diff options
Diffstat (limited to 'reftable')
54 files changed, 11701 insertions, 0 deletions
diff --git a/reftable/LICENSE b/reftable/LICENSE new file mode 100644 index 0000000..402e0f9 --- /dev/null +++ b/reftable/LICENSE @@ -0,0 +1,31 @@ +BSD License + +Copyright (c) 2020, Google LLC +All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + +* Redistributions of source code must retain the above copyright notice, +this list of conditions and the following disclaimer. + +* Redistributions in binary form must reproduce the above copyright +notice, this list of conditions and the following disclaimer in the +documentation and/or other materials provided with the distribution. + +* Neither the name of Google LLC nor the names of its contributors may +be used to endorse or promote products derived from this software +without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. diff --git a/reftable/basics.c b/reftable/basics.c new file mode 100644 index 0000000..fea711d --- /dev/null +++ b/reftable/basics.c @@ -0,0 +1,126 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "basics.h" + +void put_be24(uint8_t *out, uint32_t i) +{ + out[0] = (uint8_t)((i >> 16) & 0xff); + out[1] = (uint8_t)((i >> 8) & 0xff); + out[2] = (uint8_t)(i & 0xff); +} + +uint32_t get_be24(uint8_t *in) +{ + return (uint32_t)(in[0]) << 16 | (uint32_t)(in[1]) << 8 | + (uint32_t)(in[2]); +} + +void put_be16(uint8_t *out, uint16_t i) +{ + out[0] = (uint8_t)((i >> 8) & 0xff); + out[1] = (uint8_t)(i & 0xff); +} + +size_t binsearch(size_t sz, int (*f)(size_t k, void *args), void *args) +{ + size_t lo = 0; + size_t hi = sz; + + /* Invariants: + * + * (hi == sz) || f(hi) == true + * (lo == 0 && f(0) == true) || fi(lo) == false + */ + while (hi - lo > 1) { + size_t mid = lo + (hi - lo) / 2; + int ret = f(mid, args); + if (ret < 0) + return sz; + + if (ret > 0) + hi = mid; + else + lo = mid; + } + + if (lo) + return hi; + + return f(0, args) ? 0 : 1; +} + +void free_names(char **a) +{ + char **p; + if (!a) { + return; + } + for (p = a; *p; p++) { + reftable_free(*p); + } + reftable_free(a); +} + +size_t names_length(char **names) +{ + char **p = names; + while (*p) + p++; + return p - names; +} + +void parse_names(char *buf, int size, char ***namesp) +{ + char **names = NULL; + size_t names_cap = 0; + size_t names_len = 0; + + char *p = buf; + char *end = buf + size; + while (p < end) { + char *next = strchr(p, '\n'); + if (next && next < end) { + *next = 0; + } else { + next = end; + } + if (p < next) { + REFTABLE_ALLOC_GROW(names, names_len + 1, names_cap); + names[names_len++] = xstrdup(p); + } + p = next + 1; + } + + REFTABLE_REALLOC_ARRAY(names, names_len + 1); + names[names_len] = NULL; + *namesp = names; +} + +int names_equal(char **a, char **b) +{ + int i = 0; + for (; a[i] && b[i]; i++) { + if (strcmp(a[i], b[i])) { + return 0; + } + } + + return a[i] == b[i]; +} + +int common_prefix_size(struct strbuf *a, struct strbuf *b) +{ + int p = 0; + for (; p < a->len && p < b->len; p++) { + if (a->buf[p] != b->buf[p]) + break; + } + + return p; +} diff --git a/reftable/basics.h b/reftable/basics.h new file mode 100644 index 0000000..523ecd5 --- /dev/null +++ b/reftable/basics.h @@ -0,0 +1,74 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef BASICS_H +#define BASICS_H + +/* + * miscellaneous utilities that are not provided by Git. + */ + +#include "system.h" + +/* Bigendian en/decoding of integers */ + +void put_be24(uint8_t *out, uint32_t i); +uint32_t get_be24(uint8_t *in); +void put_be16(uint8_t *out, uint16_t i); + +/* + * find smallest index i in [0, sz) at which `f(i) > 0`, assuming that f is + * ascending. Return sz if `f(i) == 0` for all indices. The search is aborted + * and `sz` is returned in case `f(i) < 0`. + * + * Contrary to bsearch(3), this returns something useful if the argument is not + * found. + */ +size_t binsearch(size_t sz, int (*f)(size_t k, void *args), void *args); + +/* + * Frees a NULL terminated array of malloced strings. The array itself is also + * freed. + */ +void free_names(char **a); + +/* parse a newline separated list of names. `size` is the length of the buffer, + * without terminating '\0'. Empty names are discarded. */ +void parse_names(char *buf, int size, char ***namesp); + +/* compares two NULL-terminated arrays of strings. */ +int names_equal(char **a, char **b); + +/* returns the array size of a NULL-terminated array of strings. */ +size_t names_length(char **names); + +/* Allocation routines; they invoke the functions set through + * reftable_set_alloc() */ +void *reftable_malloc(size_t sz); +void *reftable_realloc(void *p, size_t sz); +void reftable_free(void *p); +void *reftable_calloc(size_t nelem, size_t elsize); + +#define REFTABLE_ALLOC_ARRAY(x, alloc) (x) = reftable_malloc(st_mult(sizeof(*(x)), (alloc))) +#define REFTABLE_CALLOC_ARRAY(x, alloc) (x) = reftable_calloc((alloc), sizeof(*(x))) +#define REFTABLE_REALLOC_ARRAY(x, alloc) (x) = reftable_realloc((x), st_mult(sizeof(*(x)), (alloc))) +#define REFTABLE_ALLOC_GROW(x, nr, alloc) \ + do { \ + if ((nr) > alloc) { \ + alloc = 2 * (alloc) + 1; \ + if (alloc < (nr)) \ + alloc = (nr); \ + REFTABLE_REALLOC_ARRAY(x, alloc); \ + } \ + } while (0) + +/* Find the longest shared prefix size of `a` and `b` */ +struct strbuf; +int common_prefix_size(struct strbuf *a, struct strbuf *b); + +#endif diff --git a/reftable/basics_test.c b/reftable/basics_test.c new file mode 100644 index 0000000..997c4d9 --- /dev/null +++ b/reftable/basics_test.c @@ -0,0 +1,105 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "system.h" + +#include "basics.h" +#include "test_framework.h" +#include "reftable-tests.h" + +struct integer_needle_lesseq_args { + int needle; + int *haystack; +}; + +static int integer_needle_lesseq(size_t i, void *_args) +{ + struct integer_needle_lesseq_args *args = _args; + return args->needle <= args->haystack[i]; +} + +static void test_binsearch(void) +{ + int haystack[] = { 2, 4, 6, 8, 10 }; + struct { + int needle; + size_t expected_idx; + } testcases[] = { + {-9000, 0}, + {-1, 0}, + {0, 0}, + {2, 0}, + {3, 1}, + {4, 1}, + {7, 3}, + {9, 4}, + {10, 4}, + {11, 5}, + {9000, 5}, + }; + size_t i = 0; + + for (i = 0; i < ARRAY_SIZE(testcases); i++) { + struct integer_needle_lesseq_args args = { + .haystack = haystack, + .needle = testcases[i].needle, + }; + size_t idx; + + idx = binsearch(ARRAY_SIZE(haystack), &integer_needle_lesseq, &args); + EXPECT(idx == testcases[i].expected_idx); + } +} + +static void test_names_length(void) +{ + char *a[] = { "a", "b", NULL }; + EXPECT(names_length(a) == 2); +} + +static void test_parse_names_normal(void) +{ + char in[] = "a\nb\n"; + char **out = NULL; + parse_names(in, strlen(in), &out); + EXPECT(!strcmp(out[0], "a")); + EXPECT(!strcmp(out[1], "b")); + EXPECT(!out[2]); + free_names(out); +} + +static void test_parse_names_drop_empty(void) +{ + char in[] = "a\n\n"; + char **out = NULL; + parse_names(in, strlen(in), &out); + EXPECT(!strcmp(out[0], "a")); + EXPECT(!out[1]); + free_names(out); +} + +static void test_common_prefix(void) +{ + struct strbuf s1 = STRBUF_INIT; + struct strbuf s2 = STRBUF_INIT; + strbuf_addstr(&s1, "abcdef"); + strbuf_addstr(&s2, "abc"); + EXPECT(common_prefix_size(&s1, &s2) == 3); + strbuf_release(&s1); + strbuf_release(&s2); +} + +int basics_test_main(int argc, const char *argv[]) +{ + RUN_TEST(test_common_prefix); + RUN_TEST(test_parse_names_normal); + RUN_TEST(test_parse_names_drop_empty); + RUN_TEST(test_binsearch); + RUN_TEST(test_names_length); + return 0; +} diff --git a/reftable/block.c b/reftable/block.c new file mode 100644 index 0000000..3e87460 --- /dev/null +++ b/reftable/block.c @@ -0,0 +1,531 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "block.h" + +#include "blocksource.h" +#include "constants.h" +#include "record.h" +#include "reftable-error.h" +#include "system.h" +#include <zlib.h> + +int header_size(int version) +{ + switch (version) { + case 1: + return 24; + case 2: + return 28; + } + abort(); +} + +int footer_size(int version) +{ + switch (version) { + case 1: + return 68; + case 2: + return 72; + } + abort(); +} + +static int block_writer_register_restart(struct block_writer *w, int n, + int is_restart, struct strbuf *key) +{ + int rlen = w->restart_len; + if (rlen >= MAX_RESTARTS) { + is_restart = 0; + } + + if (is_restart) { + rlen++; + } + if (2 + 3 * rlen + n > w->block_size - w->next) + return -1; + if (is_restart) { + REFTABLE_ALLOC_GROW(w->restarts, w->restart_len + 1, w->restart_cap); + w->restarts[w->restart_len++] = w->next; + } + + w->next += n; + + strbuf_reset(&w->last_key); + strbuf_addbuf(&w->last_key, key); + w->entries++; + return 0; +} + +void block_writer_init(struct block_writer *bw, uint8_t typ, uint8_t *buf, + uint32_t block_size, uint32_t header_off, int hash_size) +{ + bw->buf = buf; + bw->hash_size = hash_size; + bw->block_size = block_size; + bw->header_off = header_off; + bw->buf[header_off] = typ; + bw->next = header_off + 4; + bw->restart_interval = 16; + bw->entries = 0; + bw->restart_len = 0; + bw->last_key.len = 0; +} + +uint8_t block_writer_type(struct block_writer *bw) +{ + return bw->buf[bw->header_off]; +} + +/* Adds the reftable_record to the block. Returns -1 if it does not fit, 0 on + success. Returns REFTABLE_API_ERROR if attempting to write a record with + empty key. */ +int block_writer_add(struct block_writer *w, struct reftable_record *rec) +{ + struct strbuf empty = STRBUF_INIT; + struct strbuf last = + w->entries % w->restart_interval == 0 ? empty : w->last_key; + struct string_view out = { + .buf = w->buf + w->next, + .len = w->block_size - w->next, + }; + + struct string_view start = out; + + int is_restart = 0; + struct strbuf key = STRBUF_INIT; + int n = 0; + int err = -1; + + reftable_record_key(rec, &key); + if (!key.len) { + err = REFTABLE_API_ERROR; + goto done; + } + + n = reftable_encode_key(&is_restart, out, last, key, + reftable_record_val_type(rec)); + if (n < 0) + goto done; + string_view_consume(&out, n); + + n = reftable_record_encode(rec, out, w->hash_size); + if (n < 0) + goto done; + string_view_consume(&out, n); + + err = block_writer_register_restart(w, start.len - out.len, is_restart, + &key); +done: + strbuf_release(&key); + return err; +} + +int block_writer_finish(struct block_writer *w) +{ + int i; + for (i = 0; i < w->restart_len; i++) { + put_be24(w->buf + w->next, w->restarts[i]); + w->next += 3; + } + + put_be16(w->buf + w->next, w->restart_len); + w->next += 2; + put_be24(w->buf + 1 + w->header_off, w->next); + + if (block_writer_type(w) == BLOCK_TYPE_LOG) { + int block_header_skip = 4 + w->header_off; + uLongf src_len = w->next - block_header_skip; + uLongf dest_cap = src_len * 1.001 + 12; + uint8_t *compressed; + + REFTABLE_ALLOC_ARRAY(compressed, dest_cap); + + while (1) { + uLongf out_dest_len = dest_cap; + int zresult = compress2(compressed, &out_dest_len, + w->buf + block_header_skip, + src_len, 9); + if (zresult == Z_BUF_ERROR && dest_cap < LONG_MAX) { + dest_cap *= 2; + compressed = + reftable_realloc(compressed, dest_cap); + if (compressed) + continue; + } + + if (Z_OK != zresult) { + reftable_free(compressed); + return REFTABLE_ZLIB_ERROR; + } + + memcpy(w->buf + block_header_skip, compressed, + out_dest_len); + w->next = out_dest_len + block_header_skip; + reftable_free(compressed); + break; + } + } + return w->next; +} + +int block_reader_init(struct block_reader *br, struct reftable_block *block, + uint32_t header_off, uint32_t table_block_size, + int hash_size) +{ + uint32_t full_block_size = table_block_size; + uint8_t typ = block->data[header_off]; + uint32_t sz = get_be24(block->data + header_off + 1); + int err = 0; + uint16_t restart_count = 0; + uint32_t restart_start = 0; + uint8_t *restart_bytes = NULL; + + reftable_block_done(&br->block); + + if (!reftable_is_block_type(typ)) { + err = REFTABLE_FORMAT_ERROR; + goto done; + } + + if (typ == BLOCK_TYPE_LOG) { + uint32_t block_header_skip = 4 + header_off; + uLong dst_len = sz - block_header_skip; + uLong src_len = block->len - block_header_skip; + + /* Log blocks specify the *uncompressed* size in their header. */ + REFTABLE_ALLOC_GROW(br->uncompressed_data, sz, + br->uncompressed_cap); + + /* Copy over the block header verbatim. It's not compressed. */ + memcpy(br->uncompressed_data, block->data, block_header_skip); + + if (!br->zstream) { + REFTABLE_CALLOC_ARRAY(br->zstream, 1); + err = inflateInit(br->zstream); + } else { + err = inflateReset(br->zstream); + } + if (err != Z_OK) { + err = REFTABLE_ZLIB_ERROR; + goto done; + } + + br->zstream->next_in = block->data + block_header_skip; + br->zstream->avail_in = src_len; + br->zstream->next_out = br->uncompressed_data + block_header_skip; + br->zstream->avail_out = dst_len; + + /* + * We know both input as well as output size, and we know that + * the sizes should never be bigger than `uInt_MAX` because + * blocks can at most be 16MB large. We can thus use `Z_FINISH` + * here to instruct zlib to inflate the data in one go, which + * is more efficient than using `Z_NO_FLUSH`. + */ + err = inflate(br->zstream, Z_FINISH); + if (err != Z_STREAM_END) { + err = REFTABLE_ZLIB_ERROR; + goto done; + } + err = 0; + + if (br->zstream->total_out + block_header_skip != sz) { + err = REFTABLE_FORMAT_ERROR; + goto done; + } + + /* We're done with the input data. */ + reftable_block_done(block); + block->data = br->uncompressed_data; + block->len = sz; + full_block_size = src_len + block_header_skip - br->zstream->avail_in; + } else if (full_block_size == 0) { + full_block_size = sz; + } else if (sz < full_block_size && sz < block->len && + block->data[sz] != 0) { + /* If the block is smaller than the full block size, it is + padded (data followed by '\0') or the next block is + unaligned. */ + full_block_size = sz; + } + + restart_count = get_be16(block->data + sz - 2); + restart_start = sz - 2 - 3 * restart_count; + restart_bytes = block->data + restart_start; + + /* transfer ownership. */ + br->block = *block; + block->data = NULL; + block->len = 0; + + br->hash_size = hash_size; + br->block_len = restart_start; + br->full_block_size = full_block_size; + br->header_off = header_off; + br->restart_count = restart_count; + br->restart_bytes = restart_bytes; + +done: + return err; +} + +void block_reader_release(struct block_reader *br) +{ + inflateEnd(br->zstream); + reftable_free(br->zstream); + reftable_free(br->uncompressed_data); + reftable_block_done(&br->block); +} + +uint8_t block_reader_type(const struct block_reader *r) +{ + return r->block.data[r->header_off]; +} + +int block_reader_first_key(const struct block_reader *br, struct strbuf *key) +{ + int off = br->header_off + 4, n; + struct string_view in = { + .buf = br->block.data + off, + .len = br->block_len - off, + }; + uint8_t extra = 0; + + strbuf_reset(key); + + n = reftable_decode_key(key, &extra, in); + if (n < 0) + return n; + if (!key->len) + return REFTABLE_FORMAT_ERROR; + + return 0; +} + +static uint32_t block_reader_restart_offset(const struct block_reader *br, int i) +{ + return get_be24(br->restart_bytes + 3 * i); +} + +void block_iter_seek_start(struct block_iter *it, const struct block_reader *br) +{ + it->block = br->block.data; + it->block_len = br->block_len; + it->hash_size = br->hash_size; + strbuf_reset(&it->last_key); + it->next_off = br->header_off + 4; +} + +struct restart_needle_less_args { + int error; + struct strbuf needle; + const struct block_reader *reader; +}; + +static int restart_needle_less(size_t idx, void *_args) +{ + struct restart_needle_less_args *args = _args; + uint32_t off = block_reader_restart_offset(args->reader, idx); + struct string_view in = { + .buf = args->reader->block.data + off, + .len = args->reader->block_len - off, + }; + uint64_t prefix_len, suffix_len; + uint8_t extra; + int n; + + /* + * Records at restart points are stored without prefix compression, so + * there is no need to fully decode the record key here. This removes + * the need for allocating memory. + */ + n = reftable_decode_keylen(in, &prefix_len, &suffix_len, &extra); + if (n < 0 || prefix_len) { + args->error = 1; + return -1; + } + + string_view_consume(&in, n); + if (suffix_len > in.len) { + args->error = 1; + return -1; + } + + n = memcmp(args->needle.buf, in.buf, + args->needle.len < suffix_len ? args->needle.len : suffix_len); + if (n) + return n < 0; + return args->needle.len < suffix_len; +} + +int block_iter_next(struct block_iter *it, struct reftable_record *rec) +{ + struct string_view in = { + .buf = (unsigned char *) it->block + it->next_off, + .len = it->block_len - it->next_off, + }; + struct string_view start = in; + uint8_t extra = 0; + int n = 0; + + if (it->next_off >= it->block_len) + return 1; + + n = reftable_decode_key(&it->last_key, &extra, in); + if (n < 0) + return -1; + if (!it->last_key.len) + return REFTABLE_FORMAT_ERROR; + + string_view_consume(&in, n); + n = reftable_record_decode(rec, it->last_key, extra, in, it->hash_size, + &it->scratch); + if (n < 0) + return -1; + string_view_consume(&in, n); + + it->next_off += start.len - in.len; + return 0; +} + +void block_iter_reset(struct block_iter *it) +{ + strbuf_reset(&it->last_key); + it->next_off = 0; + it->block = NULL; + it->block_len = 0; + it->hash_size = 0; +} + +void block_iter_close(struct block_iter *it) +{ + strbuf_release(&it->last_key); + strbuf_release(&it->scratch); +} + +int block_iter_seek_key(struct block_iter *it, const struct block_reader *br, + struct strbuf *want) +{ + struct restart_needle_less_args args = { + .needle = *want, + .reader = br, + }; + struct reftable_record rec; + int err = 0; + size_t i; + + /* + * Perform a binary search over the block's restart points, which + * avoids doing a linear scan over the whole block. Like this, we + * identify the section of the block that should contain our key. + * + * Note that we explicitly search for the first restart point _greater_ + * than the sought-after record, not _greater or equal_ to it. In case + * the sought-after record is located directly at the restart point we + * would otherwise start doing the linear search at the preceding + * restart point. While that works alright, we would end up scanning + * too many record. + */ + i = binsearch(br->restart_count, &restart_needle_less, &args); + if (args.error) { + err = REFTABLE_FORMAT_ERROR; + goto done; + } + + /* + * Now there are multiple cases: + * + * - `i == 0`: The wanted record is smaller than the record found at + * the first restart point. As the first restart point is the first + * record in the block, our wanted record cannot be located in this + * block at all. We still need to position the iterator so that the + * next call to `block_iter_next()` will yield an end-of-iterator + * signal. + * + * - `i == restart_count`: The wanted record was not found at any of + * the restart points. As there is no restart point at the end of + * the section the record may thus be contained in the last block. + * + * - `i > 0`: The wanted record must be contained in the section + * before the found restart point. We thus do a linear search + * starting from the preceding restart point. + */ + if (i > 0) + it->next_off = block_reader_restart_offset(br, i - 1); + else + it->next_off = br->header_off + 4; + it->block = br->block.data; + it->block_len = br->block_len; + it->hash_size = br->hash_size; + + reftable_record_init(&rec, block_reader_type(br)); + + /* + * We're looking for the last entry less than the wanted key so that + * the next call to `block_reader_next()` would yield the wanted + * record. We thus don't want to position our reader at the sought + * after record, but one before. To do so, we have to go one entry too + * far and then back up. + */ + while (1) { + size_t prev_off = it->next_off; + + err = block_iter_next(it, &rec); + if (err < 0) + goto done; + if (err > 0) { + it->next_off = prev_off; + err = 0; + goto done; + } + + /* + * Check whether the current key is greater or equal to the + * sought-after key. In case it is greater we know that the + * record does not exist in the block and can thus abort early. + * In case it is equal to the sought-after key we have found + * the desired record. + * + * Note that we store the next record's key record directly in + * `last_key` without restoring the key of the preceding record + * in case we need to go one record back. This is safe to do as + * `block_iter_next()` would return the ref whose key is equal + * to `last_key` now, and naturally all keys share a prefix + * with themselves. + */ + reftable_record_key(&rec, &it->last_key); + if (strbuf_cmp(&it->last_key, want) >= 0) { + it->next_off = prev_off; + goto done; + } + } + +done: + reftable_record_release(&rec); + return err; +} + +void block_writer_release(struct block_writer *bw) +{ + FREE_AND_NULL(bw->restarts); + strbuf_release(&bw->last_key); + /* the block is not owned. */ +} + +void reftable_block_done(struct reftable_block *blockp) +{ + struct reftable_block_source source = blockp->source; + if (blockp && source.ops) + source.ops->return_block(source.arg, blockp); + blockp->data = NULL; + blockp->len = 0; + blockp->source.ops = NULL; + blockp->source.arg = NULL; +} diff --git a/reftable/block.h b/reftable/block.h new file mode 100644 index 0000000..ea4384a --- /dev/null +++ b/reftable/block.h @@ -0,0 +1,142 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef BLOCK_H +#define BLOCK_H + +#include "basics.h" +#include "record.h" +#include "reftable-blocksource.h" + +/* + * Writes reftable blocks. The block_writer is reused across blocks to minimize + * allocation overhead. + */ +struct block_writer { + uint8_t *buf; + uint32_t block_size; + + /* Offset of the global header. Nonzero in the first block only. */ + uint32_t header_off; + + /* How often to restart keys. */ + int restart_interval; + int hash_size; + + /* Offset of next uint8_t to write. */ + uint32_t next; + uint32_t *restarts; + uint32_t restart_len; + uint32_t restart_cap; + + struct strbuf last_key; + int entries; +}; + +/* + * initializes the blockwriter to write `typ` entries, using `buf` as temporary + * storage. `buf` is not owned by the block_writer. */ +void block_writer_init(struct block_writer *bw, uint8_t typ, uint8_t *buf, + uint32_t block_size, uint32_t header_off, int hash_size); + +/* returns the block type (eg. 'r' for ref records. */ +uint8_t block_writer_type(struct block_writer *bw); + +/* appends the record, or -1 if it doesn't fit. */ +int block_writer_add(struct block_writer *w, struct reftable_record *rec); + +/* appends the key restarts, and compress the block if necessary. */ +int block_writer_finish(struct block_writer *w); + +/* clears out internally allocated block_writer members. */ +void block_writer_release(struct block_writer *bw); + +struct z_stream; + +/* Read a block. */ +struct block_reader { + /* offset of the block header; nonzero for the first block in a + * reftable. */ + uint32_t header_off; + + /* the memory block */ + struct reftable_block block; + int hash_size; + + /* Uncompressed data for log entries. */ + z_stream *zstream; + unsigned char *uncompressed_data; + size_t uncompressed_cap; + + /* size of the data, excluding restart data. */ + uint32_t block_len; + uint8_t *restart_bytes; + uint16_t restart_count; + + /* size of the data in the file. For log blocks, this is the compressed + * size. */ + uint32_t full_block_size; +}; + +/* initializes a block reader. */ +int block_reader_init(struct block_reader *br, struct reftable_block *bl, + uint32_t header_off, uint32_t table_block_size, + int hash_size); + +void block_reader_release(struct block_reader *br); + +/* Returns the block type (eg. 'r' for refs) */ +uint8_t block_reader_type(const struct block_reader *r); + +/* Decodes the first key in the block */ +int block_reader_first_key(const struct block_reader *br, struct strbuf *key); + +/* Iterate over entries in a block */ +struct block_iter { + /* offset within the block of the next entry to read. */ + uint32_t next_off; + const unsigned char *block; + size_t block_len; + int hash_size; + + /* key for last entry we read. */ + struct strbuf last_key; + struct strbuf scratch; +}; + +#define BLOCK_ITER_INIT { \ + .last_key = STRBUF_INIT, \ + .scratch = STRBUF_INIT, \ +} + +/* Position `it` at start of the block */ +void block_iter_seek_start(struct block_iter *it, const struct block_reader *br); + +/* Position `it` to the `want` key in the block */ +int block_iter_seek_key(struct block_iter *it, const struct block_reader *br, + struct strbuf *want); + +/* return < 0 for error, 0 for OK, > 0 for EOF. */ +int block_iter_next(struct block_iter *it, struct reftable_record *rec); + +/* Reset the block iterator to pristine state without releasing its memory. */ +void block_iter_reset(struct block_iter *it); + +/* deallocate memory for `it`. The block reader and its block is left intact. */ +void block_iter_close(struct block_iter *it); + +/* size of file header, depending on format version */ +int header_size(int version); + +/* size of file footer, depending on format version */ +int footer_size(int version); + +/* returns a block to its source. */ +void reftable_block_done(struct reftable_block *ret); + +#endif diff --git a/reftable/block_test.c b/reftable/block_test.c new file mode 100644 index 0000000..26a9cfb --- /dev/null +++ b/reftable/block_test.c @@ -0,0 +1,123 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "block.h" + +#include "system.h" +#include "blocksource.h" +#include "basics.h" +#include "constants.h" +#include "record.h" +#include "test_framework.h" +#include "reftable-tests.h" + +static void test_block_read_write(void) +{ + const int header_off = 21; /* random */ + char *names[30]; + const int N = ARRAY_SIZE(names); + const int block_size = 1024; + struct reftable_block block = { NULL }; + struct block_writer bw = { + .last_key = STRBUF_INIT, + }; + struct reftable_record rec = { + .type = BLOCK_TYPE_REF, + }; + int i = 0; + int n; + struct block_reader br = { 0 }; + struct block_iter it = BLOCK_ITER_INIT; + int j = 0; + struct strbuf want = STRBUF_INIT; + + REFTABLE_CALLOC_ARRAY(block.data, block_size); + block.len = block_size; + block.source = malloc_block_source(); + block_writer_init(&bw, BLOCK_TYPE_REF, block.data, block_size, + header_off, hash_size(GIT_SHA1_FORMAT_ID)); + + rec.u.ref.refname = ""; + rec.u.ref.value_type = REFTABLE_REF_DELETION; + n = block_writer_add(&bw, &rec); + EXPECT(n == REFTABLE_API_ERROR); + + for (i = 0; i < N; i++) { + char name[100]; + snprintf(name, sizeof(name), "branch%02d", i); + + rec.u.ref.refname = name; + rec.u.ref.value_type = REFTABLE_REF_VAL1; + memset(rec.u.ref.value.val1, i, GIT_SHA1_RAWSZ); + + names[i] = xstrdup(name); + n = block_writer_add(&bw, &rec); + rec.u.ref.refname = NULL; + rec.u.ref.value_type = REFTABLE_REF_DELETION; + EXPECT(n == 0); + } + + n = block_writer_finish(&bw); + EXPECT(n > 0); + + block_writer_release(&bw); + + block_reader_init(&br, &block, header_off, block_size, GIT_SHA1_RAWSZ); + + block_iter_seek_start(&it, &br); + + while (1) { + int r = block_iter_next(&it, &rec); + EXPECT(r >= 0); + if (r > 0) { + break; + } + EXPECT_STREQ(names[j], rec.u.ref.refname); + j++; + } + + reftable_record_release(&rec); + block_iter_close(&it); + + for (i = 0; i < N; i++) { + struct block_iter it = BLOCK_ITER_INIT; + strbuf_reset(&want); + strbuf_addstr(&want, names[i]); + + n = block_iter_seek_key(&it, &br, &want); + EXPECT(n == 0); + + n = block_iter_next(&it, &rec); + EXPECT(n == 0); + + EXPECT_STREQ(names[i], rec.u.ref.refname); + + want.len--; + n = block_iter_seek_key(&it, &br, &want); + EXPECT(n == 0); + + n = block_iter_next(&it, &rec); + EXPECT(n == 0); + EXPECT_STREQ(names[10 * (i / 10)], rec.u.ref.refname); + + block_iter_close(&it); + } + + reftable_record_release(&rec); + reftable_block_done(&br.block); + strbuf_release(&want); + for (i = 0; i < N; i++) { + reftable_free(names[i]); + } +} + +int block_test_main(int argc, const char *argv[]) +{ + RUN_TEST(test_block_read_write); + return 0; +} diff --git a/reftable/blocksource.c b/reftable/blocksource.c new file mode 100644 index 0000000..eeed254 --- /dev/null +++ b/reftable/blocksource.c @@ -0,0 +1,144 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "system.h" + +#include "basics.h" +#include "blocksource.h" +#include "reftable-blocksource.h" +#include "reftable-error.h" + +static void strbuf_return_block(void *b, struct reftable_block *dest) +{ + if (dest->len) + memset(dest->data, 0xff, dest->len); + reftable_free(dest->data); +} + +static void strbuf_close(void *b) +{ +} + +static int strbuf_read_block(void *v, struct reftable_block *dest, uint64_t off, + uint32_t size) +{ + struct strbuf *b = v; + assert(off + size <= b->len); + REFTABLE_CALLOC_ARRAY(dest->data, size); + memcpy(dest->data, b->buf + off, size); + dest->len = size; + return size; +} + +static uint64_t strbuf_size(void *b) +{ + return ((struct strbuf *)b)->len; +} + +static struct reftable_block_source_vtable strbuf_vtable = { + .size = &strbuf_size, + .read_block = &strbuf_read_block, + .return_block = &strbuf_return_block, + .close = &strbuf_close, +}; + +void block_source_from_strbuf(struct reftable_block_source *bs, + struct strbuf *buf) +{ + assert(!bs->ops); + bs->ops = &strbuf_vtable; + bs->arg = buf; +} + +static void malloc_return_block(void *b, struct reftable_block *dest) +{ + if (dest->len) + memset(dest->data, 0xff, dest->len); + reftable_free(dest->data); +} + +static struct reftable_block_source_vtable malloc_vtable = { + .return_block = &malloc_return_block, +}; + +static struct reftable_block_source malloc_block_source_instance = { + .ops = &malloc_vtable, +}; + +struct reftable_block_source malloc_block_source(void) +{ + return malloc_block_source_instance; +} + +struct file_block_source { + uint64_t size; + unsigned char *data; +}; + +static uint64_t file_size(void *b) +{ + return ((struct file_block_source *)b)->size; +} + +static void file_return_block(void *b, struct reftable_block *dest) +{ +} + +static void file_close(void *v) +{ + struct file_block_source *b = v; + munmap(b->data, b->size); + reftable_free(b); +} + +static int file_read_block(void *v, struct reftable_block *dest, uint64_t off, + uint32_t size) +{ + struct file_block_source *b = v; + assert(off + size <= b->size); + dest->data = b->data + off; + dest->len = size; + return size; +} + +static struct reftable_block_source_vtable file_vtable = { + .size = &file_size, + .read_block = &file_read_block, + .return_block = &file_return_block, + .close = &file_close, +}; + +int reftable_block_source_from_file(struct reftable_block_source *bs, + const char *name) +{ + struct file_block_source *p; + struct stat st; + int fd; + + fd = open(name, O_RDONLY); + if (fd < 0) { + if (errno == ENOENT) + return REFTABLE_NOT_EXIST_ERROR; + return -1; + } + + if (fstat(fd, &st) < 0) { + close(fd); + return REFTABLE_IO_ERROR; + } + + REFTABLE_CALLOC_ARRAY(p, 1); + p->size = st.st_size; + p->data = xmmap(NULL, st.st_size, PROT_READ, MAP_PRIVATE, fd, 0); + close(fd); + + assert(!bs->ops); + bs->ops = &file_vtable; + bs->arg = p; + return 0; +} diff --git a/reftable/blocksource.h b/reftable/blocksource.h new file mode 100644 index 0000000..072e272 --- /dev/null +++ b/reftable/blocksource.h @@ -0,0 +1,22 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef BLOCKSOURCE_H +#define BLOCKSOURCE_H + +#include "system.h" + +struct reftable_block_source; + +/* Create an in-memory block source for reading reftables */ +void block_source_from_strbuf(struct reftable_block_source *bs, + struct strbuf *buf); + +struct reftable_block_source malloc_block_source(void); + +#endif diff --git a/reftable/constants.h b/reftable/constants.h new file mode 100644 index 0000000..5eee72c --- /dev/null +++ b/reftable/constants.h @@ -0,0 +1,21 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef CONSTANTS_H +#define CONSTANTS_H + +#define BLOCK_TYPE_LOG 'g' +#define BLOCK_TYPE_INDEX 'i' +#define BLOCK_TYPE_REF 'r' +#define BLOCK_TYPE_OBJ 'o' +#define BLOCK_TYPE_ANY 0 + +#define MAX_RESTARTS ((1 << 16) - 1) +#define DEFAULT_BLOCK_SIZE 4096 + +#endif diff --git a/reftable/dump.c b/reftable/dump.c new file mode 100644 index 0000000..26e0393 --- /dev/null +++ b/reftable/dump.c @@ -0,0 +1,105 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "git-compat-util.h" +#include "hash-ll.h" + +#include "reftable-blocksource.h" +#include "reftable-error.h" +#include "reftable-record.h" +#include "reftable-tests.h" +#include "reftable-writer.h" +#include "reftable-iterator.h" +#include "reftable-reader.h" +#include "reftable-stack.h" + +#include <stddef.h> +#include <stdio.h> +#include <stdlib.h> +#include <unistd.h> +#include <string.h> + +static int compact_stack(const char *stackdir) +{ + struct reftable_stack *stack = NULL; + struct reftable_write_options cfg = { 0 }; + + int err = reftable_new_stack(&stack, stackdir, cfg); + if (err < 0) + goto done; + + err = reftable_stack_compact_all(stack, NULL); + if (err < 0) + goto done; +done: + if (stack) { + reftable_stack_destroy(stack); + } + return err; +} + +static void print_help(void) +{ + printf("usage: dump [-cst] arg\n\n" + "options: \n" + " -c compact\n" + " -t dump table\n" + " -s dump stack\n" + " -6 sha256 hash format\n" + " -h this help\n" + "\n"); +} + +int reftable_dump_main(int argc, char *const *argv) +{ + int err = 0; + int opt_dump_table = 0; + int opt_dump_stack = 0; + int opt_compact = 0; + uint32_t opt_hash_id = GIT_SHA1_FORMAT_ID; + const char *arg = NULL, *argv0 = argv[0]; + + for (; argc > 1; argv++, argc--) + if (*argv[1] != '-') + break; + else if (!strcmp("-t", argv[1])) + opt_dump_table = 1; + else if (!strcmp("-6", argv[1])) + opt_hash_id = GIT_SHA256_FORMAT_ID; + else if (!strcmp("-s", argv[1])) + opt_dump_stack = 1; + else if (!strcmp("-c", argv[1])) + opt_compact = 1; + else if (!strcmp("-?", argv[1]) || !strcmp("-h", argv[1])) { + print_help(); + return 2; + } + + if (argc != 2) { + fprintf(stderr, "need argument\n"); + print_help(); + return 2; + } + + arg = argv[1]; + + if (opt_dump_table) { + err = reftable_reader_print_file(arg); + } else if (opt_dump_stack) { + err = reftable_stack_print_directory(arg, opt_hash_id); + } else if (opt_compact) { + err = compact_stack(arg); + } + + if (err < 0) { + fprintf(stderr, "%s: %s: %s\n", argv0, arg, + reftable_error_str(err)); + return 1; + } + return 0; +} diff --git a/reftable/error.c b/reftable/error.c new file mode 100644 index 0000000..cfb7a0f --- /dev/null +++ b/reftable/error.c @@ -0,0 +1,46 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "system.h" +#include "reftable-error.h" + +#include <stdio.h> + +const char *reftable_error_str(int err) +{ + static char buf[250]; + switch (err) { + case REFTABLE_IO_ERROR: + return "I/O error"; + case REFTABLE_FORMAT_ERROR: + return "corrupt reftable file"; + case REFTABLE_NOT_EXIST_ERROR: + return "file does not exist"; + case REFTABLE_LOCK_ERROR: + return "data is locked"; + case REFTABLE_API_ERROR: + return "misuse of the reftable API"; + case REFTABLE_ZLIB_ERROR: + return "zlib failure"; + case REFTABLE_NAME_CONFLICT: + return "file/directory conflict"; + case REFTABLE_EMPTY_TABLE_ERROR: + return "wrote empty table"; + case REFTABLE_REFNAME_ERROR: + return "invalid refname"; + case REFTABLE_ENTRY_TOO_BIG_ERROR: + return "entry too large"; + case REFTABLE_OUTDATED_ERROR: + return "data concurrently modified"; + case -1: + return "general error"; + default: + snprintf(buf, sizeof(buf), "unknown error code %d", err); + return buf; + } +} diff --git a/reftable/generic.c b/reftable/generic.c new file mode 100644 index 0000000..b9f1c7c --- /dev/null +++ b/reftable/generic.c @@ -0,0 +1,179 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "constants.h" +#include "record.h" +#include "generic.h" +#include "reftable-iterator.h" +#include "reftable-generic.h" + +int reftable_table_seek_ref(struct reftable_table *tab, + struct reftable_iterator *it, const char *name) +{ + struct reftable_record rec = { .type = BLOCK_TYPE_REF, + .u.ref = { + .refname = (char *)name, + } }; + return tab->ops->seek_record(tab->table_arg, it, &rec); +} + +int reftable_table_seek_log(struct reftable_table *tab, + struct reftable_iterator *it, const char *name) +{ + struct reftable_record rec = { .type = BLOCK_TYPE_LOG, + .u.log = { + .refname = (char *)name, + .update_index = ~((uint64_t)0), + } }; + return tab->ops->seek_record(tab->table_arg, it, &rec); +} + +int reftable_table_read_ref(struct reftable_table *tab, const char *name, + struct reftable_ref_record *ref) +{ + struct reftable_iterator it = { NULL }; + int err = reftable_table_seek_ref(tab, &it, name); + if (err) + goto