reftable/basics: provide wrappers for big endian conversion

We're using a mixture of big endian conversion functions provided by both the reftable library, but also by the Git codebase. Refactor the code so that we exclusively use reftable-provided wrappers in order to untangle us from the Git codebase. Signed-off-by: Patrick Steinhardt <ps@pks.im> Signed-off-by: Junio C Hamano <gitster@pobox.com>

Patrick Steinhardt committed Feb 18, 2025 at 10:20 UTC e676694298c4a8d9f6fdf3844cbfb03bbee552cc
7 files changed +127 -58
reftable/basics.c
-19
@@ -147,25 +147,6 @@ char *reftable_buf_detach(struct reftable_buf *buf)
147 return result;
148 }
149
150 -void put_be24(uint8_t *out, uint32_t i)
151 -{
152 - out[0] = (uint8_t)((i >> 16) & 0xff);
153 - out[1] = (uint8_t)((i >> 8) & 0xff);
154 - out[2] = (uint8_t)(i & 0xff);
155 -}
156 -
157 -uint32_t get_be24(uint8_t *in)
158 -{
159 - return (uint32_t)(in[0]) << 16 | (uint32_t)(in[1]) << 8 |
160 - (uint32_t)(in[2]);
161 -}
162 -
163 -void put_be16(uint8_t *out, uint16_t i)
164 -{
165 - out[0] = (uint8_t)((i >> 8) & 0xff);
166 - out[1] = (uint8_t)(i & 0xff);
167 -}
168 -
150 size_t binsearch(size_t sz, int (*f)(size_t k, void *args), void *args)
151 {
152 size_t lo = 0;
reftable/basics.h
+73 -3
@@ -76,9 +76,79 @@ char *reftable_buf_detach(struct reftable_buf *buf);
76
77 /* Bigendian en/decoding of integers */
78
79 -void put_be24(uint8_t *out, uint32_t i);
80 -uint32_t get_be24(uint8_t *in);
81 -void put_be16(uint8_t *out, uint16_t i);
79 +static inline void reftable_put_be16(void *out, uint16_t i)
80 +{
81 + unsigned char *p = out;
82 + p[0] = (uint8_t)((i >> 8) & 0xff);
83 + p[1] = (uint8_t)((i >> 0) & 0xff);
84 +}
85 +
86 +static inline void reftable_put_be24(void *out, uint32_t i)
87 +{
88 + unsigned char *p = out;
89 + p[0] = (uint8_t)((i >> 16) & 0xff);
90 + p[1] = (uint8_t)((i >> 8) & 0xff);
91 + p[2] = (uint8_t)((i >> 0) & 0xff);
92 +}
93 +
94 +static inline void reftable_put_be32(void *out, uint32_t i)
95 +{
96 + unsigned char *p = out;
97 + p[0] = (uint8_t)((i >> 24) & 0xff);
98 + p[1] = (uint8_t)((i >> 16) & 0xff);
99 + p[2] = (uint8_t)((i >> 8) & 0xff);
100 + p[3] = (uint8_t)((i >> 0) & 0xff);
101 +}
102 +
103 +static inline void reftable_put_be64(void *out, uint64_t i)
104 +{
105 + unsigned char *p = out;
106 + p[0] = (uint8_t)((i >> 56) & 0xff);
107 + p[1] = (uint8_t)((i >> 48) & 0xff);
108 + p[2] = (uint8_t)((i >> 40) & 0xff);
109 + p[3] = (uint8_t)((i >> 32) & 0xff);
110 + p[4] = (uint8_t)((i >> 24) & 0xff);
111 + p[5] = (uint8_t)((i >> 16) & 0xff);
112 + p[6] = (uint8_t)((i >> 8) & 0xff);
113 + p[7] = (uint8_t)((i >> 0) & 0xff);
114 +}
115 +
116 +static inline uint16_t reftable_get_be16(const void *in)
117 +{
118 + const unsigned char *p = in;
119 + return (uint16_t)(p[0]) << 8 |
120 + (uint16_t)(p[1]) << 0;
121 +}
122 +
123 +static inline uint32_t reftable_get_be24(const void *in)
124 +{
125 + const unsigned char *p = in;
