Revision f5c7f5dfbaf0d2f7d946d0fe86f08e6bcb36ed0d authored by Matt Caswell on 30 June 2016, 12:17:08 UTC, committed by Matt Caswell on 22 August 2016, 09:53:55 UTC
DTLS can handle out of order record delivery. Additionally since handshake messages can be bigger than will fit into a single packet, the messages can be fragmented across multiple records (as with normal TLS). That means that the messages can arrive mixed up, and we have to reassemble them. We keep a queue of buffered messages that are "from the future", i.e. messages we're not ready to deal with yet but have arrived early. The messages held there may not be full yet - they could be one or more fragments that are still in the process of being reassembled. The code assumes that we will eventually complete the reassembly and when that occurs the complete message is removed from the queue at the point that we need to use it. However, DTLS is also tolerant of packet loss. To get around that DTLS messages can be retransmitted. If we receive a full (non-fragmented) message from the peer after previously having received a fragment of that message, then we ignore the message in the queue and just use the non-fragmented version. At that point the queued message will never get removed. Additionally the peer could send "future" messages that we never get to in order to complete the handshake. Each message has a sequence number (starting from 0). We will accept a message fragment for the current message sequence number, or for any sequence up to 10 into the future. However if the Finished message has a sequence number of 2, anything greater than that in the queue is just left there. So, in those two ways we can end up with "orphaned" data in the queue that will never get removed - except when the connection is closed. At that point all the queues are flushed. An attacker could seek to exploit this by filling up the queues with lots of large messages that are never going to be used in order to attempt a DoS by memory exhaustion. I will assume that we are only concerned with servers here. It does not seem reasonable to be concerned about a memory exhaustion attack on a client. They are unlikely to process enough connections for this to be an issue. A "long" handshake with many messages might be 5 messages long (in the incoming direction), e.g. ClientHello, Certificate, ClientKeyExchange, CertificateVerify, Finished. So this would be message sequence numbers 0 to 4. Additionally we can buffer up to 10 messages in the future. Therefore the maximum number of messages that an attacker could send that could get orphaned would typically be 15. The maximum size that a DTLS message is allowed to be is defined by max_cert_list, which by default is 100k. Therefore the maximum amount of "orphaned" memory per connection is 1500k. Message sequence numbers get reset after the Finished message, so renegotiation will not extend the maximum number of messages that can be orphaned per connection. As noted above, the queues do get cleared when the connection is closed. Therefore in order to mount an effective attack, an attacker would have to open many simultaneous connections. Issue reported by Quan Luo. CVE-2016-2179 Reviewed-by: Richard Levitte <levitte@openssl.org>
1 parent 5dfd038
incore
#! /usr/bin/env perl
# Copyright 2011-2016 The OpenSSL Project Authors. All Rights Reserved.
#
# Licensed under the OpenSSL license (the "License"). You may not use
# this file except in compliance with the License. You can obtain a copy
# in the file LICENSE in the source distribution or at
# https://www.openssl.org/source/license.html
# The script embeds fingerprint into ELF executable object, either
# application binary or shared library.
######################################################################
#
# ELF symbol table parser by <appro@openssl.org>. The table entries
# are extended with offset within executable file...
#
{ package ELF;
use FileHandle;
sub dup { my %copy=map {$_} @_; return \%copy; }
sub Load {
my $class = shift;
my $self = {};
my $FD = FileHandle->new(); # autoclose
bless $self,$class;
sysopen($FD,shift,0) or die "$!";
binmode($FD);
#################################################
# read and parse elf_ehdr.e_ident...
#
read($FD,my $elf,16) or die "$!";
my %e_ident;
@e_ident{magic,class,data,version,osabi,abiver,pad}=
unpack("a4C*",$elf);
$!=42; # signal fipsld to revert to two-step link
die "not ELF file" if ($e_ident{magic} ne chr(0177)."ELF");
my $elf_bits = $e_ident{class}*32; # 32 or 64
my $big_endian = $e_ident{data}-1; # 0 or 1
if ($elf_bits==64) {
if (!(((1<<31)<<1) && $big_endian==(unpack("L",pack("N",1))==1))) {
die "ELF64 is supported only natively";
}
}
#################################################
# read and parse remainder of elf_ehdr...
#
read($FD,my $elfhdr,64) or die "$!";
my %elf_ehdr;
@elf_ehdr{e_type,e_machine,e_version,
e_entry,e_phoff,e_shoff,e_flags,e_ehsize,
e_phentsize,e_phnum,e_shentsize,e_shnum,e_shstrndx} =
$elf_bits==32 ?
