Skip to contentSkip to conversation

gzip: fix CVE-2026-41992

PATCH In review Jaipaul Cheernam 2 participants 10 messages Jul 7 – Jul 17

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 gzip: fix CVE-2026-41992

Backport upstream fix for a global buffer overflow in the LZH decompression logic (unlzh.c). The left[] and right[] global arrays shared across LZW and LZH decompression routines are not reinitialized between files processed in the same invocation, allowing an out-of-bounds read in the LZH decoder.

Reference: https://nvd.nist.gov/vuln/detail/CVE-2026-41992

---
 .../gzip/gzip-1.14/CVE-2026-41992.patch       | 32 +++++++++++++++++++
 meta/recipes-extended/gzip/gzip_1.14.bb       |  1 +
 2 files changed, 33 insertions(+)
 create mode 100644 meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch
meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch new file +32 -0
new file mode 100644
index 0000000000..5d3534fcb1
@@ -0,0 +1,32 @@
+From 63dbf6b3b9e6e781df1a6a64e609b10e23969681 Mon Sep 17 00:00:00 2001
+From: Paul Eggert <[email protected]>
+Date: Wed, 15 Apr 2026 12:00:17 -0700
+Subject: gzip: don't mishandle .lzh after .Z
+
+Problem reported by Michal Majchrowicz.
+* unlzh.c (read_c_len): Clear left and right when n == 0.
+
+CVE: CVE-2026-41992
+Upstream-Status: Backport [https://cgit.git.savannah.gnu.org/cgit/gzip.git/commit/?id=63dbf6b3b9e6e781df1a6a64e609b10e23969681]
+Signed-off-by: Jaipaul Cheernam <[email protected]>
+---
+ unlzh.c | 6 ++++++
+ 1 file changed, 6 insertions(+)
+
+diff --git a/unlzh.c b/unlzh.c
+index 3320196..a6cf109 100644
+--- a/unlzh.c
++++ b/unlzh.c
+@@ -232,6 +232,12 @@ read_c_len ()
+         c = getbits(CBIT);
+         for (i = 0; i < NC; i++) c_len[i] = 0;
+         for (i = 0; i < 4096; i++) c_table[i] = c;
++
++        /* Needed in case LEFT and RIGHT are reused from a previous
++           LZW decompression.  It may be overkill to clear all of both
++           arrays, but nobody has had time to analyze this carefully.  */
++        memzero(left, (2 * NC - 1) * sizeof *left);
++        memzero(right, (2 * NC - 1) * sizeof *left);
+     } else {
+         i = 0;
+         while (i < n) {
meta/recipes-extended/gzip/gzip_1.14.bb +1 -0
index c7837cdae0..99174a58c5 100644
@@ -6,6 +6,7 @@ LICENSE = "GPL-3.0-or-later"
 
 SRC_URI = "${GNU_MIRROR}/gzip/${BP}.tar.gz \
            file://run-ptest \
+           file://CVE-2026-41992.patch \
            "
 SRC_URI:append:class-target = " file://wrong-path-fix.patch"
 
Signature · 1 line
2.34.1
#1 lore raw

PATCH 1/1 gzip: fix CVE-2026-41992

Backport upstream fix for a global buffer overflow in the LZH decompression logic (unlzh.c). The left[] and right[] global arrays shared across LZW and LZH decompression routines are not reinitialized between files processed in the same invocation, allowing an out-of-bounds read in the LZH decoder.

