v11-0005-Add-benchmark-for-hashing-C-strings.patch

From 778e3bdfc761dace149cd6c136e4c2847f793c61 Mon Sep 17 00:00:00 2001
From: John Naylor <john.naylor@postgresql.org>
Date: Sun, 24 Dec 2023 09:46:44 +0700
Subject: [PATCH v11 5/8] Add benchmark for hashing C strings

---
 contrib/bench_hash/Makefile            |  23 +
 contrib/bench_hash/aligned_keywords.h  | 991 +++++++++++++++++++++++++
 contrib/bench_hash/bench_hash--1.0.sql |  21 +
 contrib/bench_hash/bench_hash.c        | 103 +++
 contrib/bench_hash/bench_hash.control  |   5 +
 contrib/bench_hash/meson.build         |  19 +
 contrib/meson.build                    |   1 +
 7 files changed, 1163 insertions(+)
 create mode 100644 contrib/bench_hash/Makefile
 create mode 100644 contrib/bench_hash/aligned_keywords.h
 create mode 100644 contrib/bench_hash/bench_hash--1.0.sql
 create mode 100644 contrib/bench_hash/bench_hash.c
 create mode 100644 contrib/bench_hash/bench_hash.control
 create mode 100644 contrib/bench_hash/meson.build

diff --git a/contrib/bench_hash/Makefile b/contrib/bench_hash/Makefile
new file mode 100644
index 0000000000..5327080376
--- /dev/null
+++ b/contrib/bench_hash/Makefile
@@ -0,0 +1,23 @@
+# src/test/modules/test_parser/Makefile
+
+MODULE_big = test_parser
+OBJS = \
+	$(WIN32RES) \
+	test_parser.o
+PGFILEDESC = "test_parser - example of a custom parser for full-text search"
+
+EXTENSION = test_parser
+DATA = test_parser--1.0.sql
+
+REGRESS = test_parser
+
+ifdef USE_PGXS
+PG_CONFIG = pg_config
+PGXS := $(shell $(PG_CONFIG) --pgxs)
+include $(PGXS)
+else
+subdir = src/test/modules/test_parser
+top_builddir = ../../../..
+include $(top_builddir)/src/Makefile.global
+include $(top_srcdir)/contrib/contrib-global.mk
+endif
diff --git a/contrib/bench_hash/aligned_keywords.h b/contrib/bench_hash/aligned_keywords.h
new file mode 100644
index 0000000000..c2bd67c856
--- /dev/null
+++ b/contrib/bench_hash/aligned_keywords.h
@@ -0,0 +1,991 @@
+/* created by copying from kwlist_d.h with this patch:
+
+--- a/src/tools/gen_keywordlist.pl
++++ b/src/tools/gen_keywordlist.pl
+@@ -97,7 +97,9 @@ while (<$kif>)
+ {
+        if (/^PG_KEYWORD\("(\w+)"/)
+        {
+-               push @keywords, $1;
++               my $len = length($1) + 1;
++               my $aligned = $1 . "\\0" . "_" x ( ($len % 8) == 0 ? 0 : (8-($len % 8)) );
++               push @keywords, $aligned;
+        }
+ }
+ 
+@@ -127,7 +129,7 @@ for my $i (0 .. $#keywords - 1)
+ # Emit the string containing all the keywords.
+ 
+ printf $kwdef qq|static const char %s_kw_string[] =\n\t"|, $varname;
+-print $kwdef join qq|\\0"\n\t"|, @keywords;
++print $kwdef join qq|"\n\t"|, @keywords;
+ print $kwdef qq|";\n\n|;
+ 
+ # Emit an array of numerical offsets which will be used to index into the
+@@ -145,7 +147,7 @@ foreach my $name (@keywords)
+ 
+        # Calculate the cumulative offset of the next keyword,
+        # taking into account the null terminator.
+-       $offset += $this_length + 1;
++       $offset += $this_length -1;
+ 
+        # Update max keyword length.
