[exim-cvs] ACL: in "regex" condition, release store every th…

Gitweb: https://git.exim.org/exim.git/commitdiff/35aacb69f5c839a4b77158464e401d86eb422ed6

Commit:     35aacb69f5c839a4b77158464e401d86eb422ed6
Parent:     b4e7527561f1c68b821d5cf25efe29ae63d1d434
Author:     Jeremy Harris <jgh146exb@???>
AuthorDate: Fri Jan 26 21:58:59 2024 +0000
Committer:  Jeremy Harris <jgh146exb@???>
CommitDate: Fri Jan 26 21:58:59 2024 +0000

    ACL: in "regex" condition, release store every thousand lines.  Bug 3047
---
 doc/doc-txt/ChangeLog           |  2 ++
 src/src/acl.c                   |  4 +--
 src/src/macros.h                |  5 +++
 src/src/regex.c                 | 77 ++++++++++++++++++++++++-----------------
 src/src/transports/appendfile.c |  9 ++---
 5 files changed, 57 insertions(+), 40 deletions(-)

diff --git a/doc/doc-txt/ChangeLog b/doc/doc-txt/ChangeLog
index e258966d8..48cc62910 100644
--- a/doc/doc-txt/ChangeLog
+++ b/doc/doc-txt/ChangeLog
@@ -99,6 +99,8 @@ JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
       when a user had over 104207 messages stored and the appendfile
       maildir_quota_directory_regex option is in use.  Release the allocated
       memory every thosand files to avoid this.
+      The same issue arises with the ACL regex condition, which is applied
+      to every line of a received message.

 Exim version 4.97
diff --git a/src/src/acl.c b/src/src/acl.c
index 9223fcec8..ecba07b25 100644
--- a/src/src/acl.c
+++ b/src/src/acl.c
@@ -3955,11 +3955,11 @@ for (; cb; cb = cb->next)
     CUSS &recipient_data);
       break;

-    #ifdef WITH_CONTENT_SCAN
+#ifdef WITH_CONTENT_SCAN
     case ACLC_REGEX:
       rc = regex(&arg, textonly);
       break;
-    #endif
+#endif

     case ACLC_REMOVE_HEADER:
       setup_remove_header(arg);
diff --git a/src/src/macros.h b/src/src/macros.h
index 5279f70d0..8aed335b5 100644
--- a/src/src/macros.h
+++ b/src/src/macros.h
@@ -1185,4 +1185,9 @@ typedef enum {
   sw_mrc_tx_fail,        /* transmit failed */
 } sw_mrc_t;

+/* Recent versions of PCRE2 are allocating 20kB per match, rather than the previous 112 B.
+When doing en extended loop of matching, release store periodically. */
+
+#define    REGEX_LOOPCOUNT_STORE_RESET    1000
+
 /* End of macros.h */
diff --git a/src/src/regex.c b/src/src/regex.c
index af425db2d..cda807716 100644
--- a/src/src/regex.c
+++ b/src/src/regex.c
@@ -31,12 +31,11 @@ extern uschar *mime_current_boundary;

static pcre_list *
-compile(const uschar * list, BOOL cacheable)
+compile(const uschar * list, BOOL cacheable, int * cntp)
{
-int sep = 0;
+int sep = 0, cnt = 0;
uschar * regex_string;
-pcre_list * re_list_head = NULL;
-pcre_list * ri;
+pcre_list * re_list_head = NULL, * ri;

 /* precompile our regexes */
 while ((regex_string = string_nextinlist(&list, &sep, NULL, 0)))
@@ -58,7 +57,9 @@ while ((regex_string = string_nextinlist(&list, &sep, NULL, 0)))
     ri->pcre_text = regex_string;
     ri->next = re_list_head;
     re_list_head = ri;
+    cnt++;
     }
+if (cntp) *cntp = cnt;
 return re_list_head;
 }

@@ -112,7 +113,8 @@ FILE * mbox_file;
pcre_list * re_list_head;
uschar * linebuffer;
long f_pos = 0;
-int ret = FAIL;
+int ret = FAIL, cnt, lcount = REGEX_LOOPCOUNT_STORE_RESET;
+rmark reset_point;

regex_vars_clear();

