postgresql/contrib/pageinspect/heapfuncs.c

478 lines
12 KiB
C
Raw Normal View History

/*-------------------------------------------------------------------------
*
* heapfuncs.c
* Functions to investigate heap pages
*
* We check the input to these functions for corrupt pointers etc. that
* might cause crashes, but at the same time we try to print out as much
* information as possible, even if it's nonsense. That's because if a
* page is corrupt, we don't know why and how exactly it is corrupt, so we
* let the user judge it.
2007-11-15 22:14:46 +01:00
*
* These functions are restricted to superusers for the fear of introducing
* security holes if the input checking isn't as water-tight as it should be.
2007-11-15 22:14:46 +01:00
* You'd need to be superuser to obtain a raw page image anyway, so
* there's hardly any use case for using these without superuser-rights
* anyway.
*
* Copyright (c) 2007-2018, PostgreSQL Global Development Group
*
* IDENTIFICATION
2010-09-20 22:08:53 +02:00
* contrib/pageinspect/heapfuncs.c
*
*-------------------------------------------------------------------------
*/
#include "postgres.h"
#include "pageinspect.h"
#include "access/htup_details.h"
#include "funcapi.h"
#include "catalog/pg_type.h"
#include "miscadmin.h"
#include "utils/array.h"
#include "utils/builtins.h"
#include "utils/rel.h"
/*
* bits_to_text
*
* Converts a bits8-array of 'len' bits to a human-readable
* c-string representation.
*/
static char *
bits_to_text(bits8 *bits, int len)
{
2007-11-15 22:14:46 +01:00
int i;
char *str;
str = palloc(len + 1);
2007-11-15 22:14:46 +01:00
for (i = 0; i < len; i++)
str[i] = (bits[(i / 8)] & (1 << (i % 8))) ? '1' : '0';
str[i] = '\0';
return str;
}
/*
* text_to_bits
*
* Converts a c-string representation of bits into a bits8-array. This is
* the reverse operation of previous routine.
*/
static bits8 *
text_to_bits(char *str, int len)
{
bits8 *bits;
int off = 0;
char byte = 0;
bits = palloc(len + 1);
while (off < len)
{
if (off % 8 == 0)
byte = 0;
if ((str[off] == '0') || (str[off] == '1'))
byte = byte | ((str[off] - '0') << off % 8);
else
ereport(ERROR,
(errcode(ERRCODE_DATA_CORRUPTED),
errmsg("illegal character '%c' in t_bits string", str[off])));
if (off % 8 == 7)
bits[off / 8] = byte;
off++;
}
return bits;
}
/*
* heap_page_items
*
* Allows inspection of line pointers and tuple headers of a heap page.
*/
PG_FUNCTION_INFO_V1(heap_page_items);
typedef struct heap_page_items_state
{
TupleDesc tupd;
Page page;
uint16 offset;
} heap_page_items_state;
Datum
heap_page_items(PG_FUNCTION_ARGS)
{
2007-11-15 22:14:46 +01:00
bytea *raw_page = PG_GETARG_BYTEA_P(0);
heap_page_items_state *inter_call_data = NULL;
FuncCallContext *fctx;
2007-11-15 22:14:46 +01:00
int raw_page_size;
if (!superuser())
ereport(ERROR,
(errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
(errmsg("must be superuser to use raw page functions"))));
raw_page_size = VARSIZE(raw_page) - VARHDRSZ;
if (SRF_IS_FIRSTCALL())
{
TupleDesc tupdesc;
MemoryContext mctx;
2007-11-15 22:14:46 +01:00
if (raw_page_size < SizeOfPageHeaderData)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("input page too small (%d bytes)", raw_page_size)));
fctx = SRF_FIRSTCALL_INIT();
mctx = MemoryContextSwitchTo(fctx->multi_call_memory_ctx);
inter_call_data = palloc(sizeof(heap_page_items_state));
/* Build a tuple descriptor for our result type */
if (get_call_result_type(fcinfo, NULL, &tupdesc) != TYPEFUNC_COMPOSITE)
elog(ERROR, "return type must be a row type");
inter_call_data->tupd = tupdesc;
inter_call_data->offset = FirstOffsetNumber;
inter_call_data->page = VARDATA(raw_page);
fctx->max_calls = PageGetMaxOffsetNumber(inter_call_data->page);
fctx->user_fctx = inter_call_data;
MemoryContextSwitchTo(mctx);
}
fctx = SRF_PERCALL_SETUP();
inter_call_data = fctx->user_fctx;
if (fctx->call_cntr < fctx->max_calls)
{
Page page = inter_call_data->page;
HeapTuple resultTuple;
Datum result;
ItemId id;
Datum values[14];
bool nulls[14];
2007-11-15 22:14:46 +01:00
uint16 lp_offset;
uint16 lp_flags;
uint16 lp_len;
memset(nulls, 0, sizeof(nulls));
/* Extract information from the line pointer */
2007-11-15 22:14:46 +01:00
id = PageGetItemId(page, inter_call_data->offset);
2007-11-15 22:14:46 +01:00
lp_offset = ItemIdGetOffset(id);
lp_flags = ItemIdGetFlags(id);
lp_len = ItemIdGetLength(id);
values[0] = UInt16GetDatum(inter_call_data->offset);
values[1] = UInt16GetDatum(lp_offset);
values[2] = UInt16GetDatum(lp_flags);
values[3] = UInt16GetDatum(lp_len);
2007-11-15 22:14:46 +01:00
/*
* We do just enough validity checking to make sure we don't reference
* data outside the page passed to us. The page could be corrupt in
* many other ways, but at least we won't crash.
