Repository navigation
Segmentation fault in pack (S_pack_rec tries to access unallocated memory) #14977
Description
Activity
From @dcollinsn
Greetings Porters,
I have compiled bleadperl with the afl-gcc compiler using:
./Configure -Dusedevel -Dprefix='/usr/local/perl-afl' -Dcc='ccache afl-gcc' -Duselongdouble -Duse64bitall -Doptimize=-g -Uversiononly -Uman1dir -Uman3dir -des
AFL_HARDEN=1 make && make testAnd then fuzzed the resulting binary using:
AFL_NO_VAR_CHECK=1 afl-fuzz -i in -o out bin/perl @@
After reducing testcases using `afl-tmin` and performing additional minimization by hand, I have located the following testcase that triggers a segmentation fault in the perl interpreter. The testcase is the file:
pack+WH200000,\0
Today I'm reporting a pair of bugs related to pack/unpack. They appeared to be related, but GDB shows that they have very different behaviours. This is by far the tamer of the two.
**GDB**
(gdb) run
Starting program: /usr/local/perl-afl/bin/perl -e pack+WH200000,\\0
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".Program received signal SIGSEGV, Segmentation fault.
S_pack_rec (cat=0x0, cat@entry=0x1205af0, symptr=0x2fcc0, beglist=0x1,
beglist@entry=0x11f3a30, endlist=0x7ffff7e1d015, endlist@entry=0x11f3a38)
at pp_pack.c:2503
2503 } else if (isALPHA(*str))
(gdb) bt
#0 S_pack_rec (cat=0x0, cat@entry=0x1205af0, symptr=0x2fcc0, beglist=0x1,
beglist@entry=0x11f3a30, endlist=0x7ffff7e1d015, endlist@entry=0x11f3a38)
at pp_pack.c:2503
#1 0x0000000000eb2f5f in Perl_packlist (endlist=0x11f3a38,
beglist=0x11f3a30, patend=<optimized out>, pat=<optimized out>,
cat=0x1205af0) at pp_pack.c:1971
#2 Perl_pp_pack () at pp_pack.c:3128
#3 0x00000000007dec7f in Perl_runops_debug () at dump.c:2224
#4 0x0000000000544469 in S_run_body (oldscope=1) at perl.c:2456
#5 perl_run (my_perl=<optimized out>) at perl.c:2379
#6 0x000000000042c6f8 in main (argc=3, argv=0x7fffffffe648,
env=0x7fffffffe668) at perlmain.c:116
(gdb) info locals
str = 0xfeb000 <error: Cannot access memory at address 0xfeb000>
utf8_flags = 195775
fromlen = 0
howlen = e_number
lookahead = {patptr = 0x120e788 "", patend = 0x120e788 "", grpbeg = 0x0,
grpend = 0x0, code = 72, length = 200000, howlen = e_number, level = 0,
flags = 9, strbeg = 0, previous = 0x0}
utf8 = 207
__PRETTY_FUNCTION__ = "S_pack_rec"
(gdb) f 1
#1 0x0000000000eb2f5f in Perl_packlist (endlist=0x11f3a38,
beglist=0x11f3a30, patend=<optimized out>, pat=<optimized out>,
cat=0x1205af0) at pp_pack.c:1971
1971 (void)pack_rec( cat, &sym, beglist, endlist );
(gdb) info locals
sym = {patptr = 0x120e788 "", patend = 0x120e788 "", grpbeg = 0x0,
grpend = 0x0, code = 72, length = 200000, howlen = e_number, level = 0,
flags = 9, strbeg = 0, previous = 0x0}**VALGRIND**
==31388== Memcheck, a memory error detector
==31388== Copyright (C) 2002-2015, and GNU GPL'd, by Julian Seward et al.
