/*
* CREATE FUNCTION utl_file.fcopy(
* src_location text,
* src_filename text,
* dest_location text,
* dest_filename text,
* start_line integer DEFAULT NULL
* end_line integer DEFAULT NULL)
*/
Datum
utl_file_fcopy(PG_FUNCTION_ARGS)
{
char *srcpath;
char *dstpath;
int start_line;
int end_line;
FILE *srcfile;
FILE *dstfile;
NOT_NULL_ARG(0);
NOT_NULL_ARG(1);
NOT_NULL_ARG(2);
NOT_NULL_ARG(3);
srcpath = get_safe_path(PG_GETARG_TEXT_P(0), PG_GETARG_TEXT_P(1));
dstpath = get_safe_path(PG_GETARG_TEXT_P(2), PG_GETARG_TEXT_P(3));
start_line = PG_GETARG_IF_EXISTS(4, INT32, 1);
if (start_line <= 0)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("start_line must be positive (%d passed)", start_line)));
end_line = PG_GETARG_IF_EXISTS(5, INT32, INT_MAX);
if (end_line <= 0)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("end_line must be positive (%d passed)", end_line)));
srcfile = AllocateFile(srcpath, "rt");
if (srcfile == NULL)
{
/* failed to open src file. */
IO_EXCEPTION();
}
dstfile = AllocateFile(dstpath, "wt");
if (dstfile == NULL)
{
/* failed to open dst file. */
fclose(srcfile);
IO_EXCEPTION();
}
if (copy_text_file(srcfile, dstfile, start_line, end_line))
IO_EXCEPTION();
FreeFile(srcfile);
FreeFile(dstfile);
PG_RETURN_VOID();
}
/*
* FindMyDatabase -- get the critical info needed to locate my database
*
* Find the named database in pg_database, return its database OID and the
* OID of its default tablespace. Return TRUE if found, FALSE if not.
*
* Since we are not yet up and running as a backend, we cannot look directly
* at pg_database (we can't obtain locks nor participate in transactions).
* So to get the info we need before starting up, we must look at the "flat
* file" copy of pg_database that is helpfully maintained by flatfiles.c.
* This is subject to various race conditions, so after we have the
* transaction infrastructure started, we have to recheck the information;
* see InitPostgres.
*/
static bool
FindMyDatabase(const char *name, Oid *db_id, Oid *db_tablespace)
{
bool result = false;
char *filename;
FILE *db_file;
char thisname[NAMEDATALEN];
TransactionId db_frozenxid;
filename = database_getflatfilename();
db_file = AllocateFile(filename, "r");
if (db_file == NULL)
ereport(FATAL,
(errcode_for_file_access(),
errmsg("could not open file \"%s\": %m", filename)));
while (read_pg_database_line(db_file, thisname, db_id,
db_tablespace, &db_frozenxid))
{
if (strcmp(thisname, name) == 0)
{
result = true;
break;
}
}
FreeFile(db_file);
pfree(filename);
return result;
}
static Source *
CreateAsyncSource(const char *path, TupleDesc desc)
{
AsyncSource *self = palloc0(sizeof(AsyncSource));
self->base.read = (SourceReadProc) AsyncSourceRead;
self->base.close = (SourceCloseProc) AsyncSourceClose;
self->size = INITIAL_BUF_LEN;
self->begin = 0;
self->end = 0;
self->buffer = palloc0(self->size);
self->errmsg[0] = '\0';
self->eof = false;
self->fd = AllocateFile(path, "r");
if (self->fd == NULL)
ereport(ERROR, (errcode_for_file_access(),
errmsg("could not open \"%s\" %m", path)));
#if defined(USE_POSIX_FADVISE)
posix_fadvise(fileno(self->fd), 0, 0, POSIX_FADV_SEQUENTIAL | POSIX_FADV_NOREUSE | POSIX_FADV_WILLNEED);
#endif
pthread_mutex_init(&self->lock, NULL);
if (pthread_create(&self->th, NULL, AsyncSourceMain, self) != 0)
elog(ERROR, "pthread_create");
return (Source *) self;
}
/* Dumps the histogram data into a file (with a md5 hash of the contents at the beginning). */
static
void pg_record_shmem_shutdown(int code, Datum arg) {
FILE * file;
/* do we need to write the queries? */
if (query_buffer->next == 0) {
return;
}
prepare_file(log_file, query_buffer, query_buffer->next);
file = AllocateFile(log_file->curr_filename, PG_BINARY_A);
if (file == NULL)
goto error;
/* now write the actual shared segment */
if (fwrite(query_buffer->buffer, query_buffer->next, 1, file) != 1)
goto error;
FreeFile(file);
return;
error:
ereport(LOG,
(errcode_for_file_access(),
errmsg("could not write query buffer to the file \"%s\": %m",
log_file->curr_filename)));
if (file)
FreeFile(file);
}
/*
* XLogArchiveNotify
*
* Create an archive notification file
*
* The name of the notification file is the message that will be picked up
* by the archiver, e.g. we write 0000000100000001000000C6.ready
* and the archiver then knows to archive XLOGDIR/0000000100000001000000C6,
* then when complete, rename it to 0000000100000001000000C6.done
*/
void
XLogArchiveNotify(const char *xlog)
{
char archiveStatusPath[MAXPGPATH];
FILE *fd;
/* insert an otherwise empty file called <XLOG>.ready */
StatusFilePath(archiveStatusPath, xlog, ".ready");
fd = AllocateFile(archiveStatusPath, "w");
if (fd == NULL)
{
ereport(LOG,
(errcode_for_file_access(),
errmsg("could not create archive status file \"%s\": %m",
archiveStatusPath)));
return;
}
if (FreeFile(fd))
{
ereport(LOG,
(errcode_for_file_access(),
errmsg("could not write archive status file \"%s\": %m",
archiveStatusPath)));
return;
}
/* Notify archiver that it's got something to do */
if (IsUnderPostmaster)
SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
}
/* Dumps the histogram data into a file (with a md5 hash of the contents at the beginning). */
static
void buffer_write() {
FILE * file;
prepare_file(log_file, query_buffer, query_buffer->next);
file = AllocateFile(log_file->curr_filename, PG_BINARY_A);
if (file == NULL)
goto error;
/* now write the actual shared segment */
if (fwrite(query_buffer->buffer, query_buffer->next, 1, file) != 1)
goto error;
FreeFile(file);
return;
error:
ereport(LOG,
(errcode_for_file_access(),
errmsg("could not write query histogram file \"%s\": %m",
log_file->curr_filename)));
if (file)
FreeFile(file);
}
/*
* Returns start time of an online exclusive backup.
*
* When there's no exclusive backup in progress, the function
* returns NULL.
*/
Datum
pg_backup_start_time(PG_FUNCTION_ARGS)
{
Datum xtime;
FILE *lfp;
char fline[MAXPGPATH];
char backup_start_time[30];
/*
* See if label file is present
*/
lfp = AllocateFile(BACKUP_LABEL_FILE, "r");
if (lfp == NULL)
{
if (errno != ENOENT)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not read file \"%s\": %m",
BACKUP_LABEL_FILE)));
PG_RETURN_NULL();
}
/*
* Parse the file to find the START TIME line.