Reference: https://nvd.nist.gov/vuln/detail/CVE-2026-41992

---
 .../gzip/gzip-1.13/CVE-2026-41992.patch       | 32 +++++++++++++++++++
 meta/recipes-extended/gzip/gzip_1.13.bb       |  1 +
 2 files changed, 33 insertions(+)
 create mode 100644 meta/recipes-extended/gzip/gzip-1.13/CVE-2026-41992.patch
meta/recipes-extended/gzip/gzip-1.13/CVE-2026-41992.patch new file +32 -0
new file mode 100644
index 0000000000..5d3534fcb1
@@ -0,0 +1,32 @@
+From 63dbf6b3b9e6e781df1a6a64e609b10e23969681 Mon Sep 17 00:00:00 2001
+From: Paul Eggert <[email protected]>
+Date: Wed, 15 Apr 2026 12:00:17 -0700
+Subject: gzip: don't mishandle .lzh after .Z
+
+Problem reported by Michal Majchrowicz.
+* unlzh.c (read_c_len): Clear left and right when n == 0.
+
+CVE: CVE-2026-41992
+Upstream-Status: Backport [https://cgit.git.savannah.gnu.org/cgit/gzip.git/commit/?id=63dbf6b3b9e6e781df1a6a64e609b10e23969681]
+Signed-off-by: Jaipaul Cheernam <[email protected]>
+---
+ unlzh.c | 6 ++++++
+ 1 file changed, 6 insertions(+)
+
+diff --git a/unlzh.c b/unlzh.c
+index 3320196..a6cf109 100644
+--- a/unlzh.c
++++ b/unlzh.c
+@@ -232,6 +232,12 @@ read_c_len ()
+         c = getbits(CBIT);
+         for (i = 0; i < NC; i++) c_len[i] = 0;
+         for (i = 0; i < 4096; i++) c_table[i] = c;
++
++        /* Needed in case LEFT and RIGHT are reused from a previous
++           LZW decompression.  It may be overkill to clear all of both
++           arrays, but nobody has had time to analyze this carefully.  */
++        memzero(left, (2 * NC - 1) * sizeof *left);
++        memzero(right, (2 * NC - 1) * sizeof *left);
+     } else {
+         i = 0;
+         while (i < n) {
meta/recipes-extended/gzip/gzip_1.13.bb +1 -0
index fd846b30a5..208220867a 100644
@@ -6,6 +6,7 @@ LICENSE = "GPL-3.0-or-later"
 
 SRC_URI = "${GNU_MIRROR}/gzip/${BP}.tar.gz \
            file://run-ptest \
+           file://CVE-2026-41992.patch \
            "
 SRC_URI:append:class-target = " file://wrong-path-fix.patch"
 
Signature · 1 line
2.34.1
#2 lore raw

PATCH v2 1/1 gzip: fix CVE-2026-41992

Backport upstream fix for a global buffer overflow in the LZH decompression logic (unlzh.c). The left[] and right[] global arrays shared across LZW and LZH decompression routines are not reinitialized between files processed in the same invocation, allowing an out-of-bounds read in the LZH decoder.

Adapted for gzip 1.13:
- Refreshed NEWS and THANKS hunks to match 1.13 release context.

Reference: https://nvd.nist.gov/vuln/detail/CVE-2026-41992

---

Changes in v2:
- Include full upstream patch (NEWS, THANKS, unlzh.c) instead of only unlzh.c
- Preserve original UTF-8 encoding in Subject and commit message
- Refreshed NEWS and THANKS hunks for target gzip version

