From ba4e26a0086dc5b49405a5ac8b495a64f1c2bb49 Mon Sep 17 00:00:00 2001 From: Neil Conway Date: Sun, 2 Jun 2024 14:00:58 -0400 Subject: [PATCH 2/2] Optimize COPY TO ... FORMAT CSV using SIMD instructions CopyAttributeOutCSV() does one or two byte-by-byte loops over the text of each attribute, depending on whether quotation is required. Implementing this loops using SIMD yields a significant speedup for long attribute values. For short attribute values, performance is roughly unchanged. Using SIMD when encoding_embeds_ascii is true seems quite complex, so for now we just use the old code path for such encodings. --- src/backend/commands/copyto.c | 145 +++++++++++++++++++++++++++++++++- 1 file changed, 142 insertions(+), 3 deletions(-) diff --git a/src/backend/commands/copyto.c b/src/backend/commands/copyto.c index 48957d8c3e..a62c4ec120 100644 --- a/src/backend/commands/copyto.c +++ b/src/backend/commands/copyto.c @@ -29,6 +29,7 @@ #include "mb/pg_wchar.h" #include "miscadmin.h" #include "pgstat.h" +#include "port/simd.h" #include "storage/fd.h" #include "tcop/tcopprot.h" #include "utils/lsyscache.h" @@ -1124,6 +1125,137 @@ CopyAttributeOutText(CopyToState cstate, const char *string) DUMPSOFAR(); } +/* + * Send text representation of one attribute, with conversion and CSV-style + * escaping. This variant uses SIMD instructions to optimize processing, but + * we can only use this approach when encoding_embeds_ascii if false. + */ +static void +CopyAttributeOutCSVFast(CopyToState cstate, const char *ptr, + bool use_quote) +{ + int len; + int vlen; + char delimc = cstate->opts.delim[0]; + char quotec = cstate->opts.quote[0]; + char escapec = cstate->opts.escape[0]; + + len = strlen(ptr); + vlen = len & (int) (~(sizeof(Vector8) - 1)); + + /* + * Make a preliminary pass to discover if it needs quoting + */ + if (!use_quote) + { + bool single_attr = (list_length(cstate->attnumlist) == 1); + + /* + * Because '\.' can be a data value, quote it if it appears alone on a + * line so it is not interpreted as the end-of-data marker. + */ + if (single_attr && strcmp(ptr, "\\.") == 0) + use_quote = true; + else + { + int i; + Vector8 chunk; + + for (i = 0; i < vlen; i += sizeof(Vector8)) + { + vector8_load(&chunk, (const uint8 *) &ptr[i]); + + if (vector8_has(chunk, (unsigned char) delimc) || + vector8_has(chunk, (unsigned char) quotec) || + vector8_has(chunk, (unsigned char) '\n') || + vector8_has(chunk, (unsigned char) '\r')) + { + use_quote = true; + break; + } + } + + /* Check the tail of the string */ + if (!use_quote) + { + for (; i < len; i++) + { + char c = ptr[i]; + + if (c == delimc || c == quotec || c == '\n' || c == '\r') + { + use_quote = true; + break; + } + } + } + } + } + + if (use_quote) + { + int i; + int start_idx = 0; + Vector8 chunk; + + CopySendChar(cstate, quotec); + + for (i = 0; i < vlen; i += sizeof(Vector8)) + { + vector8_load(&chunk, (const uint8 *) &ptr[i]); + + if (vector8_has(chunk, (unsigned char) delimc) || + vector8_has(chunk, (unsigned char) quotec)) + { + /* + * This chunk has one or more characters that require + * escaping, so switch to byte-at-a-time processing + */ + for (int j = i; j < (i + sizeof(Vector8)); j++) + { + char c = ptr[j]; + + if (c == quotec || c == escapec) + { + if (j > start_idx) + CopySendData(cstate, ptr + start_idx, j - start_idx); + + CopySendChar(cstate, escapec); + start_idx = j; + } + } + } + } + + /* Process the tail of the string */ + for (; i < len; i++) + { + char c = ptr[i]; + + if (c == quotec || c == escapec) + { + if (i > start_idx) + CopySendData(cstate, ptr + start_idx, i - start_idx); + + CopySendChar(cstate, escapec); + start_idx = i; + } + } + + /* Send any remaining text */ + if (start_idx < len) + CopySendData(cstate, ptr + start_idx, len - start_idx); + + CopySendChar(cstate, quotec); + } + else + { + /* If it doesn't need quoting, we can just dump it as-is */ + CopySendData(cstate, ptr, len); + } +} + + /* * Send text representation of one attribute, with conversion and * CSV-style escaping @@ -1138,7 +1270,6 @@ CopyAttributeOutCSV(CopyToState cstate, const char *string, char delimc = cstate->opts.delim[0]; char quotec = cstate->opts.quote[0]; char escapec = cstate->opts.escape[0]; - bool single_attr = (list_length(cstate->attnumlist) == 1); /* force quoting if it matches null_print (before conversion!) */ if (!use_quote && strcmp(string, cstate->opts.null_print) == 0) @@ -1149,11 +1280,19 @@ CopyAttributeOutCSV(CopyToState cstate, const char *string, else ptr = string; + if (!cstate->encoding_embeds_ascii) + { + CopyAttributeOutCSVFast(cstate, ptr, use_quote); + return; + } + /* * Make a preliminary pass to discover if it needs quoting */ if (!use_quote) { + bool single_attr = (list_length(cstate->attnumlist) == 1); + /* * Because '\.' can be a data value, quote it if it appears alone on a * line so it is not interpreted as the end-of-data marker. @@ -1171,7 +1310,7 @@ CopyAttributeOutCSV(CopyToState cstate, const char *string, use_quote = true; break; } - if (IS_HIGHBIT_SET(c) && cstate->encoding_embeds_ascii) + if (IS_HIGHBIT_SET(c)) tptr += pg_encoding_mblen(cstate->file_encoding, tptr); else tptr++; @@ -1195,7 +1334,7 @@ CopyAttributeOutCSV(CopyToState cstate, const char *string, CopySendChar(cstate, escapec); start = ptr; /* we include char in next run */ } - if (IS_HIGHBIT_SET(c) && cstate->encoding_embeds_ascii) + if (IS_HIGHBIT_SET(c)) ptr += pg_encoding_mblen(cstate->file_encoding, ptr); else ptr++; -- 2.39.3 (Apple Git-146)