Files
littlefs/runners/bench_runner.h
T
Christopher Haster c815c19c20 New "fragmenting" write strategy
The attempt to implement in-rbyd data slicing, being lazily coalesced
during rbyd compaction, failed pretty much completely.

Slicing is a very enticing write strategy, getting both minimal overhead
post-compaction and fast random write speeds, but the idea has some
fundamental conflicts with how we play out attrs post-compaction.

This idea might work in a more powerful filesystem, but brings back the
need to simulate rbyds in RAM, which is something I really don't want to
do (complex, bug-prone, likely adds code cost, may not even be tractable).

So, third time's the charm?

---

This new write strategy writes only datas and bptrs, and avoids dagging
by completely rewriting any regions of data larger than a configurable
crystallization threshold.

This loses most of the benefits of data crystallization, random writes
will now usually need to rewrite a full block, but as a tradeoff our
data at rest is always stored with optimal overhead.

And at least data crystallization still saves space when our data isn't
block aligned, or in sparse files. From reading up on some other
filesystem designs it seems this is a desirable optimization sometimes
referred to as "tail-packing" or "block suballocation"

Some other changes from just having more time to think about the
problem:

1. Instead of scanning to figure out our current crystal size, we can
   use a simple heuristic of 1. look up left block, 2. look up right
   block, 3. assume any data between these blocks contribute to our
   current crystal.

   This is just a heuristic, so worst case you write the first and last
   byte of a block which is enough to trigger compaction into a block.
   But on the plus side this avoids issues with small holes preventing
   blocks from being formed.

   This approach brings the number of btree lookups down from
   O(crystallize_size) to 2.

2. I've gone ahead and dropped the previous scheme of coalesce_size
   + fragment_size and instead adopted a single fragment_size that
   controls the size of, well, fragments, i.e. data elements stored
   directly in trees.

   This affects both the inlined shrub as well as fragments stored in
   the inner nodes of the btree. I believe it's very similar to what is
   often called "pages" in logging filesystems, though I'm going to
   avoid that term for now because it's a bit overloaded.

   Previously, neighboring writes that, when combined, would exceed our
   coalesce_size, they just weren't combined. Now they are combined up
   to our fragment size, potentially splitting the right fragment.

   Before (fragment_size=8):

     .---+---+---+---+---+---+---+---.
     |            8 bytes            |
     '---+---+---+---+---+---+---+---'
                         +
                         .---+---+---+---+---.
                         |      5 bytes      |
                         '---+---+---+---+---'
                         =
     .---+---+---+---+---+---+---+---+---+---.
     |      5 bytes      |      5 bytes      |
     '---+---+---+---+---+---+---+---+---+---'

   After:

     .---+---+---+---+---+---+---+---.
     |            8 bytes            |
     '---+---+---+---+---+---+---+---'
                         +
                         .---+---+---+---+---.
                         |      5 bytes      |
                         '---+---+---+---+---'
                         =
     .---+---+---+---+---+---+---+---+---+---.
     |            8 bytes            |2 bytes|
     '---+---+---+---+---+---+---+---+---+---'

   This leads to better fragment alignment (much like our block
   strategy), and minimizes tree overhead.

   Any neighboring data to the right is only coalesced if it fits in the
   current fragment, or would be rewritten (carved) anyways, to avoid
   unnecessary data rewriting.

   For example (fragment_size=8):

     .---+---+---+---+---+---+---+---+---+---+---+---+---+---.
     |        6 bytes        |        6 bytes        |2 bytes|
     '---+---+---+---+---+---+---+---+---+---+---+---+---+---'
                                 +
                         .---+---+---+---+---.
                         |      5 bytes      |
                         '---+---+---+---+---'
                                 =
     .---+---+---+---+---+---+---+---+---+---+---+---+---+---.
     |            8 bytes            |    4 bytes    |2 bytes|
     '---+---+---+---+---+---+---+---+---+---+---+---+---+---'

Other than these changes this commit is mostly a bunch of carveshrub
rewriting again, which continues to be nuanced and annoying to get
bug free.
2023-10-21 22:05:46 -05:00