 .../gzip/gzip-1.13/CVE-2026-41992.patch       | 64 +++++++++++++++++++
 meta/recipes-extended/gzip/gzip_1.13.bb       |  1 +
 2 files changed, 65 insertions(+)
 create mode 100644 meta/recipes-extended/gzip/gzip-1.13/CVE-2026-41992.patch
meta/recipes-extended/gzip/gzip-1.13/CVE-2026-41992.patch new file +64 -0
new file mode 100644
index 0000000000..4db9a1c1af
@@ -0,0 +1,64 @@
+From 63dbf6b3b9e6e781df1a6a64e609b10e23969681 Mon Sep 17 00:00:00 2001
+From: Paul Eggert <[email protected]>
+Date: Wed, 15 Apr 2026 12:00:17 -0700
+Subject: =?UTF-8?q?gzip:=20don=E2=80=99t=20mishandle=20.lzh=20after=20.Z?=
+MIME-Version: 1.0
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: 8bit
+
+Problem reported by Michał Majchrowicz.
+* unlzh.c (read_c_len): Clear left and right when n == 0.
+
+CVE: CVE-2026-41992
+Upstream-Status: Backport [https://cgit.git.savannah.gnu.org/cgit/gzip.git/commit/?id=63dbf6b3b9e6e781df1a6a64e609b10e23969681]
+Signed-off-by: Jaipaul Cheernam <[email protected]>
+---
+ NEWS    | 4 ++++
+ THANKS  | 1 +
+ unlzh.c | 6 ++++++
+ 3 files changed, 11 insertions(+)
+
+diff --git a/NEWS b/NEWS
+index bdb7cc0..af2a08f 100644
+--- a/NEWS
++++ b/NEWS
+@@ -14,6 +14,10 @@ GNU gzip NEWS                                    -*- outline -*-
+ 
+ ** Bug fixes
+ 
++  A buffer overflow has been fixed when decompressing an .lzh file
++  after decompressing a .Z file.
++  [bug present since the beginning]
++
+   'gzip -d' no longer fails to report invalid compressed data
+   that uses a dictionary distance outside the input window.
+   [bug present since the beginning]
+diff --git a/THANKS b/THANKS
+index 4e545d9..a7d25e4 100644
+--- a/THANKS
++++ b/THANKS
+@@ -186,6 +186,7 @@ Jamie Lokier            [email protected]
+ Richard Lloyd           [email protected]
+ David J. MacKenzie	[email protected]
+ John R MacMillan        [email protected]
++Michał Majchrowicz	[email protected]
+ Ron Male                [email protected]
+ Jakub Martisko		[email protected]
+ Don R. Maszle           [email protected]
+diff --git a/unlzh.c b/unlzh.c
+index 3320196..a6cf109 100644
+--- a/unlzh.c
++++ b/unlzh.c
+@@ -232,6 +232,12 @@ read_c_len ()
+         c = getbits(CBIT);
+         for (i = 0; i < NC; i++) c_len[i] = 0;
+         for (i = 0; i < 4096; i++) c_table[i] = c;
++
++        /* Needed in case LEFT and RIGHT are reused from a previous
++           LZW decompression.  It may be overkill to clear all of both
++           arrays, but nobody has had time to analyze this carefully.  */
++        memzero(left, (2 * NC - 1) * sizeof *left);
++        memzero(right, (2 * NC - 1) * sizeof *left);
+     } else {
+         i = 0;
+         while (i < n) {
meta/recipes-extended/gzip/gzip_1.13.bb +1 -0
index fd846b30a5..208220867a 100644
@@ -6,6 +6,7 @@ LICENSE = "GPL-3.0-or-later"
 
 SRC_URI = "${GNU_MIRROR}/gzip/${BP}.tar.gz \
            file://run-ptest \
+           file://CVE-2026-41992.patch \
            "
 SRC_URI:append:class-target = " file://wrong-path-fix.patch"
 
Signature · 1 line
2.34.1
#3 lore raw

PATCH 1/1 gzip: fix CVE-2026-41992

In reply to Jaipaul Cheernam

Backport upstream fix for a global buffer overflow in the LZH decompression logic (unlzh.c). The left[] and right[] global arrays shared across LZW and LZH decompression routines are not reinitialized between files processed in the same invocation, allowing an out-of-bounds read in the LZH decoder.

