From: Gao Xiang <xiang@kernel.org>
To: linux-erofs@lists.ozlabs.org
Cc: Gao Xiang <xiang@kernel.org>
Subject: [PATCH 4/4] erofs-utils: get rid of hashmap
Date: Wed, 26 Aug 2026 21:43:20 +0800 [thread overview]
Message-ID: <20260826134321.11835-4-xiang@kernel.org> (raw)
In-Reply-To: <20260826134321.11835-1-xiang@kernel.org>
This is the last piece of GPL-only 3rd-party code in
liberofs.
We may introduce a new dynamic hashmap implementation without GPL-only
restriction later.
Signed-off-by: Gao Xiang <xiang@kernel.org>
---
include/erofs/flex-array.h | 156 ---------------------
include/erofs/hashmap.h | 115 ---------------
lib/Makefile.am | 4 +-
lib/hashmap.c | 280 -------------------------------------
4 files changed, 1 insertion(+), 554 deletions(-)
delete mode 100644 include/erofs/flex-array.h
delete mode 100644 include/erofs/hashmap.h
delete mode 100644 lib/hashmap.c
diff --git a/include/erofs/flex-array.h b/include/erofs/flex-array.h
deleted file mode 100644
index 9b1642f..0000000
--- a/include/erofs/flex-array.h
+++ /dev/null
@@ -1,156 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef __EROFS_FLEX_ARRAY_H
-#define __EROFS_FLEX_ARRAY_H
-
-#ifdef __cplusplus
-extern "C"
-{
-#endif
-
-#include <stdio.h>
-#include <stdlib.h>
-#include <limits.h>
-#include <stdint.h>
-
-#include "defs.h"
-#include "print.h"
-
-/*
- * flex-array.h
- *
- * Some notes to make sense of the code.
- *
- * Flex-arrays:
- * - Flex-arrays became standard in C99 and are defined by "array[]" (at the
- * end of a struct)
- * - Pre-C99 flex-arrays can be accomplished by "array[1]"
- * - There is a GNU extension where they are defined using "array[0]"
- * Allegedly there is/was a bug in gcc whereby foo[1] generated incorrect
- * code, so it's safest to use [0] (https://lkml.org/lkml/2015/2/18/407).
- *
- * For C89 and C90, __STDC__ is 1
- * For later standards, __STDC_VERSION__ is defined according to the standard.
- * For example: 199901L or 201112L
- *
- * Whilst we're on the subject, in version 5 of gcc, the default std was
- * changed from gnu89 to gnu11. In jgmenu, CFLAGS therefore contains -std=gnu89
- * You can check your default gcc std by doing:
- * gcc -dM -E - </dev/null | grep '__STDC_VERSION__\|__STDC__'
- *
- * The code below is copied from git's git-compat-util.h in support of
- * hashmap.c
- */
-
-#ifndef FLEX_ARRAY
-#if defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) && \
- (!defined(__SUNPRO_C) || (__SUNPRO_C > 0x580))
-# define FLEX_ARRAY /* empty */
-#elif defined(__GNUC__)
-# if (__GNUC__ >= 3)
-# define FLEX_ARRAY /* empty */
-# else
-# define FLEX_ARRAY 0 /* older GNU extension */
-# endif
-#endif
-
-/* Otherwise, default to safer but a bit wasteful traditional style */
-#ifndef FLEX_ARRAY
-# define FLEX_ARRAY 1
-#endif
-#endif
-
-#define bitsizeof(x) (CHAR_BIT * sizeof(x))
-
-#define maximum_signed_value_of_type(a) \
- (INTMAX_MAX >> (bitsizeof(intmax_t) - bitsizeof(a)))
-
-#define maximum_unsigned_value_of_type(a) \
- (UINTMAX_MAX >> (bitsizeof(uintmax_t) - bitsizeof(a)))
-
-/*
- * Signed integer overflow is undefined in C, so here's a helper macro
- * to detect if the sum of two integers will overflow.