+        $max_len = $this_length if $max_len < $this_length;
+
+*/
+
+
+static const char aligned_words[] =
+	"abort\0__"
+	"absent\0_"
+	"absolute\0_______"
+	"access\0_"
+	"action\0_"
+	"add\0____"
+	"admin\0__"
+	"after\0__"
+	"aggregate\0______"
+	"all\0____"
+	"also\0___"
+	"alter\0__"
+	"always\0_"
+	"analyse\0"
+	"analyze\0"
+	"and\0____"
+	"any\0____"
+	"array\0__"
+	"as\0_____"
+	"asc\0____"
+	"asensitive\0_____"
+	"assertion\0______"
+	"assignment\0_____"
+	"asymmetric\0_____"
+	"at\0_____"
+	"atomic\0_"
+	"attach\0_"
+	"attribute\0______"
+	"authorization\0__"
+	"backward\0_______"
+	"before\0_"
+	"begin\0__"
+	"between\0"
+	"bigint\0_"
+	"binary\0_"
+	"bit\0____"
+	"boolean\0"
+	"both\0___"
+	"breadth\0"
+	"by\0_____"
+	"cache\0__"
+	"call\0___"
+	"called\0_"
+	"cascade\0"
+	"cascaded\0_______"
+	"case\0___"
+	"cast\0___"
+	"catalog\0"
+	"chain\0__"
+	"char\0___"
+	"character\0______"
+	"characteristics\0"
+	"check\0__"
+	"checkpoint\0_____"
+	"class\0__"
+	"close\0__"
+	"cluster\0"
+	"coalesce\0_______"
+	"collate\0"
+	"collation\0______"
+	"column\0_"
+	"columns\0"
+	"comment\0"
+	"comments\0_______"
+	"commit\0_"
+	"committed\0______"
+	"compression\0____"
+	"concurrently\0___"
+	"configuration\0__"
+	"conflict\0_______"
+	"connection\0_____"
+	"constraint\0_____"
+	"constraints\0____"
+	"content\0"
+	"continue\0_______"
+	"conversion\0_____"
+	"copy\0___"
+	"cost\0___"
+	"create\0_"
+	"cross\0__"
+	"csv\0____"
+	"cube\0___"
+	"current\0"
+	"current_catalog\0"
+	"current_date\0___"
+	"current_role\0___"
+	"current_schema\0_"
+	"current_time\0___"
+	"current_timestamp\0______"
+	"current_user\0___"
+	"cursor\0_"
+	"cycle\0__"
+	"data\0___"
+	"database\0_______"
+	"day\0____"
+	"deallocate\0_____"
+	"dec\0____"
+	"decimal\0"
+	"declare\0"
+	"default\0"
+	"defaults\0_______"
+	"deferrable\0_____"
+	"deferred\0_______"
+	"definer\0"
+	"delete\0_"
+	"delimiter\0______"
+	"delimiters\0_____"
+	"depends\0"
+	"depth\0__"
+	"desc\0___"
+	"detach\0_"
+	"dictionary\0_____"
+	"disable\0"
+	"discard\0"
+	"distinct\0_______"
+	"do\0_____"
+	"document\0_______"
+	"domain\0_"
+	"double\0_"
+	"drop\0___"
+	"each\0___"
+	"else\0___"
+	"enable\0_"
+	"encoding\0_______"
+	"encrypted\0______"
+	"end\0____"
+	"enum\0___"
+	"escape\0_"
+	"event\0__"
+	"except\0_"
+	"exclude\0"
+	"excluding\0______"
+	"exclusive\0______"
+	"execute\0"
+	"exists\0_"
+	"explain\0"
+	"expression\0_____"
+	"extension\0______"
+	"external\0_______"
+	"extract\0"
+	"false\0__"
+	"family\0_"
+	"fetch\0__"
+	"filter\0_"
+	"finalize\0_______"
+	"first\0__"
+	"float\0__"
+	"following\0______"
+	"for\0____"
+	"force\0__"
+	"foreign\0"
+	"format\0_"
+	"forward\0"
+	"freeze\0_"
