v5-0001-Simplify-abbreviated-key-hashing-using-murmurhash.patch
text/x-patch
Filename: v5-0001-Simplify-abbreviated-key-hashing-using-murmurhash.patch
Type: text/x-patch
Part: 1
Patch
Same data as JSON:
GET /api/v1/attachments/:id/patch
the parsed metadata as JSON — format, series position, per-file stats; never the diff bytes.
API reference →
Format: format-patch
Series: patch v5-0001
Subject: Simplify abbreviated key hashing using murmurhash64()
| File | + | − |
|---|---|---|
| src/backend/commands/async.c | 3 | 3 |
| src/backend/utils/adt/bytea.c | 2 | 7 |
| src/backend/utils/adt/network.c | 1 | 5 |
| src/backend/utils/adt/numeric.c | 1 | 4 |
| src/backend/utils/adt/uuid.c | 1 | 5 |
| src/backend/utils/adt/varlena.c | 2 | 7 |
From 395b247ee70cc2004780b6cee63088fbf89bbc32 Mon Sep 17 00:00:00 2001
From: Aleksander Alekseev <aleksander@tigerdata.com>
Date: Tue, 13 Jan 2026 14:51:21 +0300
Subject: [PATCH v5 1/2] Simplify abbreviated key hashing using murmurhash64()
Now when all Datums are 64-bit values we can simplify the code by using
murmurhash64() in *_abbrev_convert() functions. Also replace hash_uint32()
with murmurhash32() in a few other places for consistency.
Author: Aleksander Alekseev <aleksander@tigerdata.com>
Suggested-by: John Naylor <johncnaylorls@gmail.com>
Reviewed-by: John Naylor <johncnaylorls@gmail.com>
Discussion: https://postgr.es/m/CAJ7c6TMPhDRQMmkUHPv8oOK97B1mR8NRS61DgjpdaZUPAwaeZQ%40mail.gmail.com
---
src/backend/commands/async.c | 6 +++---
src/backend/utils/adt/bytea.c | 9 ++-------
src/backend/utils/adt/network.c | 6 +-----
src/backend/utils/adt/numeric.c | 5 +----
src/backend/utils/adt/uuid.c | 6 +-----
src/backend/utils/adt/varlena.c | 9 ++-------
6 files changed, 10 insertions(+), 31 deletions(-)
diff --git a/src/backend/commands/async.c b/src/backend/commands/async.c
index db6a9a6561b..4cb4b350ce3 100644
--- a/src/backend/commands/async.c
+++ b/src/backend/commands/async.c
@@ -674,9 +674,9 @@ globalChannelTableHash(const void *key, size_t size, void *arg)
const GlobalChannelKey *k = (const GlobalChannelKey *) key;
dshash_hash h;
- h = DatumGetUInt32(hash_uint32(k->dboid));
- h ^= DatumGetUInt32(hash_any((const unsigned char *) k->channel,
- strnlen(k->channel, NAMEDATALEN)));
+ h = murmurhash32(k->dboid);
+ h ^= hash_any((const unsigned char *) k->channel,
+ strnlen(k->channel, NAMEDATALEN));
return h;
}
diff --git a/src/backend/utils/adt/bytea.c b/src/backend/utils/adt/bytea.c
index f6e3266ac32..2e416a83f53 100644
--- a/src/backend/utils/adt/bytea.c
+++ b/src/backend/utils/adt/bytea.c
@@ -1111,17 +1111,12 @@ bytea_abbrev_convert(Datum original, SortSupport ssup)
Min(len, PG_CACHE_LINE_SIZE)));
if (len > PG_CACHE_LINE_SIZE)
- hash ^= DatumGetUInt32(hash_uint32((uint32) len));
+ hash ^= murmurhash32((uint32) len);
addHyperLogLog(&bss->full_card, hash);
/* Hash abbreviated key */
- {
- uint32 tmp;
-
- tmp = DatumGetUInt32(res) ^ (uint32) (DatumGetUInt64(res) >> 32);
- hash = DatumGetUInt32(hash_uint32(tmp));
- }
+ hash = (uint32) murmurhash64(DatumGetUInt64(res));
addHyperLogLog(&bss->abbr_card, hash);
diff --git a/src/backend/utils/adt/network.c b/src/backend/utils/adt/network.c
index c7e0828764e..14c639eafc6 100644
--- a/src/backend/utils/adt/network.c
+++ b/src/backend/utils/adt/network.c
@@ -739,11 +739,7 @@ network_abbrev_convert(Datum original, SortSupport ssup)
/* Hash abbreviated key */
if (uss->estimating)
{
- uint32 tmp;
-
- tmp = DatumGetUInt32(res) ^ (uint32) (DatumGetUInt64(res) >> 32);
-
- addHyperLogLog(&uss->abbr_card, DatumGetUInt32(hash_uint32(tmp)));
+ addHyperLogLog(&uss->abbr_card, (uint32) murmurhash64(DatumGetUInt64(res)));
}
return res;
diff --git a/src/backend/utils/adt/numeric.c b/src/backend/utils/adt/numeric.c
index cb23dfe9b95..797d80b99db 100644
--- a/src/backend/utils/adt/numeric.c
+++ b/src/backend/utils/adt/numeric.c
@@ -2399,10 +2399,7 @@ numeric_abbrev_convert_var(const NumericVar *var, NumericSortSupport *nss)
if (nss->estimating)
{
- uint32 tmp = ((uint32) result
- ^ (uint32) ((uint64) result >> 32));
-
- addHyperLogLog(&nss->abbr_card, DatumGetUInt32(hash_uint32(tmp)));
+ addHyperLogLog(&nss->abbr_card, (uint32) murmurhash64(result));
}
return NumericAbbrevGetDatum(result);
diff --git a/src/backend/utils/adt/uuid.c b/src/backend/utils/adt/uuid.c
index 6ee3752ac78..888802c3012 100644
--- a/src/backend/utils/adt/uuid.c
+++ b/src/backend/utils/adt/uuid.c
@@ -396,11 +396,7 @@ uuid_abbrev_convert(Datum original, SortSupport ssup)
if (uss->estimating)
{
- uint32 tmp;
-
- tmp = DatumGetUInt32(res) ^ (uint32) (DatumGetUInt64(res) >> 32);
-
- addHyperLogLog(&uss->abbr_card, DatumGetUInt32(hash_uint32(tmp)));
+ addHyperLogLog(&uss->abbr_card, (uint32) murmurhash64(DatumGetUInt64(res)));
}
/*
diff --git a/src/backend/utils/adt/varlena.c b/src/backend/utils/adt/varlena.c
index c0ff51bd2fc..9832739c846 100644
--- a/src/backend/utils/adt/varlena.c
+++ b/src/backend/utils/adt/varlena.c
@@ -2157,17 +2157,12 @@ varstr_abbrev_convert(Datum original, SortSupport ssup)
Min(len, PG_CACHE_LINE_SIZE)));
if (len > PG_CACHE_LINE_SIZE)
- hash ^= DatumGetUInt32(hash_uint32((uint32) len));
+ hash ^= murmurhash32((uint32) len);
addHyperLogLog(&sss->full_card, hash);
/* Hash abbreviated key */
- {
- uint32 tmp;
-
- tmp = DatumGetUInt32(res) ^ (uint32) (DatumGetUInt64(res) >> 32);
- hash = DatumGetUInt32(hash_uint32(tmp));
- }
+ hash = (uint32) murmurhash64(DatumGetUInt64(res));
addHyperLogLog(&sss->abbr_card, hash);
--
2.43.0