Fix a leak when compiling a typed hash dereference
This commit is contained in:
parent
d1f65a5ba7
commit
243ad0ccb9
|
@ -0,0 +1,132 @@
|
||||||
|
From 057b890a6d3201a44afd68c840f3a76d4f508d91 Mon Sep 17 00:00:00 2001
|
||||||
|
From: David Mitchell <davem@iabyn.com>
|
||||||
|
Date: Fri, 8 Mar 2019 08:40:29 +0000
|
||||||
|
Subject: [PATCH] fix leak when compiling typed hash deref
|
||||||
|
MIME-Version: 1.0
|
||||||
|
Content-Type: text/plain; charset=UTF-8
|
||||||
|
Content-Transfer-Encoding: 8bit
|
||||||
|
|
||||||
|
In something like
|
||||||
|
|
||||||
|
my Foo $h;
|
||||||
|
$h->{bad_key}
|
||||||
|
|
||||||
|
perl will croak if package Foo defines valid %FIELDS and bad_key isn't
|
||||||
|
one of them. This croak happens during the second pass in
|
||||||
|
S_maybe_multideref(), which is trying to convert $h->{bad_key} into a
|
||||||
|
single multideref op. Since the aux buffer is allocated at the end of
|
||||||
|
the first pass, the buffer leaks.
|
||||||
|
|
||||||
|
The fix is to do the check in the first pass, which has been done by
|
||||||
|
adding an extra boolean flag to S_check_hash_fields_and_hekify(),
|
||||||
|
indicating whether to just check or actually do it.
|
||||||
|
|
||||||
|
Petr Písař: Ported to 5.18.1 from
|
||||||
|
02a9632ac4bf515585a2f25b05b2939de1743ded.
|
||||||
|
|
||||||
|
Signed-off-by: Petr Písař <ppisar@redhat.com>
|
||||||
|
---
|
||||||
|
op.c | 22 +++++++++++++++-------
|
||||||
|
t/op/multideref.t | 11 ++++++++++-
|
||||||
|
2 files changed, 25 insertions(+), 8 deletions(-)
|
||||||
|
|
||||||
|
diff --git a/op.c b/op.c
|
||||||
|
index 67da715..af3c448 100644
|
||||||
|
--- a/op.c
|
||||||
|
+++ b/op.c
|
||||||
|
@@ -2416,12 +2416,13 @@ S_modkids(pTHX_ OP *o, I32 type)
|
||||||
|
|
||||||
|
/* for a helem/hslice/kvslice, if its a fixed hash, croak on invalid
|
||||||
|
* const fields. Also, convert CONST keys to HEK-in-SVs.
|
||||||
|
- * rop is the op that retrieves the hash;
|
||||||
|
+ * rop is the op that retrieves the hash;
|
||||||
|
* key_op is the first key
|
||||||
|
+ * real if false, only check (and possibly croak); don't update op
|
||||||
|
*/
|
||||||
|
|
||||||
|
STATIC void
|
||||||
|
-S_check_hash_fields_and_hekify(pTHX_ UNOP *rop, SVOP *key_op)
|
||||||
|
+S_check_hash_fields_and_hekify(pTHX_ UNOP *rop, SVOP *key_op, int real)
|
||||||
|
{
|
||||||
|
PADNAME *lexname;
|
||||||
|
GV **fields;
|
||||||
|
@@ -2471,7 +2472,8 @@ S_check_hash_fields_and_hekify(pTHX_ UNOP *rop, SVOP *key_op)
|
||||||
|
if ( !SvIsCOW_shared_hash(sv = *svp)
|
||||||
|
&& SvTYPE(sv) < SVt_PVMG
|
||||||
|
&& SvOK(sv)
|
||||||
|
- && !SvROK(sv))
|
||||||
|
+ && !SvROK(sv)
|
||||||
|
+ && real)
|
||||||
|
{
|
||||||
|
SSize_t keylen;
|
||||||
|
const char * const key = SvPV_const(sv, *(STRLEN*)&keylen);
|
||||||
|
@@ -3648,7 +3650,7 @@ S_finalize_op(pTHX_ OP* o)
|
||||||
|
check_keys:
|
||||||
|
if (o->op_private & OPpLVAL_INTRO || rop->op_type != OP_RV2HV)