Reference: https://nvd.nist.gov/vuln/detail/CVE-2026-41992

---
 .../gzip/gzip-1.14/CVE-2026-41992.patch       | 32 +++++++++++++++++++
 meta/recipes-extended/gzip/gzip_1.14.bb       |  1 +
 2 files changed, 33 insertions(+)
 create mode 100644 meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch
meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch new file +32 -0
new file mode 100644
index 0000000000..5d3534fcb1
@@ -0,0 +1,32 @@
+From 63dbf6b3b9e6e781df1a6a64e609b10e23969681 Mon Sep 17 00:00:00 2001
+From: Paul Eggert <[email protected]>
+Date: Wed, 15 Apr 2026 12:00:17 -0700
+Subject: gzip: don't mishandle .lzh after .Z
+
+Problem reported by Michal Majchrowicz.
+* unlzh.c (read_c_len): Clear left and right when n == 0.
+
+CVE: CVE-2026-41992
+Upstream-Status: Backport [https://cgit.git.savannah.gnu.org/cgit/gzip.git/commit/?id=63dbf6b3b9e6e781df1a6a64e609b10e23969681]
+Signed-off-by: Jaipaul Cheernam <[email protected]>
+---
+ unlzh.c | 6 ++++++
+ 1 file changed, 6 insertions(+)
+
+diff --git a/unlzh.c b/unlzh.c
+index 3320196..a6cf109 100644
+--- a/unlzh.c
++++ b/unlzh.c
+@@ -232,6 +232,12 @@ read_c_len ()
+         c = getbits(CBIT);
+         for (i = 0; i < NC; i++) c_len[i] = 0;
+         for (i = 0; i < 4096; i++) c_table[i] = c;
++
++        /* Needed in case LEFT and RIGHT are reused from a previous
++           LZW decompression.  It may be overkill to clear all of both
++           arrays, but nobody has had time to analyze this carefully.  */
++        memzero(left, (2 * NC - 1) * sizeof *left);
++        memzero(right, (2 * NC - 1) * sizeof *left);
+     } else {
+         i = 0;
+         while (i < n) {
meta/recipes-extended/gzip/gzip_1.14.bb +1 -0
index c7837cdae0..99174a58c5 100644
@@ -6,6 +6,7 @@ LICENSE = "GPL-3.0-or-later"
 
 SRC_URI = "${GNU_MIRROR}/gzip/${BP}.tar.gz \
            file://run-ptest \
+           file://CVE-2026-41992.patch \
            "
 SRC_URI:append:class-target = " file://wrong-path-fix.patch"
 
Signature · 1 line
2.34.1
#4 lore raw

PATCH v2 1/1 gzip: fix CVE-2026-41992

Backport upstream fix for a global buffer overflow in the LZH decompression logic (unlzh.c). The left[] and right[] global arrays shared across LZW and LZH decompression routines are not reinitialized between files processed in the same invocation, allowing an out-of-bounds read in the LZH decoder.

Adapted for gzip 1.14:
- Refreshed NEWS and THANKS hunks to match 1.14 release context.

Reference: https://nvd.nist.gov/vuln/detail/CVE-2026-41992

---

Changes in v2:
- Include full upstream patch (NEWS, THANKS, unlzh.c) instead of only unlzh.c
- Preserve original UTF-8 encoding in Subject and commit message
- Refreshed NEWS and THANKS hunks for target gzip version

 .../gzip/gzip-1.14/CVE-2026-41992.patch       | 64 +++++++++++++++++++
 meta/recipes-extended/gzip/gzip_1.14.bb       |  1 +
 2 files changed, 65 insertions(+)
 create mode 100644 meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch
meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch new file +64 -0
new file mode 100644
index 0000000000..f55c89978d
@@ -0,0 +1,64 @@
+From 63dbf6b3b9e6e781df1a6a64e609b10e23969681 Mon Sep 17 00:00:00 2001
+From: Paul Eggert <[email protected]>
+Date: Wed, 15 Apr 2026 12:00:17 -0700
+Subject: =?UTF-8?q?gzip:=20don=E2=80=99t=20mishandle=20.lzh=20after=20.Z?=
+MIME-Version: 1.0
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: 8bit
+
+Problem reported by Michał Majchrowicz.
+* unlzh.c (read_c_len): Clear left and right when n == 0.
+
+CVE: CVE-2026-41992
+Upstream-Status: Backport [https://cgit.git.savannah.gnu.org/cgit/gzip.git/commit/?id=63dbf6b3b9e6e781df1a6a64e609b10e23969681]
+Signed-off-by: Jaipaul Cheernam <[email protected]>
+---
+ NEWS    | 4 ++++
+ THANKS  | 1 +
+ unlzh.c | 6 ++++++
+ 3 files changed, 11 insertions(+)
+
+diff --git a/NEWS b/NEWS
+index 6388227..8fb8918 100644
+--- a/NEWS
++++ b/NEWS
+@@ -4,6 +4,10 @@ GNU gzip NEWS                                    -*- outline -*-
+ 
+ ** Bug fixes
+ 
++  A buffer overflow has been fixed when decompressing an .lzh file
++  after decompressing a .Z file.
++  [bug present since the beginning]
++
+   'gzip -d' no longer omits the last partial output buffer when the
+   input ends unexpectedly on an IBM Z platform.
+   [bug introduced in gzip-1.11]
+diff --git a/THANKS b/THANKS
+index 4e545d9..a7d25e4 100644
+--- a/THANKS
++++ b/THANKS
+@@ -186,6 +186,7 @@ Jamie Lokier            [email protected]
+ Richard Lloyd           [email protected]
+ David J. MacKenzie	[email protected]
+ John R MacMillan        [email protected]
++Michał Majchrowicz	[email protected]
+ Ron Male                [email protected]
+ Jakub Martisko		[email protected]
+ Don R. Maszle           [email protected]
+diff --git a/unlzh.c b/unlzh.c
+index 3320196..a6cf109 100644
+--- a/unlzh.c
++++ b/unlzh.c
+@@ -232,6 +232,12 @@ read_c_len ()
+         c = getbits(CBIT);
+         for (i = 0; i < NC; i++) c_len[i] = 0;
+         for (i = 0; i < 4096; i++) c_table[i] = c;
++
++        /* Needed in case LEFT and RIGHT are reused from a previous
++           LZW decompression.  It may be overkill to clear all of both
++           arrays, but nobody has had time to analyze this carefully.  */
++        memzero(left, (2 * NC - 1) * sizeof *left);
++        memzero(right, (2 * NC - 1) * sizeof *left);
+     } else {
+         i = 0;
+         while (i < n) {
meta/recipes-extended/gzip/gzip_1.14.bb +1 -0
index c7837cdae0..99174a58c5 100644
@@ -6,6 +6,7 @@ LICENSE = "GPL-3.0-or-later"
 