- * Requires: a >= 0, typeof(a) equals typeof(b)
- */
-#define signed_add_overflows(a, b) \
- ((b) > maximum_signed_value_of_type(a) - (a))
-
-#define unsigned_add_overflows(a, b) \
- ((b) > maximum_unsigned_value_of_type(a) - (a))
-
-static inline size_t st_add(size_t a, size_t b)
-{
- if (unsigned_add_overflows(a, b)) {
- erofs_err("size_t overflow: %llu + %llu", a | 0ULL, b | 0ULL);
- BUG_ON(1);
- return -1;
- }
- return a + b;
-}
-
-#define st_add3(a, b, c) st_add(st_add((a), (b)), (c))
-#define st_add4(a, b, c, d) st_add(st_add3((a), (b), (c)), (d))
-
-/*
- * These functions help you allocate structs with flex arrays, and copy
- * the data directly into the array. For example, if you had:
- *
- * struct foo {
- * int bar;
- * char name[FLEX_ARRAY];
- * };
- *
- * you can do:
- *
- * struct foo *f;
- * FLEX_ALLOC_MEM(f, name, src, len);
- *
- * to allocate a "foo" with the contents of "src" in the "name" field.
- * The resulting struct is automatically zero'd, and the flex-array field
- * is NUL-terminated (whether the incoming src buffer was or not).
- *
- * The FLEXPTR_* variants operate on structs that don't use flex-arrays,
- * but do want to store a pointer to some extra data in the same allocated
- * block. For example, if you have:
- *
- * struct foo {
- * char *name;
- * int bar;
- * };
- *
- * you can do:
- *
- * struct foo *f;
- * FLEXPTR_ALLOC_STR(f, name, src);
- *
- * and "name" will point to a block of memory after the struct, which will be
- * freed along with the struct (but the pointer can be repointed anywhere).
- *
- * The *_STR variants accept a string parameter rather than a ptr/len
- * combination.
- *
- * Note that these macros will evaluate the first parameter multiple
- * times, and it must be assignable as an lvalue.
- */
-#define FLEX_ALLOC_MEM(x, flexname, buf, len) do { \
- size_t flex_array_len_ = (len); \
- (x) = calloc(1, st_add3(sizeof(*(x)), flex_array_len_, 1)); \
- BUG_ON(!(x)); \
- memcpy((void *)(x)->flexname, (buf), flex_array_len_); \
-} while (0)
-#define FLEXPTR_ALLOC_MEM(x, ptrname, buf, len) do { \
- size_t flex_array_len_ = (len); \
- (x) = xcalloc(1, st_add3(sizeof(*(x)), flex_array_len_, 1)); \
- memcpy((x) + 1, (buf), flex_array_len_); \
- (x)->ptrname = (void *)((x) + 1); \
-} while (0)
-#define FLEX_ALLOC_STR(x, flexname, str) \
- FLEX_ALLOC_MEM((x), flexname, (str), strlen(str))
-#define FLEXPTR_ALLOC_STR(x, ptrname, str) \
- FLEXPTR_ALLOC_MEM((x), ptrname, (str), strlen(str))
-
-#ifdef __cplusplus
-}
-#endif
-
-#endif
diff --git a/include/erofs/hashmap.h b/include/erofs/hashmap.h
deleted file mode 100644
index 484948e..0000000
--- a/include/erofs/hashmap.h
+++ /dev/null
@@ -1,115 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef __EROFS_HASHMAP_H
-#define __EROFS_HASHMAP_H
-
-#ifdef __cplusplus
-extern "C"
-{
-#endif
-
-/* Copied from https://github.com/git/git.git */
-#include <stdio.h>
-#include <stdlib.h>
-#include <string.h>
-
-#include "flex-array.h"
-
-/*
- * Generic implementation of hash-based key-value mappings.
- * See Documentation/technical/api-hashmap.txt.
- */
-
-/* FNV-1 functions */
-unsigned int strhash(const char *str);
-unsigned int strihash(const char *str);
-unsigned int memhash(const void *buf, size_t len);
-unsigned int memihash(const void *buf, size_t len);
-
-static inline unsigned int sha1hash(const unsigned char *sha1)
-{
- /*
- * Equivalent to 'return *(unsigned int *)sha1;', but safe on
- * platforms that don't support unaligned reads.
