feat: P2-2 apply_wal_entry — standby WAL replay with LSN advance and type guards

Upgrade apply_wal_entry() from P0 static no-op stub to the real public
standby replay path:

- Reject oversized payloads (> MAX_WAL_PAYLOAD_BYTES) with -EMSGSIZE
- Skip WAL_CHECKPOINT (100) entries silently — local-only markers, never
  replicated to standbys; current_lsn is NOT advanced on skip
- Skip unknown types (> WAL_ENTRY_TYPE_MAX) with WARN log and rc=0 for
  forward compatibility with WAL written by newer controller versions
- Advance current_lsn to hdr->lsn+1 and update last_crc under p->lock
  for all accepted known entries
- Gate restore_job()/worker_from_capnp() calls on p->cluster != NULL so
  unit tests work without a full cluster wired in
- Add NULL guard returning -EINVAL for missing p or hdr

Add public declaration in persist.h with forward-declare stubs for
restore_job/worker_from_capnp in persist.c (avoids heavy cluster.h chain).

Test suite: 5 new Unity tests in test_repl_apply.c covering all branches.
Weak stubs for restore_job/worker_from_capnp in the test file satisfy the
linker without linking job.c/worker.c's heavy dependency graph.
All 5 new tests pass; full controller suite green.
This commit is contained in:
Rongsong Shen 2026-05-07 16:38:20 +00:00
parent 8ef64a7ee7
commit abefc8a128
5 changed files with 275 additions and 7 deletions