done; + + err = reftable_iterator_next_ref(&it, ref); + if (err) + goto done; + + if (strcmp(ref->refname, name) || + reftable_ref_record_is_deletion(ref)) { + reftable_ref_record_release(ref); + err = 1; + goto done; + } + +done: + reftable_iterator_destroy(&it); + return err; +} + +int reftable_table_print(struct reftable_table *tab) { + struct reftable_iterator it = { NULL }; + struct reftable_ref_record ref = { NULL }; + struct reftable_log_record log = { NULL }; + uint32_t hash_id = reftable_table_hash_id(tab); + int err = reftable_table_seek_ref(tab, &it, ""); + if (err < 0) { + return err; + } + + while (1) { + err = reftable_iterator_next_ref(&it, &ref); + if (err > 0) { + break; + } + if (err < 0) { + return err; + } + reftable_ref_record_print(&ref, hash_id); + } + reftable_iterator_destroy(&it); + reftable_ref_record_release(&ref); + + err = reftable_table_seek_log(tab, &it, ""); + if (err < 0) { + return err; + } + while (1) { + err = reftable_iterator_next_log(&it, &log); + if (err > 0) { + break; + } + if (err < 0) { + return err; + } + reftable_log_record_print(&log, hash_id); + } + reftable_iterator_destroy(&it); + reftable_log_record_release(&log); + return 0; +} + +uint64_t reftable_table_max_update_index(struct reftable_table *tab) +{ + return tab->ops->max_update_index(tab->table_arg); +} + +uint64_t reftable_table_min_update_index(struct reftable_table *tab) +{ + return tab->ops->min_update_index(tab->table_arg); +} + +uint32_t reftable_table_hash_id(struct reftable_table *tab) +{ + return tab->ops->hash_id(tab->table_arg); +} + +void reftable_iterator_destroy(struct reftable_iterator *it) +{ + if (!it->ops) { + return; + } + it->ops->close(it->iter_arg); + it->ops = NULL; + FREE_AND_NULL(it->iter_arg); +} + +int reftable_iterator_next_ref(struct reftable_iterator *it, + struct reftable_ref_record *ref) +{ + struct reftable_record rec = { + .type = BLOCK_TYPE_REF, + .u = { + .ref = *ref + }, + }; + int err = iterator_next(it, &rec); + *ref = rec.u.ref; + return err; +} + +int reftable_iterator_next_log(struct reftable_iterator *it, + struct reftable_log_record *log) +{ + struct reftable_record rec = { + .type = BLOCK_TYPE_LOG, + .u = { + .log = *log, + }, + }; + int err = iterator_next(it, &rec); + *log = rec.u.log; + return err; +} + +int iterator_next(struct reftable_iterator *it, struct reftable_record *rec) +{ + return it->ops->next(it->iter_arg, rec); +} + +static int empty_iterator_next(void *arg, struct reftable_record *rec) +{ + return 1; +} + +static void empty_iterator_close(void *arg) +{ +} + +static struct reftable_iterator_vtable empty_vtable = { + .next = &empty_iterator_next, + .close = &empty_iterator_close, +}; + +void iterator_set_empty(struct reftable_iterator *it) +{ + assert(!it->ops); + it->iter_arg = NULL; + it->ops = &empty_vtable; +} diff --git a/reftable/generic.h b/reftable/generic.h new file mode 100644 index 0000000..98886a0 --- /dev/null +++ b/reftable/generic.h @@ -0,0 +1,32 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef GENERIC_H +#define GENERIC_H + +#include "record.h" +#include "reftable-generic.h" + +/* generic interface to reftables */ +struct reftable_table_vtable { + int (*seek_record)(void *tab, struct reftable_iterator *it, + struct reftable_record *); + uint32_t (*hash_id)(void *tab); + uint64_t (*min_update_index)(void *tab); + uint64_t (*max_update_index)(void *tab); +}; + +struct reftable_iterator_vtable { + int (*next)(void *iter_arg, struct reftable_record *rec); + void (*close)(void *iter_arg); +}; + +void iterator_set_empty(struct reftable_iterator *it); +int iterator_next(struct reftable_iterator *it, struct reftable_record *rec); + +#endif diff --git a/reftable/iter.c b/reftable/iter.c new file mode 100644 index 0000000..aa9ac19 --- /dev/null +++ b/reftable/iter.c @@ -0,0 +1,188 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "iter.h" + +#include "system.h" + +#include "block.h" +#include "generic.h" +#include "constants.h" +#include "reader.h" +#include "reftable-error.h" + +static void filtering_ref_iterator_close(void *iter_arg) +{ + struct filtering_ref_iterator *fri = iter_arg; + strbuf_release(&fri->oid); + reftable_iterator_destroy(&fri->it); +} + +static int filtering_ref_iterator_next(void *iter_arg, + struct reftable_record *rec) +{ + struct filtering_ref_iterator *fri = iter_arg; + struct reftable_ref_record *ref = &rec->u.ref; + int err = 0; + while (1) { + err = reftable_iterator_next_ref(&fri->it, ref); + if (err != 0) { + break; + } + + if (fri->double_check) { + struct reftable_iterator it = { NULL }; + + err = reftable_table_seek_ref(&fri->tab, &it, + ref->refname); + if (err == 0) { + err = reftable_iterator_next_ref(&it, ref); + } + + reftable_iterator_destroy(&it); + + if (err < 0) { + break; + } + + if (err > 0) { + continue; + } + } + + if (ref->value_type == REFTABLE_REF_VAL2 && + (!memcmp(fri->oid.buf, ref->value.val2.target_value, + fri->oid.len) || + !memcmp(fri->oid.buf, ref->value.val2.value, + fri->oid.len))) + return 0; + + if (ref->value_type == REFTABLE_REF_VAL1 && + !memcmp(fri->oid.buf, ref->value.val1, fri->oid.len)) { + return 0; + } + } + + reftable_ref_record_release(ref); + return err; +} + +static struct reftable_iterator_vtable filtering_ref_iterator_vtable = { + .next = &filtering_ref_iterator_next, + .close = &filtering_ref_iterator_close, +}; + +void iterator_from_filtering_ref_iterator(struct reftable_iterator *it, + struct filtering_ref_iterator *fri) +{ + assert(!it->ops); + it->iter_arg = fri; + it->ops = &filtering_ref_iterator_vtable; +} + +static void indexed_table_ref_iter_close(void *p) +{ + struct indexed_table_ref_iter *it = p; + block_iter_close(&it->cur); + reftable_block_done(&it->block_reader.block); + reftable_free(it->offsets); + strbuf_release(&it->oid); +} + +static int indexed_table_ref_iter_next_block(struct indexed_table_ref_iter *it) +{ + uint64_t off; + int err = 0; + if (it->offset_idx == it->offset_len) { + it->is_finished = 1; + return 1; + } + + reftable_block_done(&it->block_reader.block); + + off = it->offsets[it->offset_idx++]; + err = reader_init_block_reader(it->r, &it->block_reader, off, + BLOCK_TYPE_REF); + if (err < 0) { + return err; + } + if (err > 0) { + /* indexed block does not exist. */ + return REFTABLE_FORMAT_ERROR; + } + block_iter_seek_start(&it->cur, &it->block_reader); + return 0; +} + +static int indexed_table_ref_iter_next(void *p, struct reftable_record *rec) +{ + struct indexed_table_ref_iter *it = p; + struct reftable_ref_record *ref = &rec->u.ref; + + while (1) { + int err = block_iter_next(&it->cur, rec); + if (err < 0) { + return err; + } + + if (err > 0) { + err = indexed_table_ref_iter_next_block(it); + if (err < 0) { + return err; + } + + if (it->is_finished) { + return 1; + } + continue; + } + /* BUG */ + if (!memcmp(it->oid.buf, ref->value.val2.target_value, + it->oid.len) || + !memcmp(it->oid.buf, ref->value.val2.value, it->oid.len)) { + return 0; + } + } +} + +int new_indexed_table_ref_iter(struct indexed_table_ref_iter **dest, + struct reftable_reader *r, uint8_t *oid, + int oid_len, uint64_t *offsets, int offset_len) +{ + struct indexed_table_ref_iter empty = INDEXED_TABLE_REF_ITER_INIT; + struct indexed_table_ref_iter *itr = reftable_calloc(1, sizeof(*itr)); + int err = 0; + + *itr = empty; + itr->r = r; + strbuf_add(&itr->oid, oid, oid_len); + + itr->offsets = offsets; + itr->offset_len = offset_len; + + err = indexed_table_ref_iter_next_block(itr); + if (err < 0) { + reftable_free(itr); + } else { + *dest = itr; + } + return err; +} + +static struct reftable_iterator_vtable indexed_table_ref_iter_vtable = { + .next = &indexed_table_ref_iter_next, + .close = &indexed_table_ref_iter_close, +}; + +void iterator_from_indexed_table_ref_iter(struct reftable_iterator *it, + struct indexed_table_ref_iter *itr) +{ + assert(!it->ops); + it->iter_arg = itr; + it->ops = &indexed_table_ref_iter_vtable; +} diff --git a/reftable/iter.h b/reftable/iter.h new file mode 100644 index 0000000..537431b --- /dev/null +++ b/reftable/iter.h @@ -0,0 +1,65 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef ITER_H +#define ITER_H + +#include "system.h" +#include "block.h" +#include "record.h" + +#include "reftable-iterator.h" +#include "reftable-generic.h" + +/* iterator that produces only ref records that point to `oid` */ +struct filtering_ref_iterator { + int double_check; + struct reftable_table tab; + struct strbuf oid; + struct reftable_iterator it; +}; +#define FILTERING_REF_ITERATOR_INIT \ + { \ + .oid = STRBUF_INIT \ + } + +void iterator_from_filtering_ref_iterator(struct reftable_iterator *, + struct filtering_ref_iterator *); + +/* iterator that produces only ref records that point to `oid`, + * but using the object index. + */ +struct indexed_table_ref_iter { + struct reftable_reader *r; + struct strbuf oid; + + /* mutable */ + uint64_t *offsets; + + /* Points to the next offset to read. */ + int offset_idx; + int offset_len; + struct block_reader block_reader; + struct block_iter cur; + int is_finished; +}; + +#define INDEXED_TABLE_REF_ITER_INIT { \ + .cur = BLOCK_ITER_INIT, \ + .oid = STRBUF_INIT, \ +} + +void iterator_from_indexed_table_ref_iter(struct reftable_iterator *it, + struct indexed_table_ref_iter *itr); + +/* Takes ownership of `offsets` */ +int new_indexed_table_ref_iter(struct indexed_table_ref_iter **dest, + struct reftable_reader *r, uint8_t *oid, + int oid_len, uint64_t *offsets, int offset_len); + +#endif diff --git a/reftable/merged.c b/reftable/merged.c new file mode 100644 index 0000000..f85a24c --- /dev/null +++ b/reftable/merged.c @@ -0,0 +1,355 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "merged.h" + +#include "constants.h" +#include "iter.h" +#include "pq.h" +#include "record.h" +#include "generic.h" +#include "reftable-merged.h" +#include "reftable-error.h" +#include "system.h" + +struct merged_subiter { + struct reftable_iterator iter; + struct reftable_record rec; +}; + +struct merged_iter { + struct merged_subiter *subiters; + struct merged_iter_pqueue pq; + uint32_t hash_id; + size_t stack_len; + uint8_t typ; + int suppress_deletions; + ssize_t advance_index; +}; + +static int merged_iter_init(struct merged_iter *mi) +{ + for (size_t i = 0; i < mi->stack_len; i++) { + struct pq_entry e = { + .index = i, + .rec = &mi->subiters[i].rec, + }; + int err; + + reftable_record_init(&mi->subiters[i].rec, mi->typ); + err = iterator_next(&mi->subiters[i].iter, + &mi->subiters[i].rec); + if (err < 0) + return err; + if (err > 0) + continue; + + merged_iter_pqueue_add(&mi->pq, &e); + } + + return 0; +} + +static void merged_iter_close(void *p) +{ + struct merged_iter *mi = p; + + merged_iter_pqueue_release(&mi->pq); + for (size_t i = 0; i < mi->stack_len; i++) { + reftable_iterator_destroy(&mi->subiters[i].iter); + reftable_record_release(&mi->subiters[i].rec); + } + reftable_free(mi->subiters); +} + +static int merged_iter_advance_subiter(struct merged_iter *mi, size_t idx) +{ + struct pq_entry e = { + .index = idx, + .rec = &mi->subiters[idx].rec, + }; + int err; + + err = iterator_next(&mi->subiters[idx].iter, &mi->subiters[idx].rec); + if (err) + return err; + + merged_iter_pqueue_add(&mi->pq, &e); + return 0; +} + +static int merged_iter_next_entry(struct merged_iter *mi, + struct reftable_record *rec) +{ + struct pq_entry entry = { 0 }; + int err = 0, empty; + + empty = merged_iter_pqueue_is_empty(mi->pq); + + if (mi->advance_index >= 0) { + /* + * When there are no pqueue entries then we only have a single + * subiter left. There is no need to use the pqueue in that + * case anymore as we know that the subiter will return entries + * in the correct order already. + * + * While this may sound like a very specific edge case, it may + * happen more frequently than you think. Most repositories + * will end up having a single large base table that contains + * most of the refs. It's thus likely that we exhaust all + * subiters but the one from that base ref. + */ + if (empty) + return iterator_next(&mi->subiters[mi->advance_index].iter, + rec); + + err = merged_iter_advance_subiter(mi, mi->advance_index); + if (err < 0) + return err; + if (!err) + empty = 0; + mi->advance_index = -1; + } + + if (empty) + return 1; + + entry = merged_iter_pqueue_remove(&mi->pq); + + /* + One can also use reftable as datacenter-local storage, where the ref + database is maintained in globally consistent database (eg. + CockroachDB or Spanner). In this scenario, replication delays together + with compaction may cause newer tables to contain older entries. In + such a deployment, the loop below must be changed to collect all + entries for the same key, and return new the newest one. + */ + while (!merged_iter_pqueue_is_empty(mi->pq)) { + struct pq_entry top = merged_iter_pqueue_top(mi->pq); + int cmp; + + cmp = reftable_record_cmp(top.rec, entry.rec); + if (cmp > 0) + break; + + merged_iter_pqueue_remove(&mi->pq); + err = merged_iter_advance_subiter(mi, top.index); + if (err < 0) + return err; + } + + mi->advance_index = entry.index; + SWAP(*rec, *entry.rec); + return 0; +} + +static int merged_iter_next_void(void *p, struct reftable_record *rec) +{ + struct merged_iter *mi = p; + while (1) { + int err = merged_iter_next_entry(mi, rec); + if (err) + return err; + if (mi->suppress_deletions && reftable_record_is_deletion(rec)) + continue; + return 0; + } +} + +static struct reftable_iterator_vtable merged_iter_vtable = { + .next = &merged_iter_next_void, + .close = &merged_iter_close, +}; + +static void iterator_from_merged_iter(struct reftable_iterator *it, + struct merged_iter *mi) +{ + assert(!it->ops); + it->iter_arg = mi; + it->ops = &merged_iter_vtable; +} + +int reftable_new_merged_table(struct reftable_merged_table **dest, + struct reftable_table *stack, size_t n, + uint32_t hash_id) +{ + struct reftable_merged_table *m = NULL; + uint64_t last_max = 0; + uint64_t first_min = 0; + + for (size_t i = 0; i < n; i++) { + uint64_t min = reftable_table_min_update_index(&stack[i]); + uint64_t max = reftable_table_max_update_index(&stack[i]); + + if (reftable_table_hash_id(&stack[i]) != hash_id) { + return REFTABLE_FORMAT_ERROR; + } + if (i == 0 || min < first_min) { + first_min = min; + } + if (i == 0 || max > last_max) { + last_max = max; + } + } + + REFTABLE_CALLOC_ARRAY(m, 1); + m->stack = stack; + m->stack_len = n; + m->min = first_min; + m->max = last_max; + m->hash_id = hash_id; + *dest = m; + return 0; +} + +/* clears the list of subtable, without affecting the readers themselves. */ +void merged_table_release(struct reftable_merged_table *mt) +{ + FREE_AND_NULL(mt->stack); + mt->stack_len = 0; +} + +void reftable_merged_table_free(struct reftable_merged_table *mt) +{ + if (!mt) { + return; + } + merged_table_release(mt); + reftable_free(mt); +} + +uint64_t +reftable_merged_table_max_update_index(struct reftable_merged_table *mt) +{ + return mt->max; +} + +uint64_t +reftable_merged_table_min_update_index(struct reftable_merged_table *mt) +{ + return mt->min; +} + +static int reftable_table_seek_record(struct reftable_table *tab, + struct reftable_iterator *it, + struct reftable_record *rec) +{ + return tab->ops->seek_record(tab->table_arg, it, rec); +} + +static int merged_table_seek_record(struct reftable_merged_table *mt, + struct reftable_iterator *it, + struct reftable_record *rec) +{ + struct merged_iter merged = { + .typ = reftable_record_type(rec), + .hash_id = mt->hash_id, + .suppress_deletions = mt->suppress_deletions, + .advance_index = -1, + }; + struct merged_iter *p; + int err; + + REFTABLE_CALLOC_ARRAY(merged.subiters, mt->stack_len); + for (size_t i = 0; i < mt->stack_len; i++) { + err = reftable_table_seek_record(&mt->stack[i], + &merged.subiters[merged.stack_len].iter, rec); + if (err < 0) + goto out; + if (!err) + merged.stack_len++; + } + + err = merged_iter_init(&merged); + if (err < 0) + goto out; + + p = reftable_malloc(sizeof(struct merged_iter)); + *p = merged; + iterator_from_merged_iter(it, p); + +out: + if (err < 0) + merged_iter_close(&merged); + return err; +} + +int reftable_merged_table_seek_ref(struct reftable_merged_table *mt, + struct reftable_iterator *it, + const char *name) +{ + struct reftable_record rec = { + .type = BLOCK_TYPE_REF, + .u.ref = { + .refname = (char *)name, + }, + }; + return merged_table_seek_record(mt, it, &rec); +} + +int reftable_merged_table_seek_log_at(struct reftable_merged_table *mt, + struct reftable_iterator *it, + const char *name, uint64_t update_index) +{ + struct reftable_record rec = { .type = BLOCK_TYPE_LOG, + .u.log = { + .refname = (char *)name, + .update_index = update_index, + } }; + return merged_table_seek_record(mt, it, &rec); +} + +int reftable_merged_table_seek_log(struct reftable_merged_table *mt, + struct reftable_iterator *it, + const char *name) +{ + uint64_t max = ~((uint64_t)0); + return reftable_merged_table_seek_log_at(mt, it, name, max); +} + +uint32_t reftable_merged_table_hash_id(struct reftable_merged_table *mt) +{ + return mt->hash_id; +} + +static int reftable_merged_table_seek_void(void *tab, + struct reftable_iterator *it, + struct reftable_record *rec) +{ + return merged_table_seek_record(tab, it, rec); +} + +static uint32_t reftable_merged_table_hash_id_void(void *tab) +{ + return reftable_merged_table_hash_id(tab); +} + +static uint64_t reftable_merged_table_min_update_index_void(void *tab) +{ + return reftable_merged_table_min_update_index(tab); +} + +static uint64_t reftable_merged_table_max_update_index_void(void *tab) +{ + return reftable_merged_table_max_update_index(tab); +} + +static struct reftable_table_vtable merged_table_vtable = { + .seek_record = reftable_merged_table_seek_void, + .hash_id = reftable_merged_table_hash_id_void, + .min_update_index = reftable_merged_table_min_update_index_void, + .max_update_index = reftable_merged_table_max_update_index_void, +}; + +void reftable_table_from_merged_table(struct reftable_table *tab, + struct reftable_merged_table *merged) +{ + assert(!tab->ops); + tab->ops = &merged_table_vtable; + tab->table_arg = merged; +} diff --git a/reftable/merged.h b/reftable/merged.h new file mode 100644 index 0000000..a2571db --- /dev/null +++ b/reftable/merged.h @@ -0,0 +1,29 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef MERGED_H +#define MERGED_H + +#include "system.h" + +struct reftable_merged_table { + struct reftable_table *stack; + size_t stack_len; + uint32_t hash_id; + + /* If unset, produce deletions. This is useful for compaction. For the + * full stack, deletions should be produced. */ + int suppress_deletions; + + uint64_t min; + uint64_t max; +}; + +void merged_table_release(struct reftable_merged_table *mt); + +#endif diff --git a/reftable/merged_test.c b/reftable/merged_test.c new file mode 100644 index 0000000..530fc82 --- /dev/null +++ b/reftable/merged_test.c @@ -0,0 +1,454 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "merged.h" + +#include "system.h" + +#include "basics.h" +#include "blocksource.h" +#include "reader.h" +#include "record.h" +#include "test_framework.h" +#include "reftable-merged.h" +#include "reftable-tests.h" +#include "reftable-generic.h" +#include "reftable-writer.h" + +static void write_test_table(struct strbuf *buf, + struct reftable_ref_record refs[], int n) +{ + uint64_t min = 0xffffffff; + uint64_t max = 0; + int i = 0; + int err; + + struct reftable_write_options opts = { + .block_size = 256, + }; + struct reftable_writer *w = NULL; + for (i = 0; i < n; i++) { + uint64_t ui = refs[i].update_index; + if (ui > max) { + max = ui; + } + if (ui < min) { + min = ui; + } + } + + w = reftable_new_writer(&strbuf_add_void, &noop_flush, buf, &opts); + reftable_writer_set_limits(w, min, max); + + for (i = 0; i < n; i++) { + uint64_t before = refs[i].update_index; + int n = reftable_writer_add_ref(w, &refs[i]); + EXPECT(n == 0); + EXPECT(before == refs[i].update_index); + } + + err = reftable_writer_close(w); + EXPECT_ERR(err); + + reftable_writer_free(w); +} + +static void write_test_log_table(struct strbuf *buf, + struct reftable_log_record logs[], int n, + uint64_t update_index) +{ + int i = 0; + int err; + + struct reftable_write_options opts = { + .block_size = 256, + .exact_log_message = 1, + }; + struct reftable_writer *w = NULL; + w = reftable_new_writer(&strbuf_add_void, &noop_flush, buf, &opts); + reftable_writer_set_limits(w, update_index, update_index); + + for (i = 0; i < n; i++) { + int err = reftable_writer_add_log(w, &logs[i]); + EXPECT_ERR(err); + } + + err = reftable_writer_close(w); + EXPECT_ERR(err); + + reftable_writer_free(w); +} + +static struct reftable_merged_table * +merged_table_from_records(struct reftable_ref_record **refs, + struct reftable_block_source **source, + struct reftable_reader ***readers, int *sizes, + struct strbuf *buf, size_t n) +{ + struct reftable_merged_table *mt = NULL; + struct reftable_table *tabs; + int err; + + REFTABLE_CALLOC_ARRAY(tabs, n); + REFTABLE_CALLOC_ARRAY(*readers, n); + REFTABLE_CALLOC_ARRAY(*source, n); + + for (size_t i = 0; i < n; i++) { + write_test_table(&buf[i], refs[i], sizes[i]); + block_source_from_strbuf(&(*source)[i], &buf[i]); + + err = reftable_new_reader(&(*readers)[i], &(*source)[i], + "name"); + EXPECT_ERR(err); + reftable_table_from_reader(&tabs[i], (*readers)[i]); + } + + err = reftable_new_merged_table(&mt, tabs, n, GIT_SHA1_FORMAT_ID); + EXPECT_ERR(err); + return mt; +} + +static void readers_destroy(struct reftable_reader **readers, size_t n) +{ + int i = 0; + for (; i < n; i++) + reftable_reader_free(readers[i]); + reftable_free(readers); +} + +static void test_merged_between(void) +{ + struct reftable_ref_record r1[] = { { + .refname = "b", + .update_index = 1, + .value_type = REFTABLE_REF_VAL1, + .value.val1 = { 1, 2, 3, 0 }, + } }; + struct reftable_ref_record r2[] = { { + .refname = "a", + .update_index = 2, + .value_type = REFTABLE_REF_DELETION, + } }; + + struct reftable_ref_record *refs[] = { r1, r2 }; + int sizes[] = { 1, 1 }; + struct strbuf bufs[2] = { STRBUF_INIT, STRBUF_INIT }; + struct reftable_block_source *bs = NULL; + struct reftable_reader **readers = NULL; + struct reftable_merged_table *mt = + merged_table_from_records(refs, &bs, &readers, sizes, bufs, 2); + int i; + struct reftable_ref_record ref = { NULL }; + struct reftable_iterator it = { NULL }; + int err = reftable_merged_table_seek_ref(mt, &it, "a"); + EXPECT_ERR(err); + + err = reftable_iterator_next_ref(&it, &ref); + EXPECT_ERR(err); + EXPECT(ref.update_index == 2); + reftable_ref_record_release(&ref); + reftable_iterator_destroy(&it); + readers_destroy(readers, 2); + reftable_merged_table_free(mt); + for (i = 0; i < ARRAY_SIZE(bufs); i++) { + strbuf_release(&bufs[i]); + } + reftable_free(bs); +} + +static void test_merged(void) +{ + struct reftable_ref_record r1[] = { + { + .refname = "a", + .update_index = 1, + .value_type = REFTABLE_REF_VAL1, + .value.val1 = { 1 }, + }, + { + .refname = "b", + .update_index = 1, + .value_type = REFTABLE_REF_VAL1, + .value.val1 = { 1 }, + }, + { + .refname = "c", + .update_index = 1, + .value_type = REFTABLE_REF_VAL1, + .value.val1 = { 1 }, + } + }; + struct reftable_ref_record r2[] = { { + .refname = "a", + .update_index = 2, + .value_type = REFTABLE_REF_DELETION, + } }; + struct reftable_ref_record r3[] = { + { + .refname = "c", + .update_index = 3, + .value_type = REFTABLE_REF_VAL1, + .value.val1 = { 2 }, + }, + { + .refname = "d", + .update_index = 3, + .value_type = REFTABLE_REF_VAL1, + .value.val1 = { 1 }, + }, + }; + + struct reftable_ref_record *want[] = { + &r2[0], + &r1[1], + &r3[0], + &r3[1], + }; + + struct reftable_ref_record *refs[] = { r1, r2, r3 }; + int sizes[3] = { 3, 1, 2 }; + struct strbuf bufs[3] = { STRBUF_INIT, STRBUF_INIT, STRBUF_INIT }; + struct reftable_block_source *bs = NULL; + struct reftable_reader **readers = NULL; + struct reftable_merged_table *mt = + merged_table_from_records(refs, &bs, &readers, sizes, bufs, 3); + + struct reftable_iterator it = { NULL }; + int err = reftable_merged_table_seek_ref(mt, &it, "a"); + struct reftable_ref_record *out = NULL; + size_t len = 0; + size_t cap = 0; + int i = 0; + + EXPECT_ERR(err); + EXPECT(reftable_merged_table_hash_id(mt) == GIT_SHA1_FORMAT_ID); + EXPECT(reftable_merged_table_min_update_index(mt) == 1); + + while (len < 100) { /* cap loops/recursion. */ + struct reftable_ref_record ref = { NULL }; + int err = reftable_iterator_next_ref(&it, &ref); + if (err > 0) + break; + + REFTABLE_ALLOC_GROW(out, len + 1, cap); + out[len++] = ref; + } + reftable_iterator_destroy(&it); + + EXPECT(ARRAY_SIZE(want) == len); + for (i = 0; i < len; i++) { + EXPECT(reftable_ref_record_equal(want[i], &out[i], + GIT_SHA1_RAWSZ)); + } + for (i = 0; i < len; i++) { + reftable_ref_record_release(&out[i]); + } + reftable_free(out); + + for (i = 0; i < 3; i++) { + strbuf_release(&bufs[i]); + } + readers_destroy(readers, 3); + reftable_merged_table_free(mt); + reftable_free(bs); +} + +static struct reftable_merged_table * +merged_table_from_log_records(struct reftable_log_record **logs, + struct reftable_block_source **source, + struct reftable_reader ***readers, int *sizes, + struct strbuf *buf, size_t n) +{ + struct reftable_merged_table *mt = NULL; + struct reftable_table *tabs; + int err; + + REFTABLE_CALLOC_ARRAY(tabs, n); + REFTABLE_CALLOC_ARRAY(*readers, n); + REFTABLE_CALLOC_ARRAY(*source, n); + + for (size_t i = 0; i < n; i++) { + write_test_log_table(&buf[i], logs[i], sizes[i], i + 1); + block_source_from_strbuf(&(*source)[i], &buf[i]); + + err = reftable_new_reader(&(*readers)[i], &(*source)[i], + "name"); + EXPECT_ERR(err); + reftable_table_from_reader(&tabs[i], (*readers)[i]); + } + + err = reftable_new_merged_table(&mt, tabs, n, GIT_SHA1_FORMAT_ID); + EXPECT_ERR(err); + return mt; +} + +static void test_merged_logs(void) +{ + struct reftable_log_record r1[] = { + { + .refname = "a", + .update_index = 2, + .value_type = REFTABLE_LOG_UPDATE, + .value.update = { + .old_hash = { 2 }, + /* deletion */ + .name = "jane doe", + .email = "jane@invalid", + .message = "message2", + } + }, + { + .refname = "a", + .update_index = 1, + .value_type = REFTABLE_LOG_UPDATE, + .value.update = { + .old_hash = { 1 }, + .new_hash = { 2 }, + .name = "jane doe", + .email = "jane@invalid", + .message = "message1", + } + }, + }; + struct reftable_log_record r2[] = { + { + .refname = "a", + .update_index = 3, + .value_type = REFTABLE_LOG_UPDATE, + .value.update = { + .new_hash = { 3 }, + .name = "jane doe", + .email = "jane@invalid", + .message = "message3", + } + }, + }; + struct reftable_log_record r3[] = { + { + .refname = "a", + .update_index = 2, + .value_type = REFTABLE_LOG_DELETION, + }, + }; + struct reftable_log_record *want[] = { + &r2[0], + &r3[0], + &r1[1], + }; + + struct reftable_log_record *logs[] = { r1, r2, r3 }; + int sizes[3] = { 2, 1, 1 }; + struct strbuf bufs[3] = { STRBUF_INIT, STRBUF_INIT, STRBUF_INIT }; + struct reftable_block_source *bs = NULL; + struct reftable_reader **readers = NULL; + struct reftable_merged_table *mt = merged_table_from_log_records( + logs, &bs, &readers, sizes, bufs, 3); + + struct reftable_iterator it = { NULL }; + int err = reftable_merged_table_seek_log(mt, &it, "a"); + struct reftable_log_record *out = NULL; + size_t len = 0; + size_t cap = 0; + int i = 0; + + EXPECT_ERR(err); + EXPECT(reftable_merged_table_hash_id(mt) == GIT_SHA1_FORMAT_ID); + EXPECT(reftable_merged_table_min_update_index(mt) == 1); + + while (len < 100) { /* cap loops/recursion. */ + struct reftable_log_record log = { NULL }; + int err = reftable_iterator_next_log(&it, &log); + if (err > 0) + break; + + REFTABLE_ALLOC_GROW(out, len + 1, cap); + out[len++] = log; + } + reftable_iterator_destroy(&it); + + EXPECT(ARRAY_SIZE(want) == len); + for (i = 0; i < len; i++) { + EXPECT(reftable_log_record_equal(want[i], &out[i], + GIT_SHA1_RAWSZ)); + } + + err = reftable_merged_table_seek_log_at(mt, &it, "a", 2); + EXPECT_ERR(err); + reftable_log_record_release(&out[0]); + err = reftable_iterator_next_log(&it, &out[0]); + EXPECT_ERR(err); + EXPECT(reftable_log_record_equal(&out[0], &r3[0], GIT_SHA1_RAWSZ)); + reftable_iterator_destroy(&it); + + for (i = 0; i < len; i++) { + reftable_log_record_release(&out[i]); + } + reftable_free(out); + + for (i = 0; i < 3; i++) { + strbuf_release(&bufs[i]); + } + readers_destroy(readers, 3); + reftable_merged_table_free(mt); + reftable_free(bs); +} + +static void test_default_write_opts(void) +{ + struct reftable_write_options opts = { 0 }; + struct strbuf buf = STRBUF_INIT; + struct reftable_writer *w = + reftable_new_writer(&strbuf_add_void, &noop_flush, &buf, &opts); + + struct reftable_ref_record rec = { + .refname = "master", + .update_index = 1, + }; + int err; + struct reftable_block_source source = { NULL }; + struct reftable_table *tab = reftable_calloc(1, sizeof(*tab)); + uint32_t hash_id; + struct reftable_reader *rd = NULL; + struct reftable_merged_table *merged = NULL; + + reftable_writer_set_limits(w, 1, 1); + + err = reftable_writer_add_ref(w, &rec); + EXPECT_ERR(err); + + err = reftable_writer_close(w); + EXPECT_ERR(err); + reftable_writer_free(w); + + block_source_from_strbuf(&source, &buf); + + err = reftable_new_reader(&rd, &source, "filename"); + EXPECT_ERR(err); + + hash_id = reftable_reader_hash_id(rd); + EXPECT(hash_id == GIT_SHA1_FORMAT_ID); + + reftable_table_from_reader(&tab[0], rd); + err = reftable_new_merged_table(&merged, tab, 1, GIT_SHA1_FORMAT_ID); + EXPECT_ERR(err); + + reftable_reader_free(rd); + reftable_merged_table_free(merged); + strbuf_release(&buf); +} + +/* XXX test refs_for(oid) */ + +int merged_test_main(int argc, const char *argv[]) +{ + RUN_TEST(test_merged_logs); + RUN_TEST(test_merged_between); + RUN_TEST(test_merged); + RUN_TEST(test_default_write_opts); + return 0; +} diff --git a/reftable/pq.c b/reftable/pq.c new file mode 100644 index 0000000..7fb45d8 --- /dev/null +++ b/reftable/pq.c @@ -0,0 +1,77 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "pq.h" + +#include "reftable-record.h" +#include "system.h" +#include "basics.h" + +int pq_less(struct pq_entry *a, struct pq_entry *b) +{ + int cmp = reftable_record_cmp(a->rec, b->rec); + if (cmp == 0) + return a->index > b->index; + return cmp < 0; +} + +struct pq_entry merged_iter_pqueue_remove(struct merged_iter_pqueue *pq) +{ + int i = 0; + struct pq_entry e = pq->heap[0]; + pq->heap[0] = pq->heap[pq->len - 1]; + pq->len--; + + i = 0; + while (i < pq->len) { + int min = i; + int j = 2 * i + 1; + int k = 2 * i + 2; + if (j < pq->len && pq_less(&pq->heap[j], &pq->heap[i])) { + min = j; + } + if (k < pq->len && pq_less(&pq->heap[k], &pq->heap[min])) { + min = k; + } + + if (min == i) { + break; + } + + SWAP(pq->heap[i], pq->heap[min]); + i = min; + } + + return e; +} + +void merged_iter_pqueue_add(struct merged_iter_pqueue *pq, const struct pq_entry *e) +{ + int i = 0; + + REFTABLE_ALLOC_GROW(pq->heap, pq->len + 1, pq->cap); + pq->heap[pq->len++] = *e; + + i = pq->len - 1; + while (i > 0) { + int j = (i - 1) / 2; + if (pq_less(&pq->heap[j], &pq->heap[i])) { + break; + } + + SWAP(pq->heap[j], pq->heap[i]); + + i = j; + } +} + +void merged_iter_pqueue_release(struct merged_iter_pqueue *pq) +{ + FREE_AND_NULL(pq->heap); + memset(pq, 0, sizeof(*pq)); +} diff --git a/reftable/pq.h b/reftable/pq.h new file mode 100644 index 0000000..f796c23 --- /dev/null +++ b/reftable/pq.h @@ -0,0 +1,41 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef PQ_H +#define PQ_H + +#include "record.h" + +struct pq_entry { + size_t index; + struct reftable_record *rec; +}; + +struct merged_iter_pqueue { + struct pq_entry *heap; + size_t len; + size_t cap; +}; + +void merged_iter_pqueue_check(struct merged_iter_pqueue pq); +struct pq_entry merged_iter_pqueue_remove(struct merged_iter_pqueue *pq); +void merged_iter_pqueue_add(struct merged_iter_pqueue *pq, const struct pq_entry *e); +void merged_iter_pqueue_release(struct merged_iter_pqueue *pq); +int pq_less(struct pq_entry *a, struct pq_entry *b); + +static inline struct pq_entry merged_iter_pqueue_top(struct merged_iter_pqueue pq) +{ + return pq.heap[0]; +} + +static inline int merged_iter_pqueue_is_empty(struct merged_iter_pqueue pq) +{ + return pq.len == 0; +} + +#endif diff --git a/reftable/pq_test.c b/reftable/pq_test.c new file mode 100644 index 0000000..b7d3c80 --- /dev/null +++ b/reftable/pq_test.c @@ -0,0 +1,74 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "system.h" + +#include "basics.h" +#include "constants.h" +#include "pq.h" +#include "record.h" +#include "reftable-tests.h" +#include "test_framework.h" + +void merged_iter_pqueue_check(struct merged_iter_pqueue pq) +{ + int i; + for (i = 1; i < pq.len; i++) { + int parent = (i - 1) / 2; + + EXPECT(pq_less(&pq.heap[parent], &pq.heap[i])); + } +} + +static void test_pq(void) +{ + struct merged_iter_pqueue pq = { NULL }; + struct reftable_record recs[54]; + int N = ARRAY_SIZE(recs) - 1, i; + char *last = NULL; + + for (i = 0; i < N; i++) { + struct strbuf refname = STRBUF_INIT; + strbuf_addf(&refname, "%02d", i); + + reftable_record_init(&recs[i], BLOCK_TYPE_REF); + recs[i].u.ref.refname = strbuf_detach(&refname, NULL); + } + + i = 1; + do { + struct pq_entry e = { + .rec = &recs[i], + }; + + merged_iter_pqueue_add(&pq, &e); + merged_iter_pqueue_check(pq); + + i = (i * 7) % N; + } while (i != 1); + + while (!merged_iter_pqueue_is_empty(pq)) { + struct pq_entry e = merged_iter_pqueue_remove(&pq); + merged_iter_pqueue_check(pq); + + EXPECT(reftable_record_type(e.rec) == BLOCK_TYPE_REF); + if (last) + EXPECT(strcmp(last, e.rec->u.ref.refname) < 0); + last = e.rec->u.ref.refname; + } + + for (i = 0; i < N; i++) + reftable_record_release(&recs[i]); + merged_iter_pqueue_release(&pq); +} + +int pq_test_main(int argc, const char *argv[]) +{ + RUN_TEST(test_pq); + return 0; +} diff --git a/reftable/publicbasics.c b/reftable/publicbasics.c new file mode 100644 index 0000000..44b84a1 --- /dev/null +++ b/reftable/publicbasics.c @@ -0,0 +1,66 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "system.h" +#include "reftable-malloc.h" + +#include "basics.h" + +static void *(*reftable_malloc_ptr)(size_t sz); +static void *(*reftable_realloc_ptr)(void *, size_t); +static void (*reftable_free_ptr)(void *); + +void *reftable_malloc(size_t sz) +{ + if (reftable_malloc_ptr) + return (*reftable_malloc_ptr)(sz); + return malloc(sz); +} + +void *reftable_realloc(void *p, size_t sz) +{ + if (reftable_realloc_ptr) + return (*reftable_realloc_ptr)(p, sz); + return realloc(p, sz); +} + +void reftable_free(void *p) +{ + if (reftable_free_ptr) + reftable_free_ptr(p); + else + free(p); +} + +void *reftable_calloc(size_t nelem, size_t elsize) +{ + size_t sz = st_mult(nelem, elsize); + void *p = reftable_malloc(sz); + memset(p, 0, sz); + return p; +} + +void reftable_set_alloc(void *(*malloc)(size_t), + void *(*realloc)(void *, size_t), void (*free)(void *)) +{ + reftable_malloc_ptr = malloc; + reftable_realloc_ptr = realloc; + reftable_free_ptr = free; +} + +int hash_size(uint32_t id) +{ + switch (id) { + case 0: + case GIT_SHA1_FORMAT_ID: + return GIT_SHA1_RAWSZ; + case GIT_SHA256_FORMAT_ID: + return GIT_SHA256_RAWSZ; + } + abort(); +} diff --git a/reftable/reader.c b/reftable/reader.c new file mode 100644 index 0000000..481dff1 --- /dev/null +++ b/reftable/reader.c @@ -0,0 +1,858 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "reader.h" + +#include "system.h" +#include "block.h" +#include "constants.h" +#include "generic.h" +#include "iter.h" +#include "record.h" +#include "reftable-error.h" +#include "reftable-generic.h" + +uint64_t block_source_size(struct reftable_block_source *source) +{ + return source->ops->size(source->arg); +} + +int block_source_read_block(struct reftable_block_source *source, + struct reftable_block *dest, uint64_t off, + uint32_t size) +{ + int result = source->ops->read_block(source->arg, dest, off, size); + dest->source = *source; + return result; +} + +void block_source_close(struct reftable_block_source *source) +{ + if (!source->ops) { + return; + } + + source->ops->close(source->arg); + source->ops = NULL; +} + +static struct reftable_reader_offsets * +reader_offsets_for(struct reftable_reader *r, uint8_t typ) +{ + switch (typ) { + case BLOCK_TYPE_REF: + return &r->ref_offsets; + case BLOCK_TYPE_LOG: + return &r->log_offsets; + case BLOCK_TYPE_OBJ: + return &r->obj_offsets; + } + abort(); +} + +static int reader_get_block(struct reftable_reader *r, + struct reftable_block *dest, uint64_t off, + uint32_t sz) +{ + if (off >= r->size) + return 0; + + if (off + sz > r->size) { + sz = r->size - off; + } + + return block_source_read_block(&r->source, dest, off, sz); +} + +uint32_t reftable_reader_hash_id(struct reftable_reader *r) +{ + return r->hash_id; +} + +const char *reader_name(struct reftable_reader *r) +{ + return r->name; +} + +static int parse_footer(struct reftable_reader *r, uint8_t *footer, + uint8_t *header) +{ + uint8_t *f = footer; + uint8_t first_block_typ; + int err = 0; + uint32_t computed_crc; + uint32_t file_crc; + + if (memcmp(f, "REFT", 4)) { + err = REFTABLE_FORMAT_ERROR; + goto done; + } + f += 4; + + if (memcmp(footer, header, header_size(r->version))) { + err = REFTABLE_FORMAT_ERROR; + goto done; + } + + f++; + r->block_size = get_be24(f); + + f += 3; + r->min_update_index = get_be64(f); + f += 8; + r->max_update_index = get_be64(f); + f += 8; + + if (r->version == 1) { + r->hash_id = GIT_SHA1_FORMAT_ID; + } else { + r->hash_id = get_be32(f); + switch (r->hash_id) { + case GIT_SHA1_FORMAT_ID: + break; + case GIT_SHA256_FORMAT_ID: + break; + default: + err = REFTABLE_FORMAT_ERROR; + goto done; + } + f += 4; + } + + r->ref_offsets.index_offset = get_be64(f); + f += 8; + + r->obj_offsets.offset = get_be64(f); + f += 8; + + r->object_id_len = r->obj_offsets.offset & ((1 << 5) - 1); + r->obj_offsets.offset >>= 5; + + r->obj_offsets.index_offset = get_be64(f); + f += 8; + r->log_offsets.offset = get_be64(f); + f += 8; + r->log_offsets.index_offset = get_be64(f); + f += 8; + + computed_crc = crc32(0, footer, f - footer); + file_crc = get_be32(f); + f += 4; + if (computed_crc != file_crc) { + err = REFTABLE_FORMAT_ERROR; + goto done; + } + + first_block_typ = header[header_size(r->version)]; + r->ref_offsets.is_present = (first_block_typ == BLOCK_TYPE_REF); + r->ref_offsets.offset = 0; + r->log_offsets.is_present = (first_block_typ == BLOCK_TYPE_LOG || + r->log_offsets.offset > 0); + r->obj_offsets.is_present = r->obj_offsets.offset > 0; + if (r->obj_offsets.is_present && !r->object_id_len) { + err = REFTABLE_FORMAT_ERROR; + goto done; + } + + err = 0; +done: + return err; +} + +int init_reader(struct reftable_reader *r, struct reftable_block_source *source, + const char *name) +{ + struct reftable_block footer = { NULL }; + struct reftable_block header = { NULL }; + int err = 0; + uint64_t file_size = block_source_size(source); + + /* Need +1 to read type of first block. */ + uint32_t read_size = header_size(2) + 1; /* read v2 because it's larger. */ + memset(r, 0, sizeof(struct reftable_reader)); + + if (read_size > file_size) { + err = REFTABLE_FORMAT_ERROR; + goto done; + } + + err = block_source_read_block(source, &header, 0, read_size); + if (err != read_size) { + err = REFTABLE_IO_ERROR; + goto done; + } + + if (memcmp(header.data, "REFT", 4)) { + err = REFTABLE_FORMAT_ERROR; + goto done; + } + r->version = header.data[4]; + if (r->version != 1 && r->version != 2) { + err = REFTABLE_FORMAT_ERROR; + goto done; + } + + r->size = file_size - footer_size(r->version); + r->source = *source; + r->name = xstrdup(name); + r->hash_id = 0; + + err = block_source_read_block(source, &footer, r->size, + footer_size(r->version)); + if (err != footer_size(r->version)) { + err = REFTABLE_IO_ERROR; + goto done; + } + + err = parse_footer(r, footer.data, header.data); +done: + reftable_block_done(&footer); + reftable_block_done(&header); + return err; +} + +struct table_iter { + struct reftable_reader *r; + uint8_t typ; + uint64_t block_off; + struct block_reader br; + struct block_iter bi; + int is_finished; +}; +#define TABLE_ITER_INIT { \ + .bi = BLOCK_ITER_INIT \ +} + +static int table_iter_next_in_block(struct table_iter *ti, + struct reftable_record *rec) +{ + int res = block_iter_next(&ti->bi, rec); + if (res == 0 && reftable_record_type(rec) == BLOCK_TYPE_REF) { + rec->u.ref.update_index += ti->r->min_update_index; + } + + return res; +} + +static void table_iter_block_done(struct table_iter *ti) +{ + block_reader_release(&ti->br); + block_iter_reset(&ti->bi); +} + +static int32_t extract_block_size(uint8_t *data, uint8_t *typ, uint64_t off, + int version) +{ + int32_t result = 0; + + if (off == 0) { + data += header_size(version); + } + + *typ = data[0]; + if (reftable_is_block_type(*typ)) { + result = get_be24(data + 1); + } + return result; +} + +int reader_init_block_reader(struct reftable_reader *r, struct block_reader *br, + uint64_t next_off, uint8_t want_typ) +{ + int32_t guess_block_size = r->block_size ? r->block_size : + DEFAULT_BLOCK_SIZE; + struct reftable_block block = { NULL }; + uint8_t block_typ = 0; + int err = 0; + uint32_t header_off = next_off ? 