*/
backup_start_time[0] = '\0';
while (fgets(fline, sizeof(fline), lfp) != NULL)
{
if (sscanf(fline, "START TIME: %25[^\n]\n", backup_start_time) == 1)
break;
}
/* Check for a read error. */
if (ferror(lfp))
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not read file \"%s\": %m", BACKUP_LABEL_FILE)));
/* Close the backup label file. */
if (FreeFile(lfp))
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not close file \"%s\": %m", BACKUP_LABEL_FILE)));
if (strlen(backup_start_time) == 0)
ereport(ERROR,
(errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
errmsg("invalid data in file \"%s\"", BACKUP_LABEL_FILE)));
/*
* Convert the time string read from file to TimestampTz form.
*/
xtime = DirectFunctionCall3(timestamptz_in,
CStringGetDatum(backup_start_time),
ObjectIdGetDatum(InvalidOid),
Int32GetDatum(-1));
PG_RETURN_DATUM(xtime);
}
/*
* Write into the error log file. This opens the file every time,
* so that we can keep it simple to deal with concurrent write.
*/
static void
ErrorLogWrite(CdbSreh *cdbsreh)
{
HeapTuple tuple;
char filename[MAXPGPATH];
FILE *fp;
pg_crc32 crc;
Assert(OidIsValid(cdbsreh->relid));
ErrorLogFileName(filename, MyDatabaseId, cdbsreh->relid);
tuple = FormErrorTuple(cdbsreh);
INIT_CRC32C(crc);
COMP_CRC32C(crc, tuple->t_data, tuple->t_len);
FIN_CRC32C(crc);
LWLockAcquire(ErrorLogLock, LW_EXCLUSIVE);
fp = AllocateFile(filename, "a");
if (!fp)
{
mkdir(ErrorLogDir, S_IRWXU);
fp = AllocateFile(filename, "a");
}
if (!fp)
ereport(ERROR,
(errmsg("could not open \"%s\": %m", filename)));
/*
* format:
* 0-4: length
* 5-8: crc
* 9-n: tuple data
*/
if (fwrite(&tuple->t_len, 1, sizeof(tuple->t_len), fp) != sizeof(tuple->t_len))
elog(ERROR, "could not write tuple length: %m");
if (fwrite(&crc, 1, sizeof(pg_crc32), fp) != sizeof(pg_crc32))
elog(ERROR, "could not write checksum: %m");
if (fwrite(tuple->t_data, 1, tuple->t_len, fp) != tuple->t_len)
elog(ERROR, "could not write tuple data: %m");
FreeFile(fp);
LWLockRelease(ErrorLogLock);
heap_freetuple(tuple);
}
static void
remove_duplicate(Spooler *self, Relation heap, IndexTuple itup, const char *relname)
{
HeapTupleData tuple;
BlockNumber blknum;
BlockNumber offnum;
Buffer buffer;
Page page;
ItemId itemid;
blknum = ItemPointerGetBlockNumber(&itup->t_tid);
offnum = ItemPointerGetOffsetNumber(&itup->t_tid);
buffer = ReadBuffer(heap, blknum);
LockBuffer(buffer, BUFFER_LOCK_SHARE);
page = BufferGetPage(buffer);
itemid = PageGetItemId(page, offnum);
tuple.t_data = ItemIdIsNormal(itemid)
? (HeapTupleHeader) PageGetItem(page, itemid)
: NULL;
LockBuffer(buffer, BUFFER_LOCK_UNLOCK);
if (tuple.t_data != NULL)
{
char *str;
TupleDesc tupdesc;
simple_heap_delete(heap, &itup->t_tid);
/* output duplicate bad file. */
if (self->dup_fp == NULL)
if ((self->dup_fp = AllocateFile(self->dup_badfile, "w")) == NULL)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not open duplicate bad file \"%s\": %m",
self->dup_badfile)));
tupdesc = RelationGetDescr(heap);
tuple.t_len = ItemIdGetLength(itemid);
tuple.t_self = itup->t_tid;
str = tuple_to_cstring(RelationGetDescr(heap), &tuple);
if (fprintf(self->dup_fp, "%s\n", str) < 0 || fflush(self->dup_fp))
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not write parse badfile \"%s\": %m",
self->dup_badfile)));
pfree(str);
}
ReleaseBuffer(buffer);
LoggerLog(WARNING, "Duplicate error Record " int64_FMT
": Rejected - duplicate key value violates unique constraint \"%s\"\n",
self->dup_old + self->dup_new, relname);
}
/*
* Read a section of a file, returning it as bytea
*
* Caller is responsible for all permissions checking.
*
* We read the whole of the file when bytes_to_read is negative.
*/
bytea *
read_binary_file(const char *filename, int64 seek_offset, int64 bytes_to_read)
{
bytea *buf;
size_t nbytes;
FILE *file;
if (bytes_to_read < 0)
{
if (seek_offset < 0)
bytes_to_read = -seek_offset;
else
{
struct stat fst;
if (stat(filename, &fst) < 0)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not stat file \"%s\": %m", filename)));
bytes_to_read = fst.st_size - seek_offset;
}
}
/* not sure why anyone thought that int64 length was a good idea */
if (bytes_to_read > (MaxAllocSize - VARHDRSZ))
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("requested length too large")));
if ((file = AllocateFile(filename, PG_BINARY_R)) == NULL)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not open file \"%s\" for reading: %m",
filename)));
if (fseeko(file, (off_t) seek_offset,
(seek_offset >= 0) ? SEEK_SET : SEEK_END) != 0)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not seek in file \"%s\": %m", filename)));
buf = (bytea *) palloc((Size) bytes_to_read + VARHDRSZ);
nbytes = fread(VARDATA(buf), 1, (size_t) bytes_to_read, file);
if (ferror(file))
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not read file \"%s\": %m", filename)));
SET_VARSIZE(buf, nbytes + VARHDRSZ);
FreeFile(file);
return buf;
}
/*
* shmem_shutdown hook: Dump statistics into file.
*
* Note: we don't bother with acquiring lock, because there should be no
* other processes running when this is called.