*/
if (ItemIdHasStorage(id) &&
lp_len >= MinHeapTupleSize &&
lp_offset == MAXALIGN(lp_offset) &&
lp_offset + lp_len <= raw_page_size)
{
2016-06-10 00:02:36 +02:00
HeapTupleHeader tuphdr;
bytea *tuple_data_bytea;
int tuple_data_len;
/* Extract information from the tuple header */
tuphdr = (HeapTupleHeader) PageGetItem(page, id);
2007-11-15 22:14:46 +01:00
values[4] = UInt32GetDatum(HeapTupleHeaderGetRawXmin(tuphdr));
Improve concurrency of foreign key locking This patch introduces two additional lock modes for tuples: "SELECT FOR KEY SHARE" and "SELECT FOR NO KEY UPDATE". These don't block each other, in contrast with already existing "SELECT FOR SHARE" and "SELECT FOR UPDATE". UPDATE commands that do not modify the values stored in the columns that are part of the key of the tuple now grab a SELECT FOR NO KEY UPDATE lock on the tuple, allowing them to proceed concurrently with tuple locks of the FOR KEY SHARE variety. Foreign key triggers now use FOR KEY SHARE instead of FOR SHARE; this means the concurrency improvement applies to them, which is the whole point of this patch. The added tuple lock semantics require some rejiggering of the multixact module, so that the locking level that each transaction is holding can be stored alongside its Xid. Also, multixacts now need to persist across server restarts and crashes, because they can now represent not only tuple locks, but also tuple updates. This means we need more careful tracking of lifetime of pg_multixact SLRU files; since they now persist longer, we require more infrastructure to figure out when they can be removed. pg_upgrade also needs to be careful to copy pg_multixact files over from the old server to the new, or at least part of multixact.c state, depending on the versions of the old and new servers. Tuple time qualification rules (HeapTupleSatisfies routines) need to be careful not to consider tuples with the "is multi" infomask bit set as being only locked; they might need to look up MultiXact values (i.e. possibly do pg_multixact I/O) to find out the Xid that updated a tuple, whereas they previously were assured to only use information readily available from the tuple header. This is considered acceptable, because the extra I/O would involve cases that would previously cause some commands to block waiting for concurrent transactions to finish. Another important change is the fact that locking tuples that have previously been updated causes the future versions to be marked as locked, too; this is essential for correctness of foreign key checks. This causes additional WAL-logging, also (there was previously a single WAL record for a locked tuple; now there are as many as updated copies of the tuple there exist.) With all this in place, contention related to tuples being checked by foreign key rules should be much reduced. As a bonus, the old behavior that a subtransaction grabbing a stronger tuple lock than the parent (sub)transaction held on a given tuple and later aborting caused the weaker lock to be lost, has been fixed. Many new spec files were added for isolation tester framework, to ensure overall behavior is sane. There's probably room for several more tests. There were several reviewers of this patch; in particular, Noah Misch and Andres Freund spent considerable time in it. Original idea for the patch came from Simon Riggs, after a problem report by Joel Jacobson. Most code is from me, with contributions from Marti Raudsepp, Alexander Shulgin, Noah Misch and Andres Freund. This patch was discussed in several pgsql-hackers threads; the most important start at the following message-ids: AANLkTimo9XVcEzfiBR-ut3KVNDkjm2Vxh+t8kAmWjPuv@mail.gmail.com 1290721684-sup-3951@alvh.no-ip.org 1294953201-sup-2099@alvh.no-ip.org 1320343602-sup-2290@alvh.no-ip.org 1339690386-sup-8927@alvh.no-ip.org 4FE5FF020200002500048A3D@gw.wicourts.gov 4FEAB90A0200002500048B7D@gw.wicourts.gov
2013-01-23 16:04:59 +01:00
values[5] = UInt32GetDatum(HeapTupleHeaderGetRawXmax(tuphdr));
/* shared with xvac */
values[6] = UInt32GetDatum(HeapTupleHeaderGetRawCommandId(tuphdr));
values[7] = PointerGetDatum(&tuphdr->t_ctid);
values[8] = UInt32GetDatum(tuphdr->t_infomask2);
values[9] = UInt32GetDatum(tuphdr->t_infomask);
values[10] = UInt8GetDatum(tuphdr->t_hoff);
/* Copy raw tuple data into bytea attribute */
tuple_data_len = lp_len - tuphdr->t_hoff;
tuple_data_bytea = (bytea *) palloc(tuple_data_len + VARHDRSZ);
SET_VARSIZE(tuple_data_bytea, tuple_data_len + VARHDRSZ);
memcpy(VARDATA(tuple_data_bytea), (char *) tuphdr + tuphdr->t_hoff,
2016-06-10 00:02:36 +02:00
tuple_data_len);
values[13] = PointerGetDatum(tuple_data_bytea);
2007-11-15 22:14:46 +01:00
/*
* We already checked that the item is completely within the raw
* page passed to us, with the length given in the line pointer.