0 : header_size(r->version); + int32_t block_size = 0; + + if (next_off >= r->size) + return 1; + + err = reader_get_block(r, &block, next_off, guess_block_size); + if (err < 0) + goto done; + + block_size = extract_block_size(block.data, &block_typ, next_off, + r->version); + if (block_size < 0) { + err = block_size; + goto done; + } + if (want_typ != BLOCK_TYPE_ANY && block_typ != want_typ) { + err = 1; + goto done; + } + + if (block_size > guess_block_size) { + reftable_block_done(&block); + err = reader_get_block(r, &block, next_off, block_size); + if (err < 0) { + goto done; + } + } + + err = block_reader_init(br, &block, header_off, r->block_size, + hash_size(r->hash_id)); +done: + reftable_block_done(&block); + + return err; +} + +static void table_iter_close(struct table_iter *ti) +{ + table_iter_block_done(ti); + block_iter_close(&ti->bi); +} + +static int table_iter_next_block(struct table_iter *ti) +{ + uint64_t next_block_off = ti->block_off + ti->br.full_block_size; + int err; + + err = reader_init_block_reader(ti->r, &ti->br, next_block_off, ti->typ); + if (err > 0) + ti->is_finished = 1; + if (err) + return err; + + ti->block_off = next_block_off; + ti->is_finished = 0; + block_iter_seek_start(&ti->bi, &ti->br); + + return 0; +} + +static int table_iter_next(struct table_iter *ti, struct reftable_record *rec) +{ + if (reftable_record_type(rec) != ti->typ) + return REFTABLE_API_ERROR; + + while (1) { + int err; + + if (ti->is_finished) + return 1; + + /* + * Check whether the current block still has more records. If + * so, return it. If the iterator returns positive then the + * current block has been exhausted. + */ + err = table_iter_next_in_block(ti, rec); + if (err <= 0) + return err; + + /* + * Otherwise, we need to continue to the next block in the + * table and retry. If there are no more blocks then the + * iterator is drained. + */ + err = table_iter_next_block(ti); + if (err) { + ti->is_finished = 1; + return err; + } + } +} + +static int table_iter_next_void(void *ti, struct reftable_record *rec) +{ + return table_iter_next(ti, rec); +} + +static void table_iter_close_void(void *ti) +{ + table_iter_close(ti); +} + +static struct reftable_iterator_vtable table_iter_vtable = { + .next = &table_iter_next_void, + .close = &table_iter_close_void, +}; + +static void iterator_from_table_iter(struct reftable_iterator *it, + struct table_iter *ti) +{ + assert(!it->ops); + it->iter_arg = ti; + it->ops = &table_iter_vtable; +} + +static int reader_table_iter_at(struct reftable_reader *r, + struct table_iter *ti, uint64_t off, + uint8_t typ) +{ + int err; + + err = reader_init_block_reader(r, &ti->br, off, typ); + if (err != 0) + return err; + + ti->r = r; + ti->typ = block_reader_type(&ti->br); + ti->block_off = off; + block_iter_seek_start(&ti->bi, &ti->br); + return 0; +} + +static int reader_start(struct reftable_reader *r, struct table_iter *ti, + uint8_t typ, int index) +{ + struct reftable_reader_offsets *offs = reader_offsets_for(r, typ); + uint64_t off = offs->offset; + if (index) { + off = offs->index_offset; + if (off == 0) { + return 1; + } + typ = BLOCK_TYPE_INDEX; + } + + return reader_table_iter_at(r, ti, off, typ); +} + +static int reader_seek_linear(struct table_iter *ti, + struct reftable_record *want) +{ + struct strbuf want_key = STRBUF_INIT; + struct strbuf got_key = STRBUF_INIT; + struct reftable_record rec; + int err = -1; + + reftable_record_init(&rec, reftable_record_type(want)); + reftable_record_key(want, &want_key); + + /* + * First we need to locate the block that must contain our record. To + * do so we scan through blocks linearly until we find the first block + * whose first key is bigger than our wanted key. Once we have found + * that block we know that the key must be contained in the preceding + * block. + * + * This algorithm is somewhat unfortunate because it means that we + * always have to seek one block too far and then back up. But as we + * can only decode the _first_ key of a block but not its _last_ key we + * have no other way to do this. + */ + while (1) { + struct table_iter next = *ti; + + /* + * We must be careful to not modify underlying data of `ti` + * because we may find that `next` does not contain our desired + * block, but that `ti` does. In that case, we would discard + * `next` and continue with `ti`. + * + * This also means that we cannot reuse allocated memory for + * `next` here. While it would be great if we could, it should + * in practice not be too bad given that we should only ever + * end up doing linear seeks with at most three blocks. As soon + * as we have more than three blocks we would have an index, so + * we would not do a linear search there anymore. + */ + memset(&next.br.block, 0, sizeof(next.br.block)); + next.br.zstream = NULL; + next.br.uncompressed_data = NULL; + next.br.uncompressed_cap = 0; + + err = table_iter_next_block(&next); + if (err < 0) + goto done; + if (err > 0) + break; + + err = block_reader_first_key(&next.br, &got_key); + if (err < 0) + goto done; + + if (strbuf_cmp(&got_key, &want_key) > 0) { + table_iter_block_done(&next); + break; + } + + table_iter_block_done(ti); + *ti = next; + } + + /* + * We have located the block that must contain our record, so we seek + * the wanted key inside of it. If the block does not contain our key + * we know that the corresponding record does not exist. + */ + err = block_iter_seek_key(&ti->bi, &ti->br, &want_key); + if (err < 0) + goto done; + err = 0; + +done: + reftable_record_release(&rec); + strbuf_release(&want_key); + strbuf_release(&got_key); + return err; +} + +static int reader_seek_indexed(struct reftable_reader *r, + struct reftable_iterator *it, + struct reftable_record *rec) +{ + struct reftable_record want_index = { + .type = BLOCK_TYPE_INDEX, .u.idx = { .last_key = STRBUF_INIT } + }; + struct reftable_record index_result = { + .type = BLOCK_TYPE_INDEX, + .u.idx = { .last_key = STRBUF_INIT }, + }; + struct table_iter index_iter = TABLE_ITER_INIT; + struct table_iter empty = TABLE_ITER_INIT; + struct table_iter next = TABLE_ITER_INIT; + int err = 0; + + reftable_record_key(rec, &want_index.u.idx.last_key); + err = reader_start(r, &index_iter, reftable_record_type(rec), 1); + if (err < 0) + goto done; + + /* + * The index may consist of multiple levels, where each level may have + * multiple index blocks. We start by doing a linear search in the + * highest layer that identifies the relevant index block as well as + * the record inside that block that corresponds to our wanted key. + */ + err = reader_seek_linear(&index_iter, &want_index); + if (err < 0) + goto done; + + /* + * Traverse down the levels until we find a non-index entry. + */ + while (1) { + /* + * In case we seek a record that does not exist the index iter + * will tell us that the iterator is over. This works because + * the last index entry of the current level will contain the + * last key it knows about. So in case our seeked key is larger + * than the last indexed key we know that it won't exist. + * + * There is one subtlety in the layout of the index section + * that makes this work as expected: the highest-level index is + * at end of the section and will point backwards and thus we + * start reading from the end of the index section, not the + * beginning. + * + * If that wasn't the case and the order was reversed then the + * linear seek would seek into the lower levels and traverse + * all levels of the index only to find out that the key does + * not exist. + */ + err = table_iter_next(&index_iter, &index_result); + if (err != 0) + goto done; + + err = reader_table_iter_at(r, &next, index_result.u.idx.offset, + 0); + if (err != 0) + goto done; + + err = block_iter_seek_key(&next.bi, &next.br, &want_index.u.idx.last_key); + if (err < 0) + goto done; + + if (next.typ == reftable_record_type(rec)) { + err = 0; + break; + } + + if (next.typ != BLOCK_TYPE_INDEX) { + err = REFTABLE_FORMAT_ERROR; + break; + } + + table_iter_close(&index_iter); + index_iter = next; + next = empty; + } + + if (err == 0) { + struct table_iter *malloced = reftable_calloc(1, sizeof(*malloced)); + *malloced = next; + next = empty; + iterator_from_table_iter(it, malloced); + } + +done: + table_iter_close(&next); + table_iter_close(&index_iter); + reftable_record_release(&want_index); + reftable_record_release(&index_result); + return err; +} + +static int reader_seek_internal(struct reftable_reader *r, + struct reftable_iterator *it, + struct reftable_record *rec) +{ + struct reftable_reader_offsets *offs = + reader_offsets_for(r, reftable_record_type(rec)); + uint64_t idx = offs->index_offset; + struct table_iter ti = TABLE_ITER_INIT, *p; + int err; + + if (idx > 0) + return reader_seek_indexed(r, it, rec); + + err = reader_start(r, &ti, reftable_record_type(rec), 0); + if (err < 0) + goto out; + + err = reader_seek_linear(&ti, rec); + if (err < 0) + goto out; + + REFTABLE_ALLOC_ARRAY(p, 1); + *p = ti; + iterator_from_table_iter(it, p); + +out: + if (err) + table_iter_close(&ti); + return err; +} + +static int reader_seek(struct reftable_reader *r, struct reftable_iterator *it, + struct reftable_record *rec) +{ + uint8_t typ = reftable_record_type(rec); + + struct reftable_reader_offsets *offs = reader_offsets_for(r, typ); + if (!offs->is_present) { + iterator_set_empty(it); + return 0; + } + + return reader_seek_internal(r, it, rec); +} + +int reftable_reader_seek_ref(struct reftable_reader *r, + struct reftable_iterator *it, const char *name) +{ + struct reftable_record rec = { + .type = BLOCK_TYPE_REF, + .u.ref = { + .refname = (char *)name, + }, + }; + return reader_seek(r, it, &rec); +} + +int reftable_reader_seek_log_at(struct reftable_reader *r, + struct reftable_iterator *it, const char *name, + uint64_t update_index) +{ + struct reftable_record rec = { .type = BLOCK_TYPE_LOG, + .u.log = { + .refname = (char *)name, + .update_index = update_index, + } }; + return reader_seek(r, it, &rec); +} + +int reftable_reader_seek_log(struct reftable_reader *r, + struct reftable_iterator *it, const char *name) +{ + uint64_t max = ~((uint64_t)0); + return reftable_reader_seek_log_at(r, it, name, max); +} + +void reader_close(struct reftable_reader *r) +{ + block_source_close(&r->source); + FREE_AND_NULL(r->name); +} + +int reftable_new_reader(struct reftable_reader **p, + struct reftable_block_source *src, char const *name) +{ + struct reftable_reader *rd = reftable_calloc(1, sizeof(*rd)); + int err = init_reader(rd, src, name); + if (err == 0) { + *p = rd; + } else { + block_source_close(src); + reftable_free(rd); + } + return err; +} + +void reftable_reader_free(struct reftable_reader *r) +{ + if (!r) + return; + reader_close(r); + reftable_free(r); +} + +static int reftable_reader_refs_for_indexed(struct reftable_reader *r, + struct reftable_iterator *it, + uint8_t *oid) +{ + struct reftable_record want = { + .type = BLOCK_TYPE_OBJ, + .u.obj = { + .hash_prefix = oid, + .hash_prefix_len = r->object_id_len, + }, + }; + struct reftable_iterator oit = { NULL }; + struct reftable_record got = { + .type = BLOCK_TYPE_OBJ, + .u.obj = { 0 }, + }; + int err = 0; + struct indexed_table_ref_iter *itr = NULL; + + /* Look through the reverse index. */ + err = reader_seek(r, &oit, &want); + if (err != 0) + goto done; + + /* read out the reftable_obj_record */ + err = iterator_next(&oit, &got); + if (err < 0) + goto done; + + if (err > 0 || memcmp(want.u.obj.hash_prefix, got.u.obj.hash_prefix, + r->object_id_len)) { + /* didn't find it; return empty iterator */ + iterator_set_empty(it); + err = 0; + goto done; + } + + err = new_indexed_table_ref_iter(&itr, r, oid, hash_size(r->hash_id), + got.u.obj.offsets, + got.u.obj.offset_len); + if (err < 0) + goto done; + got.u.obj.offsets = NULL; + iterator_from_indexed_table_ref_iter(it, itr); + +done: + reftable_iterator_destroy(&oit); + reftable_record_release(&got); + return err; +} + +static int reftable_reader_refs_for_unindexed(struct reftable_reader *r, + struct reftable_iterator *it, + uint8_t *oid) +{ + struct table_iter ti_empty = TABLE_ITER_INIT; + struct table_iter *ti = reftable_calloc(1, sizeof(*ti)); + struct filtering_ref_iterator *filter = NULL; + struct filtering_ref_iterator empty = FILTERING_REF_ITERATOR_INIT; + int oid_len = hash_size(r->hash_id); + int err; + + *ti = ti_empty; + err = reader_start(r, ti, BLOCK_TYPE_REF, 0); + if (err < 0) { + reftable_free(ti); + return err; + } + + filter = reftable_malloc(sizeof(struct filtering_ref_iterator)); + *filter = empty; + + strbuf_add(&filter->oid, oid, oid_len); + reftable_table_from_reader(&filter->tab, r); + filter->double_check = 0; + iterator_from_table_iter(&filter->it, ti); + + iterator_from_filtering_ref_iterator(it, filter); + return 0; +} + +int reftable_reader_refs_for(struct reftable_reader *r, + struct reftable_iterator *it, uint8_t *oid) +{ + if (r->obj_offsets.is_present) + return reftable_reader_refs_for_indexed(r, it, oid); + return reftable_reader_refs_for_unindexed(r, it, oid); +} + +uint64_t reftable_reader_max_update_index(struct reftable_reader *r) +{ + return r->max_update_index; +} + +uint64_t reftable_reader_min_update_index(struct reftable_reader *r) +{ + return r->min_update_index; +} + +/* generic table interface. */ + +static int reftable_reader_seek_void(void *tab, struct reftable_iterator *it, + struct reftable_record *rec) +{ + return reader_seek(tab, it, rec); +} + +static uint32_t reftable_reader_hash_id_void(void *tab) +{ + return reftable_reader_hash_id(tab); +} + +static uint64_t reftable_reader_min_update_index_void(void *tab) +{ + return reftable_reader_min_update_index(tab); +} + +static uint64_t reftable_reader_max_update_index_void(void *tab) +{ + return reftable_reader_max_update_index(tab); +} + +static struct reftable_table_vtable reader_vtable = { + .seek_record = reftable_reader_seek_void, + .hash_id = reftable_reader_hash_id_void, + .min_update_index = reftable_reader_min_update_index_void, + .max_update_index = reftable_reader_max_update_index_void, +}; + +void reftable_table_from_reader(struct reftable_table *tab, + struct reftable_reader *reader) +{ + assert(!tab->ops); + tab->ops = &reader_vtable; + tab->table_arg = reader; +} + + +int reftable_reader_print_file(const char *tablename) +{ + struct reftable_block_source src = { NULL }; + int err = reftable_block_source_from_file(&src, tablename); + struct reftable_reader *r = NULL; + struct reftable_table tab = { NULL }; + if (err < 0) + goto done; + + err = reftable_new_reader(&r, &src, tablename); + if (err < 0) + goto done; + + reftable_table_from_reader(&tab, r); + err = reftable_table_print(&tab); +done: + reftable_reader_free(r); + return err; +} diff --git a/reftable/reader.h b/reftable/reader.h new file mode 100644 index 0000000..e869165 --- /dev/null +++ b/reftable/reader.h @@ -0,0 +1,64 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef READER_H +#define READER_H + +#include "block.h" +#include "record.h" +#include "reftable-iterator.h" +#include "reftable-reader.h" + +uint64_t block_source_size(struct reftable_block_source *source); + +int block_source_read_block(struct reftable_block_source *source, + struct reftable_block *dest, uint64_t off, + uint32_t size); +void block_source_close(struct reftable_block_source *source); + +/* metadata for a block type */ +struct reftable_reader_offsets { + int is_present; + uint64_t offset; + uint64_t index_offset; +}; + +/* The state for reading a reftable file. */ +struct reftable_reader { + /* for convience, associate a name with the instance. */ + char *name; + struct reftable_block_source source; + + /* Size of the file, excluding the footer. */ + uint64_t size; + + /* 'sha1' for SHA1, 's256' for SHA-256 */ + uint32_t hash_id; + + uint32_t block_size; + uint64_t min_update_index; + uint64_t max_update_index; + /* Length of the OID keys in the 'o' section */ + int object_id_len; + int version; + + struct reftable_reader_offsets ref_offsets; + struct reftable_reader_offsets obj_offsets; + struct reftable_reader_offsets log_offsets; +}; + +int init_reader(struct reftable_reader *r, struct reftable_block_source *source, + const char *name); +void reader_close(struct reftable_reader *r); +const char *reader_name(struct reftable_reader *r); + +/* initialize a block reader to read from `r` */ +int reader_init_block_reader(struct reftable_reader *r, struct block_reader *br, + uint64_t next_off, uint8_t want_typ); + +#endif diff --git a/reftable/readwrite_test.c b/reftable/readwrite_test.c new file mode 100644 index 0000000..a6dbd21 --- /dev/null +++ b/reftable/readwrite_test.c @@ -0,0 +1,966 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "system.h" + +#include "basics.h" +#include "block.h" +#include "blocksource.h" +#include "reader.h" +#include "record.h" +#include "test_framework.h" +#include "reftable-tests.h" +#include "reftable-writer.h" + +static const int update_index = 5; + +static void test_buffer(void) +{ + struct strbuf buf = STRBUF_INIT; + struct reftable_block_source source = { NULL }; + struct reftable_block out = { NULL }; + int n; + uint8_t in[] = "hello"; + strbuf_add(&buf, in, sizeof(in)); + block_source_from_strbuf(&source, &buf); + EXPECT(block_source_size(&source) == 6); + n = block_source_read_block(&source, &out, 0, sizeof(in)); + EXPECT(n == sizeof(in)); + EXPECT(!memcmp(in, out.data, n)); + reftable_block_done(&out); + + n = block_source_read_block(&source, &out, 1, 2); + EXPECT(n == 2); + EXPECT(!memcmp(out.data, "el", 2)); + + reftable_block_done(&out); + block_source_close(&source); + strbuf_release(&buf); +} + +static void write_table(char ***names, struct strbuf *buf, int N, + int block_size, uint32_t hash_id) +{ + struct reftable_write_options opts = { + .block_size = block_size, + .hash_id = hash_id, + }; + struct reftable_writer *w = + reftable_new_writer(&strbuf_add_void, &noop_flush, buf, &opts); + struct reftable_ref_record ref = { NULL }; + int i = 0, n; + struct reftable_log_record log = { NULL }; + const struct reftable_stats *stats = NULL; + + REFTABLE_CALLOC_ARRAY(*names, N + 1); + + reftable_writer_set_limits(w, update_index, update_index); + for (i = 0; i < N; i++) { + char name[100]; + int n; + + snprintf(name, sizeof(name), "refs/heads/branch%02d", i); + + ref.refname = name; + ref.update_index = update_index; + ref.value_type = REFTABLE_REF_VAL1; + set_test_hash(ref.value.val1, i); + (*names)[i] = xstrdup(name); + + n = reftable_writer_add_ref(w, &ref); + EXPECT(n == 0); + } + + for (i = 0; i < N; i++) { + char name[100]; + int n; + + snprintf(name, sizeof(name), "refs/heads/branch%02d", i); + + log.refname = name; + log.update_index = update_index; + log.value_type = REFTABLE_LOG_UPDATE; + set_test_hash(log.value.update.new_hash, i); + log.value.update.message = "message"; + + n = reftable_writer_add_log(w, &log); + EXPECT(n == 0); + } + + n = reftable_writer_close(w); + EXPECT(n == 0); + + stats = reftable_writer_stats(w); + for (i = 0; i < stats->ref_stats.blocks; i++) { + int off = i * opts.block_size; + if (off == 0) { + off = header_size( + (hash_id == GIT_SHA256_FORMAT_ID) ? 2 : 1); + } + EXPECT(buf->buf[off] == 'r'); + } + + EXPECT(stats->log_stats.blocks > 0); + reftable_writer_free(w); +} + +static void test_log_buffer_size(void) +{ + struct strbuf buf = STRBUF_INIT; + struct reftable_write_options opts = { + .block_size = 4096, + }; + int err; + int i; + struct reftable_log_record + log = { .refname = "refs/heads/master", + .update_index = 0xa, + .value_type = REFTABLE_LOG_UPDATE, + .value = { .update = { + .name = "Han-Wen Nienhuys", + .email = "hanwen@google.com", + .tz_offset = 100, + .time = 0x5e430672, + .message = "commit: 9\n", + } } }; + struct reftable_writer *w = + reftable_new_writer(&strbuf_add_void, &noop_flush, &buf, &opts); + + /* This tests buffer extension for log compression. Must use a random + hash, to ensure that the compressed part is larger than the original. + */ + for (i = 0; i < GIT_SHA1_RAWSZ; i++) { + log.value.update.old_hash[i] = (uint8_t)(git_rand() % 256); + log.value.update.new_hash[i] = (uint8_t)(git_rand() % 256); + } + reftable_writer_set_limits(w, update_index, update_index); + err = reftable_writer_add_log(w, &log); + EXPECT_ERR(err); + err = reftable_writer_close(w); + EXPECT_ERR(err); + reftable_writer_free(w); + strbuf_release(&buf); +} + +static void test_log_overflow(void) +{ + struct strbuf buf = STRBUF_INIT; + char msg[256] = { 0 }; + struct reftable_write_options opts = { + .block_size = ARRAY_SIZE(msg), + }; + int err; + struct reftable_log_record log = { + .refname = "refs/heads/master", + .update_index = 0xa, + .value_type = REFTABLE_LOG_UPDATE, + .value = { + .update = { + .old_hash = { 1 }, + .new_hash = { 2 }, + .name = "Han-Wen Nienhuys", + .email = "hanwen@google.com", + .tz_offset = 100, + .time = 0x5e430672, + .message = msg, + }, + }, + }; + struct reftable_writer *w = + reftable_new_writer(&strbuf_add_void, &noop_flush, &buf, &opts); + + memset(msg, 'x', sizeof(msg) - 1); + reftable_writer_set_limits(w, update_index, update_index); + err = reftable_writer_add_log(w, &log); + EXPECT(err == REFTABLE_ENTRY_TOO_BIG_ERROR); + reftable_writer_free(w); + strbuf_release(&buf); +} + +static void test_log_write_read(void) +{ + int N = 2; + char **names = reftable_calloc(N + 1, sizeof(*names)); + int err; + struct reftable_write_options opts = { + .block_size = 256, + }; + struct reftable_ref_record ref = { NULL }; + int i = 0; + struct reftable_log_record log = { NULL }; + int n; + struct reftable_iterator it = { NULL }; + struct reftable_reader rd = { NULL }; + struct reftable_block_source source = { NULL }; + struct strbuf buf = STRBUF_INIT; + struct reftable_writer *w = + reftable_new_writer(&strbuf_add_void, &noop_flush, &buf, &opts); + const struct reftable_stats *stats = NULL; + reftable_writer_set_limits(w, 0, N); + for (i = 0; i < N; i++) { + char name[256]; + struct reftable_ref_record ref = { NULL }; + snprintf(name, sizeof(name), "b%02d%0*d", i, 130, 7); + names[i] = xstrdup(name); + ref.refname = name; + ref.update_index = i; + + err = reftable_writer_add_ref(w, &ref); + EXPECT_ERR(err); + } + for (i = 0; i < N; i++) { + struct reftable_log_record log = { NULL }; + + log.refname = names[i]; + log.update_index = i; + log.value_type = REFTABLE_LOG_UPDATE; + set_test_hash(log.value.update.old_hash, i); + set_test_hash(log.value.update.new_hash, i + 1); + + err = reftable_writer_add_log(w, &log); + EXPECT_ERR(err); + } + + n = reftable_writer_close(w); + EXPECT(n == 0); + + stats = reftable_writer_stats(w); + EXPECT(stats->log_stats.blocks > 0); + reftable_writer_free(w); + w = NULL; + + block_source_from_strbuf(&source, &buf); + + err = init_reader(&rd, &source, "file.log"); + EXPECT_ERR(err); + + err = reftable_reader_seek_ref(&rd, &it, names[N - 1]); + EXPECT_ERR(err); + + err = reftable_iterator_next_ref(&it, &ref); + EXPECT_ERR(err); + + /* end of iteration. */ + err = reftable_iterator_next_ref(&it, &ref); + EXPECT(0 < err); + + reftable_iterator_destroy(&it); + reftable_ref_record_release(&ref); + + err = reftable_reader_seek_log(&rd, &it, ""); + EXPECT_ERR(err); + + i = 0; + while (1) { + int err = reftable_iterator_next_log(&it, &log); + if (err > 0) { + break; + } + + EXPECT_ERR(err); + EXPECT_STREQ(names[i], log.refname); + EXPECT(i == log.update_index); + i++; + reftable_log_record_release(&log); + } + + EXPECT(i == N); + reftable_iterator_destroy(&it); + + /* cleanup. */ + strbuf_release(&buf); + free_names(names); + reader_close(&rd); +} + +static void test_log_zlib_corruption(void) +{ + struct reftable_write_options opts = { + .block_size = 256, + }; + struct reftable_iterator it = { 0 }; + struct reftable_reader rd = { 0 }; + struct reftable_block_source source = { 0 }; + struct strbuf buf = STRBUF_INIT; + struct reftable_writer *w = + reftable_new_writer(&strbuf_add_void, &noop_flush, &buf, &opts); + const struct reftable_stats *stats = NULL; + char message[100] = { 0 }; + int err, i, n; + struct reftable_log_record log = { + .refname = "refname", + .value_type = REFTABLE_LOG_UPDATE, + .value = { + .update = { + .new_hash = { 1 }, + .old_hash = { 2 }, + .name = "My Name", + .email = "myname@invalid", + .message = message, + }, + }, + }; + + for (i = 0; i < sizeof(message) - 1; i++) + message[i] = (uint8_t)(git_rand() % 64 + ' '); + + reftable_writer_set_limits(w, 1, 1); + + err = reftable_writer_add_log(w, &log); + EXPECT_ERR(err); + + n = reftable_writer_close(w); + EXPECT(n == 0); + + stats = reftable_writer_stats(w); + EXPECT(stats->log_stats.blocks > 0); + reftable_writer_free(w); + w = NULL; + + /* corrupt the data. */ + buf.buf[50] ^= 0x99; + + block_source_from_strbuf(&source, &buf); + + err = init_reader(&rd, &source, "file.log"); + EXPECT_ERR(err); + + err = reftable_reader_seek_log(&rd, &it, "refname"); + EXPECT(err == REFTABLE_ZLIB_ERROR); + + reftable_iterator_destroy(&it); + + /* cleanup. */ + strbuf_release(&buf); + reader_close(&rd); +} + +static void test_table_read_write_sequential(void) +{ + char **names; + struct strbuf buf = STRBUF_INIT; + int N = 50; + struct reftable_iterator it = { NULL }; + struct reftable_block_source source = { NULL }; + struct reftable_reader rd = { NULL }; + int err = 0; + int j = 0; + + write_table(&names, &buf, N, 256, GIT_SHA1_FORMAT_ID); + + block_source_from_strbuf(&source, &buf); + + err = init_reader(&rd, &source, "file.ref"); + EXPECT_ERR(err); + + err = reftable_reader_seek_ref(&rd, &it, ""); + EXPECT_ERR(err); + + while (1) { + struct reftable_ref_record ref = { NULL }; + int r = reftable_iterator_next_ref(&it, &ref); + EXPECT(r >= 0); + if (r > 0) { + break; + } + EXPECT(0 == strcmp(names[j], ref.refname)); + EXPECT(update_index == ref.update_index); + + j++; + reftable_ref_record_release(&ref); + } + EXPECT(j == N); + reftable_iterator_destroy(&it); + strbuf_release(&buf); + free_names(names); + + reader_close(&rd); +} + +static void test_table_write_small_table(void) +{ + char **names; + struct strbuf buf = STRBUF_INIT; + int N = 1; + write_table(&names, &buf, N, 4096, GIT_SHA1_FORMAT_ID); + EXPECT(buf.len < 200); + strbuf_release(&buf); + free_names(names); +} + +static void test_table_read_api(void) +{ + char **names; + struct strbuf buf = STRBUF_INIT; + int N = 50; + struct reftable_reader rd = { NULL }; + struct reftable_block_source source = { NULL }; + int err; + int i; + struct reftable_log_record log = { NULL }; + struct reftable_iterator it = { NULL }; + + write_table(&names, &buf, N, 256, GIT_SHA1_FORMAT_ID); + + block_source_from_strbuf(&source, &buf); + + err = init_reader(&rd, &source, "file.ref"); + EXPECT_ERR(err); + + err = reftable_reader_seek_ref(&rd, &it, names[0]); + EXPECT_ERR(err); + + err = reftable_iterator_next_log(&it, &log); + EXPECT(err == REFTABLE_API_ERROR); + + strbuf_release(&buf); + for (i = 0; i < N; i++) { + reftable_free(names[i]); + } + reftable_iterator_destroy(&it); + reftable_free(names); + reader_close(&rd); + strbuf_release(&buf); +} + +static void test_table_read_write_seek(int index, int hash_id) +{ + char **names; + struct strbuf buf = STRBUF_INIT; + int N = 50; + struct reftable_reader rd = { NULL }; + struct reftable_block_source source = { NULL }; + int err; + int i = 0; + + struct reftable_iterator it = { NULL }; + struct strbuf pastLast = STRBUF_INIT; + struct reftable_ref_record ref = { NULL }; + + write_table(&names, &buf, N, 256, hash_id); + + block_source_from_strbuf(&source, &buf); + + err = init_reader(&rd, &source, "file.ref"); + EXPECT_ERR(err); + EXPECT(hash_id == reftable_reader_hash_id(&rd)); + + if (!index) { + rd.ref_offsets.index_offset = 0; + } else { + EXPECT(rd.ref_offsets.index_offset > 0); + } + + for (i = 1; i < N; i++) { + int err = reftable_reader_seek_ref(&rd, &it, names[i]); + EXPECT_ERR(err); + err = reftable_iterator_next_ref(&it, &ref); + EXPECT_ERR(err); + EXPECT(0 == strcmp(names[i], ref.refname)); + EXPECT(REFTABLE_REF_VAL1 == ref.value_type); + EXPECT(i == ref.value.val1[0]); + + reftable_ref_record_release(&ref); + reftable_iterator_destroy(&it); + } + + strbuf_addstr(&pastLast, names[N - 1]); + strbuf_addstr(&pastLast, "/"); + + err = reftable_reader_seek_ref(&rd, &it, pastLast.buf); + if (err == 0) { + struct reftable_ref_record ref = { NULL }; + int err = reftable_iterator_next_ref(&it, &ref); + EXPECT(err > 0); + } else { + EXPECT(err > 0); + } + + strbuf_release(&pastLast); + reftable_iterator_destroy(&it); + + strbuf_release(&buf); + for (i = 0; i < N; i++) { + reftable_free(names[i]); + } + reftable_free(names); + reader_close(&rd); +} + +static void test_table_read_write_seek_linear(void) +{ + test_table_read_write_seek(0, GIT_SHA1_FORMAT_ID); +} + +static void test_table_read_write_seek_linear_sha256(void) +{ + test_table_read_write_seek(0, GIT_SHA256_FORMAT_ID); +} + +static void test_table_read_write_seek_index(void) +{ + test_table_read_write_seek(1, GIT_SHA1_FORMAT_ID); +} + +static void test_table_refs_for(int indexed) +{ + int N = 50; + char **want_names = reftable_calloc(N + 1, sizeof(*want_names)); + int want_names_len = 0; + uint8_t want_hash[GIT_SHA1_RAWSZ]; + + struct reftable_write_options opts = { + .block_size = 256, + }; + struct reftable_ref_record ref = { NULL }; + int i = 0; + int n; + int err; + struct reftable_reader rd; + struct reftable_block_source source = { NULL }; + + struct strbuf buf = STRBUF_INIT; + struct reftable_writer *w = + reftable_new_writer(&strbuf_add_void, &noop_flush, &buf, &opts); + + struct reftable_iterator it = { NULL }; + int j; + + set_test_hash(want_hash, 4); + + for (i = 0; i < N; i++) { + uint8_t hash[GIT_SHA1_RAWSZ]; + char fill[51] = { 0 }; + char name[100]; + struct reftable_ref_record ref = { NULL }; + + memset(hash, i, sizeof(hash)); + memset(fill, 'x', 50); + /* Put the variable part in the start */ + snprintf(name, sizeof(name), "br%02d%s", i, fill); + name[40] = 0; + ref.refname = name; + + ref.value_type = REFTABLE_REF_VAL2; + set_test_hash(ref.value.val2.value, i / 4); + set_test_hash(ref.value.val2.target_value, 3 + i / 4); + + /* 80 bytes / entry, so 3 entries per block. Yields 17 + */ + /* blocks. */ + n = reftable_writer_add_ref(w, &ref); + EXPECT(n == 0); + + if (!memcmp(ref.value.val2.value, want_hash, GIT_SHA1_RAWSZ) || + !memcmp(ref.value.val2.target_value, want_hash, GIT_SHA1_RAWSZ)) { + want_names[want_names_len++] = xstrdup(name); + } + } + + n = reftable_writer_close(w); + EXPECT(n == 0); + + reftable_writer_free(w); + w = NULL; + + block_source_from_strbuf(&source, &buf); + + err = init_reader(&rd, &source, "file.ref"); + EXPECT_ERR(err); + if (!indexed) { + rd.obj_offsets.is_present = 0; + } + + err = reftable_reader_seek_ref(&rd, &it, ""); + EXPECT_ERR(err); + reftable_iterator_destroy(&it); + + err = reftable_reader_refs_for(&rd, &it, want_hash); + EXPECT_ERR(err); + + j = 0; + while (1) { + int err = reftable_iterator_next_ref(&it, &ref); + EXPECT(err >= 0); + if (err > 0) { + break; + } + + EXPECT(j < want_names_len); + EXPECT(0 == strcmp(ref.refname, want_names[j])); + j++; + reftable_ref_record_release(&ref); + } + EXPECT(j == want_names_len); + + strbuf_release(&buf); + free_names(want_names); + reftable_iterator_destroy(&it); + reader_close(&rd); +} + +static void test_table_refs_for_no_index(void) +{ + test_table_refs_for(0); +} + +static void test_table_refs_for_obj_index(void) +{ + test_table_refs_for(1); +} + +static void test_write_empty_table(void) +{ + struct reftable_write_options opts = { 0 }; + struct strbuf buf = STRBUF_INIT; + struct reftable_writer *w = + reftable_new_writer(&strbuf_add_void, &noop_flush, &buf, &opts); + struct reftable_block_source source = { NULL }; + struct reftable_reader *rd = NULL; + struct reftable_ref_record rec = { NULL }; + struct reftable_iterator it = { NULL }; + int err; + + reftable_writer_set_limits(w, 1, 1); + + err = reftable_writer_close(w); + EXPECT(err == REFTABLE_EMPTY_TABLE_ERROR); + reftable_writer_free(w); + + EXPECT(buf.len == header_size(1) + footer_size(1)); + + block_source_from_strbuf(&source, &buf); + + err = reftable_new_reader(&rd, &source, "filename"); + EXPECT_ERR(err); + + err = reftable_reader_seek_ref(rd, &it, ""); + EXPECT_ERR(err); + + err = reftable_iterator_next_ref(&it, &rec); + EXPECT(err > 