*/
static void
pgss_shmem_shutdown(int code, Datum arg)
{
FILE *file;
HASH_SEQ_STATUS hash_seq;
int32 num_entries;
pgssEntry *entry;
/* Don't try to dump during a crash. */
if (code)
return;
/* Safety check ... shouldn't get here unless shmem is set up. */
if (!pgss || !pgss_hash)
return;
/* Don't dump if told not to. */
if (!pgss_save)
return;
file = AllocateFile(PGSS_DUMP_FILE, PG_BINARY_W);
if (file == NULL)
goto error;
if (fwrite(&PGSS_FILE_HEADER, sizeof(uint32), 1, file) != 1)
goto error;
num_entries = hash_get_num_entries(pgss_hash);
if (fwrite(&num_entries, sizeof(int32), 1, file) != 1)
goto error;
hash_seq_init(&hash_seq, pgss_hash);
while ((entry = hash_seq_search(&hash_seq)) != NULL)
{
int len = entry->key.query_len;
if (fwrite(entry, offsetof(pgssEntry, mutex), 1, file) != 1 ||
fwrite(entry->query, 1, len, file) != len)
goto error;
}
if (FreeFile(file))
{
file = NULL;
goto error;
}
return;
error:
ereport(LOG,
(errcode_for_file_access(),
errmsg("could not write pg_stat_statement file \"%s\": %m",
PGSS_DUMP_FILE)));
if (file)
FreeFile(file);
unlink(PGSS_DUMP_FILE);
}
/*
* Read a section of a file, returning it as text
*/
Datum
pg_read_file(PG_FUNCTION_ARGS)
{
text *filename_t = PG_GETARG_TEXT_P(0);
int64 seek_offset = PG_GETARG_INT64(1);
int64 bytes_to_read = PG_GETARG_INT64(2);
char *buf;
size_t nbytes;
FILE *file;
char *filename;
if (!superuser())
ereport(ERROR,
(errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
(errmsg("must be superuser to read files"))));
filename = convert_and_check_filename(filename_t);
if ((file = AllocateFile(filename, PG_BINARY_R)) == NULL)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not open file \"%s\" for reading: %m",
filename)));
if (fseeko(file, (off_t) seek_offset,
(seek_offset >= 0) ? SEEK_SET : SEEK_END) != 0)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not seek in file \"%s\": %m", filename)));
if (bytes_to_read < 0)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("requested length cannot be negative")));
/* not sure why anyone thought that int64 length was a good idea */
if (bytes_to_read > (MaxAllocSize - VARHDRSZ))
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("requested length too large")));
buf = palloc((Size) bytes_to_read + VARHDRSZ);
nbytes = fread(VARDATA(buf), 1, (size_t) bytes_to_read, file);
if (ferror(file))
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not read file \"%s\": %m", filename)));
SET_VARSIZE(buf, nbytes + VARHDRSZ);
FreeFile(file);
pfree(filename);
PG_RETURN_TEXT_P(buf);
}
/*
* Load precomputed DH parameters.
*
* To prevent "downgrade" attacks, we perform a number of checks
* to verify that the DBA-generated DH parameters file contains
* what we expect it to contain.
*/
static DH *
load_dh_file(char *filename, bool isServerStart)
{
FILE *fp;
DH *dh = NULL;
int codes;
/* attempt to open file. It's not an error if it doesn't exist. */
if ((fp = AllocateFile(filename, "r")) == NULL)
{
ereport(isServerStart ? FATAL : LOG,
(errcode_for_file_access(),
errmsg("could not open DH parameters file \"%s\": %m",
filename)));
return NULL;
}
dh = PEM_read_DHparams(fp, NULL, NULL, NULL);
FreeFile(fp);
if (dh == NULL)
{
ereport(isServerStart ? FATAL : LOG,
(errcode(ERRCODE_CONFIG_FILE_ERROR),
errmsg("could not load DH parameters file: %s",
SSLerrmessage(ERR_get_error()))));
return NULL;
}
/* make sure the DH parameters are usable */
if (DH_check(dh, &codes) == 0)
{
ereport(isServerStart ? FATAL : LOG,
(errcode(ERRCODE_CONFIG_FILE_ERROR),
errmsg("invalid DH parameters: %s",
SSLerrmessage(ERR_get_error()))));
return NULL;
}
if (codes & DH_CHECK_P_NOT_PRIME)
{
ereport(isServerStart ? FATAL : LOG,
(errcode(ERRCODE_CONFIG_FILE_ERROR),
errmsg("invalid DH parameters: p is not prime")));
return NULL;
}
if ((codes & DH_NOT_SUITABLE_GENERATOR) &&
(codes & DH_CHECK_P_NOT_SAFE_PRIME))
{
ereport(isServerStart ? FATAL : LOG,
(errcode(ERRCODE_CONFIG_FILE_ERROR),
errmsg("invalid DH parameters: neither suitable generator or safe prime")));
return NULL;
}
return dh;
}
Datum
pg_file_write(PG_FUNCTION_ARGS)
{
FILE *f;
char *filename;
text *data;
int64 count = 0;
requireSuperuser();
filename = convert_and_check_filename(PG_GETARG_TEXT_PP(0), false);
data = PG_GETARG_TEXT_PP(1);
if (!PG_GETARG_BOOL(2))
{
struct stat fst;
if (stat(filename, &fst) >= 0)
ereport(ERROR,
(ERRCODE_DUPLICATE_FILE,
errmsg("file \"%s\" exists", filename)));
f = AllocateFile(filename, "wb");
}
else
f = AllocateFile(filename, "ab");
if (!f)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not open file \"%s\" for writing: %m",
filename)));
count = fwrite(VARDATA_ANY(data), 1, VARSIZE_ANY_EXHDR(data), f);
if (count != VARSIZE_ANY_EXHDR(data) || FreeFile(f))
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not write file \"%s\": %m", filename)));
PG_RETURN_INT64(count);
}
/*
* write out the PG_VERSION file in the specified directory
*/
static void
set_short_version(const char *path)
{
char *short_version;
bool gotdot = false;
int end;
char *fullname;
FILE *version_file;
/* Construct short version string (should match initdb.c) */
short_version = pstrdup(PG_VERSION);
for (end = 0; short_version[end] != '\0'; end++)
{
if (short_version[end] == '.')
{
Assert(end != 0);
if (gotdot)
break;
else
gotdot = true;
}
else if (short_version[end] < '0' || short_version[end] > '9')
{
/* gone past digits and dots */
break;
}
}
Assert(end > 0 && short_version[end - 1] != '.' && gotdot);
short_version[end] = '\0';
/* Now write the file */
fullname = palloc(strlen(path) + 11 + 1);
sprintf(fullname, "%s/PG_VERSION", path);
version_file = AllocateFile(fullname, PG_BINARY_W);
if (version_file == NULL)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not write to file \"%s\": %m",
fullname)));
fprintf(version_file, "%s\n", short_version);
if (FreeFile(version_file))
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not write to file \"%s\": %m",
fullname)));
pfree(fullname);
pfree(short_version);
}
/*
* Set up to read a file using tsearch_readline(). This facility is
* better than just reading the file directly because it provides error
* context pointing to the specific line where a problem is detected.
*
* Expected usage is:
*
* tsearch_readline_state trst;
*
* if (!tsearch_readline_begin(&trst, filename))
* ereport(ERROR,
* (errcode(ERRCODE_CONFIG_FILE_ERROR),
* errmsg("could not open stop-word file \"%s\": %m",
* filename)));
* while ((line = tsearch_readline(&trst)) != NULL)
* process line;
* tsearch_readline_end(&trst);
*
* Note that the caller supplies the ereport() for file open failure;
* this is so that a custom message can be provided. The filename string
* passed to tsearch_readline_begin() must remain valid through
* tsearch_readline_end().
*/
bool
tsearch_readline_begin(tsearch_readline_state *stp,
const char *filename)
{
if ((stp->fp = AllocateFile(filename, "r")) == NULL)
return false;
stp->filename = filename;
stp->lineno = 0;
stp->curline = NULL;
/* Setup error traceback support for ereport() */
stp->cb.callback = tsearch_readline_callback;
stp->cb.arg = (void *) stp;
stp->cb.previous = error_context_stack;
error_context_stack = &stp->cb;
return true;
}
/*
* Identify the huge page size to use.
*
* Some Linux kernel versions have a bug causing mmap() to fail on requests
* that are not a multiple of the hugepage size. Versions without that bug
* instead silently round the request up to the next hugepage multiple ---
* and then munmap() fails when we give it a size different from that.
* So we have to round our request up to a multiple of the actual hugepage
* size to avoid trouble.
