rpms/perl
5
0
Fork 0
Code Issues Pull Requests Releases Activity

Fix a memory leak when spawning threads in a BEGIN phase

Browse Source
This commit is contained in:
Petr Písař 2019-04-05 16:45:27 +02:00
parent bf70c3addf
commit be29c61cdb
2 changed files with 56 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

49
perl-5.29.9-fix-leak-in-BEGIN-threads-new.patch Normal file
View File

@ -0,0 +1,49 @@
From 803bd7c91c63f8f263bed592a33b10cf69f567cf Mon Sep 17 00:00:00 2001
From: David Mitchell <davem@iabyn.com>
Date: Fri, 22 Mar 2019 15:43:56 +0000
Subject: [PATCH] fix leak in BEGIN { threads->new(...) }
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Normally by the time we reach perl_destruct(), PL_parser should be null
due to having its original (null) value restored by SAVEt_PARSER during
leaving scope (usually before run-time starts in fact). But if a thread
is created within a BEGIN block, the parser is duped, but the
SAVEt_PARSER savestack entry isn't. So PL_parser never gets cleaned up.
Clean it up in perl_destruct() instead. This is a bit of a hack.
Signed-off-by: Petr Písař <ppisar@redhat.com>
---
perl.c | 15 +++++++++++++++
1 file changed, 15 insertions(+)
diff --git a/perl.c b/perl.c
index cdefa99018..1ef425bb25 100644
--- a/perl.c
+++ b/perl.c
@@ -668,6 +668,21 @@ perl_destruct(pTHXx)
FREETMPS;
assert(PL_scopestack_ix == 0);
+ /* normally when we get here, PL_parser should be null due to having
+ * its original (null) value restored by SAVEt_PARSER during leaving
+ * scope (usually before run-time starts in fact).
+ * But if a thread is created within a BEGIN block, the parser is
+ * duped, but the SAVEt_PARSER savestack entry isn't. So PL_parser
+ * never gets cleaned up.
+ * Clean it up here instead. This is a bit of a hack.
+ */
+ if (PL_parser) {
+ /* stop parser_free() stomping on PL_curcop */
+ PL_parser->saved_curcop = PL_curcop;
+ parser_free(PL_parser);
+ }
+
+
/* Need to flush since END blocks can produce output */
/* flush stdout separately, since we can identify it */
#ifdef USE_PERLIO
--
2.20.1

7
perl.spec
View File

@ -279,6 +279,10 @@ Patch60: perl-5.28.1-Fix-POSIX-mblen-mbstate_t-initialization-on-threaded
# Fix a memory leak when cloning a regular expression, in upstream after 5.29.9
Patch61: perl-5.29.9-fix-leak-in-cloned-regexes.patch
# Fix a memory leak when spawning threads in a BEGIN phase,
# in upstream after 5.29.9
Patch62: perl-5.29.9-fix-leak-in-BEGIN-threads-new.patch
# Link XS modules to libperl.so with EU::CBuilder on Linux, bug #960048
Patch200: perl-5.16.3-Link-XS-modules-to-libperl.so-with-EU-CBuilder-on-Li.patch
@ -2896,6 +2900,7 @@ Perl extension for Version Objects
%patch59 -p1
%patch60 -p1
%patch61 -p1
%patch62 -p1
%patch200 -p1
%patch201 -p1
@ -2949,6 +2954,7 @@ perl -x patchlevel.h \
'Fedora Patch59: Fix a buffer overread when parsing a regular expression with an unknown character name (RT#133880)' \
'Fedora Patch60: Fix mbstate_t initialization in POSIX::mblen (RT#133928)' \
'Fedora Patch61: Fix a memory leak when cloning a regular expression' \
'Fedora Patch62: Fix a memory leak when spawning threads in a BEGIN phase' \
'Fedora Patch200: Link XS modules to libperl.so with EU::CBuilder on Linux' \
'Fedora Patch201: Link XS modules to libperl.so with EU::MM on Linux' \
%{nil}
@ -5244,6 +5250,7 @@ popd
character name (RT#133880)
- Fix mbstate_t initialization in POSIX::mblen (RT#133928)
- Fix a memory leak when cloning a regular expression
- Fix a memory leak when spawning threads in a BEGIN phase
* Tue Mar 05 2019 Björn Esser <besser82@fedoraproject.org> - 4:5.28.1-434
- Add explicit Requires: libxcrypt-devel to devel sub-package (bug #1666098)