0); + + reftable_iterator_destroy(&it); + reftable_reader_free(rd); + strbuf_release(&buf); +} + +static void test_write_object_id_min_length(void) +{ + struct reftable_write_options opts = { + .block_size = 75, + }; + struct strbuf buf = STRBUF_INIT; + struct reftable_writer *w = + reftable_new_writer(&strbuf_add_void, &noop_flush, &buf, &opts); + struct reftable_ref_record ref = { + .update_index = 1, + .value_type = REFTABLE_REF_VAL1, + .value.val1 = {42}, + }; + int err; + int i; + + reftable_writer_set_limits(w, 1, 1); + + /* Write the same hash in many refs. If there is only 1 hash, the + * disambiguating prefix is length 0 */ + for (i = 0; i < 256; i++) { + char name[256]; + snprintf(name, sizeof(name), "ref%05d", i); + ref.refname = name; + err = reftable_writer_add_ref(w, &ref); + EXPECT_ERR(err); + } + + err = reftable_writer_close(w); + EXPECT_ERR(err); + EXPECT(reftable_writer_stats(w)->object_id_len == 2); + reftable_writer_free(w); + strbuf_release(&buf); +} + +static void test_write_object_id_length(void) +{ + struct reftable_write_options opts = { + .block_size = 75, + }; + struct strbuf buf = STRBUF_INIT; + struct reftable_writer *w = + reftable_new_writer(&strbuf_add_void, &noop_flush, &buf, &opts); + struct reftable_ref_record ref = { + .update_index = 1, + .value_type = REFTABLE_REF_VAL1, + .value.val1 = {42}, + }; + int err; + int i; + + reftable_writer_set_limits(w, 1, 1); + + /* Write the same hash in many refs. If there is only 1 hash, the + * disambiguating prefix is length 0 */ + for (i = 0; i < 256; i++) { + char name[256]; + snprintf(name, sizeof(name), "ref%05d", i); + ref.refname = name; + ref.value.val1[15] = i; + err = reftable_writer_add_ref(w, &ref); + EXPECT_ERR(err); + } + + err = reftable_writer_close(w); + EXPECT_ERR(err); + EXPECT(reftable_writer_stats(w)->object_id_len == 16); + reftable_writer_free(w); + strbuf_release(&buf); +} + +static void test_write_empty_key(void) +{ + struct reftable_write_options opts = { 0 }; + struct strbuf buf = STRBUF_INIT; + struct reftable_writer *w = + reftable_new_writer(&strbuf_add_void, &noop_flush, &buf, &opts); + struct reftable_ref_record ref = { + .refname = "", + .update_index = 1, + .value_type = REFTABLE_REF_DELETION, + }; + int err; + + reftable_writer_set_limits(w, 1, 1); + err = reftable_writer_add_ref(w, &ref); + EXPECT(err == REFTABLE_API_ERROR); + + err = reftable_writer_close(w); + EXPECT(err == REFTABLE_EMPTY_TABLE_ERROR); + reftable_writer_free(w); + strbuf_release(&buf); +} + +static void test_write_key_order(void) +{ + struct reftable_write_options opts = { 0 }; + struct strbuf buf = STRBUF_INIT; + struct reftable_writer *w = + reftable_new_writer(&strbuf_add_void, &noop_flush, &buf, &opts); + struct reftable_ref_record refs[2] = { + { + .refname = "b", + .update_index = 1, + .value_type = REFTABLE_REF_SYMREF, + .value = { + .symref = "target", + }, + }, { + .refname = "a", + .update_index = 1, + .value_type = REFTABLE_REF_SYMREF, + .value = { + .symref = "target", + }, + } + }; + int err; + + reftable_writer_set_limits(w, 1, 1); + err = reftable_writer_add_ref(w, &refs[0]); + EXPECT_ERR(err); + err = reftable_writer_add_ref(w, &refs[1]); + EXPECT(err == REFTABLE_API_ERROR); + reftable_writer_close(w); + reftable_writer_free(w); + strbuf_release(&buf); +} + +static void test_write_multiple_indices(void) +{ + struct reftable_write_options opts = { + .block_size = 100, + }; + struct strbuf writer_buf = STRBUF_INIT, buf = STRBUF_INIT; + struct reftable_block_source source = { 0 }; + struct reftable_iterator it = { 0 }; + const struct reftable_stats *stats; + struct reftable_writer *writer; + struct reftable_reader *reader; + int err, i; + + writer = reftable_new_writer(&strbuf_add_void, &noop_flush, &writer_buf, &opts); + reftable_writer_set_limits(writer, 1, 1); + for (i = 0; i < 100; i++) { + struct reftable_ref_record ref = { + .update_index = 1, + .value_type = REFTABLE_REF_VAL1, + .value.val1 = {i}, + }; + + strbuf_reset(&buf); + strbuf_addf(&buf, "refs/heads/%04d", i); + ref.refname = buf.buf, + + err = reftable_writer_add_ref(writer, &ref); + EXPECT_ERR(err); + } + + for (i = 0; i < 100; i++) { + struct reftable_log_record log = { + .update_index = 1, + .value_type = REFTABLE_LOG_UPDATE, + .value.update = { + .old_hash = { i }, + .new_hash = { i }, + }, + }; + + strbuf_reset(&buf); + strbuf_addf(&buf, "refs/heads/%04d", i); + log.refname = buf.buf, + + err = reftable_writer_add_log(writer, &log); + EXPECT_ERR(err); + } + + reftable_writer_close(writer); + + /* + * The written data should be sufficiently large to result in indices + * for each of the block types. + */ + stats = reftable_writer_stats(writer); + EXPECT(stats->ref_stats.index_offset > 0); + EXPECT(stats->obj_stats.index_offset > 0); + EXPECT(stats->log_stats.index_offset > 0); + + block_source_from_strbuf(&source, &writer_buf); + err = reftable_new_reader(&reader, &source, "filename"); + EXPECT_ERR(err); + + /* + * Seeking the log uses the log index now. In case there is any + * confusion regarding indices we would notice here. + */ + err = reftable_reader_seek_log(reader, &it, ""); + EXPECT_ERR(err); + + reftable_iterator_destroy(&it); + reftable_writer_free(writer); + reftable_reader_free(reader); + strbuf_release(&writer_buf); + strbuf_release(&buf); +} + +static void test_write_multi_level_index(void) +{ + struct reftable_write_options opts = { + .block_size = 100, + }; + struct strbuf writer_buf = STRBUF_INIT, buf = STRBUF_INIT; + struct reftable_block_source source = { 0 }; + struct reftable_iterator it = { 0 }; + const struct reftable_stats *stats; + struct reftable_writer *writer; + struct reftable_reader *reader; + int err; + + writer = reftable_new_writer(&strbuf_add_void, &noop_flush, &writer_buf, &opts); + reftable_writer_set_limits(writer, 1, 1); + for (size_t i = 0; i < 200; i++) { + struct reftable_ref_record ref = { + .update_index = 1, + .value_type = REFTABLE_REF_VAL1, + .value.val1 = {i}, + }; + + strbuf_reset(&buf); + strbuf_addf(&buf, "refs/heads/%03" PRIuMAX, (uintmax_t)i); + ref.refname = buf.buf, + + err = reftable_writer_add_ref(writer, &ref); + EXPECT_ERR(err); + } + reftable_writer_close(writer); + + /* + * The written refs should be sufficiently large to result in a + * multi-level index. + */ + stats = reftable_writer_stats(writer); + EXPECT(stats->ref_stats.max_index_level == 2); + + block_source_from_strbuf(&source, &writer_buf); + err = reftable_new_reader(&reader, &source, "filename"); + EXPECT_ERR(err); + + /* + * Seeking the last ref should work as expected. + */ + err = reftable_reader_seek_ref(reader, &it, "refs/heads/199"); + EXPECT_ERR(err); + + reftable_iterator_destroy(&it); + reftable_writer_free(writer); + reftable_reader_free(reader); + strbuf_release(&writer_buf); + strbuf_release(&buf); +} + +static void test_corrupt_table_empty(void) +{ + struct strbuf buf = STRBUF_INIT; + struct reftable_block_source source = { NULL }; + struct reftable_reader rd = { NULL }; + int err; + + block_source_from_strbuf(&source, &buf); + err = init_reader(&rd, &source, "file.log"); + EXPECT(err == REFTABLE_FORMAT_ERROR); +} + +static void test_corrupt_table(void) +{ + uint8_t zeros[1024] = { 0 }; + struct strbuf buf = STRBUF_INIT; + struct reftable_block_source source = { NULL }; + struct reftable_reader rd = { NULL }; + int err; + strbuf_add(&buf, zeros, sizeof(zeros)); + + block_source_from_strbuf(&source, &buf); + err = init_reader(&rd, &source, "file.log"); + EXPECT(err == REFTABLE_FORMAT_ERROR); + strbuf_release(&buf); +} + +int readwrite_test_main(int argc, const char *argv[]) +{ + RUN_TEST(test_log_zlib_corruption); + RUN_TEST(test_corrupt_table); + RUN_TEST(test_corrupt_table_empty); + RUN_TEST(test_log_write_read); + RUN_TEST(test_write_key_order); + RUN_TEST(test_table_read_write_seek_linear_sha256); + RUN_TEST(test_log_buffer_size); + RUN_TEST(test_table_write_small_table); + RUN_TEST(test_buffer); + RUN_TEST(test_table_read_api); + RUN_TEST(test_table_read_write_sequential); + RUN_TEST(test_table_read_write_seek_linear); + RUN_TEST(test_table_read_write_seek_index); + RUN_TEST(test_table_refs_for_no_index); + RUN_TEST(test_table_refs_for_obj_index); + RUN_TEST(test_write_empty_key); + RUN_TEST(test_write_empty_table); + RUN_TEST(test_log_overflow); + RUN_TEST(test_write_object_id_length); + RUN_TEST(test_write_object_id_min_length); + RUN_TEST(test_write_multiple_indices); + RUN_TEST(test_write_multi_level_index); + return 0; +} diff --git a/reftable/record.c b/reftable/record.c new file mode 100644 index 0000000..5506f3e --- /dev/null +++ b/reftable/record.c @@ -0,0 +1,1342 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +/* record.c - methods for different types of records. */ + +#include "record.h" + +#include "system.h" +#include "constants.h" +#include "reftable-error.h" +#include "basics.h" + +static struct reftable_record_vtable * +reftable_record_vtable(struct reftable_record *rec); +static void *reftable_record_data(struct reftable_record *rec); + +int get_var_int(uint64_t *dest, struct string_view *in) +{ + int ptr = 0; + uint64_t val; + + if (in->len == 0) + return -1; + val = in->buf[ptr] & 0x7f; + + while (in->buf[ptr] & 0x80) { + ptr++; + if (ptr > in->len) { + return -1; + } + val = (val + 1) << 7 | (uint64_t)(in->buf[ptr] & 0x7f); + } + + *dest = val; + return ptr + 1; +} + +int put_var_int(struct string_view *dest, uint64_t val) +{ + uint8_t buf[10] = { 0 }; + int i = 9; + int n = 0; + buf[i] = (uint8_t)(val & 0x7f); + i--; + while (1) { + val >>= 7; + if (!val) { + break; + } + val--; + buf[i] = 0x80 | (uint8_t)(val & 0x7f); + i--; + } + + n = sizeof(buf) - i - 1; + if (dest->len < n) + return -1; + memcpy(dest->buf, &buf[i + 1], n); + return n; +} + +int reftable_is_block_type(uint8_t typ) +{ + switch (typ) { + case BLOCK_TYPE_REF: + case BLOCK_TYPE_LOG: + case BLOCK_TYPE_OBJ: + case BLOCK_TYPE_INDEX: + return 1; + } + return 0; +} + +const unsigned char *reftable_ref_record_val1(const struct reftable_ref_record *rec) +{ + switch (rec->value_type) { + case REFTABLE_REF_VAL1: + return rec->value.val1; + case REFTABLE_REF_VAL2: + return rec->value.val2.value; + default: + return NULL; + } +} + +const unsigned char *reftable_ref_record_val2(const struct reftable_ref_record *rec) +{ + switch (rec->value_type) { + case REFTABLE_REF_VAL2: + return rec->value.val2.target_value; + default: + return NULL; + } +} + +static int decode_string(struct strbuf *dest, struct string_view in) +{ + int start_len = in.len; + uint64_t tsize = 0; + int n = get_var_int(&tsize, &in); + if (n <= 0) + return -1; + string_view_consume(&in, n); + if (in.len < tsize) + return -1; + + strbuf_reset(dest); + strbuf_add(dest, in.buf, tsize); + string_view_consume(&in, tsize); + + return start_len - in.len; +} + +static int encode_string(char *str, struct string_view s) +{ + struct string_view start = s; + int l = strlen(str); + int n = put_var_int(&s, l); + if (n < 0) + return -1; + string_view_consume(&s, n); + if (s.len < l) + return -1; + memcpy(s.buf, str, l); + string_view_consume(&s, l); + + return start.len - s.len; +} + +int reftable_encode_key(int *restart, struct string_view dest, + struct strbuf prev_key, struct strbuf key, + uint8_t extra) +{ + struct string_view start = dest; + int prefix_len = common_prefix_size(&prev_key, &key); + uint64_t suffix_len = key.len - prefix_len; + int n = put_var_int(&dest, (uint64_t)prefix_len); + if (n < 0) + return -1; + string_view_consume(&dest, n); + + *restart = (prefix_len == 0); + + n = put_var_int(&dest, suffix_len << 3 | (uint64_t)extra); + if (n < 0) + return -1; + string_view_consume(&dest, n); + + if (dest.len < suffix_len) + return -1; + memcpy(dest.buf, key.buf + prefix_len, suffix_len); + string_view_consume(&dest, suffix_len); + + return start.len - dest.len; +} + +int reftable_decode_keylen(struct string_view in, + uint64_t *prefix_len, + uint64_t *suffix_len, + uint8_t *extra) +{ + size_t start_len = in.len; + int n; + + n = get_var_int(prefix_len, &in); + if (n < 0) + return -1; + string_view_consume(&in, n); + + n = get_var_int(suffix_len, &in); + if (n <= 0) + return -1; + string_view_consume(&in, n); + + *extra = (uint8_t)(*suffix_len & 0x7); + *suffix_len >>= 3; + + return start_len - in.len; +} + +int reftable_decode_key(struct strbuf *last_key, uint8_t *extra, + struct string_view in) +{ + int start_len = in.len; + uint64_t prefix_len = 0; + uint64_t suffix_len = 0; + int n; + + n = reftable_decode_keylen(in, &prefix_len, &suffix_len, extra); + if (n < 0) + return -1; + string_view_consume(&in, n); + + if (in.len < suffix_len || + prefix_len > last_key->len) + return -1; + + strbuf_setlen(last_key, prefix_len); + strbuf_add(last_key, in.buf, suffix_len); + string_view_consume(&in, suffix_len); + + return start_len - in.len; +} + +static void reftable_ref_record_key(const void *r, struct strbuf *dest) +{ + const struct reftable_ref_record *rec = + (const struct reftable_ref_record *)r; + strbuf_reset(dest); + strbuf_addstr(dest, rec->refname); +} + +static void reftable_ref_record_copy_from(void *rec, const void *src_rec, + int hash_size) +{ + struct reftable_ref_record *ref = rec; + const struct reftable_ref_record *src = src_rec; + char *refname = NULL; + size_t refname_cap = 0; + + assert(hash_size > 0); + + SWAP(refname, ref->refname); + SWAP(refname_cap, ref->refname_cap); + reftable_ref_record_release(ref); + SWAP(ref->refname, refname); + SWAP(ref->refname_cap, refname_cap); + + if (src->refname) { + size_t refname_len = strlen(src->refname); + + REFTABLE_ALLOC_GROW(ref->refname, refname_len + 1, + ref->refname_cap); + memcpy(ref->refname, src->refname, refname_len); + ref->refname[refname_len] = 0; + } + + ref->update_index = src->update_index; + ref->value_type = src->value_type; + switch (src->value_type) { + case REFTABLE_REF_DELETION: + break; + case REFTABLE_REF_VAL1: + memcpy(ref->value.val1, src->value.val1, hash_size); + break; + case REFTABLE_REF_VAL2: + memcpy(ref->value.val2.value, src->value.val2.value, hash_size); + memcpy(ref->value.val2.target_value, + src->value.val2.target_value, hash_size); + break; + case REFTABLE_REF_SYMREF: + ref->value.symref = xstrdup(src->value.symref); + break; + } +} + +static char hexdigit(int c) +{ + if (c <= 9) + return '0' + c; + return 'a' + (c - 10); +} + +static void hex_format(char *dest, const unsigned char *src, int hash_size) +{ + assert(hash_size > 0); + if (src) { + int i = 0; + for (i = 0; i < hash_size; i++) { + dest[2 * i] = hexdigit(src[i] >> 4); + dest[2 * i + 1] = hexdigit(src[i] & 0xf); + } + dest[2 * hash_size] = 0; + } +} + +static void reftable_ref_record_print_sz(const struct reftable_ref_record *ref, + int hash_size) +{ + char hex[GIT_MAX_HEXSZ + 1] = { 0 }; /* BUG */ + printf("ref{%s(%" PRIu64 ") ", ref->refname, ref->update_index); + switch (ref->value_type) { + case REFTABLE_REF_SYMREF: + printf("=> %s", ref->value.symref); + break; + case REFTABLE_REF_VAL2: + hex_format(hex, ref->value.val2.value, hash_size); + printf("val 2 %s", hex); + hex_format(hex, ref->value.val2.target_value, + hash_size); + printf("(T %s)", hex); + break; + case REFTABLE_REF_VAL1: + hex_format(hex, ref->value.val1, hash_size); + printf("val 1 %s", hex); + break; + case REFTABLE_REF_DELETION: + printf("delete"); + break; + } + printf("}\n"); +} + +void reftable_ref_record_print(const struct reftable_ref_record *ref, + uint32_t hash_id) { + reftable_ref_record_print_sz(ref, hash_size(hash_id)); +} + +static void reftable_ref_record_release_void(void *rec) +{ + reftable_ref_record_release(rec); +} + +void reftable_ref_record_release(struct reftable_ref_record *ref) +{ + switch (ref->value_type) { + case REFTABLE_REF_SYMREF: + reftable_free(ref->value.symref); + break; + case REFTABLE_REF_VAL2: + break; + case REFTABLE_REF_VAL1: + break; + case REFTABLE_REF_DELETION: + break; + default: + abort(); + } + + reftable_free(ref->refname); + memset(ref, 0, sizeof(struct reftable_ref_record)); +} + +static uint8_t reftable_ref_record_val_type(const void *rec) +{ + const struct reftable_ref_record *r = + (const struct reftable_ref_record *)rec; + return r->value_type; +} + +static int reftable_ref_record_encode(const void *rec, struct string_view s, + int hash_size) +{ + const struct reftable_ref_record *r = + (const struct reftable_ref_record *)rec; + struct string_view start = s; + int n = put_var_int(&s, r->update_index); + assert(hash_size > 0); + if (n < 0) + return -1; + string_view_consume(&s, n); + + switch (r->value_type) { + case REFTABLE_REF_SYMREF: + n = encode_string(r->value.symref, s); + if (n < 0) { + return -1; + } + string_view_consume(&s, n); + break; + case REFTABLE_REF_VAL2: + if (s.len < 2 * hash_size) { + return -1; + } + memcpy(s.buf, r->value.val2.value, hash_size); + string_view_consume(&s, hash_size); + memcpy(s.buf, r->value.val2.target_value, hash_size); + string_view_consume(&s, hash_size); + break; + case REFTABLE_REF_VAL1: + if (s.len < hash_size) { + return -1; + } + memcpy(s.buf, r->value.val1, hash_size); + string_view_consume(&s, hash_size); + break; + case REFTABLE_REF_DELETION: + break; + default: + abort(); + } + + return start.len - s.len; +} + +static int reftable_ref_record_decode(void *rec, struct strbuf key, + uint8_t val_type, struct string_view in, + int hash_size, struct strbuf *scratch) +{ + struct reftable_ref_record *r = rec; + struct string_view start = in; + uint64_t update_index = 0; + const char *refname = NULL; + size_t refname_cap = 0; + int n; + + assert(hash_size > 0); + + n = get_var_int(&update_index, &in); + if (n < 0) + return n; + string_view_consume(&in, n); + + SWAP(refname, r->refname); + SWAP(refname_cap, r->refname_cap); + reftable_ref_record_release(r); + SWAP(r->refname, refname); + SWAP(r->refname_cap, refname_cap); + + REFTABLE_ALLOC_GROW(r->refname, key.len + 1, r->refname_cap); + memcpy(r->refname, key.buf, key.len); + r->refname[key.len] = 0; + + r->update_index = update_index; + r->value_type = val_type; + switch (val_type) { + case REFTABLE_REF_VAL1: + if (in.len < hash_size) { + return -1; + } + + memcpy(r->value.val1, in.buf, hash_size); + string_view_consume(&in, hash_size); + break; + + case REFTABLE_REF_VAL2: + if (in.len < 2 * hash_size) { + return -1; + } + + memcpy(r->value.val2.value, in.buf, hash_size); + string_view_consume(&in, hash_size); + + memcpy(r->value.val2.target_value, in.buf, hash_size); + string_view_consume(&in, hash_size); + break; + + case REFTABLE_REF_SYMREF: { + int n = decode_string(scratch, in); + if (n < 0) { + return -1; + } + string_view_consume(&in, n); + r->value.symref = strbuf_detach(scratch, NULL); + } break; + + case REFTABLE_REF_DELETION: + break; + default: + abort(); + break; + } + + return start.len - in.len; +} + +static int reftable_ref_record_is_deletion_void(const void *p) +{ + return reftable_ref_record_is_deletion( + (const struct reftable_ref_record *)p); +} + +static int reftable_ref_record_equal_void(const void *a, + const void *b, int hash_size) +{ + struct reftable_ref_record *ra = (struct reftable_ref_record *) a; + struct reftable_ref_record *rb = (struct reftable_ref_record *) b; + return reftable_ref_record_equal(ra, rb, hash_size); +} + +static int reftable_ref_record_cmp_void(const void *_a, const void *_b) +{ + const struct reftable_ref_record *a = _a; + const struct reftable_ref_record *b = _b; + return strcmp(a->refname, b->refname); +} + +static void reftable_ref_record_print_void(const void *rec, + int hash_size) +{ + reftable_ref_record_print_sz((struct reftable_ref_record *) rec, hash_size); +} + +static struct reftable_record_vtable reftable_ref_record_vtable = { + .key = &reftable_ref_record_key, + .type = BLOCK_TYPE_REF, + .copy_from = &reftable_ref_record_copy_from, + .val_type = &reftable_ref_record_val_type, + .encode = &reftable_ref_record_encode, + .decode = &reftable_ref_record_decode, + .release = &reftable_ref_record_release_void, + .is_deletion = &reftable_ref_record_is_deletion_void, + .equal = &reftable_ref_record_equal_void, + .cmp = &reftable_ref_record_cmp_void, + .print = &reftable_ref_record_print_void, +}; + +static void reftable_obj_record_key(const void *r, struct strbuf *dest) +{ + const struct reftable_obj_record *rec = + (const struct reftable_obj_record *)r; + strbuf_reset(dest); + strbuf_add(dest, rec->hash_prefix, rec->hash_prefix_len); +} + +static void reftable_obj_record_release(void *rec) +{ + struct reftable_obj_record *obj = rec; + FREE_AND_NULL(obj->hash_prefix); + FREE_AND_NULL(obj->offsets); + memset(obj, 0, sizeof(struct reftable_obj_record)); +} + +static void reftable_obj_record_print(const void *rec, int hash_size) +{ + const struct reftable_obj_record *obj = rec; + char hex[GIT_MAX_HEXSZ + 1] = { 0 }; + struct strbuf offset_str = STRBUF_INIT; + int i; + + for (i = 0; i < obj->offset_len; i++) + strbuf_addf(&offset_str, "%" PRIu64 " ", obj->offsets[i]); + hex_format(hex, obj->hash_prefix, obj->hash_prefix_len); + printf("prefix %s (len %d), offsets [%s]\n", + hex, obj->hash_prefix_len, offset_str.buf); + strbuf_release(&offset_str); +} + +static void reftable_obj_record_copy_from(void *rec, const void *src_rec, + int hash_size) +{ + struct reftable_obj_record *obj = rec; + const struct reftable_obj_record *src = + (const struct reftable_obj_record *)src_rec; + + reftable_obj_record_release(obj); + + REFTABLE_ALLOC_ARRAY(obj->hash_prefix, src->hash_prefix_len); + obj->hash_prefix_len = src->hash_prefix_len; + if (src->hash_prefix_len) + memcpy(obj->hash_prefix, src->hash_prefix, obj->hash_prefix_len); + + REFTABLE_ALLOC_ARRAY(obj->offsets, src->offset_len); + obj->offset_len = src->offset_len; + COPY_ARRAY(obj->offsets, src->offsets, src->offset_len); +} + +static uint8_t reftable_obj_record_val_type(const void *rec) +{ + const struct reftable_obj_record *r = rec; + if (r->offset_len > 0 && r->offset_len < 8) + return r->offset_len; + return 0; +} + +static int reftable_obj_record_encode(const void *rec, struct string_view s, + int hash_size) +{ + const struct reftable_obj_record *r = rec; + struct string_view start = s; + int i = 0; + int n = 0; + uint64_t last = 0; + if (r->offset_len == 0 || r->offset_len >= 8) { + n = put_var_int(&s, r->offset_len); + if (n < 0) { + return -1; + } + string_view_consume(&s, n); + } + if (r->offset_len == 0) + return start.len - s.len; + n = put_var_int(&s, r->offsets[0]); + if (n < 0) + return -1; + string_view_consume(&s, n); + + last = r->offsets[0]; + for (i = 1; i < r->offset_len; i++) { + int n = put_var_int(&s, r->offsets[i] - last); + if (n < 0) { + return -1; + } + string_view_consume(&s, n); + last = r->offsets[i]; + } + return start.len - s.len; +} + +static int reftable_obj_record_decode(void *rec, struct strbuf key, + uint8_t val_type, struct string_view in, + int hash_size, struct strbuf *scratch UNUSED) +{ + struct string_view start = in; + struct reftable_obj_record *r = rec; + uint64_t count = val_type; + int n = 0; + uint64_t last; + int j; + + reftable_obj_record_release(r); + + REFTABLE_ALLOC_ARRAY(r->hash_prefix, key.len); + memcpy(r->hash_prefix, key.buf, key.len); + r->hash_prefix_len = key.len; + + if (val_type == 0) { + n = get_var_int(&count, &in); + if (n < 0) { + return n; + } + + string_view_consume(&in, n); + } + + r->offsets = NULL; + r->offset_len = 0; + if (count == 0) + return start.len - in.len; + + REFTABLE_ALLOC_ARRAY(r->offsets, count); + r->offset_len = count; + + n = get_var_int(&r->offsets[0], &in); + if (n < 0) + return n; + string_view_consume(&in, n); + + last = r->offsets[0]; + j = 1; + while (j < count) { + uint64_t delta = 0; + int n = get_var_int(&delta, &in); + if (n < 0) { + return n; + } + string_view_consume(&in, n); + + last = r->offsets[j] = (delta + last); + j++; + } + return start.len - in.len; +} + +static int not_a_deletion(const void *p) +{ + return 0; +} + +static int reftable_obj_record_equal_void(const void *a, const void *b, int hash_size) +{ + struct reftable_obj_record *ra = (struct reftable_obj_record *) a; + struct reftable_obj_record *rb = (struct reftable_obj_record *) b; + + if (ra->hash_prefix_len != rb->hash_prefix_len + || ra->offset_len != rb->offset_len) + return 0; + + if (ra->hash_prefix_len && + memcmp(ra->hash_prefix, rb->hash_prefix, ra->hash_prefix_len)) + return 0; + if (ra->offset_len && + memcmp(ra->offsets, rb->offsets, ra->offset_len * sizeof(uint64_t))) + return 0; + + return 1; +} + +static int reftable_obj_record_cmp_void(const void *_a, const void *_b) +{ + const struct reftable_obj_record *a = _a; + const struct reftable_obj_record *b = _b; + int cmp; + + cmp = memcmp(a->hash_prefix, b->hash_prefix, + a->hash_prefix_len > b->hash_prefix_len ? + a->hash_prefix_len : b->hash_prefix_len); + if (cmp) + return cmp; + + /* + * When the prefix is the same then the object record that is longer is + * considered to be bigger. + */ + return a->hash_prefix_len - b->hash_prefix_len; +} + +static struct reftable_record_vtable reftable_obj_record_vtable = { + .key = &reftable_obj_record_key, + .type = BLOCK_TYPE_OBJ, + .copy_from = &reftable_obj_record_copy_from, + .val_type = &reftable_obj_record_val_type, + .encode = &reftable_obj_record_encode, + .decode = &reftable_obj_record_decode, + .release = &reftable_obj_record_release, + .is_deletion = ¬_a_deletion, + .equal = &reftable_obj_record_equal_void, + .cmp = &reftable_obj_record_cmp_void, + .print = &reftable_obj_record_print, +}; + +static void reftable_log_record_print_sz(struct reftable_log_record *log, + int hash_size) +{ + char hex[GIT_MAX_HEXSZ + 1] = { 0 }; + + switch (log->value_type) { + case REFTABLE_LOG_DELETION: + printf("log{%s(%" PRIu64 ") delete\n", log->refname, + log->update_index); + break; + case REFTABLE_LOG_UPDATE: + printf("log{%s(%" PRIu64 ") %s <%s> %" PRIu64 " %04d\n", + log->refname, log->update_index, + log->value.update.name ? log->value.update.name : "", + log->value.update.email ? log->value.update.email : "", + log->value.update.time, + log->value.update.tz_offset); + hex_format(hex, log->value.update.old_hash, hash_size); + printf("%s => ", hex); + hex_format(hex, log->value.update.new_hash, hash_size); + printf("%s\n\n%s\n}\n", hex, + log->value.update.message ? log->value.update.message : ""); + break; + } +} + +void reftable_log_record_print(struct reftable_log_record *log, + uint32_t hash_id) +{ + reftable_log_record_print_sz(log, hash_size(hash_id)); +} + +static void reftable_log_record_key(const void *r, struct strbuf *dest) +{ + const struct reftable_log_record *rec = + (const struct reftable_log_record *)r; + int len = strlen(rec->refname); + uint8_t i64[8]; + uint64_t ts = 0; + strbuf_reset(dest); + strbuf_add(dest, (uint8_t *)rec->refname, len + 1); + + ts = (~ts) - rec->update_index; + put_be64(&i64[0], ts); + strbuf_add(dest, i64, sizeof(i64)); +} + +static void reftable_log_record_copy_from(void *rec, const void *src_rec, + int hash_size) +{ + struct reftable_log_record *dst = rec; + const struct reftable_log_record *src = + (const struct reftable_log_record *)src_rec; + + reftable_log_record_release(dst); + *dst = *src; + if (dst->refname) { + dst->refname = xstrdup(dst->refname); + } + switch (dst->value_type) { + case REFTABLE_LOG_DELETION: + break; + case REFTABLE_LOG_UPDATE: + if (dst->value.update.email) { + dst->value.update.email = + xstrdup(dst->value.update.email); + } + if (dst->value.update.name) { + dst->value.update.name = + xstrdup(dst->value.update.name); + } + if (dst->value.update.message) { + dst->value.update.message = + xstrdup(dst->value.update.message); + } + + memcpy(dst->value.update.new_hash, + src->value.update.new_hash, hash_size); + memcpy(dst->value.update.old_hash, + src->value.update.old_hash, hash_size); + break; + } +} + +static void reftable_log_record_release_void(void *rec) +{ + struct reftable_log_record *r = rec; + reftable_log_record_release(r); +} + +void reftable_log_record_release(struct reftable_log_record *r) +{ + reftable_free(r->refname); + switch (r->value_type) { + case REFTABLE_LOG_DELETION: + break; + case REFTABLE_LOG_UPDATE: + reftable_free(r->value.update.name); + reftable_free(r->value.update.email); + reftable_free(r->value.update.message); + break; + } + memset(r, 0, sizeof(struct reftable_log_record)); +} + +static uint8_t reftable_log_record_val_type(const void *rec) +{ + const struct reftable_log_record *log = + (const struct reftable_log_record *)rec; + + return reftable_log_record_is_deletion(log) ? 0 : 1; +} + +static int reftable_log_record_encode(const void *rec, struct string_view s, + int hash_size) +{ + const struct reftable_log_record *r = rec; + struct string_view start = s; + int n = 0; + if (reftable_log_record_is_deletion(r)) + return 0; + + if (s.len < 2 * hash_size) + return -1; + + memcpy(s.buf, r->value.update.old_hash, hash_size); + memcpy(s.buf + hash_size, r->value.update.new_hash, hash_size); + string_view_consume(&s, 2 * hash_size); + + n = encode_string(r->value.update.name ? r->value.update.name : "", s); + if (n < 0) + return -1; + string_view_consume(&s, n); + + n = encode_string(r->value.update.email ? r->value.update.email : "", + s); + if (n < 0) + return -1; + string_view_consume(&s, n); + + n = put_var_int(&s, r->value.update.time); + if (n < 0) + return -1; + string_view_consume(&s, n); + + if (s.len < 2) + return -1; + + put_be16(s.buf, r->value.update.tz_offset); + string_view_consume(&s, 2); + + n = encode_string( + r->value.update.message ? r->value.update.message : "", s); + if (n < 0) + return -1; + string_view_consume(&s, n); + + return start.len - s.len; +} + +static int reftable_log_record_decode(void *rec, struct strbuf key, + uint8_t val_type, struct string_view in, + int hash_size, struct strbuf *scratch) +{ + struct string_view start = in; + struct reftable_log_record *r = rec; + uint64_t max = 0; + uint64_t ts = 0; + int n; + + if (key.len <= 9 || key.buf[key.len - 9] != 0) + return REFTABLE_FORMAT_ERROR; + + REFTABLE_ALLOC_GROW(r->refname, key.len - 8, r->refname_cap); + memcpy(r->refname, key.buf, key.len - 8); + ts = get_be64(key.buf + key.len - 8); + + r->update_index = (~max) - ts; + + if (val_type != r->value_type) { + switch (r->value_type) { + case REFTABLE_LOG_UPDATE: + FREE_AND_NULL(r->value.update.message); + r->value.update.message_cap = 0; + FREE_AND_NULL(r->value.update.email); + FREE_AND_NULL(r->value.update.name); + break; + case REFTABLE_LOG_DELETION: + break; + } + } + + r->value_type = val_type; + if (val_type == REFTABLE_LOG_DELETION) + return 0; + + if (in.len < 2 * hash_size) + return REFTABLE_FORMAT_ERROR; + + memcpy(r->value.update.old_hash, in.buf, hash_size); + memcpy(r->value.update.new_hash, in.buf + hash_size, hash_size); + + string_view_consume(&in, 2 * hash_size); + + n = decode_string(scratch, in); + if (n < 0) + goto done; + string_view_consume(&in, n); + + /* + * In almost all cases we can expect the reflog name to not change for + * reflog entries as they are tied to the local identity, not to the + * target commits. As an optimization for this common case we can thus + * skip copying over the name in case it's accurate already. + */ + if (!r->value.update.name || + strcmp(r->value.update.name, scratch->buf)) { + r->value.update.name = + reftable_realloc(r->value.update.name, scratch->len + 1); + memcpy(r->value.update.name, scratch->buf, scratch->len); + r->value.update.name[scratch->len] = 0; + } + + n = decode_string(scratch, in); + if (n < 0) + goto done; + string_view_consume(&in, n); + + /* Same as above, but for the reflog email. */ + if (!r->value.update.email || + strcmp(r->value.update.email, scratch->buf)) { + r->value.update.email = + reftable_realloc(r->value.update.email, scratch->len + 1); + memcpy(r->value.update.email, scratch->buf, scratch->len); + r->value.update.email[scratch->len] = 0; + } + + ts = 0; + n = get_var_int(&ts, &in); + if (n < 0) + goto done; + string_view_consume(&in, n); + r->value.update.time = ts; + if (in.len < 2) + goto done; + + r->value.update.tz_offset = get_be16(in.buf); + string_view_consume(&in, 2); + + n = decode_string(scratch, in); + if (n < 0) + goto done; + string_view_consume(&in, n); + + REFTABLE_ALLOC_GROW(r->value.update.message, scratch->len + 1, + r->value.update.message_cap); + memcpy(r->value.update.message, scratch->buf, scratch->len); + r->value.update.message[scratch->len] = 0; + + return start.len - in.len; + +done: + return REFTABLE_FORMAT_ERROR; +} + +static int null_streq(char *a, char *b) +{ + char *empty = ""; + if (!a) + a = empty; + + if (!b) + b = empty; + + return 0 == strcmp(a, b); +} + +static int reftable_log_record_equal_void(const void *a, + const void *b, int hash_size) +{ + return reftable_log_record_equal((struct reftable_log_record *) a, + (struct reftable_log_record *) b, + hash_size); +} + +static int reftable_log_record_cmp_void(const void *_a, const void *_b) +{ + const struct reftable_log_record *a = _a; + const struct reftable_log_record *b = _b; + int cmp = strcmp(a->refname, b->refname); + if (cmp) + return cmp; + + /* + * Note that the comparison here is reversed. This is because the + * update index is reversed when comparing keys. For reference, see how + * we handle this in reftable_log_record_key()`. + */ + return b->update_index - a->update_index; +} + +int reftable_log_record_equal(const struct reftable_log_record *a, + const struct reftable_log_record *b, int hash_size) +{ + if (!