unpack($big_endian?"nnN5n6":"vvV5v6",$elfhdr)
: unpack("SSLQ3LS6",$elfhdr);
# put aside e_machine in case one has to treat specific
# platforms differently, see EM_ constants in elf.h for
# assortment...
$self->{e_machine} = $elf_ehdr{e_machine};
#################################################
# read and parse elf_shdr table...
#
my ($i,$sz,$symtab_idx,$blob,$strings);
seek($FD,$elf_ehdr{e_shoff},0) or die "$!";
read($FD,$blob,$elf_ehdr{e_shentsize}*$elf_ehdr{e_shnum}) or die "$!";
my @sections;
my $elf_shdr_struct=($elf_bits==32?($big_endian?"N10":"V10"):"L2Q4L2Q2");
for ($sz=$elf_ehdr{e_shentsize},$i=0;$i<length($blob);$i+=$sz) {
my %elf_shdr;
@elf_shdr{sh_name,sh_type,sh_flags,
sh_addr,sh_offset,sh_size,
sh_link,sh_info,sh_addalign,sh_entsize} =
unpack($elf_shdr_struct,substr($blob,$i,$sz));
push(@sections,dup(%elf_shdr));
# note SHT_SYMTAB or SHT_DYNSYM for future reference
if ($elf_shdr{sh_type}==2 || $elf_shdr{sh_type}==11) {
$symtab_idx = $#sections;
}
}
# read strings table and map section names...
seek($FD,@sections[$elf_ehdr{e_shstrndx}]->{sh_offset},0) or die "$!";
read($FD,$strings,@sections[$elf_ehdr{e_shstrndx}]->{sh_size}) or die "$!";
for (@sections) {
$_->{sh_name}=(split(chr(0),substr($strings,$_->{sh_name},64)))[0];
}
#################################################
# read symbol strings table...
#
$i=@sections[$symtab_idx]->{sh_link};
seek($FD,@sections[$i]->{sh_offset},0) or die "$!";
read($FD,$strings,@sections[$i]->{sh_size}) or die "$!";
#################################################
# read and parse elf_sym table...
#
seek($FD,@sections[$symtab_idx]->{sh_offset},0) or die "$!";
read($FD,my $blob,@sections[$symtab_idx]->{sh_size}) or die "$!";
for ($sz=@sections[$symtab_idx]->{sh_entsize},$i=0;$i<length($blob);$i+=$sz) {
my %elf_sym;
if ($elf_bits==32) {
@elf_sym{st_name,st_value,st_size,st_info,st_other,st_shndx} =
unpack($big_endian?"N3CCn":"V3CCv",substr($blob,$i,$sz));
} else {
@elf_sym{st_name,st_info,st_other,st_shndx,st_value,st_size} =
unpack("LCCSQQ",substr($blob,$i,$sz));
}
my $st_type = $elf_sym{st_info}&0xf;
my $st_bind = $elf_sym{st_info}>>4;
my $st_secn = $elf_sym{st_shndx};
my $name;
# (STT_OBJECT || STT_FUNC)
if ($st_bind<3 && ($st_type==1 || $st_type==2)
&& $st_secn <= $#sections # sane st_shndx
&& @sections[$st_secn]->{sh_type} # not SHN_UNDEF
&& ($name=(split(chr(0),substr($strings,$elf_sym{st_name},128)))[0])
) {
# synthesize st_offset, ...
$elf_sym{st_offset} = $elf_sym{st_value}
- @sections[$st_secn]->{sh_addr}
+ @sections[$st_secn]->{sh_offset};
$elf_sym{st_name} = $name;
$elf_sym{st_section} = @sections[$st_secn]->{sh_name};
# ... and add to lookup table
$self->{symbols}{$name} = dup(%elf_sym);
}
}
return $self;
}
sub Lookup {
my $self = shift;
my $name = shift;
return $self->{symbols}{$name};
}
sub Traverse {
my $self = shift;
my $code = shift;