+	"from\0___"
+	"full\0___"
+	"function\0_______"
+	"functions\0______"
+	"generated\0______"
+	"global\0_"
+	"grant\0__"
+	"granted\0"
+	"greatest\0_______"
+	"group\0__"
+	"grouping\0_______"
+	"groups\0_"
+	"handler\0"
+	"having\0_"
+	"header\0_"
+	"hold\0___"
+	"hour\0___"
+	"identity\0_______"
+	"if\0_____"
+	"ilike\0__"
+	"immediate\0______"
+	"immutable\0______"
+	"implicit\0_______"
+	"import\0_"
+	"in\0_____"
+	"include\0"
+	"including\0______"
+	"increment\0______"
+	"indent\0_"
+	"index\0__"
+	"indexes\0"
+	"inherit\0"
+	"inherits\0_______"
+	"initially\0______"
+	"inline\0_"
+	"inner\0__"
+	"inout\0__"
+	"input\0__"
+	"insensitive\0____"
+	"insert\0_"
+	"instead\0"
+	"int\0____"
+	"integer\0"
+	"intersect\0______"
+	"interval\0_______"
+	"into\0___"
+	"invoker\0"
+	"is\0_____"
+	"isnull\0_"
+	"isolation\0______"
+	"join\0___"
+	"json\0___"
+	"json_array\0_____"
+	"json_arrayagg\0__"
+	"json_object\0____"
+	"json_objectagg\0_"
+	"json_scalar\0____"
+	"json_serialize\0_"
+	"key\0____"
+	"keys\0___"
+	"label\0__"
+	"language\0_______"
+	"large\0__"
+	"last\0___"
+	"lateral\0"
+	"leading\0"
+	"leakproof\0______"
+	"least\0__"
+	"left\0___"
+	"level\0__"
+	"like\0___"
+	"limit\0__"
+	"listen\0_"
+	"load\0___"
+	"local\0__"
+	"localtime\0______"
+	"localtimestamp\0_"
+	"location\0_______"
+	"lock\0___"
+	"locked\0_"
+	"logged\0_"
+	"mapping\0"
+	"match\0__"
+	"matched\0"
+	"materialized\0___"
+	"maxvalue\0_______"
+	"merge\0__"
+	"method\0_"
+	"minute\0_"
+	"minvalue\0_______"
+	"mode\0___"
+	"month\0__"
+	"move\0___"
+	"name\0___"
+	"names\0__"
+	"national\0_______"
+	"natural\0"
+	"nchar\0__"
+	"new\0____"
+	"next\0___"
+	"nfc\0____"
+	"nfd\0____"
+	"nfkc\0___"
+	"nfkd\0___"
+	"no\0_____"
+	"none\0___"
+	"normalize\0______"
+	"normalized\0_____"
+	"not\0____"
+	"nothing\0"
+	"notify\0_"
+	"notnull\0"
+	"nowait\0_"
+	"null\0___"
+	"nullif\0_"
+	"nulls\0__"
+	"numeric\0"
+	"object\0_"
+	"of\0_____"
+	"off\0____"
+	"offset\0_"
+	"oids\0___"
+	"old\0____"
+	"on\0_____"
+	"only\0___"
+	"operator\0_______"
+	"option\0_"
+	"options\0"
+	"or\0_____"
+	"order\0__"
+	"ordinality\0_____"
+	"others\0_"
+	"out\0____"
+	"outer\0__"
+	"over\0___"
+	"overlaps\0_______"
+	"overlay\0"
+	"overriding\0_____"
+	"owned\0__"
+	"owner\0__"
+	"parallel\0_______"
+	"parameter\0______"
+	"parser\0_"
+	"partial\0"
+	"partition\0______"
+	"passing\0"
+	"password\0_______"
+	"placing\0"
+	"plans\0__"
+	"policy\0_"
+	"position\0_______"
+	"preceding\0______"
+	"precision\0______"
+	"prepare\0"
+	"prepared\0_______"
+	"preserve\0_______"
+	"primary\0"
+	"prior\0__"
+	"privileges\0_____"
+	"procedural\0_____"
+	"procedure\0______"
+	"procedures\0_____"