(null_streq(a->refname, b->refname) && + a->update_index == b->update_index && + a->value_type == b->value_type)) + return 0; + + switch (a->value_type) { + case REFTABLE_LOG_DELETION: + return 1; + case REFTABLE_LOG_UPDATE: + return null_streq(a->value.update.name, b->value.update.name) && + a->value.update.time == b->value.update.time && + a->value.update.tz_offset == b->value.update.tz_offset && + null_streq(a->value.update.email, + b->value.update.email) && + null_streq(a->value.update.message, + b->value.update.message) && + !memcmp(a->value.update.old_hash, + b->value.update.old_hash, hash_size) && + !memcmp(a->value.update.new_hash, + b->value.update.new_hash, hash_size); + } + + abort(); +} + +static int reftable_log_record_is_deletion_void(const void *p) +{ + return reftable_log_record_is_deletion( + (const struct reftable_log_record *)p); +} + +static void reftable_log_record_print_void(const void *rec, int hash_size) +{ + reftable_log_record_print_sz((struct reftable_log_record*)rec, hash_size); +} + +static struct reftable_record_vtable reftable_log_record_vtable = { + .key = &reftable_log_record_key, + .type = BLOCK_TYPE_LOG, + .copy_from = &reftable_log_record_copy_from, + .val_type = &reftable_log_record_val_type, + .encode = &reftable_log_record_encode, + .decode = &reftable_log_record_decode, + .release = &reftable_log_record_release_void, + .is_deletion = &reftable_log_record_is_deletion_void, + .equal = &reftable_log_record_equal_void, + .cmp = &reftable_log_record_cmp_void, + .print = &reftable_log_record_print_void, +}; + +static void reftable_index_record_key(const void *r, struct strbuf *dest) +{ + const struct reftable_index_record *rec = r; + strbuf_reset(dest); + strbuf_addbuf(dest, &rec->last_key); +} + +static void reftable_index_record_copy_from(void *rec, const void *src_rec, + int hash_size) +{ + struct reftable_index_record *dst = rec; + const struct reftable_index_record *src = src_rec; + + strbuf_reset(&dst->last_key); + strbuf_addbuf(&dst->last_key, &src->last_key); + dst->offset = src->offset; +} + +static void reftable_index_record_release(void *rec) +{ + struct reftable_index_record *idx = rec; + strbuf_release(&idx->last_key); +} + +static uint8_t reftable_index_record_val_type(const void *rec) +{ + return 0; +} + +static int reftable_index_record_encode(const void *rec, struct string_view out, + int hash_size) +{ + const struct reftable_index_record *r = + (const struct reftable_index_record *)rec; + struct string_view start = out; + + int n = put_var_int(&out, r->offset); + if (n < 0) + return n; + + string_view_consume(&out, n); + + return start.len - out.len; +} + +static int reftable_index_record_decode(void *rec, struct strbuf key, + uint8_t val_type, struct string_view in, + int hash_size, struct strbuf *scratch UNUSED) +{ + struct string_view start = in; + struct reftable_index_record *r = rec; + int n = 0; + + strbuf_reset(&r->last_key); + strbuf_addbuf(&r->last_key, &key); + + n = get_var_int(&r->offset, &in); + if (n < 0) + return n; + + string_view_consume(&in, n); + return start.len - in.len; +} + +static int reftable_index_record_equal(const void *a, const void *b, int hash_size) +{ + struct reftable_index_record *ia = (struct reftable_index_record *) a; + struct reftable_index_record *ib = (struct reftable_index_record *) b; + + return ia->offset == ib->offset && !strbuf_cmp(&ia->last_key, &ib->last_key); +} + +static int reftable_index_record_cmp(const void *_a, const void *_b) +{ + const struct reftable_index_record *a = _a; + const struct reftable_index_record *b = _b; + return strbuf_cmp(&a->last_key, &b->last_key); +} + +static void reftable_index_record_print(const void *rec, int hash_size) +{ + const struct reftable_index_record *idx = rec; + /* TODO: escape null chars? */ + printf("\"%s\" %" PRIu64 "\n", idx->last_key.buf, idx->offset); +} + +static struct reftable_record_vtable reftable_index_record_vtable = { + .key = &reftable_index_record_key, + .type = BLOCK_TYPE_INDEX, + .copy_from = &reftable_index_record_copy_from, + .val_type = &reftable_index_record_val_type, + .encode = &reftable_index_record_encode, + .decode = &reftable_index_record_decode, + .release = &reftable_index_record_release, + .is_deletion = ¬_a_deletion, + .equal = &reftable_index_record_equal, + .cmp = &reftable_index_record_cmp, + .print = &reftable_index_record_print, +}; + +void reftable_record_key(struct reftable_record *rec, struct strbuf *dest) +{ + reftable_record_vtable(rec)->key(reftable_record_data(rec), dest); +} + +int reftable_record_encode(struct reftable_record *rec, struct string_view dest, + int hash_size) +{ + return reftable_record_vtable(rec)->encode(reftable_record_data(rec), + dest, hash_size); +} + +void reftable_record_copy_from(struct reftable_record *rec, + struct reftable_record *src, int hash_size) +{ + assert(src->type == rec->type); + + reftable_record_vtable(rec)->copy_from(reftable_record_data(rec), + reftable_record_data(src), + hash_size); +} + +uint8_t reftable_record_val_type(struct reftable_record *rec) +{ + return reftable_record_vtable(rec)->val_type(reftable_record_data(rec)); +} + +int reftable_record_decode(struct reftable_record *rec, struct strbuf key, + uint8_t extra, struct string_view src, int hash_size, + struct strbuf *scratch) +{ + return reftable_record_vtable(rec)->decode(reftable_record_data(rec), + key, extra, src, hash_size, + scratch); +} + +void reftable_record_release(struct reftable_record *rec) +{ + reftable_record_vtable(rec)->release(reftable_record_data(rec)); +} + +int reftable_record_is_deletion(struct reftable_record *rec) +{ + return reftable_record_vtable(rec)->is_deletion( + reftable_record_data(rec)); +} + +int reftable_record_cmp(struct reftable_record *a, struct reftable_record *b) +{ + if (a->type != b->type) + BUG("cannot compare reftable records of different type"); + return reftable_record_vtable(a)->cmp( + reftable_record_data(a), reftable_record_data(b)); +} + +int reftable_record_equal(struct reftable_record *a, struct reftable_record *b, int hash_size) +{ + if (a->type != b->type) + return 0; + return reftable_record_vtable(a)->equal( + reftable_record_data(a), reftable_record_data(b), hash_size); +} + +static int hash_equal(const unsigned char *a, const unsigned char *b, int hash_size) +{ + if (a && b) + return !memcmp(a, b, hash_size); + + return a == b; +} + +int reftable_ref_record_equal(const struct reftable_ref_record *a, + const struct reftable_ref_record *b, int hash_size) +{ + assert(hash_size > 0); + if (!null_streq(a->refname, b->refname)) + return 0; + + if (a->update_index != b->update_index || + a->value_type != b->value_type) + return 0; + + switch (a->value_type) { + case REFTABLE_REF_SYMREF: + return !strcmp(a->value.symref, b->value.symref); + case REFTABLE_REF_VAL2: + return hash_equal(a->value.val2.value, b->value.val2.value, + hash_size) && + hash_equal(a->value.val2.target_value, + b->value.val2.target_value, hash_size); + case REFTABLE_REF_VAL1: + return hash_equal(a->value.val1, b->value.val1, hash_size); + case REFTABLE_REF_DELETION: + return 1; + default: + abort(); + } +} + +int reftable_ref_record_compare_name(const void *a, const void *b) +{ + return strcmp(((struct reftable_ref_record *)a)->refname, + ((struct reftable_ref_record *)b)->refname); +} + +int reftable_ref_record_is_deletion(const struct reftable_ref_record *ref) +{ + return ref->value_type == REFTABLE_REF_DELETION; +} + +int reftable_log_record_compare_key(const void *a, const void *b) +{ + const struct reftable_log_record *la = a; + const struct reftable_log_record *lb = b; + + int cmp = strcmp(la->refname, lb->refname); + if (cmp) + return cmp; + if (la->update_index > lb->update_index) + return -1; + return (la->update_index < lb->update_index) ? 1 : 0; +} + +int reftable_log_record_is_deletion(const struct reftable_log_record *log) +{ + return (log->value_type == REFTABLE_LOG_DELETION); +} + +static void *reftable_record_data(struct reftable_record *rec) +{ + switch (rec->type) { + case BLOCK_TYPE_REF: + return &rec->u.ref; + case BLOCK_TYPE_LOG: + return &rec->u.log; + case BLOCK_TYPE_INDEX: + return &rec->u.idx; + case BLOCK_TYPE_OBJ: + return &rec->u.obj; + } + abort(); +} + +static struct reftable_record_vtable * +reftable_record_vtable(struct reftable_record *rec) +{ + switch (rec->type) { + case BLOCK_TYPE_REF: + return &reftable_ref_record_vtable; + case BLOCK_TYPE_LOG: + return &reftable_log_record_vtable; + case BLOCK_TYPE_INDEX: + return &reftable_index_record_vtable; + case BLOCK_TYPE_OBJ: + return &reftable_obj_record_vtable; + } + abort(); +} + +void reftable_record_init(struct reftable_record *rec, uint8_t typ) +{ + memset(rec, 0, sizeof(*rec)); + rec->type = typ; + + switch (typ) { + case BLOCK_TYPE_REF: + case BLOCK_TYPE_LOG: + case BLOCK_TYPE_OBJ: + return; + case BLOCK_TYPE_INDEX: + strbuf_init(&rec->u.idx.last_key, 0); + return; + default: + BUG("unhandled record type"); + } +} + +void reftable_record_print(struct reftable_record *rec, int hash_size) +{ + printf("'%c': ", rec->type); + reftable_record_vtable(rec)->print(reftable_record_data(rec), hash_size); +} diff --git a/reftable/record.h b/reftable/record.h new file mode 100644 index 0000000..d778133 --- /dev/null +++ b/reftable/record.h @@ -0,0 +1,165 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef RECORD_H +#define RECORD_H + +#include "system.h" + +#include <stdint.h> + +#include "reftable-record.h" + +/* + * A substring of existing string data. This structure takes no responsibility + * for the lifetime of the data it points to. + */ +struct string_view { + uint8_t *buf; + size_t len; +}; + +/* Advance `s.buf` by `n`, and decrease length. */ +static inline void string_view_consume(struct string_view *s, int n) +{ + s->buf += n; + s->len -= n; +} + +/* utilities for de/encoding varints */ + +int get_var_int(uint64_t *dest, struct string_view *in); +int put_var_int(struct string_view *dest, uint64_t val); + +/* Methods for records. */ +struct reftable_record_vtable { + /* encode the key of to a uint8_t strbuf. */ + void (*key)(const void *rec, struct strbuf *dest); + + /* The record type of ('r' for ref). */ + uint8_t type; + + void (*copy_from)(void *dest, const void *src, int hash_size); + + /* a value of [0..7], indicating record subvariants (eg. ref vs. symref + * vs ref deletion) */ + uint8_t (*val_type)(const void *rec); + + /* encodes rec into dest, returning how much space was used. */ + int (*encode)(const void *rec, struct string_view dest, int hash_size); + + /* decode data from `src` into the record. */ + int (*decode)(void *rec, struct strbuf key, uint8_t extra, + struct string_view src, int hash_size, + struct strbuf *scratch); + + /* deallocate and null the record. */ + void (*release)(void *rec); + + /* is this a tombstone? */ + int (*is_deletion)(const void *rec); + + /* Are two records equal? This assumes they have the same type. Returns 0 for non-equal. */ + int (*equal)(const void *a, const void *b, int hash_size); + + /* + * Compare keys of two records with each other. The records must have + * the same type. + */ + int (*cmp)(const void *a, const void *b); + + /* Print on stdout, for debugging. */ + void (*print)(const void *rec, int hash_size); +}; + +/* returns true for recognized block types. Block start with the block type. */ +int reftable_is_block_type(uint8_t typ); + +/* Encode `key` into `dest`. Sets `is_restart` to indicate a restart. Returns + * number of bytes written. */ +int reftable_encode_key(int *is_restart, struct string_view dest, + struct strbuf prev_key, struct strbuf key, + uint8_t extra); + +/* Decode a record's key lengths. */ +int reftable_decode_keylen(struct string_view in, + uint64_t *prefix_len, + uint64_t *suffix_len, + uint8_t *extra); + +/* + * Decode into `last_key` and `extra` from `in`. `last_key` is expected to + * contain the decoded key of the preceding record, if any. + */ +int reftable_decode_key(struct strbuf *last_key, uint8_t *extra, + struct string_view in); + +/* reftable_index_record are used internally to speed up lookups. */ +struct reftable_index_record { + uint64_t offset; /* Offset of block */ + struct strbuf last_key; /* Last key of the block. */ +}; + +/* reftable_obj_record stores an object ID => ref mapping. */ +struct reftable_obj_record { + uint8_t *hash_prefix; /* leading bytes of the object ID */ + int hash_prefix_len; /* number of leading bytes. Constant + * across a single table. */ + uint64_t *offsets; /* a vector of file offsets. */ + int offset_len; +}; + +/* record is a generic wrapper for different types of records. It is normally + * created on the stack, or embedded within another struct. If the type is + * known, a fresh instance can be initialized explicitly. Otherwise, use + * `reftable_record_init()` to initialize generically (as the index_record is + * not valid as 0-initialized structure) + */ +struct reftable_record { + uint8_t type; + union { + struct reftable_ref_record ref; + struct reftable_log_record log; + struct reftable_obj_record obj; + struct reftable_index_record idx; + } u; +}; + +/* Initialize the reftable record for the given type */ +void reftable_record_init(struct reftable_record *rec, uint8_t typ); + +/* see struct record_vtable */ +int reftable_record_cmp(struct reftable_record *a, struct reftable_record *b); +int reftable_record_equal(struct reftable_record *a, struct reftable_record *b, int hash_size); +void reftable_record_print(struct reftable_record *rec, int hash_size); +void reftable_record_key(struct reftable_record *rec, struct strbuf *dest); +void reftable_record_copy_from(struct reftable_record *rec, + struct reftable_record *src, int hash_size); +uint8_t reftable_record_val_type(struct reftable_record *rec); +int reftable_record_encode(struct reftable_record *rec, struct string_view dest, + int hash_size); +int reftable_record_decode(struct reftable_record *rec, struct strbuf key, + uint8_t extra, struct string_view src, + int hash_size, struct strbuf *scratch); +int reftable_record_is_deletion(struct reftable_record *rec); + +static inline uint8_t reftable_record_type(struct reftable_record *rec) +{ + return rec->type; +} + +/* frees and zeroes out the embedded record */ +void reftable_record_release(struct reftable_record *rec); + +/* for qsort. */ +int reftable_ref_record_compare_name(const void *a, const void *b); + +/* for qsort. */ +int reftable_log_record_compare_key(const void *a, const void *b); + +#endif diff --git a/reftable/record_test.c b/reftable/record_test.c new file mode 100644 index 0000000..c158ee7 --- /dev/null +++ b/reftable/record_test.c @@ -0,0 +1,419 @@ +/* + Copyright 2020 Google LLC + + Use of this source code is governed by a BSD-style + license that can be found in the LICENSE file or at + https://developers.google.com/open-source/licenses/bsd +*/ + +#include "record.h" + +#include "system.h" +#include "basics.h" +#include "constants.h" +#include "test_framework.h" +#include "reftable-tests.h" + +static void test_copy(struct reftable_record *rec) +{ + struct reftable_record copy; + uint8_t typ; + + typ = reftable_record_type(rec); + reftable_record_init(©, typ); + reftable_record_copy_from(©, rec, GIT_SHA1_RAWSZ); + /* do it twice to catch memory leaks */ + reftable_record_copy_from(©, rec, GIT_SHA1_RAWSZ); + EXPECT(reftable_record_equal(rec, ©, GIT_SHA1_RAWSZ)); + + puts("testing print coverage:\n"); + reftable_record_print(©, GIT_SHA1_RAWSZ); + + reftable_record_release(©); +} + +static void test_varint_roundtrip(void) +{ + uint64_t inputs[] = { 0, + 1, + 27, + 127, + 128, + 257, + 4096, + ((uint64_t)1 << 63), + ((uint64_t)1 << 63) + ((uint64_t)1 << 63) - 1 }; + int i = 0; + for (i = 0; i < ARRAY_SIZE(inputs); i++) { + uint8_t dest[10]; + + struct string_view out = { + .buf = dest, + .len = sizeof(dest), + }; + uint64_t in = inputs[i]; + int n = put_var_int(&out, in); + uint64_t got = 0; + + EXPECT(n > 0); + out.len = n; + n = get_var_int(&got, &out); + EXPECT(n > 0); + + EXPECT(got == in); + } +} + +static void test_common_prefix(void) +{ + struct { + const char *a, *b; + int want; + } cases[] = { + { "abc", "ab", 2 }, + { "", "abc", 0 }, + { "abc", "abd", 2 }, + { "abc", "pqr", 0 }, + }; + + int i = 0; + for (i = 0; i < ARRAY_SIZE(cases); i++) { + struct strbuf a = STRBUF_INIT; + struct strbuf b = STRBUF_INIT; + strbuf_addstr(&a, cases[i].a); + strbuf_addstr(&b, cases[i].b); + EXPECT(common_prefix_size(&a, &b) == cases[i].want); + + strbuf_release(&a); + strbuf_release(&b); + } +} + +static void set_hash(uint8_t *h, int j) +{ + int i = 0; + for (i = 0; i < hash_size(GIT_SHA1_FORMAT_ID); i++) { + h[i] = (j >> i) & 0xff; + } +} + +static void test_reftable_ref_record_roundtrip(void) +{ + struct strbuf scratch = STRBUF_INIT; + int i = 0; + + for (i = REFTABLE_REF_DELETION; i < REFTABLE_NR_REF_VALUETYPES; i++) { + struct reftable_record in = { + .type = BLOCK_TYPE_REF, + }; + struct reftable_record out = { .type = BLOCK_TYPE_REF }; + struct strbuf key = STRBUF_INIT; + uint8_t buffer[1024] = { 0 }; + struct string_view dest = { + .buf = buffer, + .len = sizeof(buffer), + }; + int n, m; + + in.u.ref.value_type = i; + switch (i) { + case REFTABLE_REF_DELETION: + break; + case REFTABLE_REF_VAL1: + set_hash(in.u.ref.value.val1, 1); + break; + case REFTABLE_REF_VAL2: + set_hash(in.u.ref.value.val2.value, 1); + set_hash(in.u.ref.value.val2.target_value, 2); + break; + case REFTABLE_REF_SYMREF: + in.u.ref.value.symref = xstrdup("target"); + break; + } + in.u.ref.refname = xstrdup("refs/heads/master"); + + test_copy(&in); + + EXPECT(reftable_record_val_type(&in) == i); + + reftable_record_key(&in, &key); + n = reftable_record_encode(&in, dest, GIT_SHA1_RAWSZ); + EXPECT(n > 0); + + /* decode into a non-zero reftable_record to test for leaks. */ + m = reftable_record_decode(&out, key, i, dest, GIT_SHA1_RAWSZ, &scratch); + EXPECT(n == m); + + EXPECT(reftable_ref_record_equal(&in.u.ref, &out.u.ref, + GIT_SHA1_RAWSZ)); + reftable_record_release(&in); + + strbuf_release(&key); + reftable_record_release(&out); + } + + strbuf_release(&scratch); +} + +static void test_reftable_log_record_equal(void) +{ + struct reftable_log_record in[2] = { + { + .refname = xstrdup("refs/heads/master"), + .update_index = 42, + }, + { + .refname = xstrdup("refs/heads/master"), + .update_index = 22, + } + }; + + EXPECT(!reftable_log_record_equal(&in[0], &in[1], GIT_SHA1_RAWSZ)); + in[1].update_index = in[0].update_index; + EXPECT(reftable_log_record_equal(&in[0], &in[1], GIT_SHA1_RAWSZ)); + reftable_log_record_release(&in[0]); + reftable_log_record_release(&in[1]); +} + +static void test_reftable_log_record_roundtrip(void) +{ + int i; + struct reftable_log_record in[] = { + { + .refname = xstrdup("refs/heads/master"), + .update_index = 42, + .value_type = REFTABLE_LOG_UPDATE, + .value = { + .update = { + .name = xstrdup("han-wen"), + .email = xstrdup("hanwen@google.com"), + .message = xstrdup("test"), + .time = 1577123507, + .tz_offset = 100, + }, + } + }, + { + .refname = xstrdup("refs/heads/master"), + .update_index = 22, + .value_type = REFTABLE_LOG_DELETION, + }, + { + .refname = xstrdup("branch"), + .update_index = 33, + .value_type = REFTABLE_LOG_UPDATE, + } + }; + struct strbuf scratch = STRBUF_INIT; + + set_test_hash(in[0].value.update.new_hash, 1); + set_test_hash(in[0].value.update.old_hash, 2); + set_test_hash(in[2].value.update.new_hash, 3); + set_test_hash(in[2].value.update.old_hash, 4); + for (i = 0; i < ARRAY_SIZE(in); i++) { + struct reftable_record rec = { .type = BLOCK_TYPE_LOG }; + struct strbuf key = STRBUF_INIT; + uint8_t buffer[1024] = { 0 }; + struct string_view dest = { + .buf = buffer, + .len = sizeof(buffer), + }; + /* populate out, to check for leaks. */ + struct reftable_record out = { + .type = BLOCK_TYPE_LOG, + .u.log = { + .refname = xstrdup("old name"), + .value_type = REFTABLE_LOG_UPDATE, + .value = { + .update = { + .name = xstrdup("old name"), + .email = xstrdup("old@email"), + .message = xstrdup("old message"), + }, + }, + }, + }; + int n, m, valtype; + + rec.u.log = in[i]; + + test_copy(&rec); + + reftable_record_key(&rec, &key); + + n = reftable_record_encode(&rec, dest, GIT_SHA1_RAWSZ); + EXPECT(n >= 0); + valtype = reftable_record_val_type(&rec); + m = reftable_record_decode(&out, key, valtype, dest, + GIT_SHA1_RAWSZ, &scratch); + EXPECT(n == m); + + EXPECT(reftable_log_record_equal(&in[i], &out.u.log, + GIT_SHA1_RAWSZ)); + reftable_log_record_release(&in[i]); + strbuf_release(&key); + reftable_record_release(&out); + } + + strbuf_release(&scratch); +} + +static void test_u24_roundtrip(void) +{ + uint32_t in = 0x112233; + uint8_t dest[3]; + uint32_t out; + put_be24(dest, in); + out = get_be24(dest); + EXPECT(in == out); +} + +static void test_key_roundtrip(void) +{ + uint8_t buffer[1024] = { 0 }; + struct string_view dest = { + .buf = buffer, + .len = sizeof(buffer), + }; + struct strbuf last_key = STRBUF_INIT; + struct strbuf key = STRBUF_INIT; + struct strbuf roundtrip = STRBUF_INIT; + int restart; + uint8_t extra; + int n, m; + uint8_t rt_extra; + + strbuf_addstr(&last_key, "refs/heads/master"); + strbuf_addstr(&key, "refs/tags/bla"); + extra = 6; + n = reftable_encode_key(&restart, dest, last_key, key, extra); + EXPECT(!restart); + EXPECT(n > 0); + + strbuf_addstr(&roundtrip, "refs/heads/master"); + m = reftable_decode_key(&roundtrip, &rt_extra, dest); + EXPECT(n == m); + EXPECT(0 == strbuf_cmp(&key, &roundtrip)); + EXPECT(rt_extra == extra); + + strbuf_release(&last_key); + strbuf_release(&key); + strbuf_release(&roundtrip); +} + +static void test_reftable_obj_record_roundtrip(void) +{ + uint8_t testHash1[GIT_SHA1_RAWSZ] = { 1, 2, 3, 4, 0 }; + uint64_t till9[] = { 1, 2, 3, 4, 500, 600, 700, 800, 9000 }; + struct reftable_obj_record recs[3] = { + { + .hash_prefix = testHash1, + .hash_prefix_len = 5, + .offsets = till9, + .offset_len = 3, + }, + { + .hash_prefix = testHash1, + .hash_prefix_len = 5, + .offsets = till9, + .offset_len = 9, + }, + { + .hash_prefix = testHash1, + .hash_prefix_len = 5, + }, + }; + struct strbuf scratch = STRBUF_INIT; + int i = 0; + + for (i = 0; i < ARRAY_SIZE(recs); i++) { + uint8_t buffer[1024] = { 0 }; + struct string_view dest = { + .buf = buffer, + .len = sizeof(buffer), + }; + struct reftable_record in = { + .type = BLOCK_TYPE_OBJ, + .u = { + .obj = recs[i], + }, + }; + struct strbuf key = STRBUF_INIT; + struct reftable_record out = { .type = BLOCK_TYPE_OBJ }; + int n, m; + uint8_t extra; + + test_copy(&in); + reftable_record_key(&in, &key); + n = reftable_record_encode(&in, dest, GIT_SHA1_RAWSZ); + EXPECT(n > 0); + extra = reftable_record_val_type(&in); + m = reftable_record_decode(&out, key, extra, dest, + GIT_SHA1_RAWSZ, &scratch); + EXPECT(n == m); + + EXPECT(reftable_record_equal(&in, &out, GIT_SHA1_RAWSZ)); + strbuf_release(&key); + reftable_record_release(&out); + } + + strbuf_release(&scratch); +} + +static void test_reftable_index_record_roundtrip(void) +{ + struct reftable_record in = { + .type = BLOCK_TYPE_INDEX, + .u.idx = { + .offset = 42, + .last_key = STRBUF_INIT, + }, + }; + uint8_t buffer[1024] = { 0 }; + struct string_view dest = { + .buf = buffer, + .len = sizeof(buffer), + }; + struct strbuf scratch = STRBUF_INIT; + struct strbuf key = STRBUF_INIT; + struct reftable_record out = { + .type = BLOCK_TYPE_INDEX, + .u.idx = { .last_key = STRBUF_INIT }, + }; + int n, m; + uint8_t extra; + + strbuf_addstr(&in.u.idx.last_key, "refs/heads/master"); + reftable_record_key(&in, &key); + test_copy(&in); + + EXPECT(0 == strbuf_cmp(&key, &in.u.idx.last_key)); + n = reftable_record_encode(&in, dest, GIT_SHA1_RAWSZ); + EXPECT(n > 0); + + extra = reftable_record_val_type(&in); + m = reftable_record_decode(&out, key, extra, dest, GIT_SHA1_RAWSZ, + &scratch); + EXPECT(m == n); + + EXPECT(reftable_record_equal(&in, &out, GIT_SHA1_RAWSZ)); + + reftable_record_release(&out); + strbuf_release(&key); + strbuf_release(&scratch); + strbuf_release(&in.u.idx.last_key); +} + +int record_test_main(int argc, const char *argv[]) +{ + RUN_TEST(test_reftable_log_record_equal); + RUN_TEST(test_reftable_log_record_roundtrip); + RUN_TEST(test_reftable_ref_record_roundtrip); + RUN_TEST(test_varint_roundtrip); + RUN_TEST(test_key_roundtrip); + RUN_TEST(test_common_prefix); + RUN_TEST(test_reftable_obj_record_roundtrip); + RUN_TEST(test_reftable_index_record_roundtrip); + RUN_TEST(test_u24_roundtrip); + return 0; +} diff --git a/reftable/refname.c b/reftable/refname.c new file mode 100644 index 0000000..bbfde15 --- /dev/null +++ b/reftable/refname.c @@ -0,0 +1,206 @@ +/* + Copyright 2020 Google LLC + + Use of this source code is governed by a BSD-style + license that can be found in the LICENSE file or at + https://developers.google.com/open-source/licenses/bsd +*/ + +#include "system.h" +#include "reftable-error.h" +#include "basics.h" +#include "refname.h" +#include "reftable-iterator.h" + +struct refname_needle_lesseq_args { + char **haystack; + const char *needle; +}; + +static int refname_needle_lesseq(size_t k, void *_args) +{ + struct refname_needle_lesseq_args *args = _args; + return strcmp(args->needle, args->haystack[k]) <= 0; +} + +static int modification_has_ref(struct modification *mod, const char *name) +{ + struct reftable_ref_record ref = { NULL }; + int err = 0; + + if (mod->add_len > 0) { + struct refname_needle_lesseq_args args = { + .haystack = mod->add, + .needle = name, + }; + size_t idx = binsearch(mod->add_len, refname_needle_lesseq, &args); + if (idx < mod->add_len && !strcmp(mod->add[idx], name)) + return 0; + } + + if (mod->del_len > 0) { + struct refname_needle_lesseq_args args = { + .haystack = mod->del, + .needle = name, + }; + size_t idx = binsearch(mod->del_len, refname_needle_lesseq, &args); + if (idx < mod->del_len && !strcmp(mod->del[idx], name)) + return 1; + } + + err = reftable_table_read_ref(&mod->tab, name, &ref); + reftable_ref_record_release(&ref); + return err; +} + +static void modification_release(struct modification *mod) +{ + /* don't delete the strings themselves; they're owned by ref records. + */ + FREE_AND_NULL(mod->add); + FREE_AND_NULL(mod->del); + mod->add_len = 0; + mod->del_len = 0; +} + +static int modification_has_ref_with_prefix(struct modification *mod, + const char *prefix) +{ + struct reftable_iterator it = { NULL }; + struct reftable_ref_record ref = { NULL }; + int err = 0; + + if (mod->add_len > 0) { + struct refname_needle_lesseq_args args = { + .haystack = mod->add, + .needle = prefix, + }; + size_t idx = binsearch(mod->add_len, refname_needle_lesseq, &args); + if (idx < mod->add_len && + !strncmp(prefix, mod->add[idx], strlen(prefix))) + goto done; + } + err = reftable_table_seek_ref(&mod->tab, &it, prefix); + if (err) + goto done; + + while (1) { + err = reftable_iterator_next_ref(&it, &ref); + if (err) + goto done; + + if (mod->del_len > 0) { + struct refname_needle_lesseq_args args = { + .haystack = mod->del, + .needle = ref.refname, + }; + size_t idx = binsearch(mod->del_len, refname_needle_lesseq, &args); + if (idx < mod->del_len && + !strcmp(ref.refname, mod->del[idx])) + continue; + } + + if (strncmp(ref.refname, prefix, strlen(prefix))) { + err = 1; + goto done; + } + err = 0; + goto done; + } + +done: + reftable_ref_record_release(&ref); + reftable_iterator_destroy(&it); + return err; +} + +static int validate_refname(const char *name) +{ + while (1) { + char *next = strchr(name, '/'); + if (!*name) { + return REFTABLE_REFNAME_ERROR; + } + if (!next) { + return 0; + } + if (next - name == 0 || (next - name == 1 && *name == '.') || + (next - name == 2 && name[0] == '.' && name[1] == '.')) + return REFTABLE_REFNAME_ERROR; + name = next + 1; + } + return 0; +} + +int validate_ref_record_addition(struct reftable_table tab, + struct reftable_ref_record *recs, size_t sz) +{ + struct modification mod = { + .tab = tab, + .add = reftable_calloc(sz, sizeof(*mod.add)), + .del = reftable_calloc(sz, sizeof(*mod.del)), + }; + int i = 0; + int err = 0; + for (; i < sz; i++) { + if (reftable_ref_record_is_deletion(&recs[i])) { + mod.del[mod.del_len++] = recs[i].refname; + } else { + mod.add[mod.add_len++] = recs[i].refname; + } + } + + err = modification_validate(&mod); + modification_release(&mod); + return err; +} + +static void strbuf_trim_component(struct strbuf *sl) +{ + while (sl->len > 0) { + int is_slash = (sl->buf[sl->len - 1] == '/'); + strbuf_setlen(sl, sl->len - 1); + if (is_slash) + break; + } +} + +int modification_validate(struct modification *mod) +{ + struct strbuf slashed = STRBUF_INIT; + int err = 0; + int i = 0; + for (; i < mod->add_len; i++) { + err = validate_refname(mod->add[i]); + if (err) + goto done; + strbuf_reset(&slashed); + strbuf_addstr(&slashed, mod->add[i]); + strbuf_addstr(&slashed, "/"); + + err = modification_has_ref_with_prefix(mod, slashed.buf); + if (err == 0) { + err = REFTABLE_NAME_CONFLICT; + goto done; + } + if (err < 0) + goto done; + + strbuf_reset(&slashed); + strbuf_addstr(&slashed, mod->add[i]); + while (slashed.len) { + strbuf_trim_component(&slashed); + err = modification_has_ref(mod, slashed.buf); + if (err == 0) { + err = REFTABLE_NAME_CONFLICT; + goto done; + } + if (err < 0) + goto done; + } + } + err = 0; +done: + strbuf_release(&slashed); + return err; +} diff --git a/reftable/refname.h b/reftable/refname.h new file mode 100644 index 0000000..a24b40f --- /dev/null +++ b/reftable/refname.h @@ -0,0 +1,29 @@ +/* + Copyright 2020 Google LLC + + Use of this source code is governed by a BSD-style + license that can be found in the LICENSE file or at + https://developers.google.com/open-source/licenses/bsd +*/ +#ifndef REFNAME_H +#define REFNAME_H + +#include "reftable-record.h" +#include "reftable-generic.h" + +struct modification { + struct reftable_table tab; + + char **add; + size_t add_len; + + char **del; + size_t del_len; +}; + +int validate_ref_record_addition(struct reftable_table tab, + struct reftable_ref_record *recs, size_t sz); + +int modification_validate(struct modification *mod); + +#endif diff --git a/reftable/refname_test.c b/reftable/refname_test.c new file mode 100644 index 0000000..b9cc625 --- /dev/null +++ b/reftable/refname_test.c @@ -0,0 +1,101 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "basics.h" +#include "block.h" +#include "blocksource.h" +#include "reader.h" +#include "record.h" +#include "refname.h" +#include "reftable-error.h" +#include "reftable-writer.h" +#include "system.h" + +#include "test_framework.h" +#include "reftable-tests.h" + +struct testcase { + char *add; + char *del; + int error_code; +}; + +static void test_conflict(void) +{ + struct reftable_write_options opts = { 0 }; + struct strbuf buf = STRBUF_INIT; + struct reftable_writer *w = + reftable_new_writer(&strbuf_add_void, &noop_flush, &buf, &opts); + struct reftable_ref_record rec = { + .refname = "a/b", + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "destination", /* make sure it's not a symref. + */ + .update_index = 1, + }; + int err; + int i; + struct reftable_block_source source = { NULL }; + struct reftable_reader *rd = NULL; + struct reftable_table tab = { NULL }; + struct testcase cases[] = { + { "a/b/c", NULL, REFTABLE_NAME_CONFLICT }, + { "b", NULL, 0 }, + { "a", NULL, REFTABLE_NAME_CONFLICT }, + { "a", "a/b", 0 }, + + { "p/", NULL, REFTABLE_REFNAME_ERROR }, + { "p//q", NULL, REFTABLE_REFNAME_ERROR }, + { "p/./q", NULL, REFTABLE_REFNAME_ERROR }, + { "p/../q", NULL, REFTABLE_REFNAME_ERROR }, + + { "a/b/c", "a/b", 0 }, + { NULL, "a//b", 0 }, + }; + reftable_writer_set_limits(w, 1, 1); + + err = reftable_writer_add_ref(w, &rec); + EXPECT_ERR(err); + + err = reftable_writer_close(w); + EXPECT_ERR(err); + reftable_writer_free(w); + + block_source_from_strbuf(&source, &buf); + err = reftable_new_reader(&rd, &source, "filename"); + EXPECT_ERR(err); + + reftable_table_from_reader(&tab, rd); + + for (i = 0; i < ARRAY_SIZE(cases); i++) { + struct modification mod = { + .tab = tab, + }; + + if (cases[i].add) { + mod.add = &cases[i].add; + mod.add_len = 1; + } + if (cases[i].del) { + mod.del = &cases[i].del; + mod.del_len = 1; + } + + err = modification_validate(&mod); + EXPECT(err == cases[i].error_code); + } + + reftable_reader_free(rd); + strbuf_release(&buf); +} + +int refname_test_main(int argc, const char *argv[]) +{ + RUN_TEST(test_conflict); + return 0; +} diff --git a/reftable/reftable-blocksource.h b/reftable/reftable-blocksource.h new file mode 100644 index 0000000..5aa3990 --- /dev/null +++ b/reftable/reftable-blocksource.h @@ -0,0 +1,49 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef REFTABLE_BLOCKSOURCE_H +#define REFTABLE_BLOCKSOURCE_H + +#include <stdint.h> + +/* block_source is a generic wrapper for a seekable readable file. + */ +struct reftable_block_source { + struct reftable_block_source_vtable *ops; + void *arg; +}; + +/* a contiguous segment of bytes. It keeps track of its generating block_source + * so it can return itself into the pool. */ +struct reftable_block { + uint8_t *data; + int len; + struct reftable_block_source source; +}; + +/* block_source_vtable are the operations that make up block_source */ +struct reftable_block_source_vtable { + /* returns the size of a block source */ + uint64_t (*size)(void *source); + + /* reads a segment from the block source. It is an error to read + beyond the end of the block */ + int (*read_block)(void *source, struct reftable_block *dest, + uint64_t off, uint32_t size); + /* mark the block as read; may return the data back to malloc */ + void (*return_block)(void *source, struct reftable_block *blockp); + + /* release all resources associated with the block source */ + void (*close)(void *source); +}; + +/* opens a file on the file system as a block_source */ +int reftable_block_source_from_file(struct reftable_block_source *block_src, + const char *name); + +#endif diff --git a/reftable/reftable-error.h b/reftable/reftable-error.h new file mode 100644 index 0000000..e9b07c9 --- /dev/null +++ b/reftable/reftable-error.h @@ -0,0 +1,69 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef REFTABLE_ERROR_H +#define REFTABLE_ERROR_H + +/* + * Errors in reftable calls are signaled with negative integer return values. 0 + * means success. + */ +enum reftable_error { + /* Unexpected file system behavior */ + REFTABLE_IO_ERROR = -2, + + /* Format inconsistency on reading data */ + REFTABLE_FORMAT_ERROR = -3, + + /* File does not exist. Returned from block_source_from_file(), because + * it needs special handling in stack. + */ + REFTABLE_NOT_EXIST_ERROR = -4, + + /* Trying to access locked data. */ + REFTABLE_LOCK_ERROR = -5, + + /* Misuse of the API: + * - on writing a record with NULL refname. + * - on writing a reftable_ref_record outside the table limits + * - on writing a ref or log record before the stack's + * next_update_inde*x + * - on writing a log record with multiline message with + * exact_log_message unset + * - on reading a reftable_ref_record from log iterator, or vice versa. + * + * When a call misuses the API, the internal state of the library is + * kept unchanged. + */ + REFTABLE_API_ERROR = -6, + + /* Decompression error */ + REFTABLE_ZLIB_ERROR = -7, + + /* Wrote a table without blocks. */ + REFTABLE_EMPTY_TABLE_ERROR = -8, + + /* Dir/file conflict. */ + REFTABLE_NAME_CONFLICT = -9, + + /* Invalid ref name. */ + REFTABLE_REFNAME_ERROR = -10, + + /* Entry does not fit. This can happen when writing outsize reflog + messages. */ + REFTABLE_ENTRY_TOO_BIG_ERROR = -11, + + /* Trying to write out-of-date data. */ + REFTABLE_OUTDATED_ERROR = -12, +}; + +/* convert the numeric error code to a string. The string should not be + * deallocated. */ +const char *reftable_error_str(int err); + +#endif diff --git a/reftable/reftable-generic.h b/reftable/reftable-generic.h new file mode 100644 index 0000000..d239751 --- /dev/null +++ b/reftable/reftable-generic.h @@ -0,0 +1,47 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef REFTABLE_GENERIC_H +#define REFTABLE_GENERIC_H + +#include "reftable-iterator.h" + +struct reftable_table_vtable; + +/* + * Provides a unified API for reading tables, either merged tables, or single + * readers. */ +struct reftable_table { + struct reftable_table_vtable *ops; + void *table_arg; +}; + +int reftable_table_seek_log(struct reftable_table *tab, + struct reftable_iterator *it, const char *name); + +int reftable_table_seek_ref(struct reftable_table *tab, + struct reftable_iterator *it, const char *name); + +/* returns the hash ID from a generic reftable_table */ +uint32_t reftable_table_hash_id(struct reftable_table *tab); + +/* returns the max update_index covered by this table. */ +uint64_t reftable_table_max_update_index(struct reftable_table *tab); + +/* returns the min update_index covered by this table. */ +uint64_t reftable_table_min_update_index(struct reftable_table *tab); + +/* convenience function to read a single ref. Returns < 0 for error, 0 + for success, and 1 if ref not found. */ +int reftable_table_read_ref(struct reftable_table *tab, const char *name, + struct reftable_ref_record *ref); + +/* dump table contents onto stdout for debugging */ +int reftable_table_print(struct reftable_table *tab); + +#endif diff --git a/reftable/reftable-iterator.h b/reftable/reftable-iterator.h new file mode 100644 index 0000000..d3eee7a --- /dev/null +++ b/reftable/reftable-iterator.h @@ -0,0 +1,39 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef REFTABLE_ITERATOR_H +#define REFTABLE_ITERATOR_H + +#include "reftable-record.h" + +struct reftable_iterator_vtable; + +/* iterator is the generic interface for walking over data stored in a + * reftable. + */ +struct reftable_iterator { + struct reftable_iterator_vtable *ops; + void *iter_arg; +}; + +/* reads the next reftable_ref_record. Returns < 0 for error, 0 for OK and > 0: + * end of iteration. + */ +int reftable_iterator_next_ref(struct reftable_iterator *it, + struct reftable_ref_record *ref); + +/* reads the next reftable_log_record. Returns < 0 for error, 0 for OK and > 0: + * end of iteration. + */ +int reftable_iterator_next_log(struct reftable_iterator *it, + struct reftable_log_record *log); + +/* releases resources associated with an iterator. */ +void reftable_iterator_destroy(struct reftable_iterator *it); + +#endif diff --git a/reftable/reftable-malloc.h b/reftable/reftable-malloc.h new file mode 100644 index 0000000..5f2185f --- /dev/null +++ b/reftable/reftable-malloc.h @@ -0,0 +1,18 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef REFTABLE_H +#define REFTABLE_H + +#include <stddef.h> + +/* Overrides the functions to use for memory management. */ +void reftable_set_alloc(void *(*malloc)(size_t), + void *(*realloc)(void *, size_t), void (*free)(void *)); + +#endif diff --git a/reftable/reftable-merged.h b/reftable/reftable-merged.h new file mode 100644 index 0000000..c91a2d8 --- /dev/null +++ b/reftable/reftable-merged.h @@ -0,0 +1,72 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef REFTABLE_MERGED_H +#define REFTABLE_MERGED_H + +#include "reftable-iterator.h" + +/* + * Merged tables + * + * A ref database kept in a sequence of table files. The merged_table presents a + * unified view to reading (seeking, iterating) a sequence of immutable tables. + * + * The merged tables are on purpose kept disconnected from their actual storage + * (eg. files on disk), because it is useful to merge tables aren't files. For + * example, the per-workspace and global ref namespace can be implemented as a + * merged table of two stacks of file-backed reftables. + */ + +/* A merged table is implements seeking/iterating over a stack of tables. */ +struct reftable_merged_table; + +/* A generic reftable; see below. */ +struct reftable_table; + +/* reftable_new_merged_table creates a new merged table. It takes ownership of + the stack array. +*/ +int reftable_new_merged_table(struct reftable_merged_table **dest, + struct reftable_table *stack, size_t n, + uint32_t hash_id); + +/* returns an iterator positioned just before 'name' */ +int reftable_merged_table_seek_ref(struct