- */
- unsigned int hash;
-
- memcpy(&hash, sha1, sizeof(hash));
- return hash;
-}
-
-/* data structures */
-struct hashmap_entry {
- struct hashmap_entry *next;
- unsigned int hash;
-};
-
-typedef int (*hashmap_cmp_fn)(const void *entry, const void *entry_or_key,
- const void *keydata);
-
-struct hashmap {
- struct hashmap_entry **table;
- hashmap_cmp_fn cmpfn;
- unsigned int size, tablesize, grow_at, shrink_at;
-};
-
-struct hashmap_iter {
- struct hashmap *map;
- struct hashmap_entry *next;
- unsigned int tablepos;
-};
-
-/* hashmap functions */
-void hashmap_init(struct hashmap *map, hashmap_cmp_fn equals_function,
- size_t initial_size);
-int hashmap_free(struct hashmap *map);
-
-/* hashmap_entry functions */
-static inline void hashmap_entry_init(void *entry, unsigned int hash)
-{
- struct hashmap_entry *e = entry;
-
- e->hash = hash;
- e->next = NULL;
-}
-
-void *hashmap_get(const struct hashmap *map, const void *key, const void *keydata);
-void *hashmap_get_next(const struct hashmap *map, const void *entry);
-void hashmap_add(struct hashmap *map, void *entry);
-void *hashmap_remove(struct hashmap *map, const void *key);
-
-static inline void *hashmap_get_from_hash(const struct hashmap *map,
- unsigned int hash,
- const void *keydata)
-{
- struct hashmap_entry key;
-
- hashmap_entry_init(&key, hash);
- return hashmap_get(map, &key, keydata);
-}
-
-/* hashmap_iter functions */
-void hashmap_iter_init(struct hashmap *map, struct hashmap_iter *iter);
-void *hashmap_iter_next(struct hashmap_iter *iter);
-static inline void *hashmap_iter_first(struct hashmap *map,
- struct hashmap_iter *iter)
-{
- hashmap_iter_init(map, iter);
- return hashmap_iter_next(iter);
-}
-
-static inline void hashmap_disable_shrink(struct hashmap * map)
-{
- map->shrink_at = 0;
-}
-/* string interning */
-const void *memintern(const void *data, size_t len);
-static inline const char *strintern(const char *string)
-{
- return memintern(string, strlen(string));
-}
-
-#ifdef __cplusplus
-}
-#endif
-
-#endif
diff --git a/lib/Makefile.am b/lib/Makefile.am
index 5ef2122..25d08f2 100644
--- a/lib/Makefile.am
+++ b/lib/Makefile.am
@@ -10,8 +10,6 @@ noinst_HEADERS = $(top_srcdir)/include/erofs_fs.h \
$(top_srcdir)/include/erofs/diskbuf.h \
$(top_srcdir)/include/erofs/err.h \
$(top_srcdir)/include/erofs/exclude.h \
- $(top_srcdir)/include/erofs/flex-array.h \
- $(top_srcdir)/include/erofs/hashmap.h \
$(top_srcdir)/include/erofs/inode.h \
$(top_srcdir)/include/erofs/internal.h \
$(top_srcdir)/include/erofs/io.h \
@@ -43,7 +41,7 @@ endif
liberofs_la_SOURCES = config.c io.c cache.c super.c inode.c xattr.c exclude.c \
namei.c data.c compress.c compressor.c zmap.c decompress.c \
- compress_hints.c hashmap.c sha256.c blobchunk.c dir.c \
+ compress_hints.c sha256.c blobchunk.c dir.c \
fragments.c dedupe.c uuid_unparse.c uuid.c tar.c \
block_list.c rebuild.c diskbuf.c bitops.c dedupe_ext.c \
vmdk.c metabox.c global.c importer.c base64.c
diff --git a/lib/hashmap.c b/lib/hashmap.c
deleted file mode 100644
index 45916ae..0000000
--- a/lib/hashmap.c
+++ /dev/null
@@ -1,280 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0
-/*
- * Copied from https://github.com/git/git.git
- * Generic implementation of hash-based key value mappings.
