xref: /third_party/glib/patch/backport-gregex-set-default-max-stack-size-for-PCRE2-JIT-compiler-to-512KiB.patch

From 842a105464f6390a433da8791d7b19b65df16f47 Mon Sep 17 00:00:00 2001
From: Aleksei Rybalkin <aleksei@rybalkin.org>
Date: Mon, 14 Aug 2023 20:32:48 +0200
Subject: [PATCH 1/2] gregex: remove redundant call to
 enable_jit_with_match_options
 
There is no point to enable jit in g_regex_new, since JIT will be only
used when we do a first match, and at that point
enable_jit_with_match_options will be called again already and will
update the options set in g_regex_new. Instead just run it at first
match for the first time, to the same end result.
 
Conflict:NA
Reference:https://gitlab.gnome.org/GNOME/glib/-/commit/842a105464f6390a433da8791d7b19b65df16f47
 
---
 glib/gregex.c | 1 -
 1 file changed, 1 deletion(-)
 
diff --git a/glib/gregex.c b/glib/gregex.c
index 39b9edeecd..f6b2b716fc 100644
--- a/glib/gregex.c
+++ b/glib/gregex.c
@@ -1764,7 +1764,6 @@ G_GNUC_END_IGNORE_DEPRECATIONS
   regex->orig_compile_opts = compile_options;
   regex->match_opts = pcre_match_options;
   regex->orig_match_opts = match_options;
-  regex->jit_status = enable_jit_with_match_options (regex, regex->match_opts);
 
   return regex;
 }
-- 
GitLab
 
 
From c3ff5b8eb39f1ab31383604910ae12f325e5afee Mon Sep 17 00:00:00 2001
From: Aleksei Rybalkin <aleksei@rybalkin.org>
Date: Mon, 14 Aug 2023 20:41:40 +0200
Subject: [PATCH 2/2] gregex: set default max stack size for PCRE2 JIT compiler
 to 512KiB
 
Previous default used was 32KiB (the library default) which caused some
complex patterns to fail, see #2824. The memory will not be allocated
unless used.
 
Conflict:Move test_string to fix declaration-after-statement
Reference:https://gitlab.gnome.org/GNOME/glib/-/commit/c3ff5b8eb39f1ab31383604910ae12f325e5afee
 
---
 glib/gregex.c      | 22 ++++++++++++++--------
 glib/tests/regex.c |  9 +++++++++
 2 files changed, 23 insertions(+), 8 deletions(-)
 
diff --git a/glib/gregex.c b/glib/gregex.c
index f6b2b716fc..5ce034db41 100644
--- a/glib/gregex.c
+++ b/glib/gregex.c
@@ -232,6 +232,7 @@ struct _GMatchInfo
   gssize string_len;            /* length of string, in bytes */
   pcre2_match_context *match_context;
   pcre2_match_data *match_data;
+  pcre2_jit_stack *jit_stack;
 };
 
 typedef enum
@@ -896,22 +897,22 @@ recalc_match_offsets (GMatchInfo *match_info,
 }
 
 static JITStatus
-enable_jit_with_match_options (GRegex   *regex,
+enable_jit_with_match_options (GMatchInfo  *match_info,
                                uint32_t  match_options)
 {
   gint retval;
   uint32_t old_jit_options, new_jit_options;
 
-  if (!(regex->orig_compile_opts & G_REGEX_OPTIMIZE))
+  if (!(match_info->regex->orig_compile_opts & G_REGEX_OPTIMIZE))
     return JIT_STATUS_DISABLED;
 
-  if (regex->jit_status == JIT_STATUS_DISABLED)
+  if (match_info->regex->jit_status == JIT_STATUS_DISABLED)
     return JIT_STATUS_DISABLED;
 
   if (match_options & G_REGEX_PCRE2_JIT_UNSUPPORTED_OPTIONS)
     return JIT_STATUS_DISABLED;
 