reftable_merged_table *mt, + struct reftable_iterator *it, + const char *name); + +/* returns an iterator for log entry, at given update_index */ +int reftable_merged_table_seek_log_at(struct reftable_merged_table *mt, + struct reftable_iterator *it, + const char *name, uint64_t update_index); + +/* like reftable_merged_table_seek_log_at but look for the newest entry. */ +int reftable_merged_table_seek_log(struct reftable_merged_table *mt, + struct reftable_iterator *it, + const char *name); + +/* returns the max update_index covered by this merged table. */ +uint64_t +reftable_merged_table_max_update_index(struct reftable_merged_table *mt); + +/* returns the min update_index covered by this merged table. */ +uint64_t +reftable_merged_table_min_update_index(struct reftable_merged_table *mt); + +/* releases memory for the merged_table */ +void reftable_merged_table_free(struct reftable_merged_table *m); + +/* return the hash ID of the merged table. */ +uint32_t reftable_merged_table_hash_id(struct reftable_merged_table *m); + +/* create a generic table from reftable_merged_table */ +void reftable_table_from_merged_table(struct reftable_table *tab, + struct reftable_merged_table *table); + +#endif diff --git a/reftable/reftable-reader.h b/reftable/reftable-reader.h new file mode 100644 index 0000000..4a4bc2f --- /dev/null +++ b/reftable/reftable-reader.h @@ -0,0 +1,101 @@ +/* + Copyright 2020 Google LLC + + Use of this source code is governed by a BSD-style + license that can be found in the LICENSE file or at + https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef REFTABLE_READER_H +#define REFTABLE_READER_H + +#include "reftable-iterator.h" +#include "reftable-blocksource.h" + +/* + * Reading single tables + * + * The follow routines are for reading single files. For an + * application-level interface, skip ahead to struct + * reftable_merged_table and struct reftable_stack. + */ + +/* The reader struct is a handle to an open reftable file. */ +struct reftable_reader; + +/* Generic table. */ +struct reftable_table; + +/* reftable_new_reader opens a reftable for reading. If successful, + * returns 0 code and sets pp. The name is used for creating a + * stack. Typically, it is the basename of the file. The block source + * `src` is owned by the reader, and is closed on calling + * reftable_reader_destroy(). On error, the block source `src` is + * closed as well. + */ +int reftable_new_reader(struct reftable_reader **pp, + struct reftable_block_source *src, const char *name); + +/* reftable_reader_seek_ref returns an iterator where 'name' would be inserted + in the table. To seek to the start of the table, use name = "". + + example: + + struct reftable_reader *r = NULL; + int err = reftable_new_reader(&r, &src, "filename"); + if (err < 0) { ... } + struct reftable_iterator it = {0}; + err = reftable_reader_seek_ref(r, &it, "refs/heads/master"); + if (err < 0) { ... } + struct reftable_ref_record ref = {0}; + while (1) { + err = reftable_iterator_next_ref(&it, &ref); + if (err > 0) { + break; + } + if (err < 0) { + ..error handling.. + } + ..found.. + } + reftable_iterator_destroy(&it); + reftable_ref_record_release(&ref); +*/ +int reftable_reader_seek_ref(struct reftable_reader *r, + struct reftable_iterator *it, const char *name); + +/* returns the hash ID used in this table. */ +uint32_t reftable_reader_hash_id(struct reftable_reader *r); + +/* seek to logs for the given name, older than update_index. To seek to the + start of the table, use name = "". +*/ +int reftable_reader_seek_log_at(struct reftable_reader *r, + struct reftable_iterator *it, const char *name, + uint64_t update_index); + +/* seek to newest log entry for given name. */ +int reftable_reader_seek_log(struct reftable_reader *r, + struct reftable_iterator *it, const char *name); + +/* closes and deallocates a reader. */ +void reftable_reader_free(struct reftable_reader *); + +/* return an iterator for the refs pointing to `oid`. */ +int reftable_reader_refs_for(struct reftable_reader *r, + struct reftable_iterator *it, uint8_t *oid); + +/* return the max_update_index for a table */ +uint64_t reftable_reader_max_update_index(struct reftable_reader *r); + +/* return the min_update_index for a table */ +uint64_t reftable_reader_min_update_index(struct reftable_reader *r); + +/* creates a generic table from a file reader. */ +void reftable_table_from_reader(struct reftable_table *tab, + struct reftable_reader *reader); + +/* print table onto stdout for debugging. */ +int reftable_reader_print_file(const char *tablename); + +#endif diff --git a/reftable/reftable-record.h b/reftable/reftable-record.h new file mode 100644 index 0000000..2a2943c --- /dev/null +++ b/reftable/reftable-record.h @@ -0,0 +1,118 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef REFTABLE_RECORD_H +#define REFTABLE_RECORD_H + +#include "hash-ll.h" +#include <stdint.h> + +/* + * Basic data types + * + * Reftables store the state of each ref in struct reftable_ref_record, and they + * store a sequence of reflog updates in struct reftable_log_record. + */ + +/* reftable_ref_record holds a ref database entry target_value */ +struct reftable_ref_record { + char *refname; /* Name of the ref, malloced. */ + size_t refname_cap; + uint64_t update_index; /* Logical timestamp at which this value is + * written */ + + enum { + /* tombstone to hide deletions from earlier tables */ + REFTABLE_REF_DELETION = 0x0, + + /* a simple ref */ + REFTABLE_REF_VAL1 = 0x1, + /* a tag, plus its peeled hash */ + REFTABLE_REF_VAL2 = 0x2, + + /* a symbolic reference */ + REFTABLE_REF_SYMREF = 0x3, +#define REFTABLE_NR_REF_VALUETYPES 4 + } value_type; + union { + unsigned char val1[GIT_MAX_RAWSZ]; + struct { + unsigned char value[GIT_MAX_RAWSZ]; /* first hash */ + unsigned char target_value[GIT_MAX_RAWSZ]; /* second hash */ + } val2; + char *symref; /* referent, malloced 0-terminated string */ + } value; +}; + +/* Returns the first hash, or NULL if `rec` is not of type + * REFTABLE_REF_VAL1 or REFTABLE_REF_VAL2. */ +const unsigned char *reftable_ref_record_val1(const struct reftable_ref_record *rec); + +/* Returns the second hash, or NULL if `rec` is not of type + * REFTABLE_REF_VAL2. */ +const unsigned char *reftable_ref_record_val2(const struct reftable_ref_record *rec); + +/* returns whether 'ref' represents a deletion */ +int reftable_ref_record_is_deletion(const struct reftable_ref_record *ref); + +/* prints a reftable_ref_record onto stdout. Useful for debugging. */ +void reftable_ref_record_print(const struct reftable_ref_record *ref, + uint32_t hash_id); + +/* frees and nulls all pointer values inside `ref`. */ +void reftable_ref_record_release(struct reftable_ref_record *ref); + +/* returns whether two reftable_ref_records are the same. Useful for testing. */ +int reftable_ref_record_equal(const struct reftable_ref_record *a, + const struct reftable_ref_record *b, int hash_size); + +/* reftable_log_record holds a reflog entry */ +struct reftable_log_record { + char *refname; + size_t refname_cap; + uint64_t update_index; /* logical timestamp of a transactional update. + */ + + enum { + /* tombstone to hide deletions from earlier tables */ + REFTABLE_LOG_DELETION = 0x0, + + /* a simple update */ + REFTABLE_LOG_UPDATE = 0x1, +#define REFTABLE_NR_LOG_VALUETYPES 2 + } value_type; + + union { + struct { + unsigned char new_hash[GIT_MAX_RAWSZ]; + unsigned char old_hash[GIT_MAX_RAWSZ]; + char *name; + char *email; + uint64_t time; + int16_t tz_offset; + char *message; + size_t message_cap; + } update; + } value; +}; + +/* returns whether 'ref' represents the deletion of a log record. */ +int reftable_log_record_is_deletion(const struct reftable_log_record *log); + +/* frees and nulls all pointer values. */ +void reftable_log_record_release(struct reftable_log_record *log); + +/* returns whether two records are equal. Useful for testing. */ +int reftable_log_record_equal(const struct reftable_log_record *a, + const struct reftable_log_record *b, int hash_size); + +/* dumps a reftable_log_record on stdout, for debugging/testing. */ +void reftable_log_record_print(struct reftable_log_record *log, + uint32_t hash_id); + +#endif diff --git a/reftable/reftable-stack.h b/reftable/reftable-stack.h new file mode 100644 index 0000000..1b602dd --- /dev/null +++ b/reftable/reftable-stack.h @@ -0,0 +1,128 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef REFTABLE_STACK_H +#define REFTABLE_STACK_H + +#include "reftable-writer.h" + +/* + * The stack presents an interface to a mutable sequence of reftables. + + * A stack can be mutated by pushing a table to the top of the stack. + + * The reftable_stack automatically compacts files on disk to ensure good + * amortized performance. + * + * For windows and other platforms that cannot have open files as rename + * destinations, concurrent access from multiple processes needs the rand() + * random seed to be randomized. + */ +struct reftable_stack; + +/* open a new reftable stack. The tables along with the table list will be + * stored in 'dir'. Typically, this should be .git/reftables. + */ +int reftable_new_stack(struct reftable_stack **dest, const char *dir, + struct reftable_write_options config); + +/* returns the update_index at which a next table should be written. */ +uint64_t reftable_stack_next_update_index(struct reftable_stack *st); + +/* holds a transaction to add tables at the top of a stack. */ +struct reftable_addition; + +/* + * returns a new transaction to add reftables to the given stack. As a side + * effect, the ref database is locked. + */ +int reftable_stack_new_addition(struct reftable_addition **dest, + struct reftable_stack *st); + +/* Adds a reftable to transaction. */ +int reftable_addition_add(struct reftable_addition *add, + int (*write_table)(struct reftable_writer *wr, + void *arg), + void *arg); + +/* Commits the transaction, releasing the lock. After calling this, + * reftable_addition_destroy should still be called. + */ +int reftable_addition_commit(struct reftable_addition *add); + +/* Release all non-committed data from the transaction, and deallocate the + * transaction. Releases the lock if held. */ +void reftable_addition_destroy(struct reftable_addition *add); + +/* add a new table to the stack. The write_table function must call + * reftable_writer_set_limits, add refs and return an error value. */ +int reftable_stack_add(struct reftable_stack *st, + int (*write_table)(struct reftable_writer *wr, + void *write_arg), + void *write_arg); + +/* returns the merged_table for seeking. This table is valid until the + * next write or reload, and should not be closed or deleted. + */ +struct reftable_merged_table * +reftable_stack_merged_table(struct reftable_stack *st); + +/* frees all resources associated with the stack. */ +void reftable_stack_destroy(struct reftable_stack *st); + +/* Reloads the stack if necessary. This is very cheap to run if the stack was up + * to date */ +int reftable_stack_reload(struct reftable_stack *st); + +/* Policy for expiring reflog entries. */ +struct reftable_log_expiry_config { + /* Drop entries older than this timestamp */ + uint64_t time; + + /* Drop older entries */ + uint64_t min_update_index; +}; + +/* compacts all reftables into a giant table. Expire reflog entries if config is + * non-NULL */ +int reftable_stack_compact_all(struct reftable_stack *st, + struct reftable_log_expiry_config *config); + +/* heuristically compact unbalanced table stack. */ +int reftable_stack_auto_compact(struct reftable_stack *st); + +/* delete stale .ref tables. */ +int reftable_stack_clean(struct reftable_stack *st); + +/* convenience function to read a single ref. Returns < 0 for error, 0 for + * success, and 1 if ref not found. */ +int reftable_stack_read_ref(struct reftable_stack *st, const char *refname, + struct reftable_ref_record *ref); + +/* convenience function to read a single log. Returns < 0 for error, 0 for + * success, and 1 if ref not found. */ +int reftable_stack_read_log(struct reftable_stack *st, const char *refname, + struct reftable_log_record *log); + +/* statistics on past compactions. */ +struct reftable_compaction_stats { + uint64_t bytes; /* total number of bytes written */ + uint64_t entries_written; /* total number of entries written, including + failures. */ + int attempts; /* how often we tried to compact */ + int failures; /* failures happen on concurrent updates */ +}; + +/* return statistics for compaction up till now. */ +struct reftable_compaction_stats * +reftable_stack_compaction_stats(struct reftable_stack *st); + +/* print the entire stack represented by the directory */ +int reftable_stack_print_directory(const char *stackdir, uint32_t hash_id); + +#endif diff --git a/reftable/reftable-tests.h b/reftable/reftable-tests.h new file mode 100644 index 0000000..0019cbc --- /dev/null +++ b/reftable/reftable-tests.h @@ -0,0 +1,23 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef REFTABLE_TESTS_H +#define REFTABLE_TESTS_H + +int basics_test_main(int argc, const char **argv); +int block_test_main(int argc, const char **argv); +int merged_test_main(int argc, const char **argv); +int pq_test_main(int argc, const char **argv); +int record_test_main(int argc, const char **argv); +int refname_test_main(int argc, const char **argv); +int readwrite_test_main(int argc, const char **argv); +int stack_test_main(int argc, const char **argv); +int tree_test_main(int argc, const char **argv); +int reftable_dump_main(int argc, char *const *argv); + +#endif diff --git a/reftable/reftable-writer.h b/reftable/reftable-writer.h new file mode 100644 index 0000000..155bf0b --- /dev/null +++ b/reftable/reftable-writer.h @@ -0,0 +1,155 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef REFTABLE_WRITER_H +#define REFTABLE_WRITER_H + +#include "reftable-record.h" + +#include <stdint.h> +#include <unistd.h> /* ssize_t */ + +/* Writing single reftables */ + +/* reftable_write_options sets options for writing a single reftable. */ +struct reftable_write_options { + /* boolean: do not pad out blocks to block size. */ + unsigned unpadded : 1; + + /* the blocksize. Should be less than 2^24. */ + uint32_t block_size; + + /* boolean: do not generate a SHA1 => ref index. */ + unsigned skip_index_objects : 1; + + /* how often to write complete keys in each block. */ + int restart_interval; + + /* 4-byte identifier ("sha1", "s256") of the hash. + * Defaults to SHA1 if unset + */ + uint32_t hash_id; + + /* Default mode for creating files. If unset, use 0666 (+umask) */ + unsigned int default_permissions; + + /* boolean: do not check ref names for validity or dir/file conflicts. + */ + unsigned skip_name_check : 1; + + /* boolean: copy log messages exactly. If unset, check that the message + * is a single line, and add '\n' if missing. + */ + unsigned exact_log_message : 1; + + /* boolean: Prevent auto-compaction of tables. */ + unsigned disable_auto_compact : 1; +}; + +/* reftable_block_stats holds statistics for a single block type */ +struct reftable_block_stats { + /* total number of entries written */ + int entries; + /* total number of key restarts */ + int restarts; + /* total number of blocks */ + int blocks; + /* total number of index blocks */ + int index_blocks; + /* depth of the index */ + int max_index_level; + + /* offset of the first block for this type */ + uint64_t offset; + /* offset of the top level index block for this type, or 0 if not + * present */ + uint64_t index_offset; +}; + +/* stats holds overall statistics for a single reftable */ +struct reftable_stats { + /* total number of blocks written. */ + int blocks; + /* stats for ref data */ + struct reftable_block_stats ref_stats; + /* stats for the SHA1 to ref map. */ + struct reftable_block_stats obj_stats; + /* stats for index blocks */ + struct reftable_block_stats idx_stats; + /* stats for log blocks */ + struct reftable_block_stats log_stats; + + /* disambiguation length of shortened object IDs. */ + int object_id_len; +}; + +/* reftable_new_writer creates a new writer */ +struct reftable_writer * +reftable_new_writer(ssize_t (*writer_func)(void *, const void *, size_t), + int (*flush_func)(void *), + void *writer_arg, struct reftable_write_options *opts); + +/* Set the range of update indices for the records we will add. When writing a + table into a stack, the min should be at least + reftable_stack_next_update_index(), or REFTABLE_API_ERROR is returned. + + For transactional updates to a stack, typically min==max, and the + update_index can be obtained by inspeciting the stack. When converting an + existing ref database into a single reftable, this would be a range of + update-index timestamps. + */ +void reftable_writer_set_limits(struct reftable_writer *w, uint64_t min, + uint64_t max); + +/* + Add a reftable_ref_record. The record should have names that come after + already added records. + + The update_index must be within the limits set by + reftable_writer_set_limits(), or REFTABLE_API_ERROR is returned. It is an + REFTABLE_API_ERROR error to write a ref record after a log record. +*/ +int reftable_writer_add_ref(struct reftable_writer *w, + struct reftable_ref_record *ref); + +/* + Convenience function to add multiple reftable_ref_records; the function sorts + the records before adding them, reordering the records array passed in. +*/ +int reftable_writer_add_refs(struct reftable_writer *w, + struct reftable_ref_record *refs, int n); + +/* + adds reftable_log_records. Log records are keyed by (refname, decreasing + update_index). The key for the record added must come after the already added + log records. +*/ +int reftable_writer_add_log(struct reftable_writer *w, + struct reftable_log_record *log); + +/* + Convenience function to add multiple reftable_log_records; the function sorts + the records before adding them, reordering records array passed in. +*/ +int reftable_writer_add_logs(struct reftable_writer *w, + struct reftable_log_record *logs, int n); + +/* reftable_writer_close finalizes the reftable. The writer is retained so + * statistics can be inspected. */ +int reftable_writer_close(struct reftable_writer *w); + +/* writer_stats returns the statistics on the reftable being written. + + This struct becomes invalid when the writer is freed. + */ +const struct reftable_stats *reftable_writer_stats(struct reftable_writer *w); + +/* reftable_writer_free deallocates memory for the writer */ +void reftable_writer_free(struct reftable_writer *w); + +#endif diff --git a/reftable/stack.c b/reftable/stack.c new file mode 100644 index 0000000..80266bc --- /dev/null +++ b/reftable/stack.c @@ -0,0 +1,1518 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "stack.h" + +#include "../write-or-die.h" +#include "system.h" +#include "merged.h" +#include "reader.h" +#include "refname.h" +#include "reftable-error.h" +#include "reftable-record.h" +#include "reftable-merged.h" +#include "writer.h" +#include "tempfile.h" + +static int stack_try_add(struct reftable_stack *st, + int (*write_table)(struct reftable_writer *wr, + void *arg), + void *arg); +static int stack_write_compact(struct reftable_stack *st, + struct reftable_writer *wr, + size_t first, size_t last, + struct reftable_log_expiry_config *config); +static int stack_check_addition(struct reftable_stack *st, + const char *new_tab_name); +static void reftable_addition_close(struct reftable_addition *add); +static int reftable_stack_reload_maybe_reuse(struct reftable_stack *st, + int reuse_open); + +static void stack_filename(struct strbuf *dest, struct reftable_stack *st, + const char *name) +{ + strbuf_reset(dest); + strbuf_addstr(dest, st->reftable_dir); + strbuf_addstr(dest, "/"); + strbuf_addstr(dest, name); +} + +static ssize_t reftable_fd_write(void *arg, const void *data, size_t sz) +{ + int *fdp = (int *)arg; + return write_in_full(*fdp, data, sz); +} + +static int reftable_fd_flush(void *arg) +{ + int *fdp = (int *)arg; + + return fsync_component(FSYNC_COMPONENT_REFERENCE, *fdp); +} + +int reftable_new_stack(struct reftable_stack **dest, const char *dir, + struct reftable_write_options config) +{ + struct reftable_stack *p = reftable_calloc(1, sizeof(*p)); + struct strbuf list_file_name = STRBUF_INIT; + int err = 0; + + if (config.hash_id == 0) { + config.hash_id = GIT_SHA1_FORMAT_ID; + } + + *dest = NULL; + + strbuf_reset(&list_file_name); + strbuf_addstr(&list_file_name, dir); + strbuf_addstr(&list_file_name, "/tables.list"); + + p->list_file = strbuf_detach(&list_file_name, NULL); + p->list_fd = -1; + p->reftable_dir = xstrdup(dir); + p->config = config; + + err = reftable_stack_reload_maybe_reuse(p, 1); + if (err < 0) { + reftable_stack_destroy(p); + } else { + *dest = p; + } + return err; +} + +static int fd_read_lines(int fd, char ***namesp) +{ + off_t size = lseek(fd, 0, SEEK_END); + char *buf = NULL; + int err = 0; + if (size < 0) { + err = REFTABLE_IO_ERROR; + goto done; + } + err = lseek(fd, 0, SEEK_SET); + if (err < 0) { + err = REFTABLE_IO_ERROR; + goto done; + } + + REFTABLE_ALLOC_ARRAY(buf, size + 1); + if (read_in_full(fd, buf, size) != size) { + err = REFTABLE_IO_ERROR; + goto done; + } + buf[size] = 0; + + parse_names(buf, size, namesp); + +done: + reftable_free(buf); + return err; +} + +int read_lines(const char *filename, char ***namesp) +{ + int fd = open(filename, O_RDONLY); + int err = 0; + if (fd < 0) { + if (errno == ENOENT) { + REFTABLE_CALLOC_ARRAY(*namesp, 1); + return 0; + } + + return REFTABLE_IO_ERROR; + } + err = fd_read_lines(fd, namesp); + close(fd); + return err; +} + +struct reftable_merged_table * +reftable_stack_merged_table(struct reftable_stack *st) +{ + return st->merged; +} + +static int has_name(char **names, const char *name) +{ + while (*names) { + if (!strcmp(*names, name)) + return 1; + names++; + } + return 0; +} + +/* Close and free the stack */ +void reftable_stack_destroy(struct reftable_stack *st) +{ + char **names = NULL; + int err = 0; + if (st->merged) { + reftable_merged_table_free(st->merged); + st->merged = NULL; + } + + err = read_lines(st->list_file, &names); + if (err < 0) { + FREE_AND_NULL(names); + } + + if (st->readers) { + int i = 0; + struct strbuf filename = STRBUF_INIT; + for (i = 0; i < st->readers_len; i++) { + const char *name = reader_name(st->readers[i]); + strbuf_reset(&filename); + if (names && !has_name(names, name)) { + stack_filename(&filename, st, name); + } + reftable_reader_free(st->readers[i]); + + if (filename.len) { + /* On Windows, can only unlink after closing. */ + unlink(filename.buf); + } + } + strbuf_release(&filename); + st->readers_len = 0; + FREE_AND_NULL(st->readers); + } + + if (st->list_fd >= 0) { + close(st->list_fd); + st->list_fd = -1; + } + + FREE_AND_NULL(st->list_file); + FREE_AND_NULL(st->reftable_dir); + reftable_free(st); + free_names(names); +} + +static struct reftable_reader **stack_copy_readers(struct reftable_stack *st, + int cur_len) +{ + struct reftable_reader **cur = reftable_calloc(cur_len, sizeof(*cur)); + int i = 0; + for (i = 0; i < cur_len; i++) { + cur[i] = st->readers[i]; + } + return cur; +} + +static int reftable_stack_reload_once(struct reftable_stack *st, char **names, + int reuse_open) +{ + size_t cur_len = !st->merged ? 0 : st->merged->stack_len; + struct reftable_reader **cur = stack_copy_readers(st, cur_len); + size_t names_len = names_length(names); + struct reftable_reader **new_readers = + reftable_calloc(names_len, sizeof(*new_readers)); + struct reftable_table *new_tables = + reftable_calloc(names_len, sizeof(*new_tables)); + size_t new_readers_len = 0; + struct reftable_merged_table *new_merged = NULL; + struct strbuf table_path = STRBUF_INIT; + int err = 0; + size_t i; + + while (*names) { + struct reftable_reader *rd = NULL; + char *name = *names++; + + /* this is linear; we assume compaction keeps the number of + tables under control so this is not quadratic. */ + for (i = 0; reuse_open && i < cur_len; i++) { + if (cur[i] && 0 == strcmp(cur[i]->name, name)) { + rd = cur[i]; + cur[i] = NULL; + break; + } + } + + if (!rd) { + struct reftable_block_source src = { NULL }; + stack_filename(&table_path, st, name); + + err = reftable_block_source_from_file(&src, + table_path.buf); + if (err < 0) + goto done; + + err = reftable_new_reader(&rd, &src, name); + if (err < 0) + goto done; + } + + new_readers[new_readers_len] = rd; + reftable_table_from_reader(&new_tables[new_readers_len], rd); + new_readers_len++; + } + + /* success! */ + err = reftable_new_merged_table(&new_merged, new_tables, + new_readers_len, st->config.hash_id); + if (err < 0) + goto done; + + new_tables = NULL; + st->readers_len = new_readers_len; + if (st->merged) { + merged_table_release(st->merged); + reftable_merged_table_free(st->merged); + } + if (st->readers) { + reftable_free(st->readers); + } + st->readers = new_readers; + new_readers = NULL; + new_readers_len = 0; + + new_merged->suppress_deletions = 1; + st->merged = new_merged; + for (i = 0; i < cur_len; i++) { + if (cur[i]) { + const char *name = reader_name(cur[i]); + stack_filename(&table_path, st, name); + + reader_close(cur[i]); + reftable_reader_free(cur[i]); + + /* On Windows, can only unlink after closing. */ + unlink(table_path.buf); + } + } + +done: + for (i = 0; i < new_readers_len; i++) { + reader_close(new_readers[i]); + reftable_reader_free(new_readers[i]); + } + reftable_free(new_readers); + reftable_free(new_tables); + reftable_free(cur); + strbuf_release(&table_path); + return err; +} + +/* return negative if a before b. */ +static int tv_cmp(struct timeval *a, struct timeval *b) +{ + time_t diff = a->tv_sec - b->tv_sec; + int udiff = a->tv_usec - b->tv_usec; + + if (diff != 0) + return diff; + + return udiff; +} + +static int reftable_stack_reload_maybe_reuse(struct reftable_stack *st, + int reuse_open) +{ + char **names = NULL, **names_after = NULL; + struct timeval deadline; + int64_t delay = 0; + int tries = 0, err; + int fd = -1; + + err = gettimeofday(&deadline, NULL); + if (err < 0) + goto out; + deadline.tv_sec += 3; + + while (1) { + struct timeval now; + + err = gettimeofday(&now, NULL); + if (err < 0) + goto out; + + /* + * Only look at deadlines after the first few times. This + * simplifies debugging in GDB. + */ + tries++; + if (tries > 3 && tv_cmp(&now, &deadline) >= 0) + goto out; + + fd = open(st->list_file, O_RDONLY); + if (fd < 0) { + if (errno != ENOENT) { + err = REFTABLE_IO_ERROR; + goto out; + } + + REFTABLE_CALLOC_ARRAY(names, 1); + } else { + err = fd_read_lines(fd, &names); + if (err < 0) + goto out; + } + + err = reftable_stack_reload_once(st, names, reuse_open); + if (!err) + break; + if (err != REFTABLE_NOT_EXIST_ERROR) + goto out; + + /* + * REFTABLE_NOT_EXIST_ERROR can be caused by a concurrent + * writer. Check if there was one by checking if the name list + * changed. + */ + err = read_lines(st->list_file, &names_after); + if (err < 0) + goto out; + if (names_equal(names_after, names)) { + err = REFTABLE_NOT_EXIST_ERROR; + goto out; + } + + free_names(names); + names = NULL; + free_names(names_after); + names_after = NULL; + close(fd); + fd = -1; + + delay = delay + (delay * rand()) / RAND_MAX + 1; + sleep_millisec(delay); + } + +out: + /* + * Invalidate the stat cache. It is sufficient to only close the file + * descriptor and keep the cached stat info because we never use the + * latter when the former is negative. + */ + if (st->list_fd >= 0) { + close(st->list_fd); + st->list_fd = -1; + } + + /* + * Cache stat information in case it provides a useful signal to us. + * According to POSIX, "The st_ino and st_dev fields taken together + * uniquely identify the file within the system." That being said, + * Windows is not POSIX compliant and we do not have these fields + * available. So the information we have there is insufficient to + * determine whether two file descriptors point to the same file. + * + * While we could fall back to using other signals like the file's + * mtime, those are not sufficient to avoid races. We thus refrain from + * using the stat cache on such systems and fall back to the secondary + * caching mechanism, which is to check whether contents of the file + * have changed. + * + * On other systems which are POSIX compliant we must keep the file + * descriptor open. This is to avoid a race condition where two + * processes access the reftable stack at the same point in time: + * + * 1. A reads the reftable stack and caches its stat info. + * + * 2. B updates the stack, appending a new table to "tables.list". + * This will both use a new inode and result in a different file + * size, thus invalidating A's cache in theory. + * + * 3. B decides to auto-compact the stack and merges two tables. The + * file size now matches what A has cached again. Furthermore, the + * filesystem may decide to recycle the inode number of the file + * we have replaced in (2) because it is not in use anymore. + * + * 4. A reloads the reftable stack. Neither the inode number nor the + * file size changed. If the timestamps did not change either then + * we think the cached copy of our stack is up-to-date. + * + * By keeping the file descriptor open the inode number cannot be + * recycled, mitigating the race. + */ + if (!err && fd >= 0 && !fstat(fd, &st->list_st) && + st->list_st.st_dev && st->list_st.st_ino) { + st->list_fd = fd; + fd = -1; + } + + if (fd >= 0) + close(fd); + free_names(names); + free_names(names_after); + return err; +} + +/* -1 = error + 0 = up to date + 1 = changed. */ +static int stack_uptodate(struct reftable_stack *st) +{ + char **names = NULL; + int err; + int i = 0; + + /* + * When we have cached stat information available then we use it to + * verify whether the file has been rewritten. + * + * Note that we explicitly do not want to use `stat_validity_check()` + * and friends here because they may end up not comparing the `st_dev` + * and `st_ino` fields. These functions thus cannot guarantee that we + * indeed still have the same file. + */ + if (st->list_fd >= 0) { + struct stat list_st; + + if (stat(st->list_file, &list_st) < 0) { + /* + * It's fine for "tables.list" to not exist. In that + * case, we have to refresh when the loaded stack has + * any readers. + */ + if (errno == ENOENT) + return !!st->readers_len; + return REFTABLE_IO_ERROR; + } + + /* + * When "tables.list" refers to the same file we can assume + * that it didn't change. This is because we always use + * rename(3P) to update the file and never write to it + * directly. + */ + if (st->list_st.st_dev == list_st.st_dev && + st->list_st.st_ino == list_st.st_ino) + return 0; + } + + err = read_lines(st->list_file, &names); + if (err < 0) + return err; + + for (i = 0; i < st->readers_len; i++) { + if (!names[i]) { + err = 1; + goto done; + } + + if (strcmp(st->readers[i]->name, names[i])) { + err = 1; + goto done; + } + } + + if (names[st->merged->stack_len]) { + err = 1; + goto done; + } + +done: + free_names(names); + return err; +} + +int reftable_stack_reload(struct reftable_stack *st) +{ + int err = stack_uptodate(st); + if (err > 0) + return reftable_stack_reload_maybe_reuse(st, 1); + return err; +} + +int reftable_stack_add(struct reftable_stack *st, + int (*write)(struct reftable_writer *wr, void *arg), + void *arg) +{ + int err = stack_try_add(st, write, arg); + if (err < 0) { + if (err == REFTABLE_OUTDATED_ERROR) { + /* Ignore error return, we want to propagate + REFTABLE_OUTDATED_ERROR. + */ + reftable_stack_reload(st); + } + return err; + } + + return 0; +} + +static void format_name(struct strbuf *dest, uint64_t min, uint64_t max) +{ + char buf[100]; + uint32_t rnd = (uint32_t)git_rand(); + snprintf(buf, sizeof(buf), "0x%012" PRIx64 "-0x%012" PRIx64 "-%08x", + min, max, rnd); + strbuf_reset(dest); + strbuf_addstr(dest, buf); +} + +struct reftable_addition { + struct tempfile *lock_file; + struct reftable_stack *stack; + + char **new_tables; + size_t new_tables_len, new_tables_cap; + uint64_t next_update_index; +}; + +#define REFTABLE_ADDITION_INIT {0} + +static int reftable_stack_init_addition(struct reftable_addition *add, + struct reftable_stack *st) +{ + struct strbuf lock_file_name = STRBUF_INIT; + int err = 0; + add->stack = st; + + strbuf_addf(&lock_file_name, "%s.lock", st->list_file); + + add->lock_file = create_tempfile(lock_file_name.buf); + if (!add->lock_file) { + if (errno == EEXIST) { + err = REFTABLE_LOCK_ERROR; + } else { + err = REFTABLE_IO_ERROR; + } + goto done; + } + if (st->config.default_permissions) { + if (chmod(add->lock_file->filename.buf, st->config.default_permissions) < 0) { + err = REFTABLE_IO_ERROR; + goto done; + } + } + + err = stack_uptodate(st); + if (err < 0) + goto done; + if (err > 0) { + err = REFTABLE_OUTDATED_ERROR; + goto done; + } + + add->next_update_index = reftable_stack_next_update_index(st); +done: + if (err) { + reftable_addition_close(add); + } + strbuf_release(&lock_file_name); + return err; +} + +static void reftable_addition_close(struct reftable_addition *add) +{ + struct strbuf nm = STRBUF_INIT; + size_t i; + + for (i = 0; i < add->new_tables_len; i++) { + stack_filename(&nm, add->stack, add->new_tables[i]); + unlink(nm.buf); + reftable_free(add->new_tables[i]); + add->new_tables[i] = NULL; + } + reftable_free(add->new_tables); + add->new_tables = NULL; + add->new_tables_len = 0; + add->new_tables_cap = 0; + + delete_tempfile(&add->lock_file); + strbuf_release(&nm); +} + +void reftable_addition_destroy(struct reftable_addition *add) +{ + if (!add) { + return; + } + reftable_addition_close(add); + reftable_free(add); +} + +int reftable_addition_commit(struct reftable_addition *add) +{ + struct strbuf table_list = STRBUF_INIT; + int lock_file_fd = get_tempfile_fd(add->lock_file); + int err = 0; + size_t i; + + if (add->new_tables_len == 0) + goto done; + + for (i = 0; i < add->stack->merged->stack_len; i++) { + strbuf_addstr(&table_list, add->stack->readers[i]->name); + strbuf_addstr(&table_list, "\n"); + } + for (i = 0; i < add->new_tables_len; i++) { + strbuf_addstr(&table_list, add->new_tables[i]); + strbuf_addstr(&table_list, "\n"); + } + + err = write_in_full(lock_file_fd, table_list.buf, table_list.len); + strbuf_release(&table_list); + if (err < 0) { + err = REFTABLE_IO_ERROR; + goto done; + } + + fsync_component_or_die(FSYNC_COMPONENT_REFERENCE, lock_file_fd, + get_tempfile_path(add->lock_file)); + + err = rename_tempfile(&add->lock_file, add->stack->list_file); + if (err < 0) { + err = REFTABLE_IO_ERROR; + goto done; + } + + /* success, no more state to clean up. */ + for (i = 0; i < add->new_tables_len; i++) + reftable_free(add->new_tables[i]); + reftable_free(add->new_tables); + add->new_tables = NULL; + add->new_tables_len = 0; + add->new_tables_cap = 0; + + err = reftable_stack_reload_maybe_reuse(add->stack, 1); + if (err) + goto done; + + if (!add->stack->config.disable_auto_compact) { + /* + * Auto-compact the stack to keep the number of tables in + * control. It is possible that a concurrent writer is already + * trying to compact parts of the stack, which would lead to a + * `REFTABLE_LOCK_ERROR` because parts of the stack are locked + * already. This is a benign error though, so we ignore it. + */ + err = reftable_stack_auto_compact(add->stack); + if (err < 0 && err != REFTABLE_LOCK_ERROR) + goto done; + err = 0; + } + +done: + reftable_addition_close(add); + return err; +} + +int reftable_stack_new_addition(struct reftable_addition **dest, + struct reftable_stack *st) +{ + int err = 0; + struct reftable_addition empty = REFTABLE_ADDITION_INIT; + REFTABLE_CALLOC_ARRAY(*dest, 1); + **dest = empty; + err = reftable_stack_init_addition(*dest, st); + if (err) { + reftable_free(*dest); + *dest = NULL; + } + return err; +} + +static int stack_try_add(struct reftable_stack *st, + int (*write_table)(struct reftable_writer *wr, + void *arg), + void *arg) +{ + struct reftable_addition add = REFTABLE_ADDITION_INIT; + int err = reftable_stack_init_addition(&add, st); + if (err < 0) + goto done; + + err = reftable_addition_add(&add, write_table, arg); + if (err < 0) + goto done; + + err = reftable_addition_commit(&add); +done: + reftable_addition_close(&add); + return err; +} + +int reftable_addition_add(struct reftable_addition *add, + int (*write_table)(struct reftable_writer *wr, + void *arg), + void *arg) +{ + struct strbuf temp_tab_file_name = STRBUF_INIT; + struct strbuf tab_file_name = STRBUF_INIT; + struct strbuf next_name = STRBUF_INIT; + struct reftable_writer *wr = NULL; + struct tempfile *tab_file = NULL; + int err = 0; + int tab_fd; + + strbuf_reset(&next_name); + format_name(&next_name, add->next_update_index, add->next_update_index); + + stack_filename(&temp_tab_file_name, add->stack, next_name.buf); + strbuf_addstr(&temp_tab_file_name, ".temp.XXXXXX"); + + tab_file = mks_tempfile(temp_tab_file_name.buf); + if (!tab_file) { + err = REFTABLE_IO_ERROR; + goto done; + } + if (add->stack->config.default_permissions) { + if (chmod(get_tempfile_path(tab_file), + add->stack->config.default_permissions)) { + err = REFTABLE_IO_ERROR; + goto done; + } + } + tab_fd = get_tempfile_fd(tab_file); + + wr = reftable_new_writer(reftable_fd_write, reftable_fd_flush, &tab_fd, + &add->stack->config); + err = write_table(wr, arg); + if (err < 0) + goto done; + + err = reftable_writer_close(wr); + if (err == REFTABLE_EMPTY_TABLE_ERROR) { + err = 0; + goto done; + } + if (err < 0) + goto done; + + err = close_tempfile_gently(tab_file); + if (err < 0) { + err = REFTABLE_IO_ERROR; + goto done; + } + + err = stack_check_addition(add->stack, get_tempfile_path(tab_file)); + if (err < 0) + goto done; + + if (wr->min_update_index < add->next_update_index) { + err = REFTABLE_API_ERROR; + goto done; + } + + format_name(&next_name, wr->min_update_index, wr->max_update_index); + strbuf_addstr(&next_name, ".ref"); + stack_filename(&tab_file_name, add->stack, next_name.buf); + + /* + On windows, this relies on rand() picking a unique destination name. + Maybe we should do retry loop as well? + */ + err = rename_tempfile(&tab_file, tab_file_name.buf); + if (err < 0) { + err = REFTABLE_IO_ERROR; + goto done; + } + + REFTABLE_ALLOC_GROW(add->new_tables, add->new_tables_len + 1, + add->new_tables_cap); + add->new_tables[add->new_tables_len++] = strbuf_detach(&next_name, NULL); +done: + delete_tempfile(&tab_file); + strbuf_release(&temp_tab_file_name); + strbuf_release(&tab_file_name); + strbuf_release(&next_name); + reftable_writer_free(wr); + return err; +} + +uint64_t reftable_stack_next_update_index(struct reftable_stack *st) +{ + int sz = st->merged->stack_len; + if (sz > 0) + return reftable_reader_max_update_index(st->readers[sz - 1]) + + 1; + return 1; +} + +static int stack_compact_locked(struct reftable_stack *st, + size_t first, size_t last, + struct reftable_log_expiry_config *config, + struct tempfile **tab_file_out) +{ + struct strbuf next_name = STRBUF_INIT; + struct strbuf tab_file_path = STRBUF_INIT; + struct reftable_writer *wr = NULL; + struct tempfile *tab_file; + int tab_fd, err = 0; + + format_name(&next_name, + reftable_reader_min_update_index(st->readers[first]), + reftable_reader_max_update_index(st->readers[last])); + stack_filename(&tab_file_path, st, next_name.buf); + strbuf_addstr(&tab_file_path, ".temp.XXXXXX"); + + tab_file = mks_tempfile(tab_file_path.buf); + if (!tab_file) { + err = REFTABLE_IO_ERROR; + goto done; + } + tab_fd = get_tempfile_fd(tab_file); + + if (st->config.default_permissions && + chmod(get_tempfile_path(tab_file), st->config.default_permissions) < 0) { + err = REFTABLE_IO_ERROR; + goto done; + } + + wr = reftable_new_writer(reftable_fd_write, reftable_fd_flush, + &tab_fd, &st->config); + err = stack_write_compact(st, wr, first, last, config); + if (err < 0) + goto done; + + err = reftable_writer_close(wr); + if (err < 0) + goto done; + + err = close_tempfile_gently(tab_file); + if (err < 0) + goto done; + + *tab_file_out = tab_file; + tab_file = NULL; + +done: + delete_tempfile(&tab_file); + reftable_writer_free(wr); + strbuf_release(&next_name); + strbuf_release(&tab_file_path); + return err; +} + +static int stack_write_compact(struct reftable_stack *st, + struct reftable_writer *wr, + size_t first, size_t last, + struct reftable_log_expiry_config *config) +{ + size_t subtabs_len = last - first + 1; + struct reftable_table *subtabs = reftable_calloc( + last - first + 1, sizeof(*subtabs)); + struct reftable_merged_table *mt = NULL; + struct reftable_iterator it = { NULL }; + struct reftable_ref_record ref = { NULL }; + struct reftable_log_record log = { NULL }; + uint64_t entries = 0; + int err = 0; + + for (size_t i = first, j = 0; i <= last; i++) { + struct reftable_reader *t = st->readers[i]; + reftable_table_from_reader(&subtabs[j++], t); + st->stats.bytes += t->size; + } + reftable_writer_set_limits(wr, st->readers[first]->min_update_index, + st->readers[last]->max_update_index); + + err = reftable_new_merged_table(&mt, subtabs, subtabs_len, + st->config.hash_id); + if (err < 0) { + reftable_free(subtabs); + goto done; + } + + err = reftable_merged_table_seek_ref(mt, &it, ""); + if (err < 0) + goto done; + + while (1) { + err = reftable_iterator_next_ref(&it, &ref); + if (err > 0) { + err = 0; + break; + } + if (err < 0) + goto done; + + if (first == 0 && reftable_ref_record_is_deletion(&ref)) { + continue; + } + + err = reftable_writer_add_ref(wr, &ref); + if (err < 0) + goto done; + entries++; + } + reftable_iterator_destroy(&it); + + err = reftable_merged_table_seek_log(mt, &it, ""); + if (err < 0) + goto done; + + while (1) { + err = reftable_iterator_next_log(&it, &log); + if (err > 0) { + err = 0; + break; + } + if (err < 0) + goto done; + if (first == 0 && reftable_log_record_is_deletion(&log)) { + continue; + } + + if (config && config->min_update_index > 0 && + log.update_index < config->min_update_index) { + continue; + } + + if (config && config->time > 0 && + log.value.update.time < config->time) { + continue; + } + + err = reftable_writer_add_log(wr, &log); + if (err < 0) + goto done; + entries++; + } + +done: + reftable_iterator_destroy(&it); + if (mt) { + merged_table_release(mt); + reftable_merged_table_free(mt); + } + reftable_ref_record_release(&ref); + reftable_log_record_release(&log); + st->stats.entries_written += entries; + return err; +} + +/* + * Compact all tables in the range `[first, last)` into a single new table. + * + * This function returns `0` on success or a code `< 0` on failure. When the + * stack or any of the tables in the specified range are already locked then + * this function returns `REFTABLE_LOCK_ERROR`. This is a benign error that + * callers can either ignore, or they may choose to retry compaction after some + * amount of time. + */ +static int stack_compact_range(struct reftable_stack *st, + size_t first, size_t last, + struct reftable_log_expiry_config *expiry) +{ + struct strbuf tables_list_buf = STRBUF_INIT; + struct strbuf new_table_name = STRBUF_INIT; + struct strbuf new_table_path = STRBUF_INIT; + struct strbuf table_name = STRBUF_INIT; + struct lock_file tables_list_lock = LOCK_INIT; + struct lock_file *table_locks = NULL; + struct tempfile *new_table = NULL; + int is_empty_table = 0, err = 0; + size_t i; + + if (first > last || (!expiry && first == last)) { + err = 0; + goto done; + } + + st->stats.attempts++; + + /* + * Hold the lock so that we can read "tables.list" and lock all tables + * which are part of the user-specified range. + */ + err = hold_lock_file_for_update(&tables_list_lock, st->list_file, + LOCK_NO_DEREF); + if (err < 0) { + if (errno == EEXIST) + err = REFTABLE_LOCK_ERROR; + else + err = REFTABLE_IO_ERROR; + goto done; + } + + err = stack_uptodate(st); + if (err) + goto done; + + /* + * Lock all tables in the user-provided range. This is the slice of our + * stack which we'll compact. + */ + REFTABLE_CALLOC_ARRAY(table_locks, last - first + 1); + for (i = first; i <= last; i++) { + stack_filename(&table_name, st, reader_name(st->readers[i])); + + err = hold_lock_file_for_update(&table_locks[i - first], + table_name.buf, LOCK_NO_DEREF); + if (err < 0) { + if (errno == EEXIST) + err = REFTABLE_LOCK_ERROR; + else + err = REFTABLE_IO_ERROR; + goto done; + } + + /* + * We need to close the lockfiles as we might otherwise easily + * run into file descriptor exhaustion when we compress a lot + * of tables. + */ + err = close_lock_file_gently(&table_locks[i - first]); + if (err < 0) { + err = REFTABLE_IO_ERROR; + goto done; + } + } + + /* + * We have locked all tables in our range and can thus release the + * "tables.list" lock while compacting the locked tables. This allows + * concurrent updates to the stack to proceed. + */ + err = rollback_lock_file(&tables_list_lock); + if (err < 0) { + err = REFTABLE_IO_ERROR; + goto done; + } + + /* + * Compact the now-locked tables into a new table. Note that compacting + * these tables may end up with an empty new table in case tombstones + * end up cancelling out all refs in that range. + */ + err = stack_compact_locked(st, first, last, expiry, &new_table); + if (err < 0) { + if (err != REFTABLE_EMPTY_TABLE_ERROR) + goto done; + is_empty_table = 1; + } + + /* + * Now that we have written the new, compacted table we need to re-lock + * "tables.list". We'll then replace the compacted range of tables with + * the new table. + */ + err = hold_lock_file_for_update(&tables_list_lock, st->list_file, + LOCK_NO_DEREF); + if (err < 0) { + if (errno == EEXIST) + err = REFTABLE_LOCK_ERROR; + else + err = REFTABLE_IO_ERROR; + goto done; + } + + if (st->config.default_permissions) { + if (chmod(get_lock_file_path(&tables_list_lock), + st->config.default_permissions) < 0) { + err = REFTABLE_IO_ERROR; + goto done; + } + } + + /* + * If the resulting compacted table is not empty, then we need to move + * it into place now. + */ + if (!is_empty_table) { + format_name(&new_table_name, st->readers[first]->min_update_index, + st->readers[last]->max_update_index); + strbuf_addstr(&new_table_name, ".ref"); + stack_filename(&new_table_path, st, new_table_name.buf); + + err = rename_tempfile(&new_table, new_table_path.buf); + if (err < 0) { + err = REFTABLE_IO_ERROR; + goto done; + } + } + + /* + * Write the new "tables.list" contents with the compacted table we + * have just written. In case the compacted table became empty we + * simply skip writing it. + */ + for (i = 0; i < first; i++) + strbuf_addf(&tables_list_buf, "%s\n", st->readers[i]->name); + if (!is_empty_table) + strbuf_addf(&tables_list_buf, "%s\n", new_table_name.buf); + for (i = last + 1; i < st->merged->stack_len; i++) + strbuf_addf(&tables_list_buf, "%s\n", st->readers[i]->name); + + err = write_in_full(get_lock_file_fd(&tables_list_lock), + tables_list_buf.buf, tables_list_buf.len); + if (err < 0) { + err = REFTABLE_IO_ERROR; + unlink(new_table_path.buf); + goto done; + } + + err = fsync_component(FSYNC_COMPONENT_REFERENCE, get_lock_file_fd(&tables_list_lock)); + if (err < 0) { + err = REFTABLE_IO_ERROR; + unlink(new_table_path.buf); + goto done; + } + + err = commit_lock_file(&tables_list_lock); + if (err < 0) { + err = REFTABLE_IO_ERROR; + unlink(new_table_path.buf); + goto done; + } + + /* + * Reload the stack before deleting the compacted tables. We can only + * delete the files after we closed them on Windows, so this needs to + * happen first. + */ + err = reftable_stack_reload_maybe_reuse(st, first < last); + if (err < 0) + goto done; + + /* + * Delete the old tables. They may still be in use by concurrent + * readers, so it is expected that unlinking tables may fail. + */ + for (i = first; i <= last; i++) { + struct lock_file *table_lock = &table_locks[i - first]; + char *table_path = get_locked_file_path(table_lock); + unlink(table_path); + free(table_path); + } + +done: + rollback_lock_file(&tables_list_lock); + for (i = first; table_locks && i <= last; i++) + rollback_lock_file(&table_locks[i - first]); + reftable_free(table_locks); + + delete_tempfile(&new_table); + strbuf_release(&new_table_name); + strbuf_release(&new_table_path); + + strbuf_release(&tables_list_buf); + strbuf_release(&table_name); + return err; +} + +int reftable_stack_compact_all(struct reftable_stack *st, + struct reftable_log_expiry_config *config) +{ + return stack_compact_range(st, 0, st->merged->stack_len ? + st->merged->stack_len - 1 : 0, config); +} + +static int stack_compact_range_stats(struct reftable_stack *st, + size_t first, size_t last, + struct reftable_log_expiry_config *config) +{ + int err = stack_compact_range(st, first, last, config); + if (err == REFTABLE_LOCK_ERROR) + st->stats.failures++; + return err; +} + +static int segment_size(struct segment *s) +{ + return s->end - s->start; +} + +struct segment suggest_compaction_segment(uint64_t *sizes, size_t n) +{ + struct segment seg = { 0 }; + uint64_t bytes; + size_t i; + + /* + * If there are no tables or only a single one then we don't have to + * compact anything. The sequence is geometric by definition already. + */ + if (n <= 1) + return seg; + + /* + * Find the ending table of the compaction segment needed to restore the + * geometric sequence. Note that the segment end is exclusive. + * + * To do so, we iterate backwards starting from the most recent table + * until a valid segment end is found. If the preceding table is smaller + * than the current table multiplied by the geometric factor (2), the + * compaction segment end has been identified. + * + * Tables after the ending point are not added to the byte count because + * they are already valid members of the geometric sequence. Due to the + * properties of a geometric sequence, it is not possible for the sum of + * these tables to exceed the value of the ending point table. + * + * Example table size sequence requiring no compaction: + * 64, 32, 16, 8, 4, 2, 1 + * + * Example table size sequence where compaction segment end is set to + * the last table. Since the segment end is exclusive, the last table is + * excluded during subsequent compaction and the table with size 3 is + * the final table included: + * 64, 32, 16, 8, 4, 3, 1 + */ + for (i = n - 1; i > 0; i--) { + if (sizes[i - 1] < sizes[i] * 2) { + seg.end = i + 1; + bytes = sizes[i]; + break; + } + } + + /* + * Find the starting table of the compaction segment by iterating + * through the remaining tables and keeping track of the accumulated + * size of all tables seen from the segment end table. The previous + * table is compared to the accumulated size because the tables from the + * segment end are merged backwards recursively. + * + * Note that we keep iterating even after we have found the first + * starting point. This is because there may be tables in the stack + * preceding that first starting point which violate the geometric + * sequence. + * + * Example compaction segment start set to table with size 32: + * 128, 32, 16, 8, 4, 3, 1 + */ + for (; i > 0; i--) { + uint64_t curr = bytes; + bytes += sizes[i - 1]; + + if (sizes[i - 1] < curr * 2) { + seg.start = i - 1; + seg.bytes = bytes; + } + } + + return seg; +} + +static uint64_t *stack_table_sizes_for_compaction(struct reftable_stack *st) +{ + uint64_t *sizes = + reftable_calloc(st->merged->stack_len, sizeof(*sizes)); + int version = (st->config.hash_id == GIT_SHA1_FORMAT_ID) ? 1 : 2; + int overhead = header_size(version) - 1; + int i = 0; + for (i = 0; i < st->merged->stack_len; i++) { + sizes[i] = st->readers[i]->size - overhead; + } + return sizes; +} + +int reftable_stack_auto_compact(struct reftable_stack *st) +{ + uint64_t *sizes = stack_table_sizes_for_compaction(st); + struct segment seg = + suggest_compaction_segment(sizes, st->merged->stack_len); + reftable_free(sizes); + if (segment_size(&seg) > 0) + return stack_compact_range_stats(st, seg.start, seg.end - 1, + NULL); + + return 0; +} + +struct reftable_compaction_stats * +reftable_stack_compaction_stats(struct reftable_stack *st) +{ + return &st->stats; +} + +int reftable_stack_read_ref(struct reftable_stack *st, const char *refname, + struct reftable_ref_record *ref) +{ + struct reftable_table tab = { NULL }; + reftable_table_from_merged_table(&tab, reftable_stack_merged_table(st)); + return reftable_table_read_ref(&tab, refname, ref); +} + +int reftable_stack_read_log(struct reftable_stack *st, const char *refname, + struct reftable_log_record *log) +{ + struct reftable_iterator it = { NULL }; + struct reftable_merged_table *mt = reftable_stack_merged_table(st); + int err = reftable_merged_table_seek_log(mt, &it, refname); + if (err) + goto done; + + err = reftable_iterator_next_log(&it, log); + if (err) + goto done; + + if (strcmp(log->refname, refname) || + reftable_log_record_is_deletion(log)) { + err = 1; + goto done; + } + +done: + if (err) { + reftable_log_record_release(log); + } + reftable_iterator_destroy(&it); + return err; +} + +static int stack_check_addition(struct reftable_stack *st, + const char *new_tab_name) +{ + int err = 0; + struct reftable_block_source src = { NULL }; + struct reftable_reader *rd = NULL; + struct reftable_table tab = { NULL }; + struct reftable_ref_record *refs = NULL; + struct reftable_iterator it = { NULL }; + int cap = 0; + int len = 0; + int i = 0; + + if (st->config.skip_name_check) + return 0; + + err = reftable_block_source_from_file(&src, new_tab_name); + if (err < 0) + goto done; + + err = reftable_new_reader(&rd, &src, new_tab_name); + if (err < 0) + goto done; + + err = reftable_reader_seek_ref(rd, &it, ""); + if (err > 0) { + err = 0; + goto done; + } + if (err < 0) + goto done; + + while (1) { + struct reftable_ref_record ref = { NULL }; + err = reftable_iterator_next_ref(&it, &ref); + if (err > 0) + break; + if (err < 0) + goto done; + + REFTABLE_ALLOC_GROW(refs, len + 1, cap); + refs[len++] = ref; + } + + reftable_table_from_merged_table(&tab, reftable_stack_merged_table(st)); + + err = validate_ref_record_addition(tab, refs, len); + +done: + for (i = 0; i < len; i++) { + reftable_ref_record_release(&refs[i]); + } + + free(refs); + reftable_iterator_destroy(&it); + reftable_reader_free(rd); + return err; +} + +static int is_table_name(const char *s) +{ + const char *dot = strrchr(s, '.'); + return dot && !strcmp(dot, ".ref"); +} + +static void remove_maybe_stale_table(struct reftable_stack *st, uint64_t max, + const char *name) +{ + int err = 0; + uint64_t update_idx = 0; + struct reftable_block_source src = { NULL }; + struct reftable_reader *rd = NULL; + struct strbuf table_path = STRBUF_INIT; + stack_filename(&table_path, st, name); + + err = reftable_block_source_from_file(&src, table_path.buf); + if (err < 0) + goto done; + + err = reftable_new_reader(&rd, &src, name); + if (err < 0) + goto done; + + update_idx = reftable_reader_max_update_index(rd); + reftable_reader_free(rd); + + if (update_idx <= max) { + unlink(table_path.buf); + } +done: + strbuf_release(&table_path); +} + +static int reftable_stack_clean_locked(struct reftable_stack *st) +{ + uint64_t max = reftable_merged_table_max_update_index( + reftable_stack_merged_table(st)); + DIR *dir = opendir(st->reftable_dir); + struct dirent *d = NULL; + if (!dir) { + return REFTABLE_IO_ERROR; + } + + while ((d = readdir(dir))) { + int i = 0; + int found = 0; + if (!is_table_name(d->d_name)) + continue; + + for (i = 0; !found && i < st->readers_len; i++) { + found = !strcmp(reader_name(st->readers[i]), d->d_name); + } + if (found) + continue; + + remove_maybe_stale_table(st, max, d->d_name); + } + + closedir(dir); + return 0; +} + +int reftable_stack_clean(struct reftable_stack *st) +{ + struct reftable_addition *add = NULL; + int err = reftable_stack_new_addition(&add, st); + if (err < 0) { + goto done; + } + + err = reftable_stack_reload(st); + if (err < 0) { + goto done; + } + + err = reftable_stack_clean_locked(st); + +done: + reftable_addition_destroy(add); + return err; +} + +int reftable_stack_print_directory(const char *stackdir, uint32_t hash_id) +{ + struct reftable_stack *stack = NULL; + struct reftable_write_options cfg = { .hash_id = hash_id }; + struct reftable_merged_table *merged = NULL; + struct reftable_table table = { NULL }; + + int err = reftable_new_stack(&stack, stackdir, cfg); + if (err < 0) + goto done; + + merged = reftable_stack_merged_table(stack); + reftable_table_from_merged_table(&table, merged); + err = reftable_table_print(&table); +done: + if (stack) + reftable_stack_destroy(stack); + return err; +} diff --git a/reftable/stack.h b/reftable/stack.h new file mode 100644 index 0000000..d43efa4 --- /dev/null +++ b/reftable/stack.h @@ -0,0 +1,40 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef STACK_H +#define STACK_H + +#include "system.h" +#include "reftable-writer.h" +#include "reftable-stack.h" + +struct reftable_stack { + struct stat list_st; + char *list_file; + int list_fd; + + char *reftable_dir; + + struct reftable_write_options config; + + struct reftable_reader **readers; + size_t readers_len; + struct reftable_merged_table *merged; + struct reftable_compaction_stats stats; +}; + +int read_lines(const char *filename, char ***lines); + +struct segment { + size_t start, end; + uint64_t bytes; +}; + +struct segment suggest_compaction_segment(uint64_t *sizes, size_t n); + +#endif diff --git a/reftable/stack_test.c b/reftable/stack_test.c new file mode 100644 index 0000000..1df3ffc --- /dev/null +++ b/reftable/stack_test.c @@ -0,0 +1,1112 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "stack.h" + +#include "system.h" + +#include "reftable-reader.h" +#include "merged.h" +#include "basics.h" +#include "record.h" +#include "test_framework.h" +#include "reftable-tests.h" +#include "reader.h" + +#include <sys/types.h> +#include <dirent.h> + +static void clear_dir(const char *dirname) +{ + struct strbuf path = STRBUF_INIT; + strbuf_addstr(&path, dirname); + remove_dir_recursively(&path, 0); + strbuf_release(&path); +} + +static int count_dir_entries(const char *dirname) +{ + DIR *dir = opendir(dirname); + int len = 0; + struct dirent *d; + if (!dir) + return 0; + + while ((d = readdir(dir))) { + /* + * Besides skipping over "." and "..", we also need to + * skip over other files that have a leading ".". This + * is due to behaviour of NFS, which will rename files + * to ".nfs*" to emulate delete-on-last-close. + * + * In any case this should be fine as the reftable + * library will never write files with leading dots + * anyway. + */ + if (starts_with(d->d_name, ".")) + continue; + len++; + } + closedir(dir); + return len; +} + +/* + * Work linenumber into the tempdir, so we can see which tests forget to + * cleanup. + */ +static char *get_tmp_template(int linenumber) +{ + const char *tmp = getenv("TMPDIR"); + static char template[1024]; + snprintf(template, sizeof(template) - 1, "%s/stack_test-%d.XXXXXX", + tmp ? tmp : "/tmp", linenumber); + return template; +} + +static char *get_tmp_dir(int linenumber) +{ + char *dir = get_tmp_template(linenumber); + EXPECT(mkdtemp(dir)); + return dir; +} + +static void test_read_file(void) +{ + char *fn = get_tmp_template(__LINE__); + int fd = mkstemp(fn); + char out[1024] = "line1\n\nline2\nline3"; + int n, err; + char **names = NULL; + char *want[] = { "line1", "line2", "line3" }; + int i = 0; + + EXPECT(fd > 0); + n = write_in_full(fd, out, strlen(out)); + EXPECT(n == strlen(out)); + err = close(fd); + EXPECT(err >= 0); + + err = read_lines(fn, &names); + EXPECT_ERR(err); + + for (i = 0; names[i]; i++) { + EXPECT(0 == strcmp(want[i], names[i])); + } + free_names(names); + (void) remove(fn); +} + +static void test_parse_names(void) +{ + char buf[] = "line\n"; + char **names = NULL; + parse_names(buf, strlen(buf), &names); + + EXPECT(NULL != names[0]); + EXPECT(0 == strcmp(names[0], "line")); + EXPECT(NULL == names[1]); + free_names(names); +} + +static void test_names_equal(void) +{ + char *a[] = { "a", "b", "c", NULL }; + char *b[] = { "a", "b", "d", NULL }; + char *c[] = { "a", "b", NULL }; + + EXPECT(names_equal(a, a)); + EXPECT(!names_equal(a, b)); + EXPECT(!names_equal(a, c)); +} + +static int write_test_ref(struct reftable_writer *wr, void *arg) +{ + struct reftable_ref_record *ref = arg; + reftable_writer_set_limits(wr, ref->update_index, ref->update_index); + return reftable_writer_add_ref(wr, ref); +} + +struct write_log_arg { + struct reftable_log_record *log; + uint64_t update_index; +}; + +static int write_test_log(struct reftable_writer *wr, void *arg) +{ + struct write_log_arg *wla = arg; + + reftable_writer_set_limits(wr, wla->update_index, wla->update_index); + return reftable_writer_add_log(wr, wla->log); +} + +static void test_reftable_stack_add_one(void) +{ + char *dir = get_tmp_dir(__LINE__); + struct strbuf scratch = STRBUF_INIT; + int mask = umask(002); + struct reftable_write_options cfg = { + .default_permissions = 0660, + }; + struct reftable_stack *st = NULL; + int err; + struct reftable_ref_record ref = { + .refname = "HEAD", + .update_index = 1, + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "master", + }; + struct reftable_ref_record dest = { NULL }; + struct stat stat_result = { 0 }; + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + + err = reftable_stack_add(st, &write_test_ref, &ref); + EXPECT_ERR(err); + + err = reftable_stack_read_ref(st, ref.refname, &dest); + EXPECT_ERR(err); + EXPECT(0 == strcmp("master", dest.value.symref)); + EXPECT(st->readers_len > 0); + + printf("testing print functionality:\n"); + err = reftable_stack_print_directory(dir, GIT_SHA1_FORMAT_ID); + EXPECT_ERR(err); + + err = reftable_stack_print_directory(dir, GIT_SHA256_FORMAT_ID); + EXPECT(err == REFTABLE_FORMAT_ERROR); + +#ifndef GIT_WINDOWS_NATIVE + strbuf_addstr(&scratch, dir); + strbuf_addstr(&scratch, "/tables.list"); + err = stat(scratch.buf, &stat_result); + EXPECT(!err); + EXPECT((stat_result.st_mode & 0777) == cfg.default_permissions); + + strbuf_reset(&scratch); + strbuf_addstr(&scratch, dir); + strbuf_addstr(&scratch, "/"); + /* do not try at home; not an external API for reftable. */ + strbuf_addstr(&scratch, st->readers[0]->name); + err = stat(scratch.buf, &stat_result); + EXPECT(!err); + EXPECT((stat_result.st_mode & 0777) == cfg.default_permissions); +#else + (void) stat_result; +#endif + + reftable_ref_record_release(&dest); + reftable_stack_destroy(st); + strbuf_release(&scratch); + clear_dir(dir); + umask(mask); +} + +static void test_reftable_stack_uptodate(void) +{ + struct reftable_write_options cfg = { 0 }; + struct reftable_stack *st1 = NULL; + struct reftable_stack *st2 = NULL; + char *dir = get_tmp_dir(__LINE__); + + int err; + struct reftable_ref_record ref1 = { + .refname = "HEAD", + .update_index = 1, + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "master", + }; + struct reftable_ref_record ref2 = { + .refname = "branch2", + .update_index = 2, + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "master", + }; + + + /* simulate multi-process access to the same stack + by creating two stacks for the same directory. + */ + err = reftable_new_stack(&st1, dir, cfg); + EXPECT_ERR(err); + + err = reftable_new_stack(&st2, dir, cfg); + EXPECT_ERR(err); + + err = reftable_stack_add(st1, &write_test_ref, &ref1); + EXPECT_ERR(err); + + err = reftable_stack_add(st2, &write_test_ref, &ref2); + EXPECT(err == REFTABLE_OUTDATED_ERROR); + + err = reftable_stack_reload(st2); + EXPECT_ERR(err); + + err = reftable_stack_add(st2, &write_test_ref, &ref2); + EXPECT_ERR(err); + reftable_stack_destroy(st1); + reftable_stack_destroy(st2); + clear_dir(dir); +} + +static void test_reftable_stack_transaction_api(void) +{ + char *dir = get_tmp_dir(__LINE__); + + struct reftable_write_options cfg = { 0 }; + struct reftable_stack *st = NULL; + int err; + struct reftable_addition *add = NULL; + + struct reftable_ref_record ref = { + .refname = "HEAD", + .update_index = 1, + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "master", + }; + struct reftable_ref_record dest = { NULL }; + + + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + + reftable_addition_destroy(add); + + err = reftable_stack_new_addition(&add, st); + EXPECT_ERR(err); + + err = reftable_addition_add(add, &write_test_ref, &ref); + EXPECT_ERR(err); + + err = reftable_addition_commit(add); + EXPECT_ERR(err); + + reftable_addition_destroy(add); + + err = reftable_stack_read_ref(st, ref.refname, &dest); + EXPECT_ERR(err); + EXPECT(REFTABLE_REF_SYMREF == dest.value_type); + EXPECT(0 == strcmp("master", dest.value.symref)); + + reftable_ref_record_release(&dest); + reftable_stack_destroy(st); + clear_dir(dir); +} + +static void test_reftable_stack_transaction_api_performs_auto_compaction(void) +{ + char *dir = get_tmp_dir(__LINE__); + struct reftable_write_options cfg = {0}; + struct reftable_addition *add = NULL; + struct reftable_stack *st = NULL; + int i, n = 20, err; + + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + + for (i = 0; i <= n; i++) { + struct reftable_ref_record ref = { + .update_index = reftable_stack_next_update_index(st), + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "master", + }; + char name[100]; + + snprintf(name, sizeof(name), "branch%04d", i); + ref.refname = name; + + /* + * Disable auto-compaction for all but the last runs. Like this + * we can ensure that we indeed honor this setting and have + * better control over when exactly auto compaction runs. + */ + st->config.disable_auto_compact = i != n; + + err = reftable_stack_new_addition(&add, st); + EXPECT_ERR(err); + + err = reftable_addition_add(add, &write_test_ref, &ref); + EXPECT_ERR(err); + + err = reftable_addition_commit(add); + EXPECT_ERR(err); + + reftable_addition_destroy(add); + + /* + * The stack length should grow continuously for all runs where + * auto compaction is disabled. When enabled, we should merge + * all tables in the stack. + */ + if (i != n) + EXPECT(st->merged->stack_len == i + 1); + else + EXPECT(st->merged->stack_len == 1); + } + + reftable_stack_destroy(st); + clear_dir(dir); +} + +static void test_reftable_stack_auto_compaction_fails_gracefully(void) +{ + struct reftable_ref_record ref = { + .refname = "refs/heads/master", + .update_index = 1, + .value_type = REFTABLE_REF_VAL1, + .value.val1 = {0x01}, + }; + struct reftable_write_options cfg = {0}; + struct reftable_stack *st; + struct strbuf table_path = STRBUF_INIT; + char *dir = get_tmp_dir(__LINE__); + int err; + + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + + err = reftable_stack_add(st, write_test_ref, &ref); + EXPECT_ERR(err); + EXPECT(st->merged->stack_len == 1); + EXPECT(st->stats.attempts == 0); + EXPECT(st->stats.failures == 0); + + /* + * Lock the newly written table such that it cannot be compacted. + * Adding a new table to the stack should not be impacted by this, even + * though auto-compaction will now fail. + */ + strbuf_addf(&table_path, "%s/%s.lock", dir, st->readers[0]->name); + write_file_buf(table_path.buf, "", 0); + + ref.update_index = 2; + err = reftable_stack_add(st, write_test_ref, &ref); + EXPECT_ERR(err); + EXPECT(st->merged->stack_len == 2); + EXPECT(st->stats.attempts == 1); + EXPECT(st->stats.failures == 1); + + reftable_stack_destroy(st); + strbuf_release(&table_path); + clear_dir(dir); +} + +static void test_reftable_stack_validate_refname(void) +{ + struct reftable_write_options cfg = { 0 }; + struct reftable_stack *st = NULL; + int err; + char *dir = get_tmp_dir(__LINE__); + + int i; + struct reftable_ref_record ref = { + .refname = "a/b", + .update_index = 1, + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "master", + }; + char *additions[] = { "a", "a/b/c" }; + + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + + err = reftable_stack_add(st, &write_test_ref, &ref); + EXPECT_ERR(err); + + for (i = 0; i < ARRAY_SIZE(additions); i++) { + struct reftable_ref_record ref = { + .refname = additions[i], + .update_index = 1, + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "master", + }; + + err = reftable_stack_add(st, &write_test_ref, &ref); + EXPECT(err == REFTABLE_NAME_CONFLICT); + } + + reftable_stack_destroy(st); + clear_dir(dir); +} + +static int write_error(struct reftable_writer *wr, void *arg) +{ + return *((int *)arg); +} + +static void test_reftable_stack_update_index_check(void) +{ + char *dir = get_tmp_dir(__LINE__); + + struct reftable_write_options cfg = { 0 }; + struct reftable_stack *st = NULL; + int err; + struct reftable_ref_record ref1 = { + .refname = "name1", + .update_index = 1, + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "master", + }; + struct reftable_ref_record ref2 = { + .refname = "name2", + .update_index = 1, + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "master", + }; + + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + + err = reftable_stack_add(st, &write_test_ref, &ref1); + EXPECT_ERR(err); + + err = reftable_stack_add(st, &write_test_ref, &ref2); + EXPECT(err == REFTABLE_API_ERROR); + reftable_stack_destroy(st); + clear_dir(dir); +} + +static void test_reftable_stack_lock_failure(void) +{ + char *dir = get_tmp_dir(__LINE__); + + struct reftable_write_options cfg = { 0 }; + struct reftable_stack *st = NULL; + int err, i; + + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + for (i = -1; i != REFTABLE_EMPTY_TABLE_ERROR; i--) { + err = reftable_stack_add(st, &write_error, &i); + EXPECT(err == i); + } + + reftable_stack_destroy(st); + clear_dir(dir); +} + +static void test_reftable_stack_add(void) +{ + int i = 0; + int err = 0; + struct reftable_write_options cfg = { + .exact_log_message = 1, + .default_permissions = 0660, + .disable_auto_compact = 1, + }; + struct reftable_stack *st = NULL; + char *dir = get_tmp_dir(__LINE__); + struct reftable_ref_record refs[2] = { { NULL } }; + struct reftable_log_record logs[2] = { { NULL } }; + struct strbuf path = STRBUF_INIT; + struct stat stat_result; + int N = ARRAY_SIZE(refs); + + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + + for (i = 0; i < N; i++) { + char buf[256]; + snprintf(buf, sizeof(buf), "branch%02d", i); + refs[i].refname = xstrdup(buf); + refs[i].update_index = i + 1; + refs[i].value_type = REFTABLE_REF_VAL1; + set_test_hash(refs[i].value.val1, i); + + logs[i].refname = xstrdup(buf); + logs[i].update_index = N + i + 1; + logs[i].value_type = REFTABLE_LOG_UPDATE; + logs[i].value.update.email = xstrdup("identity@invalid"); + set_test_hash(logs[i].value.update.new_hash, i); + } + + for (i = 0; i < N; i++) { + int err = reftable_stack_add(st, &write_test_ref, &refs[i]); + EXPECT_ERR(err); + } + + for (i = 0; i < N; i++) { + struct write_log_arg arg = { + .log = &logs[i], + .update_index = reftable_stack_next_update_index(st), + }; + int err = reftable_stack_add(st, &write_test_log, &arg); + EXPECT_ERR(err); + } + + err = reftable_stack_compact_all(st, NULL); + EXPECT_ERR(err); + + for (i = 0; i < N; i++) { + struct reftable_ref_record dest = { NULL }; + + int err = reftable_stack_read_ref(st, refs[i].refname, &dest); + EXPECT_ERR(err); + EXPECT(reftable_ref_record_equal(&dest, refs + i, + GIT_SHA1_RAWSZ)); + reftable_ref_record_release(&dest); + } + + for (i = 0; i < N; i++) { + struct reftable_log_record dest = { NULL }; + int err = reftable_stack_read_log(st, refs[i].refname, &dest); + EXPECT_ERR(err); + EXPECT(reftable_log_record_equal(&dest, logs + i, + GIT_SHA1_RAWSZ)); + reftable_log_record_release(&dest); + } + +#ifndef GIT_WINDOWS_NATIVE + strbuf_addstr(&path, dir); + strbuf_addstr(&path, "/tables.list"); + err = stat(path.buf, &stat_result); + EXPECT(!err); + EXPECT((stat_result.st_mode & 0777) == cfg.default_permissions); + + strbuf_reset(&path); + strbuf_addstr(&path, dir); + strbuf_addstr(&path, "/"); + /* do not try at home; not an external API for reftable. */ + strbuf_addstr(&path, st->readers[0]->name); + err = stat(path.buf, &stat_result); + EXPECT(!err); + EXPECT((stat_result.st_mode & 0777) == cfg.default_permissions); +#else + (void) stat_result; +#endif + + /* cleanup */ + reftable_stack_destroy(st); + for (i = 0; i < N; i++) { + reftable_ref_record_release(&refs[i]); + reftable_log_record_release(&logs[i]); + } + strbuf_release(&path); + clear_dir(dir); +} + +static void test_reftable_stack_log_normalize(void) +{ + int err = 0; + struct reftable_write_options cfg = { + 0, + }; + struct reftable_stack *st = NULL; + char *dir = get_tmp_dir(__LINE__); + struct reftable_log_record input = { + .refname = "branch", + .update_index = 1, + .value_type = REFTABLE_LOG_UPDATE, + .value = { + .update = { + .new_hash = { 1 }, + .old_hash = { 2 }, + }, + }, + }; + struct reftable_log_record dest = { + .update_index = 0, + }; + struct write_log_arg arg = { + .log = &input, + .update_index = 1, + }; + + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + + input.value.update.message = "one\ntwo"; + err = reftable_stack_add(st, &write_test_log, &arg); + EXPECT(err == REFTABLE_API_ERROR); + + input.value.update.message = "one"; + err = reftable_stack_add(st, &write_test_log, &arg); + EXPECT_ERR(err); + + err = reftable_stack_read_log(st, input.refname, &dest); + EXPECT_ERR(err); + EXPECT(0 == strcmp(dest.value.update.message, "one\n")); + + input.value.update.message = "two\n"; + arg.update_index = 2; + err = reftable_stack_add(st, &write_test_log, &arg); + EXPECT_ERR(err); + err = reftable_stack_read_log(st, input.refname, &dest); + EXPECT_ERR(err); + EXPECT(0 == strcmp(dest.value.update.message, "two\n")); + + /* cleanup */ + reftable_stack_destroy(st); + reftable_log_record_release(&dest); + clear_dir(dir); +} + +static void