- */
-#include "erofs/hashmap.h"
-
-#define FNV32_BASE ((unsigned int)0x811c9dc5)
-#define FNV32_PRIME ((unsigned int)0x01000193)
-
-unsigned int strhash(const char *str)
-{
- unsigned int c, hash = FNV32_BASE;
-
- while ((c = (unsigned char)*str++))
- hash = (hash * FNV32_PRIME) ^ c;
- return hash;
-}
-
-unsigned int strihash(const char *str)
-{
- unsigned int c, hash = FNV32_BASE;
-
- while ((c = (unsigned char)*str++)) {
- if (c >= 'a' && c <= 'z')
- c -= 'a' - 'A';
- hash = (hash * FNV32_PRIME) ^ c;
- }
- return hash;
-}
-
-unsigned int memhash(const void *buf, size_t len)
-{
- unsigned int hash = FNV32_BASE;
- unsigned char *ucbuf = (unsigned char *)buf;
-
- while (len--) {
- unsigned int c = *ucbuf++;
-
- hash = (hash * FNV32_PRIME) ^ c;
- }
- return hash;
-}
-
-unsigned int memihash(const void *buf, size_t len)
-{
- unsigned int hash = FNV32_BASE;
- unsigned char *ucbuf = (unsigned char *)buf;
-
- while (len--) {
- unsigned int c = *ucbuf++;
-
- if (c >= 'a' && c <= 'z')
- c -= 'a' - 'A';
- hash = (hash * FNV32_PRIME) ^ c;
- }
- return hash;
-}
-
-#define HASHMAP_INITIAL_SIZE 64
-/* grow / shrink by 2^2 */
-#define HASHMAP_RESIZE_BITS 2
-/* load factor in percent */
-#define HASHMAP_LOAD_FACTOR 80
-
-static void alloc_table(struct hashmap *map, unsigned int size)
-{
- map->tablesize = size;
- map->table = calloc(size, sizeof(struct hashmap_entry *));
- BUG_ON(!map->table);
-
- /* calculate resize thresholds for new size */
- map->grow_at = (unsigned int)((uint64_t)size * HASHMAP_LOAD_FACTOR / 100);
- if (size <= HASHMAP_INITIAL_SIZE)
- map->shrink_at = 0;
- else
- /*
- * The shrink-threshold must be slightly smaller than
- * (grow-threshold / resize-factor) to prevent erratic resizing,
- * thus we divide by (resize-factor + 1).
- */
- map->shrink_at = map->grow_at / ((1 << HASHMAP_RESIZE_BITS) + 1);
-}
-
-static inline int entry_equals(const struct hashmap *map,
- const struct hashmap_entry *e1,
- const struct hashmap_entry *e2,
- const void *keydata)
-{
- return (e1 == e2) || (e1->hash == e2->hash && !map->cmpfn(e1, e2, keydata));
-}
-
-static inline unsigned int bucket(const struct hashmap *map,
- const struct hashmap_entry *key)
-{
- return key->hash & (map->tablesize - 1);
-}
-
-static void rehash(struct hashmap *map, unsigned int newsize)
-{
- unsigned int i, oldsize = map->tablesize;
- struct hashmap_entry **oldtable = map->table;
-
- alloc_table(map, newsize);
- for (i = 0; i < oldsize; i++) {
- struct hashmap_entry *e = oldtable[i];
-
- while (e) {
- struct hashmap_entry *next = e->next;
- unsigned int b = bucket(map, e);
-
- e->next = map->table[b];
- map->table[b] = e;
- e = next;
- }
- }
- free(oldtable);
-}
-
-static inline struct hashmap_entry **find_entry_ptr(const struct hashmap *map,
- const struct hashmap_entry *key,
- const void *keydata)
-{
- struct hashmap_entry **e = &map->table[bucket(map, key)];
-
- while (*e && !entry_equals(map, *e, key, keydata))
- e = &(*e)->next;
- return e;
-}
-
-static int always_equal(const void *unused1, const void *unused2, const void *unused3)
-{
- return 0;
-}
-
-void hashmap_init(struct hashmap *map, hashmap_cmp_fn equals_function,
- size_t initial_size)
-{
- unsigned int size = HASHMAP_INITIAL_SIZE;
-
- map->size = 0;
- map->cmpfn = equals_function ? equals_function : always_equal;
-
- /* calculate initial table size and allocate the table */
- initial_size = (unsigned int)((uint64_t)initial_size * 100
- / HASHMAP_LOAD_FACTOR);
- while (initial_size > size)
- size <<= HASHMAP_RESIZE_BITS;
- alloc_table(map, size);
-}