if (ref($code) eq 'CODE') {
for (keys(%{$self->{symbols}})) { &$code($self->{symbols}{$_}); }
}
}
}
######################################################################
#
# SHA1 and HMAC in Perl by <appro@openssl.org>.
#
{ package SHA1;
use integer;
{
################################### SHA1 block code generator
my @V = ('$A','$B','$C','$D','$E');
my $i;
sub XUpdate {
my $ret;
$ret="(\$T=\$W[($i-16)%16]^\$W[($i-14)%16]^\$W[($i-8)%16]^\$W[($i-3)%16],\n\t";
if ((1<<31)<<1) {
$ret.=" \$W[$i%16]=((\$T<<1)|(\$T>>31))&0xffffffff)\n\t ";
} else {
$ret.=" \$W[$i%16]=(\$T<<1)|((\$T>>31)&1))\n\t ";
}
}
sub tail {
my ($a,$b,$c,$d,$e)=@V;
my $ret;
if ((1<<31)<<1) {
$ret.="(($a<<5)|($a>>27));\n\t";
$ret.="$b=($b<<30)|($b>>2); $e&=0xffffffff; #$b&=0xffffffff;\n\t";
} else {
$ret.="(($a<<5)|($a>>27)&0x1f);\n\t";
$ret.="$b=($b<<30)|($b>>2)&0x3fffffff;\n\t";
}
$ret;
}
sub BODY_00_15 {
my ($a,$b,$c,$d,$e)=@V;
"$e+=\$W[$i]+0x5a827999+((($c^$d)&$b)^$d)+".tail();
}
sub BODY_16_19 {
my ($a,$b,$c,$d,$e)=@V;
"$e+=".XUpdate()."+0x5a827999+((($c^$d)&$b)^$d)+".tail();
}
sub BODY_20_39 {
my ($a,$b,$c,$d,$e)=@V;
"$e+=".XUpdate()."+0x6ed9eba1+($b^$c^$d)+".tail();
}
sub BODY_40_59 {
my ($a,$b,$c,$d,$e)=@V;
"$e+=".XUpdate()."+0x8f1bbcdc+(($b&$c)|(($b|$c)&$d))+".tail();
}
sub BODY_60_79 {
my ($a,$b,$c,$d,$e)=@V;
"$e+=".XUpdate()."+0xca62c1d6+($b^$c^$d)+".tail();
}
my $sha1_impl =
'sub block {
my $self = @_[0];
my @W = unpack("N16",@_[1]);
my ($A,$B,$C,$D,$E,$T) = @{$self->{H}};
';
$sha1_impl.='
$A &= 0xffffffff;
$B &= 0xffffffff;
' if ((1<<31)<<1);
for($i=0;$i<16;$i++){ $sha1_impl.=BODY_00_15(); unshift(@V,pop(@V)); }
for(;$i<20;$i++) { $sha1_impl.=BODY_16_19(); unshift(@V,pop(@V)); }
for(;$i<40;$i++) { $sha1_impl.=BODY_20_39(); unshift(@V,pop(@V)); }
for(;$i<60;$i++) { $sha1_impl.=BODY_40_59(); unshift(@V,pop(@V)); }
for(;$i<80;$i++) { $sha1_impl.=BODY_60_79(); unshift(@V,pop(@V)); }
$sha1_impl.='
$self->{H}[0]+=$A; $self->{H}[1]+=$B; $self->{H}[2]+=$C;
$self->{H}[3]+=$D; $self->{H}[4]+=$E; }';
#print $sha1_impl,"\n";
eval($sha1_impl); # generate code
}
sub Init {
my $class = shift; # multiple instances...
my $self = {};
bless $self,$class;
$self->{H} = [0x67452301,0xefcdab89,0x98badcfe,0x10325476,0xc3d2e1f0];
$self->{N} = 0;
return $self;
}
sub Update {
my $self = shift;
my $msg;
foreach $msg (@_) {
my $len = length($msg);
my $num = length($self->{buf});
my $off = 0;
$self->{N} += $len;
if (($num+$len)<64)
{ $self->{buf} .= $msg; next; }
elsif ($num)
{ $self->{buf} .= substr($msg,0,($off=64-$num));
$self->block($self->{buf});
}
while(($off+64) <= $len)
{ $self->block(substr($msg,$off,64));
$off += 64;
}
$self->{buf} = substr($msg,$off);
}
return $self;
}
sub Final {
my $self = shift;
my $num = length($self->{buf});
$self->{buf} .= chr(0x80); $num++;
if ($num>56)
{ $self->{buf} .= chr(0)x(64-$num);
$self->block($self->{buf});
$self->{buf}=undef;
$num=0;
}
$self->{buf} .= chr(0)x(56-$num);
$self->{buf} .= pack("N2",($self->{N}>>29)&0x7,$self->{N}<<3);
$self->block($self->{buf});
return pack("N*",@{$self->{H}});
}
sub Selftest {
my $hash;
$hash=SHA1->Init()->Update('abc')->Final();
die "SHA1 test#1" if (unpack("H*",$hash) ne 'a9993e364706816aba3e25717850c26c9cd0d89d');
$hash=SHA1->Init()->Update('abcdbcdecdefdefgefghfghighijhijkijkljklmklmnlmnomnopnopq')->Final();
die "SHA1 test#2" if (unpack("H*",$hash) ne '84983e441c3bd26ebaae4aa1f95129e5e54670f1');
#$hash=SHA1->Init()->Update('a'x1000000)->Final();
#die "SHA1 test#3" if (unpack("H*",$hash) ne '34aa973cd4c4daa4f61eeb2bdbad27316534016f');
}
}
{ package HMAC;
sub Init {
my $class = shift;
my $key = shift;
my $self = {};
bless $self,$class;
if (length($key)>64) {
$key = SHA1->Init()->Update($key)->Final();
}
$key .= chr(0x00)x(64-length($key));
my @ikey = map($_^=0x36,unpack("C*",$key));
($self->{hash} = SHA1->Init())->Update(pack("C*",@ikey));
$self->{okey} = pack("C*",map($_^=0x36^0x5c,@ikey));
return $self;
}
sub Update {
my $self = shift;
$self->{hash}->Update(@_);
return $self;
}
sub Final {
my $self = shift;
my $ihash = $self->{hash}->Final();
return SHA1->Init()->Update($self->{okey},$ihash)->Final();
}
sub Selftest {
my $hmac;
$hmac = HMAC->Init('0123456789:;<=>?@ABC')->Update('Sample #2')->Final();
die "HMAC test" if (unpack("H*",$hmac) ne '0922d3405faa3d194f82a45830737d5cc6c75d24');
}
}
######################################################################
#
# main()
#
my $legacy_mode;
if ($ARGV<0 || ($#ARGV>0 && !($legacy_mode=(@ARGV[0] =~ /^\-(dso|exe)$/)))) {
print STDERR "usage: $0 [-dso|-exe] elfbinary\n";
exit(1);
}
$exe = ELF->Load(@ARGV[$#ARGV]);
$FIPS_text_start = $exe->Lookup("FIPS_text_start") or die;
$FIPS_text_end = $exe->Lookup("FIPS_text_end") or die;
$FIPS_rodata_start = $exe->Lookup("FIPS_rodata_start") or die;
$FIPS_rodata_end = $exe->Lookup("FIPS_rodata_end") or die;
$FIPS_signature = $exe->Lookup("FIPS_signature") or die;
# new cross-compile support
$FIPS_text_startX = $exe->Lookup("FIPS_text_startX");
$FIPS_text_endX = $exe->Lookup("FIPS_text_endX");
if (!$legacy_mode) {
if (!$FIPS_text_startX || !$FIPS_text_endX) {
print STDERR "@ARGV[$#ARGV] is not cross-compiler aware.\n";
exit(42); # signal fipsld to revert to two-step link
}
$FINGERPRINT_ascii_value
= $exe->Lookup("FINGERPRINT_ascii_value");
}
if ($FIPS_text_startX && $FIPS_text_endX) {
$FIPS_text_start = $FIPS_text_startX;
$FIPS_text_end = $FIPS_text_endX;
}
sysopen(FD,@ARGV[$#ARGV],$legacy_mode?0:2) or die "$!"; # 2 is read/write
binmode(FD);
sub HMAC_Update {
my ($hmac,$off,$len) = @_;
my $blob;
seek(FD,$off,0) or die "$!";
read(FD,$blob,$len) or die "$!";
$$hmac->Update($blob);
}
# fips/fips.c:FIPS_incore_fingerprint's Perl twin
#
sub FIPS_incore_fingerprint {
my $p1 = $FIPS_text_start->{st_offset};
my $p2 = $FIPS_text_end->{st_offset};
my $p3 = $FIPS_rodata_start->{st_offset};
my $p4 = $FIPS_rodata_end->{st_offset};
my $sig = $FIPS_signature->{st_offset};
my $ctx = HMAC->Init("etaonrishdlcupfm");
# detect overlapping regions
if ($p1<=$p3 && $p2>=$p3) {
$p3 = $p1; $p4 = $p2>$p4?$p2:$p4; $p1 = 0; $p2 = 0;
} elsif ($p3<=$p1 && $p4>=$p1) {
$p3 = $p3; $p4 = $p2>$p4?$p2:$p4; $p1 = 0; $p2 = 0;
}
if ($p1) {
HMAC_Update (\$ctx,$p1,$p2-$p1);
}
if ($sig>=$p3 && $sig<$p4) {
# "punch" hole
HMAC_Update(\$ctx,$p3,$sig-$p3);
$p3 = $sig+20;
HMAC_Update(\$ctx,$p3,$p4-$p3);
} else {
HMAC_Update(\$ctx,$p3,$p4-$p3);
}
return $ctx->Final();
}
$fingerprint = FIPS_incore_fingerprint();
if ($legacy_mode) {
print unpack("H*",$fingerprint);
} elsif (defined($FINGERPRINT_ascii_value)) {
seek(FD,$FINGERPRINT_ascii_value->{st_offset},0) or die "$!";
print FD unpack("H*",$fingerprint) or die "$!";
} else {
seek(FD,$FIPS_signature->{st_offset},0) or die "$!";
print FD $fingerprint or die "$!";
}
close (FD);

Computing file changes ...