speed up hunspell inner loop

which calls into the kernel to get elapsed time, instead of using the
VDSO-based CLOCK_MONOTONIC (or at least, the C++ equivalent of that)
, which is much faster

Change-Id: I99d958b0ab64b75360db4e0c8a951c37af2505bc
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/158809
Tested-by: Jenkins
Tested-by: Caolán McNamara <caolan.mcnamara@collabora.com>
Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com>

2 files changed, 106 insertions, 0 deletions

diff --git a/external/hunspell/UnpackedTarball_hunspell.mk b/external/hunspell/UnpackedTarball_hunspell.mk
index 9075d98659e7..0d5986798265 100644
--- a/external/hunspell/UnpackedTarball_hunspell.mk
+++ b/external/hunspell/UnpackedTarball_hunspell.mk
@@ -26,6 +26,7 @@ $(eval $(call gb_UnpackedTarball_add_patches,hunspell, \
 	external/hunspell/0001-Resolves-rhbz-2158548-allow-longer-words-for-hunspel.patch \
 	external/hunspell/0001-Keep-only-REP-ph-or-2-word-dictionary-phrase-suggest.patch \
 	external/hunspell/bit_cast.patch.0 \
+	external/hunspell/clock-monotonic.patch.1 \
 ))
 
 # vim: set noet sw=4 ts=4:
diff --git a/external/hunspell/clock-monotonic.patch.1 b/external/hunspell/clock-monotonic.patch.1
new file mode 100644
index 000000000000..dd873f6c3908
--- /dev/null
+++ b/external/hunspell/clock-monotonic.patch.1
@@ -0,0 +1,105 @@
+From 5737bdb3d7e5819528e33c360a73372e0e93a6be Mon Sep 17 00:00:00 2001
+From: Noel Grandin <noel.grandin@collabora.co.uk>
+Date: Fri, 3 Nov 2023 12:04:30 +0000
+Subject: [PATCH] speed up hunspell inner loop
+
+which calls into the kernel to get elapsed time, instead of using the
+VDSO-based CLOCK_MONOTONIC (or at least, the C++ equivalent of that),
+which is much faster
+
+https://gerrit.libreoffice.org/c/core/+/158809
+---
+ src/hunspell/affixmgr.cxx | 43 +++++++++++++++++++++++----------------
+ 1 file changed, 25 insertions(+), 18 deletions(-)
+
+diff --git a/src/hunspell/affixmgr.cxx b/src/hunspell/affixmgr.cxx
+index 2cad09f..a3c93cd 100644
+--- a/src/hunspell/affixmgr.cxx
++++ b/src/hunspell/affixmgr.cxx
+@@ -75,6 +75,7 @@
+ #include <ctime>
+ 
+ #include <algorithm>
++#include <chrono>
+ #include <memory>
+ #include <limits>
+ #include <string>
+@@ -1590,17 +1591,20 @@ struct hentry* AffixMgr::compound_check(const std::string& word,
+   // add a time limit to handle possible
+   // combinatorical explosion of the overlapping words
+ 
+-  HUNSPELL_THREAD_LOCAL clock_t timelimit;
++  HUNSPELL_THREAD_LOCAL std::chrono::steady_clock::time_point clock_time_start;
++  HUNSPELL_THREAD_LOCAL bool timelimit_exceeded;
++
++  // get the current time
++  std::chrono::steady_clock::time_point clock_now = std::chrono::steady_clock::now();
+ 
+   if (wordnum == 0) {
+-      // get the start time, seeing as we're reusing this set to 0
+-      // to flag timeout, use clock() + 1 to avoid start clock()
+-      // of 0 as being a timeout
+-      timelimit = clock() + 1;
+-  }
+-  else if (timelimit != 0 && (clock() > timelimit + TIMELIMIT)) {
+-      timelimit = 0;
++      // set the start time
++      clock_time_start = clock_now;
++      timelimit_exceeded = false;
+   }
++  else if (std::chrono::duration_cast<std::chrono::milliseconds>(clock_now - clock_time_start).count()
++            > TIMELIMIT * CLOCKS_PER_SEC * 1000)
++      timelimit_exceeded = true;
+ 
+   setcminmax(&cmin, &cmax, word.c_str(), len);
+ 
+@@ -1626,7 +1630,7 @@ struct hentry* AffixMgr::compound_check(const std::string& word,
+ 
+       do {  // simplified checkcompoundpattern loop
+ 
+-        if (timelimit == 0)
++        if (timelimit_exceeded)
+           return 0;
+ 
+         if (scpd > 0) {
+@@ -2216,17 +2220,20 @@ int AffixMgr::compound_check_morph(const std::string& word,
+   // add a time limit to handle possible
+   // combinatorical explosion of the overlapping words
+ 
+-  HUNSPELL_THREAD_LOCAL clock_t timelimit;
++  HUNSPELL_THREAD_LOCAL std::chrono::steady_clock::time_point clock_time_start;
++  HUNSPELL_THREAD_LOCAL bool timelimit_exceeded;
++
++  // get the current time
++  std::chrono::steady_clock::time_point clock_now = std::chrono::steady_clock::now();
+ 
+   if (wordnum == 0) {
+-      // get the start time, seeing as we're reusing this set to 0
+-      // to flag timeout, use clock() + 1 to avoid start clock()
+-      // of 0 as being a timeout
+-      timelimit = clock() + 1;
+-  }
+-  else if (timelimit != 0 && (clock() > timelimit + TIMELIMIT)) {
+-      timelimit = 0;
++      // set the start time
++      clock_time_start = clock_now;
++      timelimit_exceeded = false;
+   }
++  else if (std::chrono::duration_cast<std::chrono::milliseconds>(clock_now - clock_time_start).count()
++            > TIMELIMIT * CLOCKS_PER_SEC * 1000)
++      timelimit_exceeded = true;
+ 
+   setcminmax(&cmin, &cmax, word.c_str(), len);
+ 
+@@ -2246,7 +2253,7 @@ int AffixMgr::compound_check_morph(const std::string& word,
+ 
+     do {  // onlycpdrule loop
+ 
+-      if (timelimit == 0)
++      if (timelimit_exceeded)
+         return 0;
+ 
+       oldnumsyllable = numsyllable;
+-- 
+2.41.0
+