*
* Doing the round-up ourselves also lets us make use of the extra memory,
* rather than just wasting it. Currently, we just increase the available
* space recorded in the shmem header, which will make the extra usable for
* purposes such as additional locktable entries. Someday, for very large
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
* Returns the (real or assumed) page size into *hugepagesize,
* and the hugepage-related mmap flags to use into *mmap_flags.
*
* Currently *mmap_flags is always just MAP_HUGETLB. Someday, on systems
* that support it, we might OR in additional bits to specify a particular
* non-default huge page size.
*/
static void
GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
/*
* If we fail to find out the system's default huge page size, assume it
* is 2MB. This will work fine when the actual size is less. If it's
* more, we might get mmap() or munmap() failures due to unaligned
* requests; but at this writing, there are no reports of any non-Linux
* systems being picky about that.
*/
*hugepagesize = 2 * 1024 * 1024;
*mmap_flags = MAP_HUGETLB;
/*
* System-dependent code to find out the default huge page size.
*
* On Linux, read /proc/meminfo looking for a line like "Hugepagesize:
* nnnn kB". Ignore any failures, falling back to the preset default.
*/
#ifdef __linux__
{
FILE *fp = AllocateFile("/proc/meminfo", "r");
char buf[128];
unsigned int sz;
char ch;
if (fp)
{
while (fgets(buf, sizeof(buf), fp))
{
if (sscanf(buf, "Hugepagesize: %u %c", &sz, &ch) == 2)
{
if (ch == 'k')
{
*hugepagesize = sz * (Size) 1024;
break;
}
/* We could accept other units besides kB, if needed */
}
}
FreeFile(fp);
}
}
#endif /* __linux__ */
}
/*
* XLogArchiveForceDone
*
* Emit notification forcibly that an XLOG segment file has been successfully
* archived, by creating <XLOG>.done regardless of whether <XLOG>.ready
* exists or not.
*/
void
XLogArchiveForceDone(const char *xlog)
{
char archiveReady[MAXPGPATH];
char archiveDone[MAXPGPATH];
struct stat stat_buf;
FILE *fd;
/* Exit if already known done */
StatusFilePath(archiveDone, xlog, ".done");
if (stat(archiveDone, &stat_buf) == 0)
return;
/* If .ready exists, rename it to .done */
StatusFilePath(archiveReady, xlog, ".ready");
if (stat(archiveReady, &stat_buf) == 0)
{
if (rename(archiveReady, archiveDone) < 0)
ereport(WARNING,
(errcode_for_file_access(),
errmsg("could not rename file \"%s\" to \"%s\": %m",
archiveReady, archiveDone)));
return;
}
/* insert an otherwise empty file called <XLOG>.done */
fd = AllocateFile(archiveDone, "w");
if (fd == NULL)
{
ereport(LOG,
(errcode_for_file_access(),
errmsg("could not create archive status file \"%s\": %m",
archiveDone)));
return;
}
if (FreeFile(fd))
{
ereport(LOG,
(errcode_for_file_access(),
errmsg("could not write archive status file \"%s\": %m",
archiveDone)));
return;
}
}
static Source *
CreateFileSource(const char *path, TupleDesc desc)
{
FileSource *self = palloc0(sizeof(FileSource));
self->base.read = (SourceReadProc) FileSourceRead;
self->base.close = (SourceCloseProc) FileSourceClose;
self->fd = AllocateFile(path, "r");
if (self->fd == NULL)
ereport(ERROR, (errcode_for_file_access(),
errmsg("could not open \"%s\" %m", path)));
#if defined(USE_POSIX_FADVISE)
posix_fadvise(fileno(self->fd), 0, 0, POSIX_FADV_SEQUENTIAL | POSIX_FADV_NOREUSE | POSIX_FADV_WILLNEED);
#endif
return (Source *) self;
}
void
write_nondefault_variables(GucContext context)
{
int elevel;
FILE *fp;
int i;
Assert(context == PGC_POSTMASTER || context == PGC_SIGHUP);
elevel = (context == PGC_SIGHUP) ? LOG : ERROR;
/*
* Open file
*/
fp = AllocateFile(CONFIG_EXEC_PARAMS_NEW, "w");
if (!fp)
{
ereport(elevel,
(errcode_for_file_access(),
errmsg("could not write to file \"%s\": %m",
CONFIG_EXEC_PARAMS_NEW)));
return;
}
for (i = 0; i < num_guc_variables; i++)
{
write_one_nondefault_variable(fp, guc_variables[i]);
}
if (FreeFile(fp))
{
ereport(elevel,
(errcode_for_file_access(),
errmsg("could not write to file \"%s\": %m",
CONFIG_EXEC_PARAMS_NEW)));
return;
}
/*
* Put new file in place. This could delay on Win32, but we don't hold
* any exclusive locks.
*/
rename(CONFIG_EXEC_PARAMS_NEW, CONFIG_EXEC_PARAMS);
}
void
CreateLogger(const char *path, bool verbose, bool writer)
{
memset(&logger, 0, sizeof(logger));
logger.verbose = verbose;
logger.writer = writer;
if (!is_absolute_path(path))
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("relative path not allowed for LOGFILE: %s", path)));
logger.logfile = pstrdup(path);
logger.fp = AllocateFile(logger.logfile, "at");
if (logger.fp == NULL)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not open loader log file \"%s\": %m",
logger.logfile)));
}
/*
* CStoreWriteFooter writes the given footer to given file. First, the function
* serializes and writes the footer to the file. Then, the function serializes
* and writes the postscript. Then, the function writes the postscript size as
* the last byte of the file. Last, the function syncs and closes the footer file.
*/
static void
CStoreWriteFooter(StringInfo tableFooterFilename, TableFooter *tableFooter)
{
FILE *tableFooterFile = NULL;
StringInfo tableFooterBuffer = NULL;
StringInfo postscriptBuffer = NULL;
uint8 postscriptSize = 0;
tableFooterFile = AllocateFile(tableFooterFilename->data, PG_BINARY_W);
if (tableFooterFile == NULL)
{
ereport(ERROR, (errcode_for_file_access(),
errmsg("could not open file \"%s\" for writing: %m",
tableFooterFilename->data)));
}
/* write the footer */
tableFooterBuffer = SerializeTableFooter(tableFooter);
WriteToFile(tableFooterFile, tableFooterBuffer->data, tableFooterBuffer->len);
/* write the postscript */
postscriptBuffer = SerializePostScript(tableFooterBuffer->len);
WriteToFile(tableFooterFile, postscriptBuffer->data, postscriptBuffer->len);
/* write the 1-byte postscript size */
Assert(postscriptBuffer->len < CSTORE_POSTSCRIPT_SIZE_MAX);
postscriptSize = postscriptBuffer->len;
WriteToFile(tableFooterFile, &postscriptSize, CSTORE_POSTSCRIPT_SIZE_LENGTH);
SyncAndCloseFile(tableFooterFile);
pfree(tableFooterBuffer->data);
pfree(tableFooterBuffer);
pfree(postscriptBuffer->data);
pfree(postscriptBuffer);
}
/* Dumps the histogram data into a file (with a md5 hash of the contents at the beginning). */
static
void query_write(double duration, const char * query, int len, const char * header, int hlen) {
FILE * file;
/* write the buffer first */
buffer_write();
/* now write the query */
prepare_file(log_file, query_buffer, hlen + len);
file = AllocateFile(log_file->curr_filename, PG_BINARY_A);
if (file == NULL)
goto error;
/* now write the actual shared segment */
if (fwrite(header, hlen, 1, file) != 1)
goto error;
/* now write the actual shared segment */
if (fwrite(query, len, 1, file) != 1)
goto error;
FreeFile(file);
return;
error:
ereport(LOG,
(errcode_for_file_access(),
errmsg("could not write query to the file \"%s\": %m",
log_file->curr_filename)));
if (file)
FreeFile(file);
}
/*
* Load rules from the file.
*
* Parses the pg_limits.conf file and loads all the connection rules that
* are defined in it. A syntax error should not result in a failure,
* only a WARNING message (and skipping the row). If there are too many
* rules in the file (exceeding MAX_RULES), that fails with an ERROR.
*
* FIXME The current implementation modifies the segment in-place, so
* if a config reload fails, the backends will see the result of
* the failed reload. That's not really nice. This should use a
* local buffer an only copy it in place if everything went OK.
*
* FIXME The other issue is that we're holding ProcArrayLock while
* parsing the file - at the moment this is necessary because of
* the in-place reload. Once this is fixed, we can hold the lock
* only for the final copy (in case of success).
*/
static void
load_rules()
{
FILE *file;
char line[LINE_MAXLEN];
char dbname[NAMEDATALEN], user[NAMEDATALEN], ip[NAMEDATALEN], mask[NAMEDATALEN];
int limit;
int line_number = 0;
file = AllocateFile(LIMITS_FILE, "r");
if (file == NULL)
{
ereport(WARNING,
(errcode_for_file_access(),
errmsg("could not open configuration file \"%s\": %m",
LIMITS_FILE)));
return;
}
/*
* Use the same lock as when checking the rules (when opening the
* connection) etc. This is probably the right thing to do.
*/
LWLockAcquire(ProcArrayLock, LW_EXCLUSIVE);
/* make sure there are no rules (keep the backend info though) */
memset(rules, 0, RULES_SEGMENT_SIZE);
while (fgets(line, LINE_MAXLEN, file) != NULL)
{
/* remove the comment from the line */
char * comment = strchr(line, '#');
if (comment != NULL)
(*comment) = '\0';
/* remove all white-space chars from the end of the line */
comment--;
while (isspace(comment) && (comment >= line))
{
*comment = '\0';
comment--;
}
++line_number;
/* database user ip mask limit */
if (sscanf(line, "%s %s %s %s %d", dbname, user, ip, mask, &limit) == 5)
load_rule(line_number, dbname, user, ip, mask, limit);
/* database user ip/mask limit */
else if (sscanf(line, "%s %s %s %d", dbname, user, ip, &limit) == 4)
load_rule(line_number, dbname, user, ip, NULL, limit);
/* non-empty line with invalid format */
else if (strlen(line) > 0)
elog(WARNING, "invalid rule at line %d", line_number);
}
FreeFile(file);
LWLockRelease(ProcArrayLock);
elog(DEBUG1, "loaded %d connection limit rule(s)", rules->n_rules);
}
/*
* shmem_startup hook: allocate or attach to shared memory,
* then load any pre-existing statistics from file.
*/
static void
pgss_shmem_startup(void)
{
bool found;
HASHCTL info;
FILE *file;
uint32 header;
int32 num;
int32 i;
int query_size;
int buffer_size;
char *buffer = NULL;
if (prev_shmem_startup_hook)
prev_shmem_startup_hook();
/* reset in case this is a restart within the postmaster */
pgss = NULL;
pgss_hash = NULL;
/*
* Create or attach to the shared memory state, including hash table
*/
LWLockAcquire(AddinShmemInitLock, LW_EXCLUSIVE);
pgss = ShmemInitStruct("pg_stat_statements",
sizeof(pgssSharedState),
&found);
if (!found)
{
/* First time through ... */
pgss->lock = LWLockAssign();
pgss->query_size = pgstat_track_activity_query_size;
}
/* Be sure everyone agrees on the hash table entry size */
query_size = pgss->query_size;
memset(&info, 0, sizeof(info));
info.keysize = sizeof(pgssHashKey);
info.entrysize = offsetof(pgssEntry, query) +query_size;
info.hash = pgss_hash_fn;
info.match = pgss_match_fn;
pgss_hash = ShmemInitHash("pg_stat_statements hash",
pgss_max, pgss_max,
&info,
HASH_ELEM | HASH_FUNCTION | HASH_COMPARE);
LWLockRelease(AddinShmemInitLock);
/*
* If we're in the postmaster (or a standalone backend...), set up a shmem
* exit hook to dump the statistics to disk.
*/
if (!IsUnderPostmaster)
on_shmem_exit(pgss_shmem_shutdown, (Datum) 0);
/*
* Attempt to load old statistics from the dump file, if this is the first
* time through and we weren't told not to.
*/
if (found || !pgss_save)
return;
/*
* Note: we don't bother with locks here, because there should be no other
* processes running when this code is reached.
*/
file = AllocateFile(PGSS_DUMP_FILE, PG_BINARY_R);
if (file == NULL)
{
if (errno == ENOENT)
return; /* ignore not-found error */
goto error;
}
buffer_size = query_size;
buffer = (char *) palloc(buffer_size);
if (fread(&header, sizeof(uint32), 1, file) != 1 ||
header != PGSS_FILE_HEADER ||
fread(&num, sizeof(int32), 1, file) != 1)
goto error;
for (i = 0; i < num; i++)
{
pgssEntry temp;
pgssEntry *entry;
if (fread(&temp, offsetof(pgssEntry, mutex), 1, file) != 1)
goto error;
/* Encoding is the only field we can easily sanity-check */
if (!PG_VALID_BE_ENCODING(temp.key.encoding))
goto error;
/* Previous incarnation might have had a larger query_size */
if (temp.key.query_len >= buffer_size)
{
buffer = (char *) repalloc(buffer, temp.key.query_len + 1);
buffer_size = temp.key.query_len + 1;
}
if (fread(buffer, 1, temp.key.query_len, file) != temp.key.query_len)
goto error;
buffer[temp.key.query_len] = '\0';
/* Clip to available length if needed */
if (temp.key.query_len >= query_size)
temp.key.query_len = pg_encoding_mbcliplen(temp.key.encoding,
buffer,
temp.key.query_len,
query_size - 1);
temp.key.query_ptr = buffer;
/* make the hashtable entry (discards old entries if too many) */
entry = entry_alloc(&temp.key);
/* copy in the actual stats */
entry->counters = temp.counters;
}
pfree(buffer);
FreeFile(file);
return;
error:
ereport(LOG,
(errcode_for_file_access(),
errmsg("could not read pg_stat_statement file \"%s\": %m",
PGSS_DUMP_FILE)));
if (buffer)
pfree(buffer);
if (file)
FreeFile(file);
/* If possible, throw away the bogus file; ignore any error */
unlink(PGSS_DUMP_FILE);
}
/*
* write out the PG_VERSION file in the specified directory. If mirror is true,
* mirror the file creation to our segment mirror.
*
* XXX: API is terrible, make it cleaner
*/
void
set_short_version(const char *path, DbDirNode *dbDirNode, bool mirror)
{
char *short_version;
bool gotdot = false;
int end;
char *fullname;
FILE *version_file;
/* Construct short version string (should match initdb.c) */
short_version = pstrdup(PG_VERSION);
for (end = 0; short_version[end] != '\0'; end++)
{
if (short_version[end] == '.')
{
Assert(end != 0);
if (gotdot)
break;
else
gotdot = true;
}
else if (short_version[end] < '0' || short_version[end] > '9')
{
/* gone past digits and dots */
break;
}
}
Assert(end > 0 && short_version[end - 1] != '.' && gotdot);
short_version[end++] = '\n';
short_version[end] = '\0';
if (mirror)
{
MirroredFlatFileOpen mirroredOpen;
Insist(!PointerIsValid(path));
Insist(PointerIsValid(dbDirNode));
MirroredFlatFile_OpenInDbDir(&mirroredOpen, dbDirNode, "PG_VERSION",
O_CREAT | O_WRONLY | PG_BINARY, S_IRUSR | S_IWUSR,
/* suppressError */ false);
MirroredFlatFile_Append(&mirroredOpen, short_version,
end,
/* suppressError */ false);
MirroredFlatFile_Flush(&mirroredOpen, /* suppressError */ false);
MirroredFlatFile_Close(&mirroredOpen);
}
else
{
Insist(!PointerIsValid(dbDirNode));
Insist(PointerIsValid(path));
/* Now write the file */
fullname = palloc(strlen(path) + 11 + 1);
sprintf(fullname, "%s/PG_VERSION", path);
version_file = AllocateFile(fullname, PG_BINARY_W);
if (version_file == NULL)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not write to file \"%s\": %m",
fullname)));
fprintf(version_file, "%s", short_version);
if (FreeFile(version_file))
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not write to file \"%s\": %m",
fullname)));
pfree(fullname);
}
pfree(short_version);
}
/*
* gp_read_error_log
*
* Returns set of error log tuples.
*/
Datum
gp_read_error_log(PG_FUNCTION_ARGS)
{
FuncCallContext *funcctx;
ReadErrorLogContext *context;
HeapTuple tuple;
Datum result;
/*
* First call setup
*/
if (SRF_IS_FIRSTCALL())
{
MemoryContext oldcontext;
FILE *fp;
text *relname;
funcctx = SRF_FIRSTCALL_INIT();
relname = PG_GETARG_TEXT_P(0);
oldcontext = MemoryContextSwitchTo(funcctx->multi_call_memory_ctx);
context = palloc0(sizeof(ReadErrorLogContext));
funcctx->user_fctx = (void *) context;
funcctx->tuple_desc = BlessTupleDesc(GetErrorTupleDesc());
/*
* Though this function is usually executed on segment, we dispatch
* the execution if it happens to be on QD, and combine the results
* into one set.
*/
if (Gp_role == GP_ROLE_DISPATCH)
{
struct CdbPgResults cdb_pgresults = {NULL, 0};
StringInfoData sql;
int i;
initStringInfo(&sql);
/*
* construct SQL
*/
appendStringInfo(&sql,
"SELECT * FROM pg_catalog.gp_read_error_log(%s) ",
quote_literal_internal(text_to_cstring(relname)));
CdbDispatchCommand(sql.data, DF_WITH_SNAPSHOT, &cdb_pgresults);
for (i = 0; i < cdb_pgresults.numResults; i++)
{
if (PQresultStatus(cdb_pgresults.pg_results[i]) != PGRES_TUPLES_OK)
{
cdbdisp_clearCdbPgResults(&cdb_pgresults);
elog(ERROR, "unexpected result from segment: %d",
PQresultStatus(cdb_pgresults.pg_results[i]));
}
context->numTuples += PQntuples(cdb_pgresults.pg_results[i]);
}
pfree(sql.data);
context->segResults = cdb_pgresults.pg_results;
context->numSegResults = cdb_pgresults.numResults;
}
else
{
/*
* In QE, read the error log.
*/
RangeVar *relrv;
Oid relid;
relrv = makeRangeVarFromNameList(textToQualifiedNameList(relname));
relid = RangeVarGetRelid(relrv, true);
/*
* If the relation has gone, silently return no tuples.
*/
if (OidIsValid(relid))
{
AclResult aclresult;
/*
* Requires SELECT priv to read error log.
*/
aclresult = pg_class_aclcheck(relid, GetUserId(), ACL_SELECT);
if (aclresult != ACLCHECK_OK)
aclcheck_error(aclresult, ACL_KIND_CLASS, relrv->relname);
ErrorLogFileName(context->filename, MyDatabaseId, relid);
fp = AllocateFile(context->filename, "r");
context->fp = fp;
}
}
MemoryContextSwitchTo(oldcontext);
if (Gp_role != GP_ROLE_DISPATCH && !context->fp)
{
pfree(context);
SRF_RETURN_DONE(funcctx);
}
}
funcctx = SRF_PERCALL_SETUP();
context = (ReadErrorLogContext *) funcctx->user_fctx;
/*
* Read error log, probably on segments. We don't check Gp_role, however,
* in case master also wants to read the file.
*/
if (context->fp)
{
pg_crc32 crc, written_crc;
tuple = ErrorLogRead(context->fp, &written_crc);
/*
* CRC check.
*/
if (HeapTupleIsValid(tuple))
{
INIT_CRC32C(crc);
COMP_CRC32C(crc, tuple->t_data, tuple->t_len);
FIN_CRC32C(crc);
if (!EQ_CRC32C(crc, written_crc))
{
elog(LOG, "incorrect checksum in error log %s",
context->filename);
tuple = NULL;
}
}
/*
* If we found a valid tuple, return it. Otherwise, fall through
* in the DONE routine.
*/
if (HeapTupleIsValid(tuple))
{
/*
* We need to set typmod for the executor to understand
* its type we just blessed.
*/
HeapTupleHeaderSetTypMod(tuple->t_data,
funcctx->tuple_desc->tdtypmod);
result = HeapTupleGetDatum(tuple);
SRF_RETURN_NEXT(funcctx, result);
}
}
/*
* If we got results from dispatch, return all the tuples.
*/
while (context->currentResult < context->numSegResults)
{
Datum values[NUM_ERRORTABLE_ATTR];
bool isnull[NUM_ERRORTABLE_ATTR];
PGresult *segres = context->segResults[context->currentResult];
int row = context->currentRow;
if (row >= PQntuples(segres))
{
context->currentRow = 0;
context->currentResult++;
continue;
}
context->currentRow++;
MemSet(isnull, false, sizeof(isnull));
values[0] = ResultToDatum(segres, row, 0, timestamptz_in, &isnull[0]);
values[1] = ResultToDatum(segres, row, 1, textin, &isnull[1]);
values[2] = ResultToDatum(segres, row, 2, textin, &isnull[2]);
values[3] = ResultToDatum(segres, row, 3, int4in, &isnull[3]);
values[4] = ResultToDatum(segres, row, 4, int4in, &isnull[4]);
values[5] = ResultToDatum(segres, row, 5, textin, &isnull[5]);
values[6] = ResultToDatum(segres, row, 6, textin, &isnull[6]);
values[7] = ResultToDatum(segres, row, 7, byteain, &isnull[7]);
tuple = heap_form_tuple(funcctx->tuple_desc, values, isnull);
result = HeapTupleGetDatum(tuple);
SRF_RETURN_NEXT(funcctx, result);
}
if (context->segResults != NULL)
{
int i;
for (i = 0; i < context->numSegResults; i++)
PQclear(context->segResults[i]);
/* XXX: better to copy to palloc'ed area */
free(context->segResults);
}
/*
* Close the file, if we have opened it.