View file

@ -0,0 +1,197 @@
/*
* test_repl_apply.c — P2-2: Tests for apply_wal_entry() standby WAL replay.
*
* TDD RED phase: written BEFORE the implementation.
* These tests verify:
* - LSN advances past replayed entry
* - WAL_CHECKPOINT entries are skipped (rc=0, lsn unchanged)
* - Oversized payloads are rejected (rc != 0)
* - Unknown types (> WAL_ENTRY_TYPE_MAX) are skipped with rc=0
*/
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unity/unity.h>
#include "config.h"
#include "persist.h"
/* ------------------------------------------------------------------ */
/* Weak stubs for P0 entity-restore functions (job.c / worker.c). */
/* The unit test binary does not link those translation units — they */
/* pull in the full cluster.h / netio.h chain. apply_wal_entry() only */
/* calls these when p->cluster != NULL; tests use NULL cluster, so the */
/* weak stubs are never reached. The linker needs the symbols to link. */
/* ------------------------------------------------------------------ */
typedef struct cluster_s cluster_t;
__attribute__((weak))
int restore_job(cluster_t* c, uint64_t jobid,
const void* capnp_payload, size_t len)
{
(void)c; (void)jobid; (void)capnp_payload; (void)len;
return 0;
}
__attribute__((weak))
int worker_from_capnp(cluster_t* c, const void* capnp_payload, size_t len)
{
(void)c; (void)capnp_payload; (void)len;
return 0;
}
/* ------------------------------------------------------------------ */
/* Test helpers */
/* ------------------------------------------------------------------ */
static persist_t* make_test_persist_for_apply(void)
{
persist_t* p = calloc(1, sizeof(persist_t));
if (!p) {
fprintf(stderr, "calloc failed\n");
exit(1);
}
persist_config_t cfg = persist_config_defaults();
cfg.wal_sync_ms = 5000; /* slow fsync — we don't need it in unit tests */
char tmpdir[256];
snprintf(tmpdir, sizeof(tmpdir), "/tmp/xce_apply_XXXXXX");
if (!mkdtemp(tmpdir)) {
fprintf(stderr, "mkdtemp failed\n");
free(p);
exit(1);
}
persist_set_workdir(p, tmpdir, "ctrl-apply");
persist_init(p, &cfg, NULL); /* NULL cluster — entity stubs gate on p->cluster */
persist_start(p);
return p;
}
static void destroy_test_persist(persist_t* p)
{
persist_stop(p);
persist_fini(p);
free(p);
}
/* ------------------------------------------------------------------ */
/* Tests */
/* ------------------------------------------------------------------ */
static void test_apply_wal_entry_advances_lsn(void)
{
persist_t* p = make_test_persist_for_apply();
wal_entry_hdr_t hdr;
memset(&hdr, 0, sizeof(hdr));
hdr.type = WAL_JOB_UPDATE;
hdr.lsn = 42;
hdr.term = 1;
hdr.payload_len = 4;
hdr.crc = 0xDEADBEEF;
hdr.prev_crc = WAL_CRC_NONE;
uint8_t payload[4] = {1, 2, 3, 4};
int rc = apply_wal_entry(p, &hdr, payload);
TEST_ASSERT_EQUAL_INT(0, rc);
/* current_lsn must have advanced to at least lsn+1 */
TEST_ASSERT_TRUE(p->current_lsn > 42);
/* last_crc must reflect the entry's crc */
TEST_ASSERT_EQUAL_UINT32(0xDEADBEEF, p->last_crc);
destroy_test_persist(p);
printf("test_apply_wal_entry_advances_lsn: PASS\n");
}
static void test_apply_wal_entry_skips_checkpoint(void)
{
persist_t* p = make_test_persist_for_apply();
uint64_t lsn_before = p->current_lsn;
wal_entry_hdr_t hdr;
memset(&hdr, 0, sizeof(hdr));
hdr.type = WAL_CHECKPOINT;
hdr.lsn = 100;
hdr.term = 1;
hdr.payload_len = 0;
int rc = apply_wal_entry(p, &hdr, NULL);
TEST_ASSERT_EQUAL_INT(0, rc); /* skip = success, not error */
/* Checkpoint entries are local-only markers — standbys never receive them.
* current_lsn must NOT advance: skip means no-op. */
TEST_ASSERT_EQUAL_UINT64(lsn_before, p->current_lsn);
destroy_test_persist(p);
printf("test_apply_wal_entry_skips_checkpoint: PASS\n");
}
static void test_apply_wal_entry_rejects_oversized(void)
{
persist_t* p = make_test_persist_for_apply();
wal_entry_hdr_t hdr;
memset(&hdr, 0, sizeof(hdr));
hdr.type = WAL_JOB_SUBMIT;
hdr.lsn = 1;
hdr.term = 1;
hdr.payload_len = MAX_WAL_PAYLOAD_BYTES + 1; /* too large */
int rc = apply_wal_entry(p, &hdr, NULL);
TEST_ASSERT_NOT_EQUAL(0, rc); /* must be rejected */
destroy_test_persist(p);
printf("test_apply_wal_entry_rejects_oversized: PASS\n");
}
static void test_apply_wal_entry_skips_unknown_type(void)
{
persist_t* p = make_test_persist_for_apply();
uint64_t lsn_before = p->current_lsn;
/*
* Type 200 is above WAL_ENTRY_TYPE_MAX (100).
* Spec: log WARN and skip with rc=0 (forward compatibility).
*/
wal_entry_hdr_t hdr;
memset(&hdr, 0, sizeof(hdr));
hdr.type = 200;
hdr.lsn = 5;
hdr.term = 1;
hdr.payload_len = 0;
int rc = apply_wal_entry(p, &hdr, NULL);
TEST_ASSERT_EQUAL_INT(0, rc); /* unknown type: skip not error */
/* LSN must NOT advance for a skipped-unknown type */
TEST_ASSERT_EQUAL_UINT64(lsn_before, p->current_lsn);
destroy_test_persist(p);
printf("test_apply_wal_entry_skips_unknown_type: PASS\n");
}
static void test_apply_wal_entry_null_guard(void)
{
/* Both NULL inputs must return non-zero (EINVAL) */
int rc = apply_wal_entry(NULL, NULL, NULL);
TEST_ASSERT_NOT_EQUAL(0, rc);
printf("test_apply_wal_entry_null_guard: PASS\n");
}
/* ------------------------------------------------------------------ */
/* Suite entry point (registered in controller_test.c) */
/* ------------------------------------------------------------------ */
void test_repl_apply(void)
{
RUN_TEST(test_apply_wal_entry_advances_lsn);
RUN_TEST(test_apply_wal_entry_skips_checkpoint);
RUN_TEST(test_apply_wal_entry_rejects_oversized);
RUN_TEST(test_apply_wal_entry_skips_unknown_type);
RUN_TEST(test_apply_wal_entry_null_guard);
}