181 lines
5.9 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 LFS_TRACE
void bench_trace(const char *fmt, ...);
#define LFS_TRACE_(fmt, ...) \
bench_trace("%s:%d:trace: " fmt "%s\n", \
__FILE__, \
__LINE__, \
__VA_ARGS__)
#define LFS_TRACE(...) LFS_TRACE_(__VA_ARGS__, "")
#define LFS_EMUBD_TRACE(...) LFS_TRACE_(__VA_ARGS__, "")
// provide BENCH_START/BENCH_STOP macros
void bench_start(void);
void bench_stop(void);
#define BENCH_START() bench_start()
#define BENCH_STOP() bench_stop()
// note these are indirectly included in any generated files
#include "bd/lfs_emubd.h"
#include <stdio.h>
// give source a chance to define feature macros
#undef _FEATURES_H
#undef _STDIO_H
// generated bench configurations
struct lfs_config;
enum bench_flags {
BENCH_INTERNAL = 0x1,
};
typedef uint8_t bench_flags_t;
typedef struct bench_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 lfs_config *cfg);
};
struct bench_suite {
const char *name;
const char *path;
bench_flags_t flags;
const char *const *define_names;
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)
// access generated bench defines
intmax_t bench_define(size_t define);
#define BENCH_DEFINE(i) bench_define(i)
// a few preconfigured defines that control how benches run
#define BENCH_IMPLICIT_DEFINE_COUNT 15
#define BENCH_GEOMETRY_DEFINE_COUNT 3
#define READ_SIZE_i 0
#define PROG_SIZE_i 1
#define BLOCK_SIZE_i 2
#define BLOCK_COUNT_i 3
#define DISK_SIZE_i 4
#define CACHE_SIZE_i 5
#define INLINE_SIZE_i 6
#define FRAGMENT_SIZE_i 7
#define CRYSTALLIZE_SIZE_i 8
#define LOOKAHEAD_SIZE_i 9
#define BLOCK_CYCLES_i 10
#define ERASE_VALUE_i 11
#define ERASE_CYCLES_i 12
#define BADBLOCK_BEHAVIOR_i 13
#define POWERLOSS_BEHAVIOR_i 14
#define READ_SIZE bench_define(READ_SIZE_i)
#define PROG_SIZE bench_define(PROG_SIZE_i)
#define BLOCK_SIZE bench_define(BLOCK_SIZE_i)
#define BLOCK_COUNT bench_define(BLOCK_COUNT_i)
#define DISK_SIZE bench_define(DISK_SIZE_i)
#define CACHE_SIZE bench_define(CACHE_SIZE_i)
#define INLINE_SIZE bench_define(INLINE_SIZE_i)
#define FRAGMENT_SIZE bench_define(FRAGMENT_SIZE_i)
#define CRYSTALLIZE_SIZE bench_define(CRYSTALLIZE_SIZE_i)
#define LOOKAHEAD_SIZE bench_define(LOOKAHEAD_SIZE_i)
#define BLOCK_CYCLES bench_define(BLOCK_CYCLES_i)
#define ERASE_VALUE bench_define(ERASE_VALUE_i)
#define ERASE_CYCLES bench_define(ERASE_CYCLES_i)
#define BADBLOCK_BEHAVIOR bench_define(BADBLOCK_BEHAVIOR_i)
#define POWERLOSS_BEHAVIOR bench_define(POWERLOSS_BEHAVIOR_i)
#define BENCH_IMPLICIT_DEFINES \
/* name value (overridable) */ \
BENCH_DEF(READ_SIZE, PROG_SIZE ) \
BENCH_DEF(PROG_SIZE, BLOCK_SIZE ) \
BENCH_DEF(BLOCK_SIZE, 0 ) \
BENCH_DEF(BLOCK_COUNT, DISK_SIZE/BLOCK_SIZE ) \
BENCH_DEF(DISK_SIZE, 1024*1024 ) \
BENCH_DEF(CACHE_SIZE, lfs_max(16, lfs_max(READ_SIZE, PROG_SIZE))) \
BENCH_DEF(INLINE_SIZE, BLOCK_SIZE/8 ) \
BENCH_DEF(FRAGMENT_SIZE, CACHE_SIZE ) \
BENCH_DEF(CRYSTALLIZE_SIZE, BLOCK_SIZE/4 ) \
BENCH_DEF(LOOKAHEAD_SIZE, 16 ) \
BENCH_DEF(BLOCK_CYCLES, -1 ) \
BENCH_DEF(ERASE_VALUE, 0xff ) \
BENCH_DEF(ERASE_CYCLES, 0 ) \
BENCH_DEF(BADBLOCK_BEHAVIOR, LFS_EMUBD_BADBLOCK_PROGERROR ) \
BENCH_DEF(POWERLOSS_BEHAVIOR, LFS_EMUBD_POWERLOSS_NOOP )
#define BENCH_GEOMETRIES \
/* name read_size prog_size block_size */ \
BENCH_GEO("default", 16, 16, 512 ) \
BENCH_GEO("eeprom", 1, 1, 512 ) \
BENCH_GEO("emmc", 512, 512, 512 ) \
BENCH_GEO("nor", 1, 1, 4096 ) \
BENCH_GEO("nand", 4096, 4096, 32768 )
#define BENCH_CFG \
.read_size = READ_SIZE, \
.prog_size = PROG_SIZE, \
.block_size = BLOCK_SIZE, \
.block_count = BLOCK_COUNT, \
.block_cycles = BLOCK_CYCLES, \
.cache_size = CACHE_SIZE, \
.inline_size = INLINE_SIZE, \
.fragment_size = FRAGMENT_SIZE, \
.crystallize_size = CRYSTALLIZE_SIZE, \
.lookahead_size = LOOKAHEAD_SIZE,
#define BENCH_BDCFG \
.erase_value = ERASE_VALUE, \
.erase_cycles = ERASE_CYCLES, \
.badblock_behavior = BADBLOCK_BEHAVIOR,
#endif