|
||||||
|
rop = NULL;
|
||||||
|
- S_check_hash_fields_and_hekify(aTHX_ rop, key_op);
|
||||||
|
+ S_check_hash_fields_and_hekify(aTHX_ rop, key_op, 1);
|
||||||
|
break;
|
||||||
|
}
|
||||||
|
case OP_NULL:
|
||||||
|
@@ -14605,12 +14607,13 @@ S_maybe_multideref(pTHX_ OP *start, OP *orig_o, UV orig_action, U8 hints)
|
||||||
|
* the extra hassle for those edge cases */
|
||||||
|
break;
|
||||||
|
|
||||||
|
- if (pass) {
|
||||||
|
+ {
|
||||||
|
UNOP *rop = NULL;
|
||||||
|
OP * helem_op = o->op_next;
|
||||||
|
|
||||||
|
ASSUME( helem_op->op_type == OP_HELEM
|
||||||
|
- || helem_op->op_type == OP_NULL);
|
||||||
|
+ || helem_op->op_type == OP_NULL
|
||||||
|
+ || pass == 0);
|
||||||
|
if (helem_op->op_type == OP_HELEM) {
|
||||||
|
rop = (UNOP*)(((BINOP*)helem_op)->op_first);
|
||||||
|
if ( helem_op->op_private & OPpLVAL_INTRO
|
||||||
|
@@ -14618,9 +14621,14 @@ S_maybe_multideref(pTHX_ OP *start, OP *orig_o, UV orig_action, U8 hints)
|
||||||
|
)
|
||||||
|
rop = NULL;
|
||||||
|
}
|
||||||
|
- S_check_hash_fields_and_hekify(aTHX_ rop, cSVOPo);
|
||||||
|
+ /* on first pass just check; on second pass
|
||||||
|
+ * hekify */
|
||||||
|
+ S_check_hash_fields_and_hekify(aTHX_ rop, cSVOPo,
|
||||||
|
+ pass);
|
||||||
|
+ }
|
||||||
|
|
||||||
|
#ifdef USE_ITHREADS
|
||||||
|
+ if (pass) {
|
||||||
|
/* Relocate sv to the pad for thread safety */
|
||||||
|
op_relocate_sv(&cSVOPo->op_sv, &o->op_targ);
|
||||||
|
arg->pad_offset = o->op_targ;
|
||||||
|
diff --git a/t/op/multideref.t b/t/op/multideref.t
|
||||||
|
index 20ba1ca..12b0453 100644
|
||||||
|
--- a/t/op/multideref.t
|
||||||
|
+++ b/t/op/multideref.t
|
||||||
|
@@ -18,7 +18,7 @@ BEGIN {
|
||||||
|
use warnings;
|
||||||
|
use strict;
|
||||||
|
|
||||||
|
-plan 63;
|
||||||
|
+plan 64;
|
||||||
|
|
||||||
|
|
||||||
|
# check that strict refs hint is handled
|
||||||
|
@@ -233,3 +233,12 @@ sub defer {}
|
||||||
|
is $x[qw(rt131627)->$*], 11, 'RT #131627: $a[qw(var)->$*]';
|
||||||
|
}
|
||||||
|
|
||||||
|
+# this used to leak - run the code for ASan to spot any problems
|
||||||
|
+{
|
||||||
|
+ package Foo;
|
||||||
|
+ our %FIELDS = ();
|
||||||
|
+ my Foo $f;
|
||||||
|
+ eval q{ my $x = $f->{c}; };
|
||||||
|
+ ::pass("S_maybe_multideref() shouldn't leak on croak");
|
||||||
|
+}
|
||||||
|
+
|
||||||
|
--
|
||||||
|
2.20.1
|
||||||
|
|
|
@ -0,0 +1,34 @@
|
||||||
|
From eef8d518b95b0221f81805d75bd63fbbf2995f3b Mon Sep 17 00:00:00 2001
|
||||||
|
From: David Mitchell <davem@iabyn.com>
|
||||||
|
Date: Tue, 12 Mar 2019 07:10:10 +0000
|
||||||
|
Subject: [PATCH] fix blead on non-threaded builds
|
||||||
|
MIME-Version: 1.0
|
||||||
|
Content-Type: text/plain; charset=UTF-8
|
||||||
|
Content-Transfer-Encoding: 8bit
|
||||||
|
|
||||||
|
My recent v5.29.8-64-g02a9632ac4 commit broke unthreaded builds.
|
||||||
|
This is the obvious fix. I've heard a report that unthreaded perl
|
||||||
|
SEGVs now but can't reproduce.