@@ -136,26 +138,34 @@ else
mbox_file = mime_stream;
}

-/* precompile our regexes */
-if (!(re_list_head = compile(*listptr, cacheable)))
-  return FAIL;            /* no regexes -> nothing to do */
-
-/* match each line against all regexes */
-linebuffer = store_get(32767, GET_TAINTED);
-while (fgets(CS linebuffer, 32767, mbox_file))
+reset_point = store_mark();
   {
-  if (  mime_stream && mime_current_boundary        /* check boundary */
-     && Ustrncmp(linebuffer, "--", 2) == 0
-     && Ustrncmp((linebuffer+2), mime_current_boundary,
-          Ustrlen(mime_current_boundary)) == 0)
-      break;                        /* found boundary */
-
-  if ((ret = matcher(re_list_head, linebuffer, (int)Ustrlen(linebuffer))) == OK)
-    goto done;
+  /* precompile our regexes */
+  if ((re_list_head = compile(*listptr, cacheable, &cnt)))
+    {
+    /* match each line against all regexes */
+    linebuffer = store_get(32767, GET_TAINTED);
+    while (fgets(CS linebuffer, 32767, mbox_file))
+      {
+      if (  mime_stream && mime_current_boundary        /* check boundary */
+     && Ustrncmp(linebuffer, "--", 2) == 0
+     && Ustrncmp((linebuffer+2), mime_current_boundary,
+              Ustrlen(mime_current_boundary)) == 0)
+    break;                        /* found boundary */
+
+      if ((ret = matcher(re_list_head, linebuffer, (int)Ustrlen(linebuffer))) == OK)
+    break;
+
+      if ((lcount -= cnt) <= 0)
+    {
+    store_reset(reset_point); reset_point = store_mark();
+    lcount = REGEX_LOOPCOUNT_STORE_RESET;
+    }
+      }
+    }
   }
-/* no matches ... */
+store_reset(reset_point);

-done:
if (!mime_stream)
(void)fclose(mbox_file);
else
@@ -180,14 +190,11 @@ pcre_list * re_list_head = NULL;
FILE * f;
uschar * mime_subject = NULL;
int mime_subject_len = 0;
-int ret;
+int ret = FAIL;
+rmark reset_point;

regex_vars_clear();

-/* precompile our regexes */
-if (!(re_list_head = compile(*listptr, cacheable)))
-  return FAIL;            /* no regexes -> nothing to do */
-
 /* check if the file is already decoded */
 if (!mime_decoded_filename)
   {                /* no, decode it first */
@@ -210,12 +217,20 @@ if (!(f = fopen(CS mime_decoded_filename, "rb")))
   return DEFER;
   }

-/* get 32k memory, tainted */
-mime_subject = store_get(32767, GET_TAINTED);
+reset_point = store_mark();
+  {
+  /* precompile our regexes */
+  if ((re_list_head = compile(*listptr, cacheable, NULL)))
+    {
+    /* get 32k memory, tainted */
+    mime_subject = store_get(32767, GET_TAINTED);

-mime_subject_len = fread(mime_subject, 1, 32766, f);
+    mime_subject_len = fread(mime_subject, 1, 32766, f);

-ret = matcher(re_list_head, mime_subject, mime_subject_len);
+    ret = matcher(re_list_head, mime_subject, mime_subject_len);
+    }
+  }
+store_reset(reset_point);
 (void)fclose(f);
 return ret;
 }
diff --git a/src/src/transports/appendfile.c b/src/src/transports/appendfile.c
index 91b353079..ce52cc6ff 100644
--- a/src/src/transports/appendfile.c
+++ b/src/src/transports/appendfile.c
@@ -153,11 +153,6 @@ static const char *mailbox_formats[] = {
   (!ob->quota_warn_threshold_is_percent || ob->quota_value > 0))

-/* Free memory allocated by PCRE2 every so often, because a recent version
-is now using 20kB for every match call */
-
-#define RESET_STORE_FILECNT    1000
-
 /*************************************************
 *              Setup entry point                 *
 *************************************************/
@@ -667,7 +662,7 @@ check_dir_size(const uschar * dirname, int * countptr, const pcre2_code * re)
 {
 DIR * dir;
 off_t sum = 0;
-int count = *countptr, lcount = RESET_STORE_FILECNT;
+int count = *countptr, lcount = REGEX_LOOPCOUNT_STORE_RESET;
 rmark reset_point = store_mark();

 if (!(dir = exim_opendir(dirname))) return 0;
@@ -683,7 +678,7 @@ for (struct dirent * ent; ent = readdir(dir); )
   if (--lcount == 0)
     {
     store_reset(reset_point); reset_point = store_mark();
-    lcount = RESET_STORE_FILECNT;
+    lcount = REGEX_LOOPCOUNT_STORE_RESET;
     }

/* If there's a regex, try to find the size using it */

--
## subscription configuration (requires account):
## https://lists.exim.org/mailman3/postorius/lists/exim-cvs.lists.exim.org/
## unsubscribe (doesn't require an account):
## exim-cvs-unsubscribe@???
## Exim details at http://www.exim.org/
## Please use the Wiki with this list - http://wiki.exim.org/