* Let's check that t_hoff doesn't point over lp_len, before using
* it to access t_bits and oid.
*/
if (tuphdr->t_hoff >= SizeofHeapTupleHeader &&
tuphdr->t_hoff <= lp_len &&
tuphdr->t_hoff == MAXALIGN(tuphdr->t_hoff))
{
if (tuphdr->t_infomask & HEAP_HASNULL)
{
int bits_len;
bits_len =
BITMAPLEN(HeapTupleHeaderGetNatts(tuphdr)) * BITS_PER_BYTE;
values[11] = CStringGetTextDatum(
bits_to_text(tuphdr->t_bits, bits_len));
2007-11-15 22:14:46 +01:00
}
else
nulls[11] = true;
if (tuphdr->t_infomask & HEAP_HASOID)
values[12] = HeapTupleHeaderGetOid(tuphdr);
else
nulls[12] = true;
}
else
{
nulls[11] = true;
nulls[12] = true;
}
}
else
{
2007-11-15 22:14:46 +01:00
/*
* The line pointer is not used, or it's invalid. Set the rest of
* the fields to NULL
*/
int i;
for (i = 4; i <= 13; i++)
nulls[i] = true;
}
2007-11-15 22:14:46 +01:00
/* Build and return the result tuple. */
resultTuple = heap_form_tuple(inter_call_data->tupd, values, nulls);
result = HeapTupleGetDatum(resultTuple);
inter_call_data->offset++;
SRF_RETURN_NEXT(fctx, result);
}
else
SRF_RETURN_DONE(fctx);
}
/*
* tuple_data_split_internal
*
* Split raw tuple data taken directly from a page into an array of bytea
* elements. This routine does a lookup on NULL values and creates array
2016-10-26 10:10:13 +02:00
* elements accordingly. This is a reimplementation of nocachegetattr()
* in heaptuple.c simplified for educational purposes.
*/
static Datum
tuple_data_split_internal(Oid relid, char *tupdata,
2016-06-10 00:02:36 +02:00
uint16 tupdata_len, uint16 t_infomask,
uint16 t_infomask2, bits8 *t_bits,
bool do_detoast)
{
2016-06-10 00:02:36 +02:00
ArrayBuildState *raw_attrs;
int nattrs;
int i;
int off = 0;
Relation rel;
TupleDesc tupdesc;
/* Get tuple descriptor from relation OID */
rel = relation_open(relid, AccessShareLock);
tupdesc = RelationGetDescr(rel);
raw_attrs = initArrayResult(BYTEAOID, CurrentMemoryContext, false);
nattrs = tupdesc->natts;
if (nattrs < (t_infomask2 & HEAP_NATTS_MASK))
ereport(ERROR,
(errcode(ERRCODE_DATA_CORRUPTED),
errmsg("number of attributes in tuple header is greater than number of attributes in tuple descriptor")));
for (i = 0; i < nattrs; i++)
{
2016-06-10 00:02:36 +02:00
Form_pg_attribute attr;
bool is_null;
bytea *attr_data = NULL;
attr = TupleDescAttr(tupdesc, i);
/*
2016-06-10 00:02:36 +02:00
* Tuple header can specify less attributes than tuple descriptor as
* ALTER TABLE ADD COLUMN without DEFAULT keyword does not actually
* change tuples in pages, so attributes with numbers greater than
* (t_infomask2 & HEAP_NATTS_MASK) should be treated as NULL.
*/
if (i >= (t_infomask2 & HEAP_NATTS_MASK))
is_null = true;
else
is_null = (t_infomask & HEAP_HASNULL) && att_isnull(i, t_bits);
if (!is_null)
{
2016-06-10 00:02:36 +02:00
int len;
if (attr->attlen == -1)
{
off = att_align_pointer(off, attr->attalign, -1,
tupdata + off);
2016-06-10 00:02:36 +02:00
/*
* As VARSIZE_ANY throws an exception if it can't properly
* detect the type of external storage in macros VARTAG_SIZE,
* this check is repeated to have a nicer error handling.
*/
if (VARATT_IS_EXTERNAL(tupdata + off) &&
!VARATT_IS_EXTERNAL_ONDISK(tupdata + off) &&
!VARATT_IS_EXTERNAL_INDIRECT(tupdata + off))
ereport(ERROR,
2016-06-10 00:02:36 +02:00
(errcode(ERRCODE_DATA_CORRUPTED),
errmsg("first byte of varlena attribute is incorrect for attribute %d", i)));
len = VARSIZE_ANY(tupdata + off);
}
else
{
off = att_align_nominal(off, attr->attalign);
len = attr->attlen;
}
if (tupdata_len < off + len)
ereport(ERROR,
(errcode(ERRCODE_DATA_CORRUPTED),
errmsg("unexpected end of tuple data")));
if (attr->attlen == -1 && do_detoast)
attr_data = DatumGetByteaPCopy(tupdata + off);
else
{
attr_data = (bytea *) palloc(len + VARHDRSZ);
SET_VARSIZE(attr_data, len + VARHDRSZ);
memcpy(VARDATA(attr_data), tupdata + off, len);
}
off = att_addlength_pointer(off, attr->attlen,
tupdata + off);
}
raw_attrs = accumArrayResult(raw_attrs, PointerGetDatum(attr_data),
is_null, BYTEAOID, CurrentMemoryContext);
if (attr_data)
pfree(attr_data);
}
if (tupdata_len != off)
ereport(ERROR,
(errcode(ERRCODE_DATA_CORRUPTED),
errmsg("end of tuple reached without looking at all its data")));
relation_close(rel, AccessShareLock);
return makeArrayResult(raw_attrs, CurrentMemoryContext);
}
/*
* tuple_data_split
*
* Split raw tuple data taken directly from page into distinct elements
* taking into account null values.
*/
PG_FUNCTION_INFO_V1(tuple_data_split);
Datum
tuple_data_split(PG_FUNCTION_ARGS)
{
2016-06-10 00:02:36 +02:00
Oid relid;
bytea *raw_data;
uint16 t_infomask;
uint16 t_infomask2;
char *t_bits_str;
bool do_detoast = false;
bits8 *t_bits = NULL;
Datum res;
relid = PG_GETARG_OID(0);
raw_data = PG_ARGISNULL(1) ? NULL : PG_GETARG_BYTEA_P(1);
t_infomask = PG_GETARG_INT16(2);
t_infomask2 = PG_GETARG_INT16(3);
t_bits_str = PG_ARGISNULL(4) ? NULL :
text_to_cstring(PG_GETARG_TEXT_PP(4));
if (PG_NARGS() >= 6)
do_detoast = PG_GETARG_BOOL(5);
if (!superuser())
ereport(ERROR,
(errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
errmsg("must be superuser to use raw page functions")));
if (!raw_data)
PG_RETURN_NULL();
/*
* Convert t_bits string back to the bits8 array as represented in the
* tuple header.
*/
if (t_infomask & HEAP_HASNULL)
{
2016-06-10 00:02:36 +02:00
int bits_str_len;
int bits_len;
bits_len = BITMAPLEN(t_infomask2 & HEAP_NATTS_MASK) * BITS_PER_BYTE;
if (!t_bits_str)
ereport(ERROR,
(errcode(ERRCODE_DATA_CORRUPTED),
errmsg("argument of t_bits is null, but it is expected to be null and %d character long",
bits_len)));
bits_str_len = strlen(t_bits_str);
if (bits_len != bits_str_len)
ereport(ERROR,
(errcode(ERRCODE_DATA_CORRUPTED),
errmsg("unexpected length of t_bits %u, expected %d",
bits_str_len, bits_len)));
/* do the conversion */
t_bits = text_to_bits(t_bits_str, bits_str_len);
}
else
{
if (t_bits_str)
ereport(ERROR,
(errcode(ERRCODE_DATA_CORRUPTED),
2016-04-08 18:31:44 +02:00
errmsg("t_bits string is expected to be NULL, but instead it is %zu bytes length",
strlen(t_bits_str))));
}
/* Split tuple data */
res = tuple_data_split_internal(relid, (char *) raw_data + VARHDRSZ,
VARSIZE(raw_data) - VARHDRSZ,
t_infomask, t_infomask2, t_bits,
do_detoast);
if (t_bits)
pfree(t_bits);
PG_RETURN_ARRAYTYPE_P(res);
}