*/
if (context->fp != NULL)
{
FreeFile(context->fp);
context->fp = NULL;
}
SRF_RETURN_DONE(funcctx);
}
/*
* ImportSnapshot
* Import a previously exported snapshot. The argument should be a
* filename in SNAPSHOT_EXPORT_DIR. Load the snapshot from that file.
* This is called by "SET TRANSACTION SNAPSHOT 'foo'".
*/
void
ImportSnapshot(const char *idstr)
{
char path[MAXPGPATH];
FILE *f;
struct stat stat_buf;
char *filebuf;
int xcnt;
int i;
TransactionId src_xid;
Oid src_dbid;
int src_isolevel;
bool src_readonly;
SnapshotData snapshot;
/*
* Must be at top level of a fresh transaction. Note in particular that
* we check we haven't acquired an XID --- if we have, it's conceivable
* that the snapshot would show it as not running, making for very
* screwy behavior.
*/
if (FirstSnapshotSet ||
GetTopTransactionIdIfAny() != InvalidTransactionId ||
IsSubTransaction())
ereport(ERROR,
(errcode(ERRCODE_ACTIVE_SQL_TRANSACTION),
errmsg("SET TRANSACTION SNAPSHOT must be called before any query")));
/*
* If we are in read committed mode then the next query would execute
* with a new snapshot thus making this function call quite useless.
*/
if (!IsolationUsesXactSnapshot())
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("a snapshot-importing transaction must have isolation level SERIALIZABLE or REPEATABLE READ")));
/*
* Verify the identifier: only 0-9, A-F and hyphens are allowed. We do
* this mainly to prevent reading arbitrary files.
*/
if (strspn(idstr, "0123456789ABCDEF-") != strlen(idstr))
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("invalid snapshot identifier \"%s\"", idstr)));
/* OK, read the file */
snprintf(path, MAXPGPATH, SNAPSHOT_EXPORT_DIR "/%s", idstr);
f = AllocateFile(path, PG_BINARY_R);
if (!f)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("invalid snapshot identifier \"%s\"", idstr)));
/* get the size of the file so that we know how much memory we need */
if (fstat(fileno(f), &stat_buf))
elog(ERROR, "could not stat file \"%s\": %m", path);
/* and read the file into a palloc'd string */
filebuf = (char *) palloc(stat_buf.st_size + 1);
if (fread(filebuf, stat_buf.st_size, 1, f) != 1)
elog(ERROR, "could not read file \"%s\": %m", path);
filebuf[stat_buf.st_size] = '\0';
FreeFile(f);
/*
* Construct a snapshot struct by parsing the file content.
*/
memset(&snapshot, 0, sizeof(snapshot));
src_xid = parseXidFromText("xid:", &filebuf, path);
/* we abuse parseXidFromText a bit here ... */
src_dbid = parseXidFromText("dbid:", &filebuf, path);
src_isolevel = parseIntFromText("iso:", &filebuf, path);
src_readonly = parseIntFromText("ro:", &filebuf, path);
snapshot.xmin = parseXidFromText("xmin:", &filebuf, path);
snapshot.xmax = parseXidFromText("xmax:", &filebuf, path);
snapshot.xcnt = xcnt = parseIntFromText("xcnt:", &filebuf, path);
/* sanity-check the xid count before palloc */
if (xcnt < 0 || xcnt > GetMaxSnapshotXidCount())
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid snapshot data in file \"%s\"", path)));
snapshot.xip = (TransactionId *) palloc(xcnt * sizeof(TransactionId));
for (i = 0; i < xcnt; i++)
snapshot.xip[i] = parseXidFromText("xip:", &filebuf, path);
snapshot.suboverflowed = parseIntFromText("sof:", &filebuf, path);
if (!snapshot.suboverflowed)
{
snapshot.subxcnt = xcnt = parseIntFromText("sxcnt:", &filebuf, path);
/* sanity-check the xid count before palloc */
if (xcnt < 0 || xcnt > GetMaxSnapshotSubxidCount())
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid snapshot data in file \"%s\"", path)));
snapshot.subxip = (TransactionId *) palloc(xcnt * sizeof(TransactionId));
for (i = 0; i < xcnt; i++)
snapshot.subxip[i] = parseXidFromText("sxp:", &filebuf, path);
}
else
{
snapshot.subxcnt = 0;
snapshot.subxip = NULL;
}
snapshot.takenDuringRecovery = parseIntFromText("rec:", &filebuf, path);
/*
* Do some additional sanity checking, just to protect ourselves. We
* don't trouble to check the array elements, just the most critical
* fields.
*/
if (!TransactionIdIsNormal(src_xid) ||
!OidIsValid(src_dbid) ||
!TransactionIdIsNormal(snapshot.xmin) ||
!TransactionIdIsNormal(snapshot.xmax))
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid snapshot data in file \"%s\"", path)));
/*
* If we're serializable, the source transaction must be too, otherwise
* predicate.c has problems (SxactGlobalXmin could go backwards). Also,
* a non-read-only transaction can't adopt a snapshot from a read-only
* transaction, as predicate.c handles the cases very differently.
*/
if (IsolationIsSerializable())
{
if (src_isolevel != XACT_SERIALIZABLE)
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("a serializable transaction cannot import a snapshot from a non-serializable transaction")));
if (src_readonly && !XactReadOnly)
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("a non-read-only serializable transaction cannot import a snapshot from a read-only transaction")));
}
/*
* We cannot import a snapshot that was taken in a different database,
* because vacuum calculates OldestXmin on a per-database basis; so the
* source transaction's xmin doesn't protect us from data loss. This
* restriction could be removed if the source transaction were to mark
* its xmin as being globally applicable. But that would require some
* additional syntax, since that has to be known when the snapshot is
* initially taken. (See pgsql-hackers discussion of 2011-10-21.)
*/
if (src_dbid != MyDatabaseId)
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("cannot import a snapshot from a different database")));
/* OK, install the snapshot */
SetTransactionSnapshot(&snapshot, src_xid);
}
/*
* ExportSnapshot
* Export the snapshot to a file so that other backends can import it.
* Returns the token (the file name) that can be used to import this
* snapshot.
*/
static char *
ExportSnapshot(Snapshot snapshot)
{
TransactionId topXid;
TransactionId *children;
int nchildren;
int addTopXid;
StringInfoData buf;
FILE *f;
int i;
MemoryContext oldcxt;
char path[MAXPGPATH];
char pathtmp[MAXPGPATH];
/*
* It's tempting to call RequireTransactionChain here, since it's not
* very useful to export a snapshot that will disappear immediately
* afterwards. However, we haven't got enough information to do that,
* since we don't know if we're at top level or not. For example, we
* could be inside a plpgsql function that is going to fire off other
* transactions via dblink. Rather than disallow perfectly legitimate
* usages, don't make a check.
*
* Also note that we don't make any restriction on the transaction's
* isolation level; however, importers must check the level if they
* are serializable.
*/
/*
* This will assign a transaction ID if we do not yet have one.
*/
topXid = GetTopTransactionId();
/*
* We cannot export a snapshot from a subtransaction because there's no
* easy way for importers to verify that the same subtransaction is still
* running.
*/
if (IsSubTransaction())
ereport(ERROR,
(errcode(ERRCODE_ACTIVE_SQL_TRANSACTION),
errmsg("cannot export a snapshot from a subtransaction")));
/*
* We do however allow previous committed subtransactions to exist.