126 + return (uint32_t)(p[0]) << 16 |
127 + (uint32_t)(p[1]) << 8 |
128 + (uint32_t)(p[2]) << 0;
129 +}
130 +
131 +static inline uint32_t reftable_get_be32(const void *in)
132 +{
133 + const unsigned char *p = in;
134 + return (uint32_t)(p[0]) << 24 |
135 + (uint32_t)(p[1]) << 16 |
136 + (uint32_t)(p[2]) << 8|
137 + (uint32_t)(p[3]) << 0;
138 +}
139 +
140 +static inline uint64_t reftable_get_be64(const void *in)
141 +{
142 + const unsigned char *p = in;
143 + return (uint64_t)(p[0]) << 56 |
144 + (uint64_t)(p[1]) << 48 |
145 + (uint64_t)(p[2]) << 40 |
146 + (uint64_t)(p[3]) << 32 |
147 + (uint64_t)(p[4]) << 24 |
148 + (uint64_t)(p[5]) << 16 |
149 + (uint64_t)(p[6]) << 8 |
150 + (uint64_t)(p[7]) << 0;
151 +}
152
153 /*
154 * find smallest index i in [0, sz) at which `f(i) > 0`, assuming that f is
reftable/block.c
+6 -6
@@ -147,13 +147,13 @@ done:
147 int block_writer_finish(struct block_writer *w)
148 {
149 for (uint32_t i = 0; i < w->restart_len; i++) {
150 - put_be24(w->block + w->next, w->restarts[i]);
150 + reftable_put_be24(w->block + w->next, w->restarts[i]);
151 w->next += 3;
152 }
153
154 - put_be16(w->block + w->next, w->restart_len);
154 + reftable_put_be16(w->block + w->next, w->restart_len);
155 w->next += 2;
156 - put_be24(w->block + 1 + w->header_off, w->next);
156 + reftable_put_be24(w->block + 1 + w->header_off, w->next);
157
158 /*
159 * Log records are stored zlib-compressed. Note that the compression
@@ -215,7 +215,7 @@ int block_reader_init(struct block_reader *br, struct reftable_block *block,
215 {
216 uint32_t full_block_size = table_block_size;
217 uint8_t typ = block->data[header_off];
218 - uint32_t sz = get_be24(block->data + header_off + 1);
218 + uint32_t sz = reftable_get_be24(block->data + header_off + 1);
219 int err = 0;
220 uint16_t restart_count = 0;
221 uint32_t restart_start = 0;
@@ -299,7 +299,7 @@ int block_reader_init(struct block_reader *br, struct reftable_block *block,
299 full_block_size = sz;
300 }
301
302 - restart_count = get_be16(block->data + sz - 2);
302 + restart_count = reftable_get_be16(block->data + sz - 2);
303 restart_start = sz - 2 - 3 * restart_count;
304 restart_bytes = block->data + restart_start;
305
@@ -354,7 +354,7 @@ int block_reader_first_key(const struct block_reader *br, struct reftable_buf *k
354
355 static uint32_t block_reader_restart_offset(const struct block_reader *br, size_t idx)
356 {
357 - return get_be24(br->restart_bytes + 3 * idx);
357 + return reftable_get_be24(br->restart_bytes + 3 * idx);
358 }
359
360 void block_iter_seek_start(struct block_iter *it, const struct block_reader *br)
reftable/reader.c
+11 -11
@@ -101,18 +101,18 @@ static int parse_footer(struct reftable_reader *r, uint8_t *footer,
101 }
102
103 f++;
104 - r->block_size = get_be24(f);
104 + r->block_size = reftable_get_be24(f);
105
106 f += 3;
107 - r->min_update_index = get_be64(f);
107 + r->min_update_index = reftable_get_be64(f);
108 f += 8;