+	"program\0"
+	"publication\0____"
+	"quote\0__"
+	"range\0__"
+	"read\0___"
+	"real\0___"
+	"reassign\0_______"
+	"recheck\0"
+	"recursive\0______"
+	"ref\0____"
+	"references\0_____"
+	"referencing\0____"
+	"refresh\0"
+	"reindex\0"
+	"relative\0_______"
+	"release\0"
+	"rename\0_"
+	"repeatable\0_____"
+	"replace\0"
+	"replica\0"
+	"reset\0__"
+	"restart\0"
+	"restrict\0_______"
+	"return\0_"
+	"returning\0______"
+	"returns\0"
+	"revoke\0_"
+	"right\0__"
+	"role\0___"
+	"rollback\0_______"
+	"rollup\0_"
+	"routine\0"
+	"routines\0_______"
+	"row\0____"
+	"rows\0___"
+	"rule\0___"
+	"savepoint\0______"
+	"scalar\0_"
+	"schema\0_"
+	"schemas\0"
+	"scroll\0_"
+	"search\0_"
+	"second\0_"
+	"security\0_______"
+	"select\0_"
+	"sequence\0_______"
+	"sequences\0______"
+	"serializable\0___"
+	"server\0_"
+	"session\0"
+	"session_user\0___"
+	"set\0____"
+	"setof\0__"
+	"sets\0___"
+	"share\0__"
+	"show\0___"
+	"similar\0"
+	"simple\0_"
+	"skip\0___"
+	"smallint\0_______"
+	"snapshot\0_______"
+	"some\0___"
+	"sql\0____"
+	"stable\0_"
+	"standalone\0_____"
+	"start\0__"
+	"statement\0______"
+	"statistics\0_____"
+	"stdin\0__"
+	"stdout\0_"
+	"storage\0"
+	"stored\0_"
+	"strict\0_"
+	"strip\0__"
+	"subscription\0___"
+	"substring\0______"
+	"support\0"
+	"symmetric\0______"
+	"sysid\0__"
+	"system\0_"
+	"system_user\0____"
+	"table\0__"
+	"tables\0_"
+	"tablesample\0____"
+	"tablespace\0_____"
+	"temp\0___"
+	"template\0_______"
+	"temporary\0______"
+	"text\0___"
+	"then\0___"
+	"ties\0___"
+	"time\0___"
+	"timestamp\0______"
+	"to\0_____"
+	"trailing\0_______"
+	"transaction\0____"
+	"transform\0______"
+	"treat\0__"
+	"trigger\0"
+	"trim\0___"
+	"true\0___"
+	"truncate\0_______"
+	"trusted\0"
+	"type\0___"
+	"types\0__"
+	"uescape\0"
+	"unbounded\0______"
+	"uncommitted\0____"
+	"unencrypted\0____"
+	"union\0__"
+	"unique\0_"
+	"unknown\0"
+	"unlisten\0_______"
+	"unlogged\0_______"
+	"until\0__"
+	"update\0_"
+	"user\0___"
+	"using\0__"
+	"vacuum\0_"
+	"valid\0__"
+	"validate\0_______"
+	"validator\0______"
+	"value\0__"
+	"values\0_"
+	"varchar\0"
+	"variadic\0_______"
+	"varying\0"
+	"verbose\0"
+	"version\0"
+	"view\0___"
+	"views\0__"
+	"volatile\0_______"
+	"when\0___"
+	"where\0__"
+	"whitespace\0_____"
+	"window\0_"
+	"with\0___"
+	"within\0_"
+	"without\0"
+	"work\0___"
+	"wrapper\0"
+	"write\0__"
+	"xml\0____"
+	"xmlattributes\0__"
+	"xmlconcat\0______"
+	"xmlelement\0_____"
+	"xmlexists\0______"
+	"xmlforest\0______"
+	"xmlnamespaces\0__"
+	"xmlparse\0_______"
+	"xmlpi\0__"
+	"xmlroot\0"
+	"xmlserialize\0___"
+	"xmltable\0_______"
+	"year\0___"
+	"yes\0____"
+	"zone\0___";
+
+static const uint16 word_offsets[] = {
+	0,
+	8,
+	16,
+	32,
+	40,
+	48,
+	56,
+	64,
+	72,
+	88,
+	96,
+	104,
+	112,
+	120,
+	128,
+	136,
+	144,
+	152,
+	160,
+	168,
+	176,
+	192,
+	208,
+	224,
+	240,
+	248,
+	256,
+	264,
+	280,
+	296,
+	312,
+	320,
+	328,
+	336,
+	344,
+	352,
+	360,
+	368,
+	376,
+	384,
+	392,
+	400,
+	408,
+	416,
+	424,
+	440,
+	448,
+	456,
+	464,
+	472,
+	480,
+	496,
+	512,
+	520,
+	536,
+	544,
+	552,
+	560,
+	576,
+	584,
+	600,
+	608,
+	616,
+	624,
+	640,
+	648,
+	664,
+	680,
+	696,
+	712,
+	728,
+	744,
+	760,
+	776,
+	784,
+	800,
+	816,
+	824,
+	832,
+	840,
+	848,
+	856,
+	864,
+	872,
+	888,
+	904,
+	920,
+	936,
+	952,
+	976,
+	992,
+	1000,
+	1008,
+	1016,
+	1032,
+	1040,
+	1056,
+	1064,
+	1072,
+	1080,
+	1088,
+	1104,
+	1120,
+	1136,
+	1144,
+	1152,
+	1168,
+	1184,
+	1192,
+	1200,
+	1208,
+	1216,
+	1232,
+	1240,
+	1248,
+	1264,
+	1272,
+	1288,
+	1296,
+	1304,
+	1312,
+	1320,
+	1328,
+	1336,
+	1352,
+	1368,
+	1376,
+	1384,
+	1392,
+	1400,
+	1408,
+	1416,
+	1432,
+	1448,
+	1456,
+	1464,
+	1472,
+	1488,
+	1504,
+	1520,
+	1528,
+	1536,
+	1544,
+	1552,
+	1560,
+	1576,
+	1584,
+	1592,
+	1608,
+	1616,
+	1624,
+	1632,
+	1640,
+	1648,
+	1656,
+	1664,
+	1672,
+	1688,
+	1704,
+	1720,
+	1728,
+	1736,
+	1744,
+	1760,
+	1768,
+	1784,
+	1792,
+	1800,
+	1808,
+	1816,
+	1824,
+	1832,
+	1848,
+	1856,
+	1864,
+	1880,
+	1896,
+	1912,
+	1920,
+	1928,
+	1936,
+	1952,
+	1968,
+	1976,
+	1984,
+	1992,
+	2000,
+	2016,
+	2032,
+	2040,
+	2048,
+	2056,
+	2064,
+	2080,
+	2088,
+	2096,
+	2104,
+	2112,
+	2128,
+	2144,
+	2152,
+	2160,
+	2168,
+	2176,
+	2192,
+	2200,
+	2208,
+	2224,
+	2240,
+	2256,
+	2272,
+	2288,
+	2304,
+	2312,
+	2320,
+	2328,
+	2344,
+	2352,
+	2360,
+	2368,
+	2376,
+	2392,
+	2400,
+	2408,
+	2416,
+	2424,
+	2432,
+	2440,
+	2448,
+	2456,
+	2472,
+	2488,
+	2504,
+	2512,
+	2520,
+	2528,
+	2536,
+	2544,
+	2552,
+	2568,
+	2584,
+	2592,
+	2600,
+	2608,
+	2624,
+	2632,
+	2640,
+	2648,
+	2656,
+	2664,
+	2680,
+	2688,
+	2696,
+	2704,
+	2712,
+	2720,
+	2728,
+	2736,
+	2744,
+	2752,
+	2760,
+	2776,
+	2792,
+	2800,
+	2808,
+	2816,
+	2824,
+	2832,
+	2840,
+	2848,
+	2856,
+	2864,
+	2872,
+	2880,
+	2888,
+	2896,
+	2904,
+	2912,
+	2920,
+	2928,
+	2944,
+	2952,
+	2960,
+	2968,
+	2976,
+	2992,
+	3000,
+	3008,
+	3016,
+	3024,
+	3040,
+	3048,
+	3064,
+	3072,
+	3080,
+	3096,
+	3112,
+	3120,
+	3128,
+	3144,
+	3152,
+	3168,
+	3176,
+	3184,
+	3192,
+	3208,
+	3224,
+	3240,
+	3248,
+	3264,
+	3280,
+	3288,
+	3296,
+	3312,
+	3328,
+	3344,
+	3360,
+	3368,
+	3384,
+	3392,
+	3400,
+	3408,
+	3416,
+	3432,
+	3440,
+	3456,
+	3464,
+	3480,
+	3496,
+	3504,
+	3512,
+	3528,
+	3536,
+	3544,
+	3560,
+	3568,
+	3576,
+	3584,
+	3592,
+	3608,
+	3616,
+	3632,
+	3640,
+	3648,
+	3656,
+	3664,
+	3680,
+	3688,
+	3696,
