = PROG_SIZE, \
|
|
.block_size = BLOCK_SIZE, \
|
|
.block_count = BLOCK_COUNT, \
|
|
.block_recycles = BLOCK_RECYCLES, \
|
|
.rcache_size = RCACHE_SIZE, \
|
|
.pcache_size = PCACHE_SIZE, \
|
|
.fcache_size = FCACHE_SIZE, \
|
|
.lookahead_size = LOOKAHEAD_SIZE, \
|
|
BENCH_GBMAP_CFG \
|
|
BENCH_PREERASE_CFG \
|
|
BENCH_GC_CFG \
|
|
.gc_lookahead_thresh = GC_LOOKAHEAD_THRESH, \
|
|
.gc_compact_thresh = GC_COMPACT_THRESH, \
|
|
.shrub_size = SHRUB_SIZE, \
|
|
.fragment_size = FRAGMENT_SIZE, \
|
|
.crystal_thresh = CRYSTAL_THRESH,
|
|
|
|
#ifdef LFS3_GBMAP
|
|
#define BENCH_GBMAP_CFG \
|
|
.gc_lookgbmap_thresh = GC_LOOKGBMAP_THRESH, \
|
|
.lookgbmap_thresh = LOOKGBMAP_THRESH,
|
|
#else
|
|
#define BENCH_GBMAP_CFG
|
|
#endif
|
|
|
|
#if defined(LFS3_GBMAP) && !defined(LFS3_NO_PREERASE)
|
|
#define BENCH_PREERASE_CFG \
|
|
.gc_preerase_count = GC_PREERASE_COUNT,
|
|
#else
|
|
#define BENCH_PREERASE_CFG
|
|
#endif
|
|
|
|
#ifdef LFS3_GC
|
|
#define BENCH_GC_CFG \
|
|
.gc_flags = GC_FLAGS, \
|
|
.gc_steps = GC_STEPS,
|
|
#else
|
|
#define BENCH_GC_CFG
|
|
#endif
|
|
|
|
#define BENCH_BDCFG \
|
|
.erase_value = ERASE_VALUE, \
|
|
.erase_cycles = ERASE_CYCLES, \
|
|
.badblock_behavior = BADBLOCK_BEHAVIOR, \
|
|
.powerloss_behavior = POWERLOSS_BEHAVIOR, \
|
|
.seed = EMUBD_SEED,
|
|
|
|
|
|
#endif
|