Optimize hex_encode() and hex_decode() using SIMD.

The hex_encode() and hex_decode() functions serve as the workhorses
for hexadecimal data for bytea's text format conversion functions,
and some workloads are sensitive to their performance.  This commit
adds new implementations that use routines from port/simd.h, which
testing indicates are much faster for larger inputs.  For small or
invalid inputs, we fall back on the existing scalar versions.
Since we are using port/simd.h, these optimizations apply to both
x86-64 and AArch64.

Author: Nathan Bossart <nathandbossart@gmail.com>
Co-authored-by: Chiranmoy Bhattacharya <chiranmoy.bhattacharya@fujitsu.com>
Co-authored-by: Susmitha Devanga <devanga.susmitha@fujitsu.com>
Reviewed-by: John Naylor <johncnaylorls@gmail.com>
Discussion: https://postgr.es/m/aLhVWTRy0QPbW2tl%40nathan

2 files changed, 74 insertions, 0 deletions

diff --git a/src/test/regress/expected/strings.out b/src/test/regress/expected/strings.out
index 691e475bce3..b9dc08d5f61 100644
--- a/src/test/regress/expected/strings.out
+++ b/src/test/regress/expected/strings.out
@@ -260,6 +260,64 @@ SELECT reverse('\xabcd'::bytea);
  \xcdab
 (1 row)
 
+SELECT ('\x' || repeat(' ', 32))::bytea;
+ bytea 
+-------
+ \x
+(1 row)
+
+SELECT ('\x' || repeat('!', 32))::bytea;
+ERROR:  invalid hexadecimal digit: "!"
+SELECT ('\x' || repeat('/', 34))::bytea;
+ERROR:  invalid hexadecimal digit: "/"
+SELECT ('\x' || repeat('0', 34))::bytea;
+                bytea                 
+--------------------------------------
+ \x0000000000000000000000000000000000
+(1 row)
+
+SELECT ('\x' || repeat('9', 32))::bytea;
+               bytea                
+------------------------------------
+ \x99999999999999999999999999999999
+(1 row)
+
+SELECT ('\x' || repeat(':', 32))::bytea;
+ERROR:  invalid hexadecimal digit: ":"
+SELECT ('\x' || repeat('@', 34))::bytea;
+ERROR:  invalid hexadecimal digit: "@"
+SELECT ('\x' || repeat('A', 34))::bytea;
+                bytea                 
+--------------------------------------
+ \xaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
+(1 row)
+
+SELECT ('\x' || repeat('F', 32))::bytea;
+               bytea                
+------------------------------------
+ \xffffffffffffffffffffffffffffffff
+(1 row)
+
+SELECT ('\x' || repeat('G', 32))::bytea;
+ERROR:  invalid hexadecimal digit: "G"
+SELECT ('\x' || repeat('`', 34))::bytea;
+ERROR:  invalid hexadecimal digit: "`"
+SELECT ('\x' || repeat('a', 34))::bytea;
+                bytea                 
+--------------------------------------
+ \xaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
+(1 row)
+
+SELECT ('\x' || repeat('f', 32))::bytea;
+               bytea                
+------------------------------------
+ \xffffffffffffffffffffffffffffffff
+(1 row)
+
+SELECT ('\x' || repeat('g', 32))::bytea;
+ERROR:  invalid hexadecimal digit: "g"
+SELECT ('\x' || repeat('~', 34))::bytea;
+ERROR:  invalid hexadecimal digit: "~"
 SET bytea_output TO escape;
 SELECT E'\\xDeAdBeEf'::bytea;
       bytea       
diff --git a/src/test/regress/sql/strings.sql b/src/test/regress/sql/strings.sql
index c05f3413699..a2a91523404 100644
--- a/src/test/regress/sql/strings.sql
+++ b/src/test/regress/sql/strings.sql
@@ -82,6 +82,22 @@ SELECT reverse(''::bytea);
 SELECT reverse('\xaa'::bytea);
 SELECT reverse('\xabcd'::bytea);
 
+SELECT ('\x' || repeat(' ', 32))::bytea;
+SELECT ('\x' || repeat('!', 32))::bytea;
+SELECT ('\x' || repeat('/', 34))::bytea;
+SELECT ('\x' || repeat('0', 34))::bytea;
+SELECT ('\x' || repeat('9', 32))::bytea;
+SELECT ('\x' || repeat(':', 32))::bytea;
+SELECT ('\x' || repeat('@', 34))::bytea;
+SELECT ('\x' || repeat('A', 34))::bytea;
+SELECT ('\x' || repeat('F', 32))::bytea;
+SELECT ('\x' || repeat('G', 32))::bytea;
+SELECT ('\x' || repeat('`', 34))::bytea;
+SELECT ('\x' || repeat('a', 34))::bytea;
+SELECT ('\x' || repeat('f', 32))::bytea;
+SELECT ('\x' || repeat('g', 32))::bytea;
+SELECT ('\x' || repeat('~', 34))::bytea;
+
 SET bytea_output TO escape;
 SELECT E'\\xDeAdBeEf'::bytea;
 SELECT E'\\x De Ad Be Ef '::bytea;