 SRC_URI = "${GNU_MIRROR}/gzip/${BP}.tar.gz \
            file://run-ptest \
+           file://CVE-2026-41992.patch \
            "
 SRC_URI:append:class-target = " file://wrong-path-fix.patch"
 
Signature · 1 line
2.34.1
#5 lore raw

Hi Yoann,

Master commit has been merged and wrynose commit marked as Awaiting upstream, Please check and let me know if I need to send another patch.

Regards,
Jaipaul

#6 lore raw
Quoted 5 lines · Jaipaul Cheernam via lists.openembedded.org

Hi Yoann,

Master commit has been merged and wrynose commit marked as Awaiting upstream, Please check and let me know if I need to send another patch.

Noted, thanks for the ping.
I've put both scarhgap/wrynose patches in the review lists.

Regards,

Quoted 14 lines

Regards,
Jaipaul

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group. View/Reply Online (#241156):
https://lists.openembedded.org/g/openembedded-core/message/241156
Mute This Topic: https://lists.openembedded.org/mt/120178637/4316185
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [ [email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Signature · 2 lines
Yoann Congal
Smile ECS
#7 lore raw
In reply to Jaipaul Cheernam
Quoted 21 lines · Jaipaul Cheernam via lists.openembedded.org

Backport upstream fix for a global buffer overflow in the LZH decompression logic (unlzh.c). The left[] and right[] global arrays shared across LZW and LZH decompression routines are not reinitialized between files processed in the same invocation, allowing an out-of-bounds read in the LZH decoder.

Reference: https://nvd.nist.gov/vuln/detail/CVE-2026-41992

---
 .../gzip/gzip-1.14/CVE-2026-41992.patch       | 32 +++++++++++++++++++
 meta/recipes-extended/gzip/gzip_1.14.bb       |  1 +
 2 files changed, 33 insertions(+)
 create mode 100644 meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch
meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch new file +1 -0
new file mode 100644
index 0000000000..5d3534fcb1
@@ -0,0 +1,32 @@
+From 63dbf6b3b9e6e781df1a6a64e609b10e23969681 Mon Sep 17 00:00:00 2001

meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch line 11

Quoted 10 lines
meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch new file +10 -0
+From: Paul Eggert <[email protected]>
+Date: Wed, 15 Apr 2026 12:00:17 -0700
+Subject: gzip: don't mishandle .lzh after .Z
+
+Problem reported by Michal Majchrowicz.
+* unlzh.c (read_c_len): Clear left and right when n == 0.
+
+CVE: CVE-2026-41992
+Upstream-Status: Backport [https://cgit.git.savannah.gnu.org/cgit/gzip.git/commit/?id=63dbf6b3b9e6e781df1a6a64e609b10e23969681]
+Signed-off-by: Jaipaul Cheernam <[email protected]>