-
-int hashmap_free(struct hashmap *map)
-{
- if (map && map->table) {
- struct hashmap_iter iter;
- struct hashmap_entry *e;
-
- hashmap_iter_init(map, &iter);
- e = hashmap_iter_next(&iter);
- if (e)
- return -EBUSY;
-
- free(map->table);
- memset(map, 0, sizeof(*map));
- }
- return 0;
-}
-
-void *hashmap_get(const struct hashmap *map, const void *key, const void *keydata)
-{
- return *find_entry_ptr(map, key, keydata);
-}
-
-void *hashmap_get_next(const struct hashmap *map, const void *entry)
-{
- struct hashmap_entry *e = ((struct hashmap_entry *)entry)->next;
-
- for (; e; e = e->next)
- if (entry_equals(map, entry, e, NULL))
- return e;
- return NULL;
-}
-
-void hashmap_add(struct hashmap *map, void *entry)
-{
- unsigned int b = bucket(map, entry);
-
- /* add entry */
- ((struct hashmap_entry *)entry)->next = map->table[b];
- map->table[b] = entry;
-
- /* fix size and rehash if appropriate */
- map->size++;
- if (map->size > map->grow_at)
- rehash(map, map->tablesize << HASHMAP_RESIZE_BITS);
-}
-
-void *hashmap_remove(struct hashmap *map, const void *entry)
-{
- struct hashmap_entry *old;
- struct hashmap_entry **e = &map->table[bucket(map, entry)];
-
- while (*e && *e != entry)
- e = &(*e)->next;
-
- if (!*e)
- return NULL;
-
- /* remove existing entry */
- old = *e;
- *e = old->next;
- old->next = NULL;
-
- /* fix size and rehash if appropriate */
- map->size--;
- if (map->size < map->shrink_at)
- rehash(map, map->tablesize >> HASHMAP_RESIZE_BITS);
- return old;
-}
-
-void hashmap_iter_init(struct hashmap *map, struct hashmap_iter *iter)
-{
- iter->map = map;
- iter->tablepos = 0;
- iter->next = NULL;
-}
-
-void *hashmap_iter_next(struct hashmap_iter *iter)
-{
- struct hashmap_entry *current = iter->next;
-
- for (;;) {
- if (current) {
- iter->next = current->next;
- return current;
- }
-
- if (iter->tablepos >= iter->map->tablesize)
- return NULL;
-
- current = iter->map->table[iter->tablepos++];
- }
-}
-
-struct pool_entry {
- struct hashmap_entry ent;
- size_t len;
- unsigned char data[FLEX_ARRAY];
-};
-
-static int pool_entry_cmp(const struct pool_entry *e1,
- const struct pool_entry *e2,
- const unsigned char *keydata)
-{
- return e1->data != keydata &&
- (e1->len != e2->len || memcmp(e1->data, keydata, e1->len));
-}
-
-const void *memintern(const void *data, size_t len)
-{
- static struct hashmap map;
- struct pool_entry key, *e;
-
- /* initialize string pool hashmap */
- if (!map.tablesize)
- hashmap_init(&map, (hashmap_cmp_fn)pool_entry_cmp, 0);
-
- /* lookup interned string in pool */
- hashmap_entry_init(&key, memhash(data, len));
- key.len = len;
- e = hashmap_get(&map, &key, data);
- if (!e) {
- /* not found: create it */
- FLEX_ALLOC_MEM(e, data, data, len);
- hashmap_entry_init(e, key.ent.hash);
- e->len = len;
- hashmap_add(&map, e);
- }
- return e->data;
-}
--
2.47.3
prev parent reply other threads:[~2026-08-26 13:44 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-08-26 13:43 [PATCH 1/4] erofs-utils: lib: fix wrong forward declaration of `erofs_metamgr` Gao Xiang
2026-08-26 13:43 ` [PATCH 2/4] erofs-utils: lib: use `struct erofs_device_info` for the primary device Gao Xiang
2026-08-27 1:30 ` Yifan Zhao
2026-08-27 7:59 ` Gao Xiang
2026-08-26 13:43 ` [PATCH 3/4] erofs-utils: lib: refactor unencoded chunk handling Gao Xiang
2026-08-26 13:43 ` Gao Xiang [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260826134321.11835-4-xiang@kernel.org \
--to=xiang@kernel.org \
--cc=linux-erofs@lists.ozlabs.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox