Skip to content

Commit d5406de

Browse files
robertmhaasrobertmhaas
committed
Code cleanup for toast_fetch_datum and toast_fetch_datum_slice.
Rework some of the checks for bad TOAST chunks to be a bit simpler and easier to understand. These checks verify that (1) we get all and only the chunk numbers we expect to see and (2) each chunk has the expected size. However, the existing code was a bit hard to understand, at least for me; try to make it clearer. As part of that, have toast_fetch_datum_slice check the relationship between endchunk and totalchunks only with an Assert() rather than checking every chunk number against both values. There's no need to check that relationship in production builds because it's not a function of whether on-disk corruption is present; it's just a question of whether the code does the right math. Also, have toast_fetch_datum_slice() use ereport(ERROR) rather than elog(ERROR). Commit fd6ec93 made the two functions inconsistent with each other. Rename assorted variables for better clarity and consistency, and move assorted variables from function scope to the function's main loop. Remove a few variables that are used only once entirely. Patch by me, reviewed by Peter Eisentraut. Discussion: http://postgr.es/m/CA+TgmobBzxwFojJ0zV0Own3dr09y43hp+OzU2VW+nos4PMXWEg@mail.gmail.com
1 parent da41d71 commit d5406de

File tree

1 file changed

+102
-118
lines changed

1 file changed

+102
-118
lines changed

‎src/backend/access/common/detoast.c‎

Lines changed: 102 additions & 118 deletions
Original file line numberDiff line numberDiff line change
@@ -25,7 +25,8 @@
2525

2626
static struct varlena *toast_fetch_datum(struct varlena *attr);
2727
static struct varlena *toast_fetch_datum_slice(struct varlena *attr,
28-
int32 sliceoffset, int32 length);
28+
int32 sliceoffset,
29+
int32 slicelength);
2930
static struct varlena *toast_decompress_datum(struct varlena *attr);
3031
static struct varlena *toast_decompress_datum_slice(struct varlena *attr, int32 slicelength);
3132

@@ -331,14 +332,9 @@ toast_fetch_datum(struct varlena *attr)
331332
TupleDesc toasttupDesc;
332333
struct varlena *result;
333334
struct varatt_external toast_pointer;
334-
int32 ressize;
335-
int32 residx,
336-
nextidx;
337-
int32 numchunks;
338-
Pointer chunk;
339-
bool isnull;
340-
char *chunkdata;
341-
int32 chunksize;
335+
int32 attrsize;
336+
int32 expectedchunk;
337+
int32 totalchunks;
342338
int num_indexes;
343339
int validIndex;
344340
SnapshotData SnapshotToast;
@@ -349,15 +345,15 @@ toast_fetch_datum(struct varlena *attr)
349345
/* Must copy to access aligned fields */
350346
VARATT_EXTERNAL_GET_POINTER(toast_pointer, attr);
351347

352-
ressize = toast_pointer.va_extsize;
353-
numchunks = ((ressize - 1) / TOAST_MAX_CHUNK_SIZE) + 1;
348+
attrsize = toast_pointer.va_extsize;
349+
totalchunks = ((attrsize - 1) / TOAST_MAX_CHUNK_SIZE) + 1;
354350

355-
result = (struct varlena *) palloc(ressize + VARHDRSZ);
351+
result = (struct varlena *) palloc(attrsize + VARHDRSZ);
356352

357353
if (VARATT_EXTERNAL_IS_COMPRESSED(toast_pointer))
358-
SET_VARSIZE_COMPRESSED(result, ressize + VARHDRSZ);
354+
SET_VARSIZE_COMPRESSED(result, attrsize + VARHDRSZ);
359355
else
360-
SET_VARSIZE(result, ressize + VARHDRSZ);
356+
SET_VARSIZE(result, attrsize + VARHDRSZ);
361357

362358
/*
363359
* Open the toast relation and its indexes
@@ -386,17 +382,24 @@ toast_fetch_datum(struct varlena *attr)
386382
* see the chunks in chunkidx order, even though we didn't explicitly ask
387383
* for it.
388384
*/
389-
nextidx = 0;
385+
expectedchunk = 0;
390386

391387
init_toast_snapshot(&SnapshotToast);
392388
toastscan = systable_beginscan_ordered(toastrel, toastidxs[validIndex],
393389
&SnapshotToast, 1, &toastkey);
394390
while ((ttup = systable_getnext_ordered(toastscan, ForwardScanDirection)) != NULL)
395391
{
392+
int32 curchunk;
393+
Pointer chunk;
394+
bool isnull;
395+
char *chunkdata;
396+
int32 chunksize;
397+
int32 expected_size;
398+
396399
/*
397400
* Have a chunk, extract the sequence number and the data
398401
*/
399-
residx = DatumGetInt32(fastgetattr(ttup, 2, toasttupDesc, &isnull));
402+
curchunk = DatumGetInt32(fastgetattr(ttup, 2, toasttupDesc, &isnull));
400403
Assert(!isnull);
401404
chunk = DatumGetPointer(fastgetattr(ttup, 3, toasttupDesc, &isnull));
402405
Assert(!isnull);
@@ -424,63 +427,50 @@ toast_fetch_datum(struct varlena *attr)
424427
/*
425428
* Some checks on the data we've found
426429
*/
427-
if (residx != nextidx)
430+
if (curchunk != expectedchunk)
428431
ereport(ERROR,
429432
(errcode(ERRCODE_DATA_CORRUPTED),
430433
errmsg_internal("unexpected chunk number %d (expected %d) for toast value %u in %s",
431-
residx, nextidx,
434+
curchunk, expectedchunk,
432435
toast_pointer.va_valueid,
433436
RelationGetRelationName(toastrel))));
434-
if (residx < numchunks - 1)
435-
{
436-
if (chunksize != TOAST_MAX_CHUNK_SIZE)
437-
ereport(ERROR,
438-
(errcode(ERRCODE_DATA_CORRUPTED),
439-
errmsg_internal("unexpected chunk size %d (expected %d) in chunk %d of %d for toast value %u in %s",
440-
chunksize, (int) TOAST_MAX_CHUNK_SIZE,
441-
residx, numchunks,
442-
toast_pointer.va_valueid,
443-
RelationGetRelationName(toastrel))));
444-
}
445-
else if (residx == numchunks - 1)
446-
{
447-
if ((residx * TOAST_MAX_CHUNK_SIZE + chunksize) != ressize)
448-
ereport(ERROR,
449-
(errcode(ERRCODE_DATA_CORRUPTED),
450-
errmsg_internal("unexpected chunk size %d (expected %d) in final chunk %d for toast value %u in %s",
451-
chunksize,
452-
(int) (ressize - residx * TOAST_MAX_CHUNK_SIZE),
453-
residx,
454-
toast_pointer.va_valueid,
455-
RelationGetRelationName(toastrel))));
456-
}
457-
else
437+
if (curchunk > totalchunks - 1)
458438
ereport(ERROR,
459439
(errcode(ERRCODE_DATA_CORRUPTED),
460440
errmsg_internal("unexpected chunk number %d (out of range %d..%d) for toast value %u in %s",
461-
residx,
462-
0, numchunks - 1,
441+
curchunk,
442+
0, totalchunks - 1,
443+
toast_pointer.va_valueid,
444+
RelationGetRelationName(toastrel))));
445+
expected_size = curchunk < totalchunks - 1 ? TOAST_MAX_CHUNK_SIZE
446+
: attrsize % TOAST_MAX_CHUNK_SIZE;
447+
if (chunksize != expected_size)
448+
ereport(ERROR,
449+
(errcode(ERRCODE_DATA_CORRUPTED),
450+
errmsg_internal("unexpected chunk size %d (expected %d) in chunk %d of %d for toast value %u in %s",
451+
chunksize, expected_size,
452+
curchunk, totalchunks,
463453
toast_pointer.va_valueid,
464454
RelationGetRelationName(toastrel))));
465455

466456
/*
467457
* Copy the data into proper place in our result
468458
*/
469-
memcpy(VARDATA(result) + residx * TOAST_MAX_CHUNK_SIZE,
459+
memcpy(VARDATA(result) + curchunk * TOAST_MAX_CHUNK_SIZE,
470460
chunkdata,
471461
chunksize);
472462

473-
nextidx++;
463+
expectedchunk++;
474464
}
475465

476466
/*
477467
* Final checks that we successfully fetched the datum
478468
*/
479-
if (nextidx != numchunks)
469+
if (expectedchunk != totalchunks)
480470
ereport(ERROR,
481471
(errcode(ERRCODE_DATA_CORRUPTED),
482472
errmsg_internal("missing chunk number %d for toast value %u in %s",
483-
nextidx,
473+
expectedchunk,
484474
toast_pointer.va_valueid,
485475
RelationGetRelationName(toastrel))));
486476

@@ -506,7 +496,8 @@ toast_fetch_datum(struct varlena *attr)
506496
* ----------
507497
*/
508498
static struct varlena *
509-
toast_fetch_datum_slice(struct varlena *attr, int32 sliceoffset, int32 length)
499+
toast_fetch_datum_slice(struct varlena *attr, int32 sliceoffset,
500+
int32 slicelength)
510501
{
511502
Relation toastrel;
512503
Relation *toastidxs;
@@ -518,20 +509,10 @@ toast_fetch_datum_slice(struct varlena *attr, int32 sliceoffset, int32 length)
518509
struct varlena *result;
519510
struct varatt_external toast_pointer;
520511
int32 attrsize;
521-
int32 residx;
522-
int32 nextidx;
523-
int numchunks;
512+
int32 expectedchunk;
524513
int startchunk;
525514
int endchunk;
526-
int32 startoffset;
527-
int32 endoffset;
528515
int totalchunks;
529-
Pointer chunk;
530-
bool isnull;
531-
char *chunkdata;
532-
int32 chunksize;
533-
int32 chcpystrt;
534-
int32 chcpyend;
535516
int num_indexes;
536517
int validIndex;
537518
SnapshotData SnapshotToast;
@@ -555,36 +536,33 @@ toast_fetch_datum_slice(struct varlena *attr, int32 sliceoffset, int32 length)
555536
if (sliceoffset >= attrsize)
556537
{
557538
sliceoffset = 0;
558-
length = 0;
539+
slicelength = 0;
559540
}
560541

561542
/*
562543
* When fetching a prefix of a compressed external datum, account for the
563544
* rawsize tracking amount of raw data, which is stored at the beginning
564545
* as an int32 value).
565546
*/
566-
if (VARATT_EXTERNAL_IS_COMPRESSED(toast_pointer) && length > 0)
567-
length = length + sizeof(int32);
547+
if (VARATT_EXTERNAL_IS_COMPRESSED(toast_pointer) && slicelength > 0)
548+
slicelength = slicelength + sizeof(int32);
568549

569-
if (((sliceoffset + length) > attrsize) || length < 0)
570-
length = attrsize - sliceoffset;
550+
if (((sliceoffset + slicelength) > attrsize) || slicelength < 0)
551+
slicelength = attrsize - sliceoffset;
571552

572-
result = (struct varlena *) palloc(length + VARHDRSZ);
553+
result = (struct varlena *) palloc(slicelength + VARHDRSZ);
573554

574555
if (VARATT_EXTERNAL_IS_COMPRESSED(toast_pointer))
575-
SET_VARSIZE_COMPRESSED(result, length + VARHDRSZ);
556+
SET_VARSIZE_COMPRESSED(result, slicelength + VARHDRSZ);
576557
else
577-
SET_VARSIZE(result, length + VARHDRSZ);
558+
SET_VARSIZE(result, slicelength + VARHDRSZ);
578559

579-
if (length == 0)
560+
if (slicelength == 0)
580561
return result; /* Can save a lot of work at this point! */
581562

582563
startchunk = sliceoffset / TOAST_MAX_CHUNK_SIZE;
583-
endchunk = (sliceoffset + length - 1) / TOAST_MAX_CHUNK_SIZE;
584-
numchunks = (endchunk - startchunk) + 1;
585-
586-
startoffset = sliceoffset % TOAST_MAX_CHUNK_SIZE;
587-
endoffset = (sliceoffset + length - 1) % TOAST_MAX_CHUNK_SIZE;
564+
endchunk = (sliceoffset + slicelength - 1) / TOAST_MAX_CHUNK_SIZE;
565+
Assert(endchunk < totalchunks);
588566

589567
/*
590568
* Open the toast relation and its indexes
@@ -610,7 +588,7 @@ toast_fetch_datum_slice(struct varlena *attr, int32 sliceoffset, int32 length)
610588
/*
611589
* Use equality condition for one chunk, a range condition otherwise:
612590
*/
613-
if (numchunks == 1)
591+
if (startchunk == endchunk)
614592
{
615593
ScanKeyInit(&toastkey[1],
616594
(AttrNumber) 2,
@@ -637,15 +615,24 @@ toast_fetch_datum_slice(struct varlena *attr, int32 sliceoffset, int32 length)
637615
* The index is on (valueid, chunkidx) so they will come in order
638616
*/
639617
init_toast_snapshot(&SnapshotToast);
640-
nextidx = startchunk;
618+
expectedchunk = startchunk;
641619
toastscan = systable_beginscan_ordered(toastrel, toastidxs[validIndex],
642620
&SnapshotToast, nscankeys, toastkey);
643621
while ((ttup = systable_getnext_ordered(toastscan, ForwardScanDirection)) != NULL)
644622
{
623+
int32 curchunk;
624+
Pointer chunk;
625+
bool isnull;
626+
char *chunkdata;
627+
int32 chunksize;
628+
int32 expected_size;
629+
int32 chcpystrt;
630+
int32 chcpyend;
631+
645632
/*
646633
* Have a chunk, extract the sequence number and the data
647634
*/
648-
residx = DatumGetInt32(fastgetattr(ttup, 2, toasttupDesc, &isnull));
635+
curchunk = DatumGetInt32(fastgetattr(ttup, 2, toasttupDesc, &isnull));
649636
Assert(!isnull);
650637
chunk = DatumGetPointer(fastgetattr(ttup, 3, toasttupDesc, &isnull));
651638
Assert(!isnull);
@@ -673,63 +660,60 @@ toast_fetch_datum_slice(struct varlena *attr, int32 sliceoffset, int32 length)
673660
/*
674661
* Some checks on the data we've found
675662
*/
676-
if ((residx != nextidx) || (residx > endchunk) || (residx < startchunk))
677-
elog(ERROR, "unexpected chunk number %d (expected %d) for toast value %u in %s",
678-
residx, nextidx,
679-
toast_pointer.va_valueid,
680-
RelationGetRelationName(toastrel));
681-
if (residx < totalchunks - 1)
682-
{
683-
if (chunksize != TOAST_MAX_CHUNK_SIZE)
684-
elog(ERROR, "unexpected chunk size %d (expected %d) in chunk %d of %d for toast value %u in %s when fetching slice",
685-
chunksize, (int) TOAST_MAX_CHUNK_SIZE,
686-
residx, totalchunks,
687-
toast_pointer.va_valueid,
688-
RelationGetRelationName(toastrel));
689-
}
690-
else if (residx == totalchunks - 1)
691-
{
692-
if ((residx * TOAST_MAX_CHUNK_SIZE + chunksize) != attrsize)
693-
elog(ERROR, "unexpected chunk size %d (expected %d) in final chunk %d for toast value %u in %s when fetching slice",
694-
chunksize,
695-
(int) (attrsize - residx * TOAST_MAX_CHUNK_SIZE),
696-
residx,
697-
toast_pointer.va_valueid,
698-
RelationGetRelationName(toastrel));
699-
}
700-
else
701-
elog(ERROR, "unexpected chunk number %d (out of range %d..%d) for toast value %u in %s",
702-
residx,
703-
0, totalchunks - 1,
704-
toast_pointer.va_valueid,
705-
RelationGetRelationName(toastrel));
663+
if (curchunk != expectedchunk)
664+
ereport(ERROR,
665+
(errcode(ERRCODE_DATA_CORRUPTED),
666+
errmsg_internal("unexpected chunk number %d (expected %d) for toast value %u in %s",
667+
curchunk, expectedchunk,
668+
toast_pointer.va_valueid,
669+
RelationGetRelationName(toastrel))));
670+
if (curchunk > endchunk)
671+
ereport(ERROR,
672+
(errcode(ERRCODE_DATA_CORRUPTED),
673+
errmsg_internal("unexpected chunk number %d (out of range %d..%d) for toast value %u in %s",
674+
curchunk,
675+
startchunk, endchunk,
676+
toast_pointer.va_valueid,
677+
RelationGetRelationName(toastrel))));
678+
expected_size = curchunk < totalchunks - 1 ? TOAST_MAX_CHUNK_SIZE
679+
: attrsize % TOAST_MAX_CHUNK_SIZE;
680+
if (chunksize != expected_size)
681+
ereport(ERROR,
682+
(errcode(ERRCODE_DATA_CORRUPTED),
683+
errmsg_internal("unexpected chunk size %d (expected %d) in chunk %d of %d for toast value %u in %s",
684+
chunksize, expected_size,
685+
curchunk, totalchunks,
686+
toast_pointer.va_valueid,
687+
RelationGetRelationName(toastrel))));
706688

707689
/*
708690
* Copy the data into proper place in our result
709691
*/
710692
chcpystrt = 0;
711693
chcpyend = chunksize - 1;
712-
if (residx == startchunk)
713-
chcpystrt = startoffset;
714-
if (residx == endchunk)
715-
chcpyend = endoffset;
694+
if (curchunk == startchunk)
695+
chcpystrt = sliceoffset % TOAST_MAX_CHUNK_SIZE;
696+
if (curchunk == endchunk)
697+
chcpyend = (sliceoffset + slicelength - 1) % TOAST_MAX_CHUNK_SIZE;
716698

717699
memcpy(VARDATA(result) +
718-
(residx * TOAST_MAX_CHUNK_SIZE - sliceoffset) + chcpystrt,
700+
(curchunk * TOAST_MAX_CHUNK_SIZE - sliceoffset) + chcpystrt,
719701
chunkdata + chcpystrt,
720702
(chcpyend - chcpystrt) + 1);
721703

722-
nextidx++;
704+
expectedchunk++;
723705
}
724706

725707
/*
726708
* Final checks that we successfully fetched the datum
727709
*/
728-
if (nextidx != (endchunk + 1))
729-
elog(ERROR, "missing chunk number %d for toast value %u in %s",
730-
nextidx,
731-
toast_pointer.va_valueid,
732-
RelationGetRelationName(toastrel));
710+
if (expectedchunk != (endchunk + 1))
711+
ereport(ERROR,
712+
(errcode(ERRCODE_DATA_CORRUPTED),
713+
errmsg_internal("missing chunk number %d for toast value %u in %s",
714+
expectedchunk,
715+
toast_pointer.va_valueid,
716+
RelationGetRelationName(toastrel))));
733717

734718
/*
735719
* End scan and close relations

0 commit comments

Comments
 (0)