* Importers of the snapshot must see them as still running, so get their
* XIDs to add them to the snapshot.
*/
nchildren = xactGetCommittedChildren(&children);
/*
* Copy the snapshot into TopTransactionContext, add it to the
* exportedSnapshots list, and mark it pseudo-registered. We do this to
* ensure that the snapshot's xmin is honored for the rest of the
* transaction. (Right now, because SnapshotResetXmin is so stupid, this
* is overkill; but later we might make that routine smarter.)
*/
snapshot = CopySnapshot(snapshot);
oldcxt = MemoryContextSwitchTo(TopTransactionContext);
exportedSnapshots = lappend(exportedSnapshots, snapshot);
MemoryContextSwitchTo(oldcxt);
snapshot->regd_count++;
RegisteredSnapshots++;
/*
* Fill buf with a text serialization of the snapshot, plus identification
* data about this transaction. The format expected by ImportSnapshot
* is pretty rigid: each line must be fieldname:value.
*/
initStringInfo(&buf);
appendStringInfo(&buf, "xid:%u\n", topXid);
appendStringInfo(&buf, "dbid:%u\n", MyDatabaseId);
appendStringInfo(&buf, "iso:%d\n", XactIsoLevel);
appendStringInfo(&buf, "ro:%d\n", XactReadOnly);
appendStringInfo(&buf, "xmin:%u\n", snapshot->xmin);
appendStringInfo(&buf, "xmax:%u\n", snapshot->xmax);
/*
* We must include our own top transaction ID in the top-xid data, since
* by definition we will still be running when the importing transaction
* adopts the snapshot, but GetSnapshotData never includes our own XID in
* the snapshot. (There must, therefore, be enough room to add it.)
*
* However, it could be that our topXid is after the xmax, in which case
* we shouldn't include it because xip[] members are expected to be before
* xmax. (We need not make the same check for subxip[] members, see
* snapshot.h.)
*/
addTopXid = TransactionIdPrecedes(topXid, snapshot->xmax) ? 1 : 0;
appendStringInfo(&buf, "xcnt:%d\n", snapshot->xcnt + addTopXid);
for (i = 0; i < snapshot->xcnt; i++)
appendStringInfo(&buf, "xip:%u\n", snapshot->xip[i]);
if (addTopXid)
appendStringInfo(&buf, "xip:%u\n", topXid);
/*
* Similarly, we add our subcommitted child XIDs to the subxid data.
* Here, we have to cope with possible overflow.
*/
if (snapshot->suboverflowed ||
snapshot->subxcnt + nchildren > GetMaxSnapshotSubxidCount())
appendStringInfoString(&buf, "sof:1\n");
else
{
appendStringInfoString(&buf, "sof:0\n");
appendStringInfo(&buf, "sxcnt:%d\n", snapshot->subxcnt + nchildren);
for (i = 0; i < snapshot->subxcnt; i++)
appendStringInfo(&buf, "sxp:%u\n", snapshot->subxip[i]);
for (i = 0; i < nchildren; i++)
appendStringInfo(&buf, "sxp:%u\n", children[i]);
}
appendStringInfo(&buf, "rec:%u\n", snapshot->takenDuringRecovery);
/*
* Now write the text representation into a file. We first write to a
* ".tmp" filename, and rename to final filename if no error. This
* ensures that no other backend can read an incomplete file
* (ImportSnapshot won't allow it because of its valid-characters check).
*/
XactExportFilePath(pathtmp, topXid, list_length(exportedSnapshots), ".tmp");
if (!(f = AllocateFile(pathtmp, PG_BINARY_W)))
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not create file \"%s\": %m", pathtmp)));
if (fwrite(buf.data, buf.len, 1, f) != 1)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not write to file \"%s\": %m", pathtmp)));
/* no fsync() since file need not survive a system crash */
if (FreeFile(f))
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not write to file \"%s\": %m", pathtmp)));
/*
* Now that we have written everything into a .tmp file, rename the file
* to remove the .tmp suffix.
*/
XactExportFilePath(path, topXid, list_length(exportedSnapshots), "");
if (rename(pathtmp, path) < 0)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not rename file \"%s\" to \"%s\": %m",
pathtmp, path)));
/*
* The basename of the file is what we return from pg_export_snapshot().
* It's already in path in a textual format and we know that the path
* starts with SNAPSHOT_EXPORT_DIR. Skip over the prefix and the slash
* and pstrdup it so as not to return the address of a local variable.
*/
return pstrdup(path + strlen(SNAPSHOT_EXPORT_DIR) + 1);
}
/**
* @brief Read the next tuple from parser.
* @param rd [in/out] reader
* @return type
*/
HeapTuple
ReaderNext(Reader *rd)
{
HeapTuple tuple;
MemoryContext ccxt;
bool eof;
Parser *parser = rd->parser;
ccxt = CurrentMemoryContext;
eof = false;
do
{
tuple = NULL;
parser->parsing_field = -1;
PG_TRY();
{
tuple = ParserRead(parser, &rd->checker);
if (tuple == NULL)
eof = true;
else
{
tuple = CheckerTuple(&rd->checker, tuple,
&parser->parsing_field);
CheckerConstraints(&rd->checker, tuple, &parser->parsing_field);
}
}
PG_CATCH();
{
ErrorData *errdata;
MemoryContext ecxt;
char *message;
StringInfoData buf;
if (parser->parsing_field < 0)
PG_RE_THROW(); /* should not ignore */
ecxt = MemoryContextSwitchTo(ccxt);
errdata = CopyErrorData();
/* We cannot ignore query aborts. */
switch (errdata->sqlerrcode)
{
case ERRCODE_ADMIN_SHUTDOWN:
case ERRCODE_QUERY_CANCELED:
MemoryContextSwitchTo(ecxt);
PG_RE_THROW();
break;
}
/* Absorb parse errors. */
rd->parse_errors++;
if (errdata->message)
message = pstrdup(errdata->message);
else
message = "<no error message>";
FlushErrorState();
FreeErrorData(errdata);
initStringInfo(&buf);
appendStringInfo(&buf, "Parse error Record " int64_FMT
": Input Record " int64_FMT ": Rejected",
rd->parse_errors, parser->count);
if (parser->parsing_field > 0)
appendStringInfo(&buf, " - column %d", parser->parsing_field);
appendStringInfo(&buf, ". %s\n", message);
LoggerLog(WARNING, buf.data);
/* Terminate if PARSE_ERRORS has been reached. */
if (rd->parse_errors > rd->max_parse_errors)
{
eof = true;
LoggerLog(WARNING,
"Maximum parse error count exceeded - " int64_FMT
" error(s) found in input file\n",
rd->parse_errors);
}
/* output parse bad file. */
if (rd->parse_fp == NULL)
if ((rd->parse_fp = AllocateFile(rd->parse_badfile, "w")) == NULL)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not open parse bad file \"%s\": %m",
rd->parse_badfile)));
ParserDumpRecord(parser, rd->parse_fp, rd->parse_badfile);
MemoryContextReset(ccxt);
// Without the below line, the regression tests shows the different result on debug-build mode.
tuple = NULL;
}
PG_END_TRY();
} while (!eof && !tuple);
BULKLOAD_PROFILE(&prof_reader_parser);
return tuple;
}