test_reftable_stack_tombstone(void) +{ + int i = 0; + char *dir = get_tmp_dir(__LINE__); + + struct reftable_write_options cfg = { 0 }; + struct reftable_stack *st = NULL; + int err; + struct reftable_ref_record refs[2] = { { NULL } }; + struct reftable_log_record logs[2] = { { NULL } }; + int N = ARRAY_SIZE(refs); + struct reftable_ref_record dest = { NULL }; + struct reftable_log_record log_dest = { NULL }; + + + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + + /* even entries add the refs, odd entries delete them. */ + for (i = 0; i < N; i++) { + const char *buf = "branch"; + refs[i].refname = xstrdup(buf); + refs[i].update_index = i + 1; + if (i % 2 == 0) { + refs[i].value_type = REFTABLE_REF_VAL1; + set_test_hash(refs[i].value.val1, i); + } + + logs[i].refname = xstrdup(buf); + /* update_index is part of the key. */ + logs[i].update_index = 42; + if (i % 2 == 0) { + logs[i].value_type = REFTABLE_LOG_UPDATE; + set_test_hash(logs[i].value.update.new_hash, i); + logs[i].value.update.email = + xstrdup("identity@invalid"); + } + } + for (i = 0; i < N; i++) { + int err = reftable_stack_add(st, &write_test_ref, &refs[i]); + EXPECT_ERR(err); + } + + for (i = 0; i < N; i++) { + struct write_log_arg arg = { + .log = &logs[i], + .update_index = reftable_stack_next_update_index(st), + }; + int err = reftable_stack_add(st, &write_test_log, &arg); + EXPECT_ERR(err); + } + + err = reftable_stack_read_ref(st, "branch", &dest); + EXPECT(err == 1); + reftable_ref_record_release(&dest); + + err = reftable_stack_read_log(st, "branch", &log_dest); + EXPECT(err == 1); + reftable_log_record_release(&log_dest); + + err = reftable_stack_compact_all(st, NULL); + EXPECT_ERR(err); + + err = reftable_stack_read_ref(st, "branch", &dest); + EXPECT(err == 1); + + err = reftable_stack_read_log(st, "branch", &log_dest); + EXPECT(err == 1); + reftable_ref_record_release(&dest); + reftable_log_record_release(&log_dest); + + /* cleanup */ + reftable_stack_destroy(st); + for (i = 0; i < N; i++) { + reftable_ref_record_release(&refs[i]); + reftable_log_record_release(&logs[i]); + } + clear_dir(dir); +} + +static void test_reftable_stack_hash_id(void) +{ + char *dir = get_tmp_dir(__LINE__); + + struct reftable_write_options cfg = { 0 }; + struct reftable_stack *st = NULL; + int err; + + struct reftable_ref_record ref = { + .refname = "master", + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "target", + .update_index = 1, + }; + struct reftable_write_options cfg32 = { .hash_id = GIT_SHA256_FORMAT_ID }; + struct reftable_stack *st32 = NULL; + struct reftable_write_options cfg_default = { 0 }; + struct reftable_stack *st_default = NULL; + struct reftable_ref_record dest = { NULL }; + + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + + err = reftable_stack_add(st, &write_test_ref, &ref); + EXPECT_ERR(err); + + /* can't read it with the wrong hash ID. */ + err = reftable_new_stack(&st32, dir, cfg32); + EXPECT(err == REFTABLE_FORMAT_ERROR); + + /* check that we can read it back with default config too. */ + err = reftable_new_stack(&st_default, dir, cfg_default); + EXPECT_ERR(err); + + err = reftable_stack_read_ref(st_default, "master", &dest); + EXPECT_ERR(err); + + EXPECT(reftable_ref_record_equal(&ref, &dest, GIT_SHA1_RAWSZ)); + reftable_ref_record_release(&dest); + reftable_stack_destroy(st); + reftable_stack_destroy(st_default); + clear_dir(dir); +} + +static void test_suggest_compaction_segment(void) +{ + uint64_t sizes[] = { 512, 64, 17, 16, 9, 9, 9, 16, 2, 16 }; + struct segment min = + suggest_compaction_segment(sizes, ARRAY_SIZE(sizes)); + EXPECT(min.start == 1); + EXPECT(min.end == 10); +} + +static void test_suggest_compaction_segment_nothing(void) +{ + uint64_t sizes[] = { 64, 32, 16, 8, 4, 2 }; + struct segment result = + suggest_compaction_segment(sizes, ARRAY_SIZE(sizes)); + EXPECT(result.start == result.end); +} + +static void test_reflog_expire(void) +{ + char *dir = get_tmp_dir(__LINE__); + + struct reftable_write_options cfg = { 0 }; + struct reftable_stack *st = NULL; + struct reftable_log_record logs[20] = { { NULL } }; + int N = ARRAY_SIZE(logs) - 1; + int i = 0; + int err; + struct reftable_log_expiry_config expiry = { + .time = 10, + }; + struct reftable_log_record log = { NULL }; + + + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + + for (i = 1; i <= N; i++) { + char buf[256]; + snprintf(buf, sizeof(buf), "branch%02d", i); + + logs[i].refname = xstrdup(buf); + logs[i].update_index = i; + logs[i].value_type = REFTABLE_LOG_UPDATE; + logs[i].value.update.time = i; + logs[i].value.update.email = xstrdup("identity@invalid"); + set_test_hash(logs[i].value.update.new_hash, i); + } + + for (i = 1; i <= N; i++) { + struct write_log_arg arg = { + .log = &logs[i], + .update_index = reftable_stack_next_update_index(st), + }; + int err = reftable_stack_add(st, &write_test_log, &arg); + EXPECT_ERR(err); + } + + err = reftable_stack_compact_all(st, NULL); + EXPECT_ERR(err); + + err = reftable_stack_compact_all(st, &expiry); + EXPECT_ERR(err); + + err = reftable_stack_read_log(st, logs[9].refname, &log); + EXPECT(err == 1); + + err = reftable_stack_read_log(st, logs[11].refname, &log); + EXPECT_ERR(err); + + expiry.min_update_index = 15; + err = reftable_stack_compact_all(st, &expiry); + EXPECT_ERR(err); + + err = reftable_stack_read_log(st, logs[14].refname, &log); + EXPECT(err == 1); + + err = reftable_stack_read_log(st, logs[16].refname, &log); + EXPECT_ERR(err); + + /* cleanup */ + reftable_stack_destroy(st); + for (i = 0; i <= N; i++) { + reftable_log_record_release(&logs[i]); + } + clear_dir(dir); + reftable_log_record_release(&log); +} + +static int write_nothing(struct reftable_writer *wr, void *arg) +{ + reftable_writer_set_limits(wr, 1, 1); + return 0; +} + +static void test_empty_add(void) +{ + struct reftable_write_options cfg = { 0 }; + struct reftable_stack *st = NULL; + int err; + char *dir = get_tmp_dir(__LINE__); + + struct reftable_stack *st2 = NULL; + + + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + + err = reftable_stack_add(st, &write_nothing, NULL); + EXPECT_ERR(err); + + err = reftable_new_stack(&st2, dir, cfg); + EXPECT_ERR(err); + clear_dir(dir); + reftable_stack_destroy(st); + reftable_stack_destroy(st2); +} + +static int fastlog2(uint64_t sz) +{ + int l = 0; + if (sz == 0) + return 0; + for (; sz; sz /= 2) + l++; + return l - 1; +} + +static void test_reftable_stack_auto_compaction(void) +{ + struct reftable_write_options cfg = { + .disable_auto_compact = 1, + }; + struct reftable_stack *st = NULL; + char *dir = get_tmp_dir(__LINE__); + + int err, i; + int N = 100; + + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + + for (i = 0; i < N; i++) { + char name[100]; + struct reftable_ref_record ref = { + .refname = name, + .update_index = reftable_stack_next_update_index(st), + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "master", + }; + snprintf(name, sizeof(name), "branch%04d", i); + + err = reftable_stack_add(st, &write_test_ref, &ref); + EXPECT_ERR(err); + + err = reftable_stack_auto_compact(st); + EXPECT_ERR(err); + EXPECT(i < 3 || st->merged->stack_len < 2 * fastlog2(i)); + } + + EXPECT(reftable_stack_compaction_stats(st)->entries_written < + (uint64_t)(N * fastlog2(N))); + + reftable_stack_destroy(st); + clear_dir(dir); +} + +static void test_reftable_stack_add_performs_auto_compaction(void) +{ + struct reftable_write_options cfg = { 0 }; + struct reftable_stack *st = NULL; + struct strbuf refname = STRBUF_INIT; + char *dir = get_tmp_dir(__LINE__); + int err, i, n = 20; + + err = reftable_new_stack(&st, dir, cfg); + EXPECT_ERR(err); + + for (i = 0; i <= n; i++) { + struct reftable_ref_record ref = { + .update_index = reftable_stack_next_update_index(st), + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "master", + }; + + /* + * Disable auto-compaction for all but the last runs. Like this + * we can ensure that we indeed honor this setting and have + * better control over when exactly auto compaction runs. + */ + st->config.disable_auto_compact = i != n; + + strbuf_reset(&refname); + strbuf_addf(&refname, "branch-%04d", i); + ref.refname = refname.buf; + + err = reftable_stack_add(st, &write_test_ref, &ref); + EXPECT_ERR(err); + + /* + * The stack length should grow continuously for all runs where + * auto compaction is disabled. When enabled, we should merge + * all tables in the stack. + */ + if (i != n) + EXPECT(st->merged->stack_len == i + 1); + else + EXPECT(st->merged->stack_len == 1); + } + + reftable_stack_destroy(st); + strbuf_release(&refname); + clear_dir(dir); +} + +static void test_reftable_stack_compaction_concurrent(void) +{ + struct reftable_write_options cfg = { 0 }; + struct reftable_stack *st1 = NULL, *st2 = NULL; + char *dir = get_tmp_dir(__LINE__); + + int err, i; + int N = 3; + + err = reftable_new_stack(&st1, dir, cfg); + EXPECT_ERR(err); + + for (i = 0; i < N; i++) { + char name[100]; + struct reftable_ref_record ref = { + .refname = name, + .update_index = reftable_stack_next_update_index(st1), + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "master", + }; + snprintf(name, sizeof(name), "branch%04d", i); + + err = reftable_stack_add(st1, &write_test_ref, &ref); + EXPECT_ERR(err); + } + + err = reftable_new_stack(&st2, dir, cfg); + EXPECT_ERR(err); + + err = reftable_stack_compact_all(st1, NULL); + EXPECT_ERR(err); + + reftable_stack_destroy(st1); + reftable_stack_destroy(st2); + + EXPECT(count_dir_entries(dir) == 2); + clear_dir(dir); +} + +static void unclean_stack_close(struct reftable_stack *st) +{ + /* break abstraction boundary to simulate unclean shutdown. */ + int i = 0; + for (; i < st->readers_len; i++) { + reftable_reader_free(st->readers[i]); + } + st->readers_len = 0; + FREE_AND_NULL(st->readers); +} + +static void test_reftable_stack_compaction_concurrent_clean(void) +{ + struct reftable_write_options cfg = { 0 }; + struct reftable_stack *st1 = NULL, *st2 = NULL, *st3 = NULL; + char *dir = get_tmp_dir(__LINE__); + + int err, i; + int N = 3; + + err = reftable_new_stack(&st1, dir, cfg); + EXPECT_ERR(err); + + for (i = 0; i < N; i++) { + char name[100]; + struct reftable_ref_record ref = { + .refname = name, + .update_index = reftable_stack_next_update_index(st1), + .value_type = REFTABLE_REF_SYMREF, + .value.symref = "master", + }; + snprintf(name, sizeof(name), "branch%04d", i); + + err = reftable_stack_add(st1, &write_test_ref, &ref); + EXPECT_ERR(err); + } + + err = reftable_new_stack(&st2, dir, cfg); + EXPECT_ERR(err); + + err = reftable_stack_compact_all(st1, NULL); + EXPECT_ERR(err); + + unclean_stack_close(st1); + unclean_stack_close(st2); + + err = reftable_new_stack(&st3, dir, cfg); + EXPECT_ERR(err); + + err = reftable_stack_clean(st3); + EXPECT_ERR(err); + EXPECT(count_dir_entries(dir) == 2); + + reftable_stack_destroy(st1); + reftable_stack_destroy(st2); + reftable_stack_destroy(st3); + + clear_dir(dir); +} + +int stack_test_main(int argc, const char *argv[]) +{ + RUN_TEST(test_empty_add); + RUN_TEST(test_names_equal); + RUN_TEST(test_parse_names); + RUN_TEST(test_read_file); + RUN_TEST(test_reflog_expire); + RUN_TEST(test_reftable_stack_add); + RUN_TEST(test_reftable_stack_add_one); + RUN_TEST(test_reftable_stack_auto_compaction); + RUN_TEST(test_reftable_stack_add_performs_auto_compaction); + RUN_TEST(test_reftable_stack_compaction_concurrent); + RUN_TEST(test_reftable_stack_compaction_concurrent_clean); + RUN_TEST(test_reftable_stack_hash_id); + RUN_TEST(test_reftable_stack_lock_failure); + RUN_TEST(test_reftable_stack_log_normalize); + RUN_TEST(test_reftable_stack_tombstone); + RUN_TEST(test_reftable_stack_transaction_api); + RUN_TEST(test_reftable_stack_transaction_api_performs_auto_compaction); + RUN_TEST(test_reftable_stack_auto_compaction_fails_gracefully); + RUN_TEST(test_reftable_stack_update_index_check); + RUN_TEST(test_reftable_stack_uptodate); + RUN_TEST(test_reftable_stack_validate_refname); + RUN_TEST(test_suggest_compaction_segment); + RUN_TEST(test_suggest_compaction_segment_nothing); + return 0; +} diff --git a/reftable/system.h b/reftable/system.h new file mode 100644 index 0000000..5d8b6de --- /dev/null +++ b/reftable/system.h @@ -0,0 +1,23 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef SYSTEM_H +#define SYSTEM_H + +/* This header glues the reftable library to the rest of Git */ + +#include "git-compat-util.h" +#include "lockfile.h" +#include "strbuf.h" +#include "tempfile.h" +#include "hash-ll.h" /* hash ID, sizes.*/ +#include "dir.h" /* remove_dir_recursively, for tests.*/ + +int hash_size(uint32_t id); + +#endif diff --git a/reftable/test_framework.c b/reftable/test_framework.c new file mode 100644 index 0000000..4066924 --- /dev/null +++ b/reftable/test_framework.c @@ -0,0 +1,27 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "system.h" +#include "test_framework.h" + + +void set_test_hash(uint8_t *p, int i) +{ + memset(p, (uint8_t)i, hash_size(GIT_SHA1_FORMAT_ID)); +} + +ssize_t strbuf_add_void(void *b, const void *data, size_t sz) +{ + strbuf_add(b, data, sz); + return sz; +} + +int noop_flush(void *arg) +{ + return 0; +} diff --git a/reftable/test_framework.h b/reftable/test_framework.h new file mode 100644 index 0000000..687390f --- /dev/null +++ b/reftable/test_framework.h @@ -0,0 +1,61 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef TEST_FRAMEWORK_H +#define TEST_FRAMEWORK_H + +#include "system.h" +#include "reftable-error.h" + +#define EXPECT_ERR(c) \ + do { \ + if (c != 0) { \ + fflush(stderr); \ + fflush(stdout); \ + fprintf(stderr, "%s: %d: error == %d (%s), want 0\n", \ + __FILE__, __LINE__, c, reftable_error_str(c)); \ + abort(); \ + } \ + } while (0) + +#define EXPECT_STREQ(a, b) \ + do { \ + if (strcmp(a, b)) { \ + fflush(stderr); \ + fflush(stdout); \ + fprintf(stderr, "%s:%d: %s (%s) != %s (%s)\n", __FILE__, \ + __LINE__, #a, a, #b, b); \ + abort(); \ + } \ + } while (0) + +#define EXPECT(c) \ + do { \ + if (!(c)) { \ + fflush(stderr); \ + fflush(stdout); \ + fprintf(stderr, "%s: %d: failed assertion %s\n", __FILE__, \ + __LINE__, #c); \ + abort(); \ + } \ + } while (0) + +#define RUN_TEST(f) \ + fprintf(stderr, "running %s\n", #f); \ + fflush(stderr); \ + f(); + +void set_test_hash(uint8_t *p, int i); + +/* Like strbuf_add, but suitable for passing to reftable_new_writer + */ +ssize_t strbuf_add_void(void *b, const void *data, size_t sz); + +int noop_flush(void *); + +#endif diff --git a/reftable/tree.c b/reftable/tree.c new file mode 100644 index 0000000..528f33a --- /dev/null +++ b/reftable/tree.c @@ -0,0 +1,63 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "system.h" +#include "tree.h" + +#include "basics.h" + +struct tree_node *tree_search(void *key, struct tree_node **rootp, + int (*compare)(const void *, const void *), + int insert) +{ + int res; + if (!*rootp) { + if (!insert) { + return NULL; + } else { + struct tree_node *n; + REFTABLE_CALLOC_ARRAY(n, 1); + n->key = key; + *rootp = n; + return *rootp; + } + } + + res = compare(key, (*rootp)->key); + if (res < 0) + return tree_search(key, &(*rootp)->left, compare, insert); + else if (res > 0) + return tree_search(key, &(*rootp)->right, compare, insert); + return *rootp; +} + +void infix_walk(struct tree_node *t, void (*action)(void *arg, void *key), + void *arg) +{ + if (t->left) { + infix_walk(t->left, action, arg); + } + action(arg, t->key); + if (t->right) { + infix_walk(t->right, action, arg); + } +} + +void tree_free(struct tree_node *t) +{ + if (!t) { + return; + } + if (t->left) { + tree_free(t->left); + } + if (t->right) { + tree_free(t->right); + } + reftable_free(t); +} diff --git a/reftable/tree.h b/reftable/tree.h new file mode 100644 index 0000000..fbdd002 --- /dev/null +++ b/reftable/tree.h @@ -0,0 +1,34 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef TREE_H +#define TREE_H + +/* tree_node is a generic binary search tree. */ +struct tree_node { + void *key; + struct tree_node *left, *right; +}; + +/* looks for `key` in `rootp` using `compare` as comparison function. If insert + * is set, insert the key if it's not found. Else, return NULL. + */ +struct tree_node *tree_search(void *key, struct tree_node **rootp, + int (*compare)(const void *, const void *), + int insert); + +/* performs an infix walk of the tree. */ +void infix_walk(struct tree_node *t, void (*action)(void *arg, void *key), + void *arg); + +/* + * deallocates the tree nodes recursively. Keys should be deallocated separately + * by walking over the tree. */ +void tree_free(struct tree_node *t); + +#endif diff --git a/reftable/tree_test.c b/reftable/tree_test.c new file mode 100644 index 0000000..6961a65 --- /dev/null +++ b/reftable/tree_test.c @@ -0,0 +1,60 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "system.h" +#include "tree.h" + +#include "test_framework.h" +#include "reftable-tests.h" + +static int test_compare(const void *a, const void *b) +{ + return (char *)a - (char *)b; +} + +struct curry { + void *last; +}; + +static void check_increasing(void *arg, void *key) +{ + struct curry *c = arg; + if (c->last) { + EXPECT(test_compare(c->last, key) < 0); + } + c->last = key; +} + +static void test_tree(void) +{ + struct tree_node *root = NULL; + + void *values[11] = { NULL }; + struct tree_node *nodes[11] = { NULL }; + int i = 1; + struct curry c = { NULL }; + do { + nodes[i] = tree_search(values + i, &root, &test_compare, 1); + i = (i * 7) % 11; + } while (i != 1); + + for (i = 1; i < ARRAY_SIZE(nodes); i++) { + EXPECT(values + i == nodes[i]->key); + EXPECT(nodes[i] == + tree_search(values + i, &root, &test_compare, 0)); + } + + infix_walk(root, check_increasing, &c); + tree_free(root); +} + +int tree_test_main(int argc, const char *argv[]) +{ + RUN_TEST(test_tree); + return 0; +} diff --git a/reftable/writer.c b/reftable/writer.c new file mode 100644 index 0000000..1d9ff0f --- /dev/null +++ b/reftable/writer.c @@ -0,0 +1,713 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#include "writer.h" + +#include "system.h" + +#include "block.h" +#include "constants.h" +#include "record.h" +#include "tree.h" +#include "reftable-error.h" + +/* finishes a block, and writes it to storage */ +static int writer_flush_block(struct reftable_writer *w); + +/* deallocates memory related to the index */ +static void writer_clear_index(struct reftable_writer *w); + +/* finishes writing a 'r' (refs) or 'g' (reflogs) section */ +static int writer_finish_public_section(struct reftable_writer *w); + +static struct reftable_block_stats * +writer_reftable_block_stats(struct reftable_writer *w, uint8_t typ) +{ + switch (typ) { + case 'r': + return &w->stats.ref_stats; + case 'o': + return &w->stats.obj_stats; + case 'i': + return &w->stats.idx_stats; + case 'g': + return &w->stats.log_stats; + } + abort(); + return NULL; +} + +/* write data, queuing the padding for the next write. Returns negative for + * error. */ +static int padded_write(struct reftable_writer *w, uint8_t *data, size_t len, + int padding) +{ + int n = 0; + if (w->pending_padding > 0) { + uint8_t *zeroed = reftable_calloc(w->pending_padding, sizeof(*zeroed)); + int n = w->write(w->write_arg, zeroed, w->pending_padding); + if (n < 0) + return n; + + w->pending_padding = 0; + reftable_free(zeroed); + } + + w->pending_padding = padding; + n = w->write(w->write_arg, data, len); + if (n < 0) + return n; + n += padding; + return 0; +} + +static void options_set_defaults(struct reftable_write_options *opts) +{ + if (opts->restart_interval == 0) { + opts->restart_interval = 16; + } + + if (opts->hash_id == 0) { + opts->hash_id = GIT_SHA1_FORMAT_ID; + } + if (opts->block_size == 0) { + opts->block_size = DEFAULT_BLOCK_SIZE; + } +} + +static int writer_version(struct reftable_writer *w) +{ + return (w->opts.hash_id == 0 || w->opts.hash_id == GIT_SHA1_FORMAT_ID) ? + 1 : + 2; +} + +static int writer_write_header(struct reftable_writer *w, uint8_t *dest) +{ + memcpy(dest, "REFT", 4); + + dest[4] = writer_version(w); + + put_be24(dest + 5, w->opts.block_size); + put_be64(dest + 8, w->min_update_index); + put_be64(dest + 16, w->max_update_index); + if (writer_version(w) == 2) { + put_be32(dest + 24, w->opts.hash_id); + } + return header_size(writer_version(w)); +} + +static void writer_reinit_block_writer(struct reftable_writer *w, uint8_t typ) +{ + int block_start = 0; + if (w->next == 0) { + block_start = header_size(writer_version(w)); + } + + strbuf_release(&w->last_key); + block_writer_init(&w->block_writer_data, typ, w->block, + w->opts.block_size, block_start, + hash_size(w->opts.hash_id)); + w->block_writer = &w->block_writer_data; + w->block_writer->restart_interval = w->opts.restart_interval; +} + +static struct strbuf reftable_empty_strbuf = STRBUF_INIT; + +struct reftable_writer * +reftable_new_writer(ssize_t (*writer_func)(void *, const void *, size_t), + int (*flush_func)(void *), + void *writer_arg, struct reftable_write_options *opts) +{ + struct reftable_writer *wp = reftable_calloc(1, sizeof(*wp)); + strbuf_init(&wp->block_writer_data.last_key, 0); + options_set_defaults(opts); + if (opts->block_size >= (1 << 24)) { + /* TODO - error return? */ + abort(); + } + wp->last_key = reftable_empty_strbuf; + REFTABLE_CALLOC_ARRAY(wp->block, opts->block_size); + wp->write = writer_func; + wp->write_arg = writer_arg; + wp->opts = *opts; + wp->flush = flush_func; + writer_reinit_block_writer(wp, BLOCK_TYPE_REF); + + return wp; +} + +void reftable_writer_set_limits(struct reftable_writer *w, uint64_t min, + uint64_t max) +{ + w->min_update_index = min; + w->max_update_index = max; +} + +void reftable_writer_free(struct reftable_writer *w) +{ + if (!w) + return; + reftable_free(w->block); + reftable_free(w); +} + +struct obj_index_tree_node { + struct strbuf hash; + uint64_t *offsets; + size_t offset_len; + size_t offset_cap; +}; + +#define OBJ_INDEX_TREE_NODE_INIT \ + { \ + .hash = STRBUF_INIT \ + } + +static int obj_index_tree_node_compare(const void *a, const void *b) +{ + return strbuf_cmp(&((const struct obj_index_tree_node *)a)->hash, + &((const struct obj_index_tree_node *)b)->hash); +} + +static void writer_index_hash(struct reftable_writer *w, struct strbuf *hash) +{ + uint64_t off = w->next; + + struct obj_index_tree_node want = { .hash = *hash }; + + struct tree_node *node = tree_search(&want, &w->obj_index_tree, + &obj_index_tree_node_compare, 0); + struct obj_index_tree_node *key = NULL; + if (!node) { + struct obj_index_tree_node empty = OBJ_INDEX_TREE_NODE_INIT; + key = reftable_malloc(sizeof(struct obj_index_tree_node)); + *key = empty; + + strbuf_reset(&key->hash); + strbuf_addbuf(&key->hash, hash); + tree_search((void *)key, &w->obj_index_tree, + &obj_index_tree_node_compare, 1); + } else { + key = node->key; + } + + if (key->offset_len > 0 && key->offsets[key->offset_len - 1] == off) { + return; + } + + REFTABLE_ALLOC_GROW(key->offsets, key->offset_len + 1, key->offset_cap); + key->offsets[key->offset_len++] = off; +} + +static int writer_add_record(struct reftable_writer *w, + struct reftable_record *rec) +{ + struct strbuf key = STRBUF_INIT; + int err = -1; + reftable_record_key(rec, &key); + if (strbuf_cmp(&w->last_key, &key) >= 0) { + err = REFTABLE_API_ERROR; + goto done; + } + + strbuf_reset(&w->last_key); + strbuf_addbuf(&w->last_key, &key); + if (!w->block_writer) { + writer_reinit_block_writer(w, reftable_record_type(rec)); + } + + assert(block_writer_type(w->block_writer) == reftable_record_type(rec)); + + if (block_writer_add(w->block_writer, rec) == 0) { + err = 0; + goto done; + } + + err = writer_flush_block(w); + if (err < 0) { + goto done; + } + + writer_reinit_block_writer(w, reftable_record_type(rec)); + err = block_writer_add(w->block_writer, rec); + if (err == -1) { + /* we are writing into memory, so an error can only mean it + * doesn't fit. */ + err = REFTABLE_ENTRY_TOO_BIG_ERROR; + goto done; + } + +done: + strbuf_release(&key); + return err; +} + +int reftable_writer_add_ref(struct reftable_writer *w, + struct reftable_ref_record *ref) +{ + struct reftable_record rec = { + .type = BLOCK_TYPE_REF, + .u = { + .ref = *ref + }, + }; + int err = 0; + + if (!ref->refname) + return REFTABLE_API_ERROR; + if (ref->update_index < w->min_update_index || + ref->update_index > w->max_update_index) + return REFTABLE_API_ERROR; + + rec.u.ref.update_index -= w->min_update_index; + + err = writer_add_record(w, &rec); + if (err < 0) + return err; + + if (!w->opts.skip_index_objects && reftable_ref_record_val1(ref)) { + struct strbuf h = STRBUF_INIT; + strbuf_add(&h, (char *)reftable_ref_record_val1(ref), + hash_size(w->opts.hash_id)); + writer_index_hash(w, &h); + strbuf_release(&h); + } + + if (!w->opts.skip_index_objects && reftable_ref_record_val2(ref)) { + struct strbuf h = STRBUF_INIT; + strbuf_add(&h, reftable_ref_record_val2(ref), + hash_size(w->opts.hash_id)); + writer_index_hash(w, &h); + strbuf_release(&h); + } + return 0; +} + +int reftable_writer_add_refs(struct reftable_writer *w, + struct reftable_ref_record *refs, int n) +{ + int err = 0; + int i = 0; + QSORT(refs, n, reftable_ref_record_compare_name); + for (i = 0; err == 0 && i < n; i++) { + err = reftable_writer_add_ref(w, &refs[i]); + } + return err; +} + +static int reftable_writer_add_log_verbatim(struct reftable_writer *w, + struct reftable_log_record *log) +{ + struct reftable_record rec = { + .type = BLOCK_TYPE_LOG, + .u = { + .log = *log, + }, + }; + if (w->block_writer && + block_writer_type(w->block_writer) == BLOCK_TYPE_REF) { + int err = writer_finish_public_section(w); + if (err < 0) + return err; + } + + w->next -= w->pending_padding; + w->pending_padding = 0; + return writer_add_record(w, &rec); +} + +int reftable_writer_add_log(struct reftable_writer *w, + struct reftable_log_record *log) +{ + char *input_log_message = NULL; + struct strbuf cleaned_message = STRBUF_INIT; + int err = 0; + + if (log->value_type == REFTABLE_LOG_DELETION) + return reftable_writer_add_log_verbatim(w, log); + + if (!log->refname) + return REFTABLE_API_ERROR; + + input_log_message = log->value.update.message; + if (!w->opts.exact_log_message && log->value.update.message) { + strbuf_addstr(&cleaned_message, log->value.update.message); + while (cleaned_message.len && + cleaned_message.buf[cleaned_message.len - 1] == '\n') + strbuf_setlen(&cleaned_message, + cleaned_message.len - 1); + if (strchr(cleaned_message.buf, '\n')) { + /* multiple lines not allowed. */ + err = REFTABLE_API_ERROR; + goto done; + } + strbuf_addstr(&cleaned_message, "\n"); + log->value.update.message = cleaned_message.buf; + } + + err = reftable_writer_add_log_verbatim(w, log); + log->value.update.message = input_log_message; +done: + strbuf_release(&cleaned_message); + return err; +} + +int reftable_writer_add_logs(struct reftable_writer *w, + struct reftable_log_record *logs, int n) +{ + int err = 0; + int i = 0; + QSORT(logs, n, reftable_log_record_compare_key); + + for (i = 0; err == 0 && i < n; i++) { + err = reftable_writer_add_log(w, &logs[i]); + } + return err; +} + +static int writer_finish_section(struct reftable_writer *w) +{ + struct reftable_block_stats *bstats = NULL; + uint8_t typ = block_writer_type(w->block_writer); + uint64_t index_start = 0; + int max_level = 0; + size_t threshold = w->opts.unpadded ? 1 : 3; + int before_blocks = w->stats.idx_stats.blocks; + int err; + + err = writer_flush_block(w); + if (err < 0) + return err; + + /* + * When the section we are about to index has a lot of blocks then the + * index itself may span across multiple blocks, as well. This would + * require a linear scan over index blocks only to find the desired + * indexed block, which is inefficient. Instead, we write a multi-level + * index where index records of level N+1 will refer to index blocks of + * level N. This isn't constant time, either, but at least logarithmic. + * + * This loop handles writing this multi-level index. Note that we write + * the lowest-level index pointing to the indexed blocks first. We then + * continue writing additional index levels until the current level has + * less blocks than the threshold so that the highest level will be at + * the end of the index section. + * + * Readers are thus required to start reading the index section from + * its end, which is why we set `index_start` to the beginning of the + * last index section. + */ + while (w->index_len > threshold) { + struct reftable_index_record *idx = NULL; + size_t i, idx_len; + + max_level++; + index_start = w->next; + writer_reinit_block_writer(w, BLOCK_TYPE_INDEX); + + idx = w->index; + idx_len = w->index_len; + + w->index = NULL; + w->index_len = 0; + w->index_cap = 0; + for (i = 0; i < idx_len; i++) { + struct reftable_record rec = { + .type = BLOCK_TYPE_INDEX, + .u = { + .idx = idx[i], + }, + }; + + err = writer_add_record(w, &rec); + if (err < 0) + return err; + } + + err = writer_flush_block(w); + if (err < 0) + return err; + + for (i = 0; i < idx_len; i++) + strbuf_release(&idx[i].last_key); + reftable_free(idx); + } + + /* + * The index may still contain a number of index blocks lower than the + * threshold. Clear it so that these entries don't leak into the next + * index section. + */ + writer_clear_index(w); + + bstats = writer_reftable_block_stats(w, typ); + bstats->index_blocks = w->stats.idx_stats.blocks - before_blocks; + bstats->index_offset = index_start; + bstats->max_index_level = max_level; + + /* Reinit lastKey, as the next section can start with any key. */ + w->last_key.len = 0; + + return 0; +} + +struct common_prefix_arg { + struct strbuf *last; + int max; +}; + +static void update_common(void *void_arg, void *key) +{ + struct common_prefix_arg *arg = void_arg; + struct obj_index_tree_node *entry = key; + if (arg->last) { + int n = common_prefix_size(&entry->hash, arg->last); + if (n > arg->max) { + arg->max = n; + } + } + arg->last = &entry->hash; +} + +struct write_record_arg { + struct reftable_writer *w; + int err; +}; + +static void write_object_record(void *void_arg, void *key) +{ + struct write_record_arg *arg = void_arg; + struct obj_index_tree_node *entry = key; + struct reftable_record + rec = { .type = BLOCK_TYPE_OBJ, + .u.obj = { + .hash_prefix = (uint8_t *)entry->hash.buf, + .hash_prefix_len = arg->w->stats.object_id_len, + .offsets = entry->offsets, + .offset_len = entry->offset_len, + } }; + if (arg->err < 0) + goto done; + + arg->err = block_writer_add(arg->w->block_writer, &rec); + if (arg->err == 0) + goto done; + + arg->err = writer_flush_block(arg->w); + if (arg->err < 0) + goto done; + + writer_reinit_block_writer(arg->w, BLOCK_TYPE_OBJ); + arg->err = block_writer_add(arg->w->block_writer, &rec); + if (arg->err == 0) + goto done; + + rec.u.obj.offset_len = 0; + arg->err = block_writer_add(arg->w->block_writer, &rec); + + /* Should be able to write into a fresh block. */ + assert(arg->err == 0); + +done:; +} + +static void object_record_free(void *void_arg, void *key) +{ + struct obj_index_tree_node *entry = key; + + FREE_AND_NULL(entry->offsets); + strbuf_release(&entry->hash); + reftable_free(entry); +} + +static int writer_dump_object_index(struct reftable_writer *w) +{ + struct write_record_arg closure = { .w = w }; + struct common_prefix_arg common = { + .max = 1, /* obj_id_len should be >= 2. */ + }; + if (w->obj_index_tree) { + infix_walk(w->obj_index_tree, &update_common, &common); + } + w->stats.object_id_len = common.max + 1; + + writer_reinit_block_writer(w, BLOCK_TYPE_OBJ); + + if (w->obj_index_tree) { + infix_walk(w->obj_index_tree, &write_object_record, &closure); + } + + if (closure.err < 0) + return closure.err; + return writer_finish_section(w); +} + +static int writer_finish_public_section(struct reftable_writer *w) +{ + uint8_t typ = 0; + int err = 0; + + if (!w->block_writer) + return 0; + + typ = block_writer_type(w->block_writer); + err = writer_finish_section(w); + if (err < 0) + return err; + if (typ == BLOCK_TYPE_REF && !w->opts.skip_index_objects && + w->stats.ref_stats.index_blocks > 0) { + err = writer_dump_object_index(w); + if (err < 0) + return err; + } + + if (w->obj_index_tree) { + infix_walk(w->obj_index_tree, &object_record_free, NULL); + tree_free(w->obj_index_tree); + w->obj_index_tree = NULL; + } + + w->block_writer = NULL; + return 0; +} + +int reftable_writer_close(struct reftable_writer *w) +{ + uint8_t footer[72]; + uint8_t *p = footer; + int err = writer_finish_public_section(w); + int empty_table = w->next == 0; + if (err != 0) + goto done; + w->pending_padding = 0; + if (empty_table) { + /* Empty tables need a header anyway. */ + uint8_t header[28]; + int n = writer_write_header(w, header); + err = padded_write(w, header, n, 0); + if (err < 0) + goto done; + } + + p += writer_write_header(w, footer); + put_be64(p, w->stats.ref_stats.index_offset); + p += 8; + put_be64(p, (w->stats.obj_stats.offset) << 5 | w->stats.object_id_len); + p += 8; + put_be64(p, w->stats.obj_stats.index_offset); + p += 8; + + put_be64(p, w->stats.log_stats.offset); + p += 8; + put_be64(p, w->stats.log_stats.index_offset); + p += 8; + + put_be32(p, crc32(0, footer, p - footer)); + p += 4; + + err = w->flush(w->write_arg); + if (err < 0) { + err = REFTABLE_IO_ERROR; + goto done; + } + + err = padded_write(w, footer, footer_size(writer_version(w)), 0); + if (err < 0) + goto done; + + if (empty_table) { + err = REFTABLE_EMPTY_TABLE_ERROR; + goto done; + } + +done: + /* free up memory. */ + block_writer_release(&w->block_writer_data); + writer_clear_index(w); + strbuf_release(&w->last_key); + return err; +} + +static void writer_clear_index(struct reftable_writer *w) +{ + for (size_t i = 0; i < w->index_len; i++) + strbuf_release(&w->index[i].last_key); + FREE_AND_NULL(w->index); + w->index_len = 0; + w->index_cap = 0; +} + +static const int debug = 0; + +static int writer_flush_nonempty_block(struct reftable_writer *w) +{ + uint8_t typ = block_writer_type(w->block_writer); + struct reftable_block_stats *bstats = + writer_reftable_block_stats(w, typ); + uint64_t block_typ_off = (bstats->blocks == 0) ? w->next : 0; + int raw_bytes = block_writer_finish(w->block_writer); + int padding = 0; + int err = 0; + struct reftable_index_record ir = { .last_key = STRBUF_INIT }; + if (raw_bytes < 0) + return raw_bytes; + + if (!w->opts.unpadded && typ != BLOCK_TYPE_LOG) { + padding = w->opts.block_size - raw_bytes; + } + + if (block_typ_off > 0) { + bstats->offset = block_typ_off; + } + + bstats->entries += w->block_writer->entries; + bstats->restarts += w->block_writer->restart_len; + bstats->blocks++; + w->stats.blocks++; + + if (debug) { + fprintf(stderr, "block %c off %" PRIu64 " sz %d (%d)\n", typ, + w->next, raw_bytes, + get_be24(w->block + w->block_writer->header_off + 1)); + } + + if (w->next == 0) { + writer_write_header(w, w->block); + } + + err = padded_write(w, w->block, raw_bytes, padding); + if (err < 0) + return err; + + REFTABLE_ALLOC_GROW(w->index, w->index_len + 1, w->index_cap); + + ir.offset = w->next; + strbuf_reset(&ir.last_key); + strbuf_addbuf(&ir.last_key, &w->block_writer->last_key); + w->index[w->index_len] = ir; + + w->index_len++; + w->next += padding + raw_bytes; + w->block_writer = NULL; + return 0; +} + +static int writer_flush_block(struct reftable_writer *w) +{ + if (!w->block_writer) + return 0; + if (w->block_writer->entries == 0) + return 0; + return writer_flush_nonempty_block(w); +} + +const struct reftable_stats *reftable_writer_stats(struct reftable_writer *w) +{ + return &w->stats; +} diff --git a/reftable/writer.h b/reftable/writer.h new file mode 100644 index 0000000..8d0df9c --- /dev/null +++ b/reftable/writer.h @@ -0,0 +1,51 @@ +/* +Copyright 2020 Google LLC + +Use of this source code is governed by a BSD-style +license that can be found in the LICENSE file or at +https://developers.google.com/open-source/licenses/bsd +*/ + +#ifndef WRITER_H +#define WRITER_H + +#include "basics.h" +#include "block.h" +#include "tree.h" +#include "reftable-writer.h" + +struct reftable_writer { + ssize_t (*write)(void *, const void *, size_t); + int (*flush)(void *); + void *write_arg; + int pending_padding; + struct strbuf last_key; + + /* offset of next block to write. */ + uint64_t next; + uint64_t min_update_index, max_update_index; + struct reftable_write_options opts; + + /* memory buffer for writing */ + uint8_t *block; + + /* writer for the current section. NULL or points to + * block_writer_data */ + struct block_writer *block_writer; + + struct block_writer block_writer_data; + + /* pending index records for the current section */ + struct reftable_index_record *index; + size_t index_len; + size_t index_cap; + + /* + * tree for use with tsearch; used to populate the 'o' inverse OID + * map */ + struct tree_node *obj_index_tree; + + struct reftable_stats stats; +}; + +#endif |