+	3712,
+	3720,
+	3728,
+	3736,
+	3752,
+	3760,
+	3768,
+	3776,
+	3784,
+	3792,
+	3800,
+	3816,
+	3824,
+	3840,
+	3856,
+	3872,
+	3880,
+	3888,
+	3904,
+	3912,
+	3920,
+	3928,
+	3936,
+	3944,
+	3952,
+	3960,
+	3968,
+	3984,
+	4000,
+	4008,
+	4016,
+	4024,
+	4040,
+	4048,
+	4064,
+	4080,
+	4088,
+	4096,
+	4104,
+	4112,
+	4120,
+	4128,
+	4144,
+	4160,
+	4168,
+	4184,
+	4192,
+	4200,
+	4216,
+	4224,
+	4232,
+	4248,
+	4264,
+	4272,
+	4288,
+	4304,
+	4312,
+	4320,
+	4328,
+	4336,
+	4352,
+	4360,
+	4376,
+	4392,
+	4408,
+	4416,
+	4424,
+	4432,
+	4440,
+	4456,
+	4464,
+	4472,
+	4480,
+	4488,
+	4504,
+	4520,
+	4536,
+	4544,
+	4552,
+	4560,
+	4576,
+	4592,
+	4600,
+	4608,
+	4616,
+	4624,
+	4632,
+	4640,
+	4656,
+	4672,
+	4680,
+	4688,
+	4696,
+	4712,
+	4720,
+	4728,
+	4736,
+	4744,
+	4752,
+	4768,
+	4776,
+	4784,
+	4800,
+	4808,
+	4816,
+	4824,
+	4832,
+	4840,
+	4848,
+	4856,
+	4864,
+	4880,
+	4896,
+	4912,
+	4928,
+	4944,
+	4960,
+	4976,
+	4984,
+	4992,
+	5008,
+	5024,
+	5032,
+	5040,
+};
+
+
+#define SCANKEYWORDS_NUM_KEYWORDS 473
+
diff --git a/contrib/bench_hash/bench_hash--1.0.sql b/contrib/bench_hash/bench_hash--1.0.sql
new file mode 100644
index 0000000000..b3a5747432
--- /dev/null
+++ b/contrib/bench_hash/bench_hash--1.0.sql
@@ -0,0 +1,21 @@
+/* src/test/modules/bench_hash/bench_hash--1.0.sql */
+
+-- complain if script is sourced in psql, rather than via CREATE EXTENSION
+\echo Use "CREATE EXTENSION bench_hash" to load this file. \quit
+
+CREATE FUNCTION bench_string_hash(int4)
+RETURNS int
+AS 'MODULE_PATHNAME'
+LANGUAGE C STRICT;
+
+
+CREATE FUNCTION bench_cstring_hash_unaligned(int4)
+RETURNS int
+AS 'MODULE_PATHNAME'
+LANGUAGE C STRICT;
+
+CREATE FUNCTION bench_cstring_hash_aligned(int4)
+RETURNS int
+AS 'MODULE_PATHNAME'
+LANGUAGE C STRICT;
+
diff --git a/contrib/bench_hash/bench_hash.c b/contrib/bench_hash/bench_hash.c
new file mode 100644
index 0000000000..9c9dba93f0
--- /dev/null
+++ b/contrib/bench_hash/bench_hash.c
@@ -0,0 +1,103 @@
+/*-------------------------------------------------------------------------
+ *
+ * bench_hash.c
+ *
+ * Copyright (c) 2023, PostgreSQL Global Development Group
+ *
+ * IDENTIFICATION
+ *	  src/test/modules/bench_hash/bench_hash.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "fmgr.h"
+
+PG_MODULE_MAGIC;
+
+#include "aligned_keywords.h"
+
+#include "common/hashfn.h"
+#include "common/hashfn_unstable.h"
+#include "miscadmin.h"
+#include "utils/memutils.h"
+
+
+PG_FUNCTION_INFO_V1(bench_string_hash);
+Datum
+bench_string_hash(PG_FUNCTION_ARGS)
+{
+	int32		count = PG_GETARG_INT32(0);
+	uint32 		hash = 0;
+
+	while (count-- > 0)
+	{
+		for (int i=0; i< SCANKEYWORDS_NUM_KEYWORDS; i++)
+		{