==31388== Using Valgrind-3.11.0 and LibVEX; rerun with -h for copyright info
==31388== Command: ../bin/perl -e pack+WH200000,\\0
==31388==
==31388== Invalid read of size 1
==31388== at 0xEA3014: S_pack_rec (pp_pack.c:2503)
==31388== by 0xEB2F5E: Perl_packlist (pp_pack.c:1971)
==31388== by 0xEB2F5E: Perl_pp_pack (pp_pack.c:3128)
==31388== by 0x7DEC7E: Perl_runops_debug (dump.c:2224)
==31388== by 0x544468: S_run_body (perl.c:2456)
==31388== by 0x544468: perl_run (perl.c:2379)
==31388== by 0x42C6F7: main (perlmain.c:116)
==31388== Address 0xfeb000 is not stack'd, malloc'd or (recently) free'd
==31388==
==31388==
==31388== Process terminating with default action of signal 11 (SIGSEGV)
==31388== Access not within mapped region at address 0xFEB000
==31388== at 0xEA3014: S_pack_rec (pp_pack.c:2503)
==31388== by 0xEB2F5E: Perl_packlist (pp_pack.c:1971)
==31388== by 0xEB2F5E: Perl_pp_pack (pp_pack.c:3128)
==31388== by 0x7DEC7E: Perl_runops_debug (dump.c:2224)
==31388== by 0x544468: S_run_body (perl.c:2456)
==31388== by 0x544468: perl_run (perl.c:2379)
==31388== by 0x42C6F7: main (perlmain.c:116)
==31388== If you believe this happened as a result of a stack
==31388== overflow in your program's main thread (unlikely but
==31388== possible), you can try to increase the size of the
==31388== main thread stack using the --main-stacksize= flag.
==31388== The main thread stack size used in this run was 8388608.
==31388==
==31388== HEAP SUMMARY:
==31388== in use at exit: 310,063 bytes in 540 blocks
==31388== total heap usage: 639 allocs, 99 frees, 327,593 bytes allocated
==31388==
==31388== LEAK SUMMARY:
==31388== definitely lost: 176 bytes in 1 blocks
==31388== indirectly lost: 1,989 bytes in 20 blocks
==31388== possibly lost: 0 bytes in 0 blocks
==31388== still reachable: 307,898 bytes in 519 blocks
==31388== suppressed: 0 bytes in 0 blocks
==31388== Rerun with --leak-check=full to see details of leaked memory
==31388==
==31388== For counts of detected and suppressed errors, rerun with: -v
==31388== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 0 from 0)
Segmentation fault**PERL -V**
Summary of my perl5 (revision 5 version 23 subversion 4) configuration:
Commit id: 94757bf
Platform:
osname=linux, osvers=3.16.0-4-amd64, archname=x86_64-linux-ld
uname='linux nightshade64 3.16.0-4-amd64 #1 smp debian 3.16.7-ckt11-1+deb8u4 (2015-09-19) x86_64 gnulinux '
config_args='-Dusedevel -Dprefix=/usr/local/perl-afl -Dcc=ccache afl-gcc -Duselongdouble -Duse64bitall -Doptimize=-g -Uversiononly -Uman1dir -Uman3dir -DDEBUGGING -DPERL_POISON -des'
hint=recommended, useposix=true, d_sigaction=define
useithreads=undef, usemultiplicity=undef
use64bitint=define, use64bitall=define, uselongdouble=define
usemymalloc=n, bincompat5005=undef
Compiler:
cc='ccache afl-gcc', ccflags ='-fwrapv -DDEBUGGING -fno-strict-aliasing -pipe -fstack-protector-strong -I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64',
optimize='-g',
cppflags='-fwrapv -DDEBUGGING -fno-strict-aliasing -pipe -fstack-protector-strong -I/usr/local/include'
ccversion='', gccversion='4.9.2', gccosandvers=''
intsize=4, longsize=8, ptrsize=8, doublesize=8, byteorder=12345678, doublekind=3
d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=16, longdblkind=3
ivtype='long', ivsize=8, nvtype='long double', nvsize=16, Off_t='off_t', lseeksize=8
alignbytes=16, prototype=define
Linker and Libraries:
ld='ccache afl-gcc', ldflags =' -fstack-protector-strong -L/usr/local/lib'
libpth=/usr/local/lib /usr/lib/gcc/x86_64-linux-gnu/4.9/include-fixed /usr/include/x86_64-linux-gnu /usr/lib /lib/x86_64-linux-gnu /lib/../lib /usr/lib/x86_64-linux-gnu /usr/lib/../lib /lib
libs=-lpthread -lnsl -ldl -lm -lcrypt -lutil -lc
perllibs=-lpthread -lnsl -ldl -lm -lcrypt -lutil -lc
libc=libc-2.19.so, so=so, useshrplib=false, libperl=libperl.a
gnulibc_version='2.19'
Dynamic Linking:
dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=undef, ccdlflags='-Wl,-E'
cccdlflags='-fPIC', lddlflags='-shared -g -L/usr/local/lib -fstack-protector-strong'Characteristics of this binary (from libperl):
Compile-time options: DEBUGGING HAS_TIMES PERLIO_LAYERS PERL_COPY_ON_WRITE
PERL_DONT_CREATE_GVSV
PERL_HASH_FUNC_ONE_AT_A_TIME_HARD PERL_MALLOC_WRAP
PERL_PRESERVE_IVUV PERL_USE_DEVEL USE_64_BIT_ALL
USE_64_BIT_INT USE_LARGE_FILES USE_LOCALE
USE_LOCALE_COLLATE USE_LOCALE_CTYPE
USE_LOCALE_NUMERIC USE_LOCALE_TIME USE_LONG_DOUBLE
USE_PERLIO USE_PERL_ATOF
Built under linux
Compiled at Sep 30 2015 16:39:59
@INC:
/usr/local/perl-afl/lib/site_perl/5.23.4/x86_64-linux-ld
/usr/local/perl-afl/lib/site_perl/5.23.4
/usr/local/perl-afl/lib/5.23.4/x86_64-linux-ld
/usr/local/perl-afl/lib/5.23.4
.From @tonycoz
On Sun Oct 11 04:49:37 2015, dcollinsn@gmail.com wrote:
After reducing testcases using `afl-tmin` and performing additional
minimization by hand, I have located the following testcase that
triggers a segmentation fault in the perl interpreter. The testcase is
the file:pack+WH200000,\0
The attached seems to fix it.