|
||||||
|
|
||||||
|
Signed-off-by: Petr Písař <ppisar@redhat.com>
|
||||||
|
---
|
||||||
|
op.c | 2 +-
|
||||||
|
1 file changed, 1 insertion(+), 1 deletion(-)
|
||||||
|
|
||||||
|
diff --git a/op.c b/op.c
|
||||||
|
index 4e49eeeedf..b4ba9c8f83 100644
|
||||||
|
--- a/op.c
|
||||||
|
+++ b/op.c
|
||||||
|
@@ -14716,8 +14716,8 @@ S_maybe_multideref(pTHX_ OP *start, OP *orig_o, UV orig_action, U8 hints)
|
||||||
|
pass);
|
||||||
|
}
|
||||||
|
|
||||||
|
-#ifdef USE_ITHREADS
|
||||||
|
if (pass) {
|
||||||
|
+#ifdef USE_ITHREADS
|
||||||
|
/* Relocate sv to the pad for thread safety */
|
||||||
|
op_relocate_sv(&cSVOPo->op_sv, &o->op_targ);
|
||||||
|
arg->pad_offset = o->op_targ;
|
||||||
|
--
|
||||||
|
2.20.1
|
||||||
|
|
12
perl.spec
12
perl.spec
|
@ -83,7 +83,7 @@ License: GPL+ or Artistic
|
||||||
Epoch: %{perl_epoch}
|
Epoch: %{perl_epoch}
|
||||||
Version: %{perl_version}
|
Version: %{perl_version}
|
||||||
# release number must be even higher, because dual-lived modules will be broken otherwise
|
# release number must be even higher, because dual-lived modules will be broken otherwise
|
||||||
Release: 434%{?dist}
|
Release: 435%{?dist}
|
||||||
Summary: Practical Extraction and Report Language
|
Summary: Practical Extraction and Report Language
|
||||||
Url: https://www.perl.org/
|
Url: https://www.perl.org/
|
||||||
Source0: https://www.cpan.org/src/5.0/perl-%{perl_version}.tar.xz
|
Source0: https://www.cpan.org/src/5.0/perl-%{perl_version}.tar.xz
|
||||||
|
@ -260,6 +260,10 @@ Patch54: perl-5.29.7-Perl_my_cxt_init-fix-potential-race-condition.patch
|
||||||
# Fix extending a stack in Perl parser, RT#133778, in upstream after 5.29.8
|
# Fix extending a stack in Perl parser, RT#133778, in upstream after 5.29.8
|
||||||
Patch55: perl-5.29.8-perl-133778-adjust-MARK-if-we-extend-the-stack-in-pp.patch
|
Patch55: perl-5.29.8-perl-133778-adjust-MARK-if-we-extend-the-stack-in-pp.patch
|
||||||
|
|
||||||
|
# Fix a leak when compiling a typed hash dereference, in upstream after 5.29.8
|
||||||
|
Patch56: perl-5.28.1-fix-leak-when-compiling-typed-hash-deref.patch
|
||||||
|
Patch57: perl-5.29.8-fix-blead-on-non-threaded-builds.patch
|
||||||
|
|
||||||
# Link XS modules to libperl.so with EU::CBuilder on Linux, bug #960048
|
# 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
|
Patch200: perl-5.16.3-Link-XS-modules-to-libperl.so-with-EU-CBuilder-on-Li.patch
|
||||||
|
|
||||||
|
@ -2871,6 +2875,8 @@ Perl extension for Version Objects
|
||||||
%patch53 -p1
|
%patch53 -p1
|
||||||
%patch54 -p1
|
%patch54 -p1
|
||||||
%patch55 -p1
|
%patch55 -p1
|
||||||
|
%patch56 -p1
|
||||||
|
%patch57 -p1
|
||||||
%patch200 -p1
|
%patch200 -p1
|
||||||
%patch201 -p1
|
%patch201 -p1
|
||||||
|
|
||||||
|
@ -2919,6 +2925,7 @@ perl -x patchlevel.h \
|
||||||
'Fedora Patch53: Fix setting magic when changing $^R (RT#133782)' \
|
'Fedora Patch53: Fix setting magic when changing $^R (RT#133782)' \
|
||||||
'Fedora Patch54: Fix a race when loading XS modules' \
|
'Fedora Patch54: Fix a race when loading XS modules' \
|
||||||
'Fedora Patch55: Fix extending a stack in Perl parser (RT#133778)' \
|
'Fedora Patch55: Fix extending a stack in Perl parser (RT#133778)' \
|
||||||
|
'Fedora Patch56: Fix a leak when compiling a typed hash dereference' \
|
||||||
'Fedora Patch200: Link XS modules to libperl.so with EU::CBuilder on Linux' \
|
'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' \
|
'Fedora Patch201: Link XS modules to libperl.so with EU::MM on Linux' \
|
||||||
%{nil}
|
%{nil}
|
||||||
|
@ -5207,6 +5214,9 @@ popd
|
||||||
|
|
||||||
# Old changelog entries are preserved in CVS.
|
# Old changelog entries are preserved in CVS.
|
||||||
%changelog
|
%changelog
|
||||||
|
* Fri Apr 05 2019 Petr Pisar <ppisar@redhat.com> - 4:5.28.1-435
|
||||||
|
- Fix a leak when compiling a typed hash dereference
|
||||||
|
|
||||||
* Tue Mar 05 2019 Björn Esser <besser82@fedoraproject.org> - 4:5.28.1-434
|
* 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)
|
- Add explicit Requires: libxcrypt-devel to devel sub-package (bug #1666098)
|
||||||
|
|
||||||
|
|
Loading…
Reference in New Issue