Skip to contentSkip to conversation

package/exiv2: annotate CVE-2019-13504

PATCH New Fabrice Fontaine 3 participants 8 messages Feb 29, 2020 – Mar 14, 2020

lore
Share & tools
lore
#0 lore raw

PATCH 1/3 package/exiv2: annotate CVE-2019-13504

CVE-2019-13504 is misclassified (by our CVE tracker) as affecting version 0.27.2, while in fact both commits that fixed this issue are already in this version.

---
 package/exiv2/exiv2.mk | 5 +++++
 1 file changed, 5 insertions(+)
package/exiv2/exiv2.mk +5 -0
index 228b3a980e..09988f49b2 100644
@@ -10,6 +10,11 @@ EXIV2_INSTALL_STAGING = YES
 EXIV2_LICENSE = GPL-2.0+, BSD-3-Clause
 EXIV2_LICENSE_FILES = COPYING COPYING-CMAKE-SCRIPTS
 
+# CVE-2019-13504 is misclassified (by our CVE tracker) as affecting version
+# 0.27.2, while in fact both commits that fixed this issue are already in this
+# version.
+EXIV2_IGNORE_CVES += CVE-2019-13504
+
 EXIV2_CONF_OPTS += -DEXIV2_ENABLE_BUILD_SAMPLES=OFF
 
 # The following CMake variable disables a TRY_RUN call in the -pthread
Signature · 1 line
2.25.0
#1 lore raw

PATCH 2/3 package/exiv2: fix CVE-2019-17402

Exiv2 0.27.2 allows attackers to trigger a crash in Exiv2::getULong in types.cpp when called from Exiv2::Internal::CiffDirectory::readDirectory in crwimage_int.cpp, because there is no validation of the relationship of the total size to the offset and size.

---
 ...k-offset-and-size-against-total-size.patch | 39 +++++++++++++++++++
 package/exiv2/exiv2.mk                        |  3 ++
 2 files changed, 42 insertions(+)
 create mode 100644 package/exiv2/0001-crwimage-Check-offset-and-size-against-total-size.patch
package/exiv2/0001-crwimage-Check-offset-and-size-against-total-size.patch new file +39 -0
new file mode 100644
index 0000000000..78adb2e9d6
@@ -0,0 +1,39 @@
+From b7890776c62398ca1005e8edc32786859d60fcf7 Mon Sep 17 00:00:00 2001
+From: Jens Georg <[email protected]>
+Date: Sun, 6 Oct 2019 15:05:20 +0200
+Subject: [PATCH] crwimage: Check offset and size against total size
+
+Corrupted or specially crafted CRW images might exceed the overall
+buffersize.
+
+Fixes #1019
+
+(cherry picked from commit 683451567284005cd24e1ccb0a76ca401000968b)
+[Retrieved (and slightly updated to keep only the fix) from:
+https://github.com/Exiv2/exiv2/commit/50e9dd964a439da357798344ed1dd86edcadf0ec]
+Signed-off-by: Fabrice Fontaine <[email protected]>
+---
+ src/crwimage_int.cpp | 3 +++
+ 1 file changed, 3 insertions(+)
+
+diff --git a/src/crwimage_int.cpp b/src/crwimage_int.cpp
+index 29311fdb7..c0d955350 100644
+--- a/src/crwimage_int.cpp
++++ b/src/crwimage_int.cpp
+@@ -268,6 +268,9 @@ namespace Exiv2 {
+ #ifdef EXIV2_DEBUG_MESSAGES
+         std::cout << "Reading directory 0x" << std::hex << tag() << "\n";
+ #endif
++        if (this->offset() + this->size() > size)
++            throw Error(kerOffsetOutOfRange);
++
+         readDirectory(pData + offset(), this->size(), byteOrder);
+ #ifdef EXIV2_DEBUG_MESSAGES
+         std::cout << "<---- 0x" << std::hex << tag() << "\n";
+
+From 0a4fdae360c15c0f6e8d0670f8fa7acfa4526497 Mon Sep 17 00:00:00 2001
+From: =?UTF-8?q?Dan=20=C4=8Cerm=C3=A1k?= <[email protected]>
+Date: Mon, 7 Oct 2019 23:25:00 +0200
+Subject: [PATCH 2/2] [tests] Add regression test for #1019
+
+(cherry picked from commit 73b874fb14d02578f876aa7dd404cf7c07b6dc4e)
package/exiv2/exiv2.mk +3 -0
index 09988f49b2..83c880a109 100644
@@ -15,6 +15,9 @@ EXIV2_LICENSE_FILES = COPYING COPYING-CMAKE-SCRIPTS
 # version.
 EXIV2_IGNORE_CVES += CVE-2019-13504
 
+# 0001-crwimage-Check-offset-and-size-against-total-size.patch
+EXIV2_IGNORE_CVES += CVE-2019-17402
+
 EXIV2_CONF_OPTS += -DEXIV2_ENABLE_BUILD_SAMPLES=OFF
 
 # The following CMake variable disables a TRY_RUN call in the -pthread
Signature · 1 line
2.25.0
#2 lore raw

PATCH 2/3 package/exiv2: fix CVE-2019-17402

Quoted 8 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Fabrice" == Fabrice Fontaine <[email protected]> writes:

Exiv2 0.27.2 allows attackers to trigger a crash in Exiv2::getULong in types.cpp when called from Exiv2::Internal::CiffDirectory::readDirectory in crwimage_int.cpp, because there is no validation of the relationship of the total size to the offset and size.

Committed to 2019.02.x and 2019.11.x, thanks.

Signature · 1 line
Bye, Peter Korsgaard
#3 lore raw

PATCH 3/3 package/exiv2: fix CVE-2019-20421

In reply to Fabrice Fontaine

In Jp2Image::readMetadata() in jp2image.cpp in Exiv2 0.27.2, an input file can result in an infinite loop and hang, with high CPU consumption. Remote attackers could leverage this vulnerability to cause a denial of service via a crafted file.

---
 .../0002-fix_1011_jp2_readmetadata_loop.patch | 86 +++++++++++++++++++
 package/exiv2/exiv2.mk                        |  3 +
 2 files changed, 89 insertions(+)
 create mode 100644 package/exiv2/0002-fix_1011_jp2_readmetadata_loop.patch
package/exiv2/0002-fix_1011_jp2_readmetadata_loop.patch new file +86 -0
new file mode 100644
index 0000000000..400bf342ce
@@ -0,0 +1,86 @@
+From 1b917c3f7dd86336a9f6fda4456422c419dfe88c Mon Sep 17 00:00:00 2001
+From: clanmills <[email protected]>
+Date: Tue, 1 Oct 2019 17:39:44 +0100
+Subject: [PATCH] Fix #1011 fix_1011_jp2_readmetadata_loop
+
+[Retrieved (and slighlty updated to keep only the fix) from:
+https://github.com/Exiv2/exiv2/commit/a82098f4f90cd86297131b5663c3dec6a34470e8]
+Signed-off-by: Fabrice Fontaine <[email protected]>
+---
+ src/jp2image.cpp                             |  25 +++++++++++++++----
+ test/data/Jp2Image_readMetadata_loop.poc     | Bin 0 -> 738 bytes
+ tests/bugfixes/github/test_CVE_2017_17725.py |   4 +--
+ tests/bugfixes/github/test_issue_1011.py     |  13 ++++++++++
+ 4 files changed, 35 insertions(+), 7 deletions(-)
+ create mode 100755 test/data/Jp2Image_readMetadata_loop.poc
+ create mode 100644 tests/bugfixes/github/test_issue_1011.py
+
+diff --git a/src/jp2image.cpp b/src/jp2image.cpp
+index d5cd1340a..0de088d62 100644
+--- a/src/jp2image.cpp
++++ b/src/jp2image.cpp
+@@ -18,10 +18,6 @@
+  * Foundation, Inc., 51 Franklin Street, 5th Floor, Boston, MA 02110-1301 USA.
+  */
+ 
+-/*
+-  File:      jp2image.cpp
+-*/
+-
+ // *****************************************************************************
+ 
+ // included header files
+@@ -197,6 +193,16 @@ namespace Exiv2
+         return result;
+     }
+ 
++static void boxes_check(size_t b,size_t m)
++{
++    if ( b > m ) {
++#ifdef EXIV2_DEBUG_MESSAGES
++        std::cout << "Exiv2::Jp2Image::readMetadata box maximum exceeded" << std::endl;
++#endif
++        throw Error(kerCorruptedMetadata);
++    }
++}
++
+     void Jp2Image::readMetadata()
+     {
+ #ifdef EXIV2_DEBUG_MESSAGES
+@@ -219,9 +225,12 @@ namespace Exiv2
+         Jp2BoxHeader      subBox    = {0,0};
+         Jp2ImageHeaderBox ihdr      = {0,0,0,0,0,0,0,0};
+         Jp2UuidBox        uuid      = {{0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0}};
++        size_t            boxes     = 0 ;
++        size_t            boxem     = 1000 ; // boxes max
+ 
+         while (io_->read((byte*)&box, sizeof(box)) == sizeof(box))
+         {
++            boxes_check(boxes++,boxem );
+             position   = io_->tell();
+             box.length = getLong((byte*)&box.length, bigEndian);
+             box.type   = getLong((byte*)&box.type, bigEndian);
+@@ -251,8 +260,12 @@ namespace Exiv2
+ 
+                     while (io_->read((byte*)&subBox, sizeof(subBox)) == sizeof(subBox) && subBox.length )
+                     {
++                        boxes_check(boxes++, boxem) ;
+                         subBox.length = getLong((byte*)&subBox.length, bigEndian);
+                         subBox.type   = getLong((byte*)&subBox.type, bigEndian);
++                        if (subBox.length > io_->size() ) {
++                            throw Error(kerCorruptedMetadata);
++                        }
+ #ifdef EXIV2_DEBUG_MESSAGES
+                         std::cout << "Exiv2::Jp2Image::readMetadata: "
+                         << "subBox = " << toAscii(subBox.type) << " length = " << subBox.length << std::endl;
+@@ -308,7 +321,9 @@ namespace Exiv2
+                         }
+ 
+                         io_->seek(restore,BasicIo::beg);
+-                        io_->seek(subBox.length, Exiv2::BasicIo::cur);
++                        if ( io_->seek(subBox.length, Exiv2::BasicIo::cur) != 0 ) {
++                            throw Error(kerCorruptedMetadata);
++                        }
+                         restore = io_->tell();
+                     }
+                     break;
package/exiv2/exiv2.mk +3 -0
index 83c880a109..5ca16c4747 100644
@@ -18,6 +18,9 @@ EXIV2_IGNORE_CVES += CVE-2019-13504
 # 0001-crwimage-Check-offset-and-size-against-total-size.patch
 EXIV2_IGNORE_CVES += CVE-2019-17402
 
+# 0002-fix_1011_jp2_readmetadata_loop.patch
+EXIV2_IGNORE_CVES += CVE-2019-20421
+
 EXIV2_CONF_OPTS += -DEXIV2_ENABLE_BUILD_SAMPLES=OFF
 
 # The following CMake variable disables a TRY_RUN call in the -pthread
Signature · 1 line
2.25.0
#4 lore raw

PATCH 3/3 package/exiv2: fix CVE-2019-20421

Quoted 8 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Fabrice" == Fabrice Fontaine <[email protected]> writes:

In Jp2Image::readMetadata() in jp2image.cpp in Exiv2 0.27.2, an input file can result in an infinite loop and hang, with high CPU consumption. Remote attackers could leverage this vulnerability to cause a denial of service via a crafted file.

Committed to 2019.02.x and 2019.11.x, thanks.

Signature · 1 line
Bye, Peter Korsgaard
#5 lore raw

PATCH 1/3 package/exiv2: annotate CVE-2019-13504

In reply to Fabrice Fontaine

Fabrice, All,

On 2020-02-29 22:32 +0100, Fabrice Fontaine spake thusly:

Quoted 5 lines

CVE-2019-13504 is misclassified (by our CVE tracker) as affecting version 0.27.2, while in fact both commits that fixed this issue are already in this version.

I've applied patches 2 and 3, as I could follow the commit to upstream you provided in the patches.

However, for patch 1, I have been able to track only one commit, but you noted two. It would be nice if youc ould provide the sha1 for those two commits.

Thanks! :-)

Regards,
Yann E. MORIN.

Quoted 27 lines
---
 package/exiv2/exiv2.mk | 5 +++++
 1 file changed, 5 insertions(+)
package/exiv2/exiv2.mk +5 -0
index 228b3a980e..09988f49b2 100644
@@ -10,6 +10,11 @@ EXIV2_INSTALL_STAGING = YES
 EXIV2_LICENSE = GPL-2.0+, BSD-3-Clause
 EXIV2_LICENSE_FILES = COPYING COPYING-CMAKE-SCRIPTS
 
+# CVE-2019-13504 is misclassified (by our CVE tracker) as affecting version
+# 0.27.2, while in fact both commits that fixed this issue are already in this
+# version.
+EXIV2_IGNORE_CVES += CVE-2019-13504
+
 EXIV2_CONF_OPTS += -DEXIV2_ENABLE_BUILD_SAMPLES=OFF
 
 # The following CMake variable disables a TRY_RUN call in the -pthread
Signature · 6 lines
2.25.0

_______________________________________________
buildroot mailing list
buildroot at busybox.net
http://lists.busybox.net/mailman/listinfo/buildroot
Signature · 6 lines
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 561 099 427 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'
#6 lore raw

PATCH 1/3 package/exiv2: annotate CVE-2019-13504

Le sam. 29 f?vr. 2020 ? 23:21, Yann E. MORIN <[email protected]> a ?crit :

Quoted 8 lines

Fabrice, All,

On 2020-02-29 22:32 +0100, Fabrice Fontaine spake thusly:

Quoted 4 lines

CVE-2019-13504 is misclassified (by our CVE tracker) as affecting version 0.27.2, while in fact both commits that fixed this issue are already in this version.

Quoted 8 lines
Quoted 1 line

I've applied patches 2 and 3, as I could follow the commit to upstream you provided in the patches.

However, for patch 1, I have been able to track only one commit, but you noted two. It would be nice if youc ould provide the sha1 for those two commits.

Sure, here it is (from
https://security-tracker.debian.org/tracker/CVE-2019-13504):
- https://github.com/Exiv2/exiv2/commit/bd0afe0390439b2c424d881c8c6eb0c5624e31d9
- https://github.com/Exiv2/exiv2/commit/54f0bebca032d0286a0e48f47e67dfc6141fedff

Quoted 36 lines

Thanks! :-)

Regards,
Yann E. MORIN.

Quoted 23 lines
---
 package/exiv2/exiv2.mk | 5 +++++
 1 file changed, 5 insertions(+)
package/exiv2/exiv2.mk +5 -1
index 228b3a980e..09988f49b2 100644
@@ -10,6 +10,11 @@ EXIV2_INSTALL_STAGING = YES
 EXIV2_LICENSE = GPL-2.0+, BSD-3-Clause
 EXIV2_LICENSE_FILES = COPYING COPYING-CMAKE-SCRIPTS
 
+# CVE-2019-13504 is misclassified (by our CVE tracker) as affecting version
+# 0.27.2, while in fact both commits that fixed this issue are already in this
+# version.
+EXIV2_IGNORE_CVES += CVE-2019-13504
+
 EXIV2_CONF_OPTS += -DEXIV2_ENABLE_BUILD_SAMPLES=OFF
 
 # The following CMake variable disables a TRY_RUN call in the -pthread
--

2.25.0

Quoted 4 lines

_______________________________________________
buildroot mailing list
buildroot at busybox.net
http://lists.busybox.net/mailman/listinfo/buildroot

--
.-----------------.--------------------.------------------.--------------------.

Quoted 5 lines
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 561 099 427 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |

'------------------------------^-------^------------------^--------------------'

Best Regards,

Fabrice

#7 lore raw

PATCH 1/3 package/exiv2: annotate CVE-2019-13504

Fabrice, All,

On 2020-02-29 23:28 +0100, Fabrice Fontaine spake thusly:

Quoted 5 lines

Le sam. 29 f?vr. 2020 ? 23:21, Yann E. MORIN <[email protected]> a ?crit :

Quoted 4 lines

On 2020-02-29 22:32 +0100, Fabrice Fontaine spake thusly:

Quoted 3 lines

CVE-2019-13504 is misclassified (by our CVE tracker) as affecting version 0.27.2, while in fact both commits that fixed this issue are already in this version.

[--SNIP--]

Quoted 7 lines
Quoted 3 lines

However, for patch 1, I have been able to track only one commit, but you noted two. It would be nice if youc ould provide the sha1 for those two commits.

Sure, here it is (from
https://security-tracker.debian.org/tracker/CVE-2019-13504):
- https://github.com/Exiv2/exiv2/commit/bd0afe0390439b2c424d881c8c6eb0c5624e31d9
- https://github.com/Exiv2/exiv2/commit/54f0bebca032d0286a0e48f47e67dfc6141fedff

Applied to master with those references added to the commit log. Thanks! :-)

Regards,
Yann E. MORIN.

Signature · 6 lines
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 561 099 427 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'