109 - r->max_update_index = get_be64(f);
109 + r->max_update_index = reftable_get_be64(f);
110 f += 8;
111
112 if (r->version == 1) {
113 r->hash_id = REFTABLE_HASH_SHA1;
114 } else {
115 - switch (get_be32(f)) {
115 + switch (reftable_get_be32(f)) {
116 case REFTABLE_FORMAT_ID_SHA1:
117 r->hash_id = REFTABLE_HASH_SHA1;
118 break;
@@ -127,24 +127,24 @@ static int parse_footer(struct reftable_reader *r, uint8_t *footer,
127 f += 4;
128 }
129
130 - r->ref_offsets.index_offset = get_be64(f);
130 + r->ref_offsets.index_offset = reftable_get_be64(f);
131 f += 8;
132
133 - r->obj_offsets.offset = get_be64(f);
133 + r->obj_offsets.offset = reftable_get_be64(f);
134 f += 8;
135
136 r->object_id_len = r->obj_offsets.offset & ((1 << 5) - 1);
137 r->obj_offsets.offset >>= 5;
138
139 - r->obj_offsets.index_offset = get_be64(f);
139 + r->obj_offsets.index_offset = reftable_get_be64(f);
140 f += 8;
141 - r->log_offsets.offset = get_be64(f);
141 + r->log_offsets.offset = reftable_get_be64(f);
142 f += 8;
143 - r->log_offsets.index_offset = get_be64(f);
143 + r->log_offsets.index_offset = reftable_get_be64(f);
144 f += 8;
145
146 computed_crc = crc32(0, footer, f - footer);
147 - file_crc = get_be32(f);
147 + file_crc = reftable_get_be32(f);
148 f += 4;
149 if (computed_crc != file_crc) {
150 err = REFTABLE_FORMAT_ERROR;
@@ -214,7 +214,7 @@ static int32_t extract_block_size(uint8_t *data, uint8_t *typ, uint64_t off,
214
215 *typ = data[0];
216 if (reftable_is_block_type(*typ)) {
217 - result = get_be24(data + 1);
217 + result = reftable_get_be24(data + 1);
218 }
219 return result;
220 }
reftable/record.c
+4 -4
@@ -689,7 +689,7 @@ static int reftable_log_record_key(const void *r, struct reftable_buf *dest)
689 return err;
690
691 ts = (~ts) - rec->update_index;
692 - put_be64(&i64[0], ts);
692 + reftable_put_be64(&i64[0], ts);
693
694 err = reftable_buf_add(dest, i64, sizeof(i64));
695 if (err < 0)
@@ -814,7 +814,7 @@ static int reftable_log_record_encode(const void *rec, struct string_view s,
814 if (s.len < 2)
815 return -1;
816
817 - put_be16(s.buf, r->value.update.tz_offset);
817 + reftable_put_be16(s.buf, r->value.update.tz_offset);
818 string_view_consume(&s, 2);
819
820 n = encode_string(
@@ -846,7 +846,7 @@ static int reftable_log_record_decode(void *rec, struct reftable_buf key,
846 }
847
848 memcpy(r->refname, key.buf, key.len - 8);
849 - ts = get_be64(key.buf + key.len - 8);
849 + ts = reftable_get_be64((unsigned char *)key.buf + key.len - 8);
850
851 r->update_index = (~max) - ts;
852
@@ -937,7 +937,7 @@ static int reftable_log_record_decode(void *rec, struct reftable_buf key,
937 goto done;
938 }
939
940 - r->value.update.tz_offset = get_be16(in.buf);
940 + r->value.update.tz_offset = reftable_get_be16(in.buf);
941 string_view_consume(&in, 2);
942
943 n = decode_string(scratch, in);
reftable/writer.c
+10 -10
@@ -99,9 +99,9 @@ static int writer_write_header(struct reftable_writer *w, uint8_t *dest)
99