+			int idx = word_offsets[i];
+			int s_len = strlen(&aligned_words[idx]);
+			hash += hash_bytes((const unsigned char *) &aligned_words[idx], s_len);
+		}
+		CHECK_FOR_INTERRUPTS();
+	}
+
+	PG_RETURN_INT32(hash);
+}
+
+
+PG_FUNCTION_INFO_V1(bench_cstring_hash_unaligned);
+Datum
+bench_cstring_hash_unaligned(PG_FUNCTION_ARGS)
+{
+	int32		count = PG_GETARG_INT32(0);
+	uint32 		hash = 0;
+
+	char* p = (char*) palloc(5048);
+	memcpy(p, aligned_words, 5048);
+
+	while (count-- > 0)
+	{
+		for (int i=0; i< SCANKEYWORDS_NUM_KEYWORDS; i++)
+		{
+			int idx = word_offsets[i];
+			int s_len;
+			fasthash_state hs;
+
+			fasthash_init(&hs, FH_UNKNOWN_LENGTH, 0);
+			s_len = fasthash_accum_cstring_unaligned(&hs, &p[idx]);
+			hash += fasthash_final32(&hs, s_len);
+		}
+		CHECK_FOR_INTERRUPTS();
+	}
+
+	PG_RETURN_INT32(hash);
+}
+
+
+PG_FUNCTION_INFO_V1(bench_cstring_hash_aligned);
+Datum
+bench_cstring_hash_aligned(PG_FUNCTION_ARGS)
+{
+	int32		count = PG_GETARG_INT32(0);
+	uint32 		hash = 0;
+
+	char* p = (char*) palloc(5048);
+	memcpy(p, aligned_words, 5048);
+
+	while (count-- > 0)
+	{
+		for (int i=0; i< SCANKEYWORDS_NUM_KEYWORDS; i++)
+		{
+			int idx = word_offsets[i];
+			int s_len;
+			fasthash_state hs;
+
+			fasthash_init(&hs, FH_UNKNOWN_LENGTH, 0);
+			s_len = fasthash_accum_cstring_aligned(&hs, &p[idx]);
+			hash += fasthash_final32(&hs, s_len);
+		}
+		CHECK_FOR_INTERRUPTS();
+	}
+
+	PG_RETURN_INT32(hash);
+}
diff --git a/contrib/bench_hash/bench_hash.control b/contrib/bench_hash/bench_hash.control
new file mode 100644
index 0000000000..ffc63858d2
--- /dev/null
+++ b/contrib/bench_hash/bench_hash.control
@@ -0,0 +1,5 @@
+# bench_hash extension
+comment = 'benchmark some hash functions'
+default_version = '1.0'
+module_pathname = '$libdir/bench_hash'
+relocatable = true
diff --git a/contrib/bench_hash/meson.build b/contrib/bench_hash/meson.build
new file mode 100644
index 0000000000..f8d88d8b5c
--- /dev/null
+++ b/contrib/bench_hash/meson.build
@@ -0,0 +1,19 @@
+# Copyright (c) 2022-2023, PostgreSQL Global Development Group
+
+bench_hash_sources = files(
+  'bench_hash.c',
+)
+
+bench_hash = shared_module('bench_hash',
+  bench_hash_sources,
+#  link_with: pgport_srv,
+  kwargs: contrib_mod_args,
+)
+contrib_targets += bench_hash
+
+install_data(
+  'bench_hash.control',
+  'bench_hash--1.0.sql',
+  kwargs: contrib_data_args,
+)
+
diff --git a/contrib/meson.build b/contrib/meson.build
index c0b267c632..0e99195476 100644
--- a/contrib/meson.build
+++ b/contrib/meson.build
@@ -12,6 +12,7 @@ contrib_doc_args = {
   'install_dir': contrib_doc_dir,
 }
 
+subdir('bench_hash')
 subdir('adminpack')
 subdir('amcheck')
 subdir('auth_delay')
-- 
2.43.0