Fix a bug in the compilation of PCRE2 patterns under JIT (the most
common runtime configuration). Any pattern with a (*NO_JIT) verb would
segfault in any currently released PCRE2 version:
$ git grep -P '(*NO_JIT)hi.*there'
Segmentation fault
That this segfaulted was a bug in PCRE2 itself, after reporting it[1]
on pcre-dev it's been fixed in a yet-to-be-released version of
PCRE (presumably released first as 10.31). Now it'll die with:
$ git grep -P '(*NO_JIT)hi.*there'
fatal: pcre2_jit_match failed with error code -45: bad JIT option
But the cause of the bug is in our own code dating back to my
94da9193a6 ("grep: add support for PCRE v2", 2017-06-01).
As explained at more length in the comment being added here, it isn't
sufficient to just check pcre2_config() to see whether the JIT should
be used, pcre2_pattern_info() also has to be asked.
This is something I discovered myself when fiddling around with PCRE2
verbs in patterns passed to git. I don't expect that any user of git
has encountered this given the obscurity of passing PCRE2 verbs
through to the library, along with the relative obscurity of (*NO_JIT)
itself.
1. "How am I supposed to use PCRE2 JIT in the face of (*NO_JIT) ?"
(<CACBZZX5mMqDuWuFmi7sRBp3wH6CFyd-ghACukd=v0NN=rBMnJg@xxxxxxxxxxxxxx> &
https://lists.exim.org/lurker/thread/20171123.101502.7f0d38ca.en.html)
on the pcre-dev mailing list
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@xxxxxxxxx>
---
Incorporates feedback from Eric & Simon. Thanks both. I also amended
the commit message / comment to note that this was also a bug in PCRE2
upstream, which has been fixed after I reported it.
grep.c | 26 ++++++++++++++++++++++++++
t/t7810-grep.sh | 6 ++++++
2 files changed, 32 insertions(+)
diff --git a/grep.c b/grep.c
index d0b9b6cdfa..e8ae0b5d8f 100644
--- a/grep.c
+++ b/grep.c
@@ -477,6 +477,8 @@ static void compile_pcre2_pattern(struct grep_pat *p, const struct grep_opt *opt
int options = PCRE2_MULTILINE;
const uint8_t *character_tables = NULL;
int jitret;
+ int patinforet;
+ size_t jitsizearg;
assert(opt->pcre2);
@@ -511,6 +513,30 @@ static void compile_pcre2_pattern(struct grep_pat *p, const struct grep_opt *opt
jitret = pcre2_jit_compile(p->pcre2_pattern, PCRE2_JIT_COMPLETE);
if (jitret)
die("Couldn't JIT the PCRE2 pattern '%s', got '%d'\n", p->pattern, jitret);
+
+ /*
+ * The pcre2_config(PCRE2_CONFIG_JIT, ...) call just
+ * tells us whether the library itself supports JIT,
+ * but to see whether we're going to be actually using
+ * JIT we need to extract PCRE2_INFO_JITSIZE from the
+ * pattern *after* we do pcre2_jit_compile() above.
+ *
+ * This is because if the pattern contains the
+ * (*NO_JIT) verb (see pcre2syntax(3))
+ * pcre2_jit_compile() will exit early with 0. If we
+ * then proceed to call pcre2_jit_match() further down
+ * the line instead of pcre2_match() we'll either
+ * segfault (pre PCRE 10.31) or run into a fatal error
+ * (post PCRE2 10.31)
+ */
+ patinforet = pcre2_pattern_info(p->pcre2_pattern, PCRE2_INFO_JITSIZE, &jitsizearg);
+ if (patinforet)
+ BUG("pcre2_pattern_info() failed: %d", patinforet);
+ if (jitsizearg == 0) {
+ p->pcre2_jit_on = 0;
+ return;
+ }
+
p->pcre2_jit_stack = pcre2_jit_stack_create(1, 1024 * 1024, NULL);
if (!p->pcre2_jit_stack)
die("Couldn't allocate PCRE2 JIT stack");
diff --git a/t/t7810-grep.sh b/t/t7810-grep.sh
index 2a6679c2f5..c8ff50cc30 100755
--- a/t/t7810-grep.sh
+++ b/t/t7810-grep.sh
@@ -1110,6 +1110,12 @@ test_expect_success PCRE 'grep -P pattern' '
test_cmp expected actual
'
+test_expect_success LIBPCRE2 "grep -P with (*NO_JIT) doesn't error out" '
+ git grep -P "(*NO_JIT)\p{Ps}.*?\p{Pe}" hello.c >actual &&
+ test_cmp expected actual
+
+'
+
test_expect_success !PCRE 'grep -P pattern errors without PCRE' '
test_must_fail git grep -P "foo.*bar"
'
--
2.15.0.403.gc27cc4dac6