-  old_jit_options = regex->jit_options;
+  old_jit_options = match_info->regex->jit_options;
   new_jit_options = old_jit_options | PCRE2_JIT_COMPLETE;
   if (match_options & PCRE2_PARTIAL_HARD)
     new_jit_options |= PCRE2_JIT_PARTIAL_HARD;
@@ -920,13 +921,16 @@ enable_jit_with_match_options (GRegex   *regex,
 
   /* no new options enabled */
   if (new_jit_options == old_jit_options)
-    return regex->jit_status;
+    return match_info->regex->jit_status;
 
-  retval = pcre2_jit_compile (regex->pcre_re, new_jit_options);
+  retval = pcre2_jit_compile (match_info->regex->pcre_re, new_jit_options);
   switch (retval)
     {
     case 0: /* JIT enabled successfully */
-      regex->jit_options = new_jit_options;
+      match_info->regex->jit_options = new_jit_options;
+      /* Set min stack size for JIT to 32KiB and max to 512KiB */
+      match_info->jit_stack = pcre2_jit_stack_create (1 << 15, 1 << 19, NULL);
+      pcre2_jit_stack_assign (match_info->match_context, NULL, match_info->jit_stack);
       return JIT_STATUS_ENABLED;
     case PCRE2_ERROR_NOMEMORY:
       g_debug ("JIT compilation was requested with G_REGEX_OPTIMIZE, "
@@ -1023,6 +1027,8 @@ g_match_info_unref (GMatchInfo *match_info)
       g_regex_unref (match_info->regex);
       if (match_info->match_context)
         pcre2_match_context_free (match_info->match_context);
+      if (match_info->jit_stack)
+        pcre2_jit_stack_free (match_info->jit_stack);
       if (match_info->match_data)
         pcre2_match_data_free (match_info->match_data);
       g_free (match_info->offsets);
@@ -1091,7 +1097,7 @@ g_match_info_next (GMatchInfo  *match_info,
 
   opts = match_info->regex->match_opts | match_info->match_opts;
 
-  jit_status = enable_jit_with_match_options (match_info->regex, opts);
+  jit_status = enable_jit_with_match_options (match_info, opts);
   if (jit_status == JIT_STATUS_ENABLED)
     {
       match_info->matches = pcre2_jit_match (match_info->regex->pcre_re,
diff --git a/glib/tests/regex.c b/glib/tests/regex.c
index cf2bb8199d..821fc59608 100644
--- a/glib/tests/regex.c
+++ b/glib/tests/regex.c
@@ -51,6 +51,9 @@
 /* A random value use to mark untouched integer variables. */
 #define UNTOUCHED -559038737
 
+/* A length of the test string in JIT stack test */
+#define TEST_STRING_LEN 20000
+
 static gint total;
 
 typedef struct {
@@ -2481,6 +2484,7 @@ test_jit_unsupported_matching_options (void)
 int
 main (int argc, char *argv[])
 {
+  char test_string[TEST_STRING_LEN];
   setlocale (LC_ALL, "");
 
   g_test_init (&argc, &argv, NULL);
@@ -2702,6 +2706,11 @@ G_GNUC_END_IGNORE_DEPRECATIONS
   TEST_MATCH_SIMPLE("\\", "a", 0, 0, FALSE);
   TEST_MATCH_SIMPLE("[", "", 0, 0, FALSE);
 
+  /* Test that JIT compiler has enough stack */
+  memset (test_string, '*', TEST_STRING_LEN);
+  test_string[TEST_STRING_LEN - 1] = '\0';
+  TEST_MATCH_SIMPLE ("^(?:[ \t\n]|[^[:cntrl:]])*$", test_string, 0, 0, TRUE);
+
   /* TEST_MATCH(pattern, compile_opts, match_opts, string,
    * 		string_len, start_position, match_opts2, expected) */
   TEST_MATCH("a", 0, 0, "a", -1, 0, 0, TRUE);
-- 
GitLab