100 dest[4] = writer_version(w);
101
102 - put_be24(dest + 5, w->opts.block_size);
103 - put_be64(dest + 8, w->min_update_index);
104 - put_be64(dest + 16, w->max_update_index);
102 + reftable_put_be24(dest + 5, w->opts.block_size);
103 + reftable_put_be64(dest + 8, w->min_update_index);
104 + reftable_put_be64(dest + 16, w->max_update_index);
105 if (writer_version(w) == 2) {
106 uint32_t hash_id;
107
@@ -116,7 +116,7 @@ static int writer_write_header(struct reftable_writer *w, uint8_t *dest)
116 return -1;
117 }
118
119 - put_be32(dest + 24, hash_id);
119 + reftable_put_be32(dest + 24, hash_id);
120 }
121
122 return header_size(writer_version(w));
@@ -730,19 +730,19 @@ int reftable_writer_close(struct reftable_writer *w)
730 }
731
732 p += writer_write_header(w, footer);
733 - put_be64(p, w->stats.ref_stats.index_offset);
733 + reftable_put_be64(p, w->stats.ref_stats.index_offset);
734 p += 8;
735 - put_be64(p, (w->stats.obj_stats.offset) << 5 | w->stats.object_id_len);
735 + reftable_put_be64(p, (w->stats.obj_stats.offset) << 5 | w->stats.object_id_len);
736 p += 8;
737 - put_be64(p, w->stats.obj_stats.index_offset);
737 + reftable_put_be64(p, w->stats.obj_stats.index_offset);
738 p += 8;
739
740 - put_be64(p, w->stats.log_stats.offset);
740 + reftable_put_be64(p, w->stats.log_stats.offset);
741 p += 8;
742 - put_be64(p, w->stats.log_stats.index_offset);
742 + reftable_put_be64(p, w->stats.log_stats.index_offset);
743 p += 8;
744
745 - put_be32(p, crc32(0, footer, p - footer));
745 + reftable_put_be32(p, crc32(0, footer, p - footer));
746 p += 4;
747
748 err = w->flush(w->write_arg);
t/unit-tests/t-reftable-basics.c
+23 -5
@@ -128,12 +128,30 @@ int cmd_main(int argc UNUSED, const char *argv[] UNUSED)
128 reftable_buf_release(&b);
129 }
130
131 - if_test ("put_be24 and get_be24 work") {
131 + if_test ("reftable_put_be64 and reftable_get_be64 work") {
132 + uint64_t in = 0x1122334455667788;
133 + uint8_t dest[8];
134 + uint64_t out;
135 + reftable_put_be64(dest, in);
136 + out = reftable_get_be64(dest);
137 + check_int(in, ==, out);
138 + }
139 +
140 + if_test ("reftable_put_be32 and reftable_get_be32 work") {
141 + uint32_t in = 0x11223344;
142 + uint8_t dest[4];
143 + uint32_t out;
144 + reftable_put_be32(dest, in);
145 + out = reftable_get_be32(dest);
146 + check_int(in, ==, out);
147 + }
148 +
149 + if_test ("reftable_put_be24 and reftable_get_be24 work") {
150 uint32_t in = 0x112233;
151 uint8_t dest[3];
152 uint32_t out;
135 - put_be24(dest, in);
136 - out = get_be24(dest);
153 + reftable_put_be24(dest, in);
154 + out = reftable_get_be24(dest);
155 check_int(in, ==, out);
156 }
157
@@ -141,8 +159,8 @@ int cmd_main(int argc UNUSED, const char *argv[] UNUSED)
159 uint32_t in = 0xfef1;
160 uint8_t dest[3];
161 uint32_t out;
144 - put_be16(dest, in);
145 - out = get_be16(dest);
162 + reftable_put_be16(dest, in);
163 + out = reftable_get_be16(dest);
164 check_int(in, ==, out);
165 }
166