I'm going to review the other pack handlers to check for similar issues.
Tony
From @tonycoz
0001-perl-126325-don-t-read-past-the-end-of-the-source-fo.patch
From 45f3993798e8339daceeb6b1a0ce606da9e8a613 Mon Sep 17 00:00:00 2001 From: Tony Cook <tony@develop-help.com> Date: Thu, 22 Oct 2015 12:03:05 +1100 Subject: [perl #126325] don't read past the end of the source for pack [Hh] With a utf8 target but a non-utf8 source, pack Hh would read past the end of the source when given a length, due to an incorrect condition. --- pp_pack.c | 2 +- t/op/pack.t | 22 +++++++++++++++++++++- 2 files changed, 22 insertions(+), 2 deletions(-) diff --git a/pp_pack.c b/pp_pack.c index 96dfd20..044ea7f 100644 --- a/pp_pack.c +++ b/pp_pack.c @@ -2488,7 +2488,7 @@ S_pack_rec(pTHX_ SV *cat, tempsym_t* symptr, SV **beglist, SV **endlist ) if (howlen == e_star) len = fromlen; field_len = (len+1)/2; GROWING(utf8, cat, start, cur, field_len); - if (!utf8 && len > (I32)fromlen) len = fromlen; + if (!utf8_source && len > (I32)fromlen) len = fromlen; bits = 0; l = 0; if (datumtype == 'H') diff --git a/t/op/pack.t b/t/op/pack.t index e348693..a2da636 100644 --- a/t/op/pack.t +++ b/t/op/pack.t @@ -12,7 +12,7 @@ my $no_endianness = $] > 5.009 ? '' : my $no_signedness = $] > 5.009 ? '' : "Signed/unsigned pack modifiers not available on this perl"; -plan tests => 14708; +plan tests => 14712; use strict; use warnings qw(FATAL all); @@ -2024,3 +2024,23 @@ is $o::num, 1, 'pack "c" does call num overloading'; #[perl #123874]: argument underflow leads to corrupt length eval q{ pack "pi/x" }; ok(1, "argument underflow did not crash"); + +{ + # [perl #126325] pack [hH] with a unicode string + # the hex encoders would read past the end of the string, using + # invalid source bytes + my $twenty_nuls = "\0" x 20; + # This is the case that failed + is(pack("WH40", 0x100, ""), "\x{100}$twenty_nuls", + "check pack H zero fills (utf8 target)"); + my $up_nul = "\0"; + + utf8::upgrade($up_nul); + # check the other combinations too + is(pack("WH40", 0x100, $up_nul), "\x{100}$twenty_nuls", + "check pack H zero fills (utf8 target/source)"); + is(pack("H40", ""), $twenty_nuls, + "check pack H zero fills (utf8 none)"); + is(pack("H40", $up_nul), $twenty_nuls, + "check pack H zero fills (utf8 source)"); +} -- 2.1.4
The RT System itself - Status changed from 'new' to 'open'
From @tonycoz
On Wed Oct 21 18:05:47 2015, tonyc wrote:
On Sun Oct 11 04:49:37 2015, dcollinsn@gmail.com wrote:
After reducing testcases using `afl-tmin` and performing additional
minimization by hand, I have located the following testcase that
triggers a segmentation fault in the perl interpreter. The testcase is
the file:pack+WH200000,\0
The attached seems to fix it.
Pushed as 0403a1a.
I'm going to review the other pack handlers to check for similar issues.
The others seemed ok to me.
Tony
Migrated from rt.perl.org#126325 (status was 'resolved')
Searchable as RT126325$