Hello,

This patch is slightly different from upstream (without explanation):
-Subject: =?UTF-8?q?gzip:=20don=E2=80=99t=20mishandle=20.lzh=20after=20.Z?= -MIME-Version: 1.0
-Content-Type: text/plain; charset=UTF-8 -Content-Transfer-Encoding: 8bit
+Subject: gzip: don't mishandle .lzh after .Z

-Problem reported by Michał Majchrowicz.
+Problem reported by Michal Majchrowicz.

... and the THANKS file was dropped.

Michał name was transcribed as Michal. I don't know how "nitpickery" it is...

Regards,

Signature · 2 lines
Yoann Congal
Smile ECS
#8 lore raw

Hi Yoann,

Something went wrong from myside. I will check and share v2 .

Apologies for the confusion.

Regards,
Jaipaul
________________________________
From: Yoann Congal <[email protected]>
Sent: Wednesday, July 8, 2026 5:19 PM
To: Jaipaul Cheernam <[email protected]>; [email protected] <[email protected]>
Subject: Re: [OE-core] [PATCH] gzip: fix CVE-2026-41992

Quoted 21 lines · Jaipaul Cheernam via lists.openembedded.org

Backport upstream fix for a global buffer overflow in the LZH decompression logic (unlzh.c). The left[] and right[] global arrays shared across LZW and LZH decompression routines are not reinitialized between files processed in the same invocation, allowing an out-of-bounds read in the LZH decoder.

Reference: https://nvd.nist.gov/vuln/detail/CVE-2026-41992

---
 .../gzip/gzip-1.14/CVE-2026-41992.patch       | 32 +++++++++++++++++++
 meta/recipes-extended/gzip/gzip_1.14.bb       |  1 +
 2 files changed, 33 insertions(+)
 create mode 100644 meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch
meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch new file +1 -0
new file mode 100644
index 0000000000..5d3534fcb1
@@ -0,0 +1,32 @@
+From 63dbf6b3b9e6e781df1a6a64e609b10e23969681 Mon Sep 17 00:00:00 2001

meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch line 11

Quoted 10 lines
meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch new file +10 -0
+From: Paul Eggert <[email protected]>
+Date: Wed, 15 Apr 2026 12:00:17 -0700
+Subject: gzip: don't mishandle .lzh after .Z
+
+Problem reported by Michal Majchrowicz.
+* unlzh.c (read_c_len): Clear left and right when n == 0.
+
+CVE: CVE-2026-41992
+Upstream-Status: Backport [https://cgit.git.savannah.gnu.org/cgit/gzip.git/commit/?id=63dbf6b3b9e6e781df1a6a64e609b10e23969681]
+Signed-off-by: Jaipaul Cheernam <[email protected]>

Hello,

This patch is slightly different from upstream (without explanation):
-Subject: =?UTF-8?q?gzip:=20don=E2=80=99t=20mishandle=20.lzh=20after=20.Z?= -MIME-Version: 1.0
-Content-Type: text/plain; charset=UTF-8 -Content-Transfer-Encoding: 8bit
+Subject: gzip: don't mishandle .lzh after .Z

-Problem reported by Michał Majchrowicz.
+Problem reported by Michal Majchrowicz.

... and the THANKS file was dropped.

Michał name was transcribed as Michal. I don't know how "nitpickery" it is...

Regards,
--
Yoann Congal
Smile ECS

#9 lore raw

PATCH v2 1/1 gzip: fix CVE-2026-41992

In reply to Jaipaul Cheernam

Backport upstream fix for a global buffer overflow in the LZH decompression logic (unlzh.c). The left[] and right[] global arrays shared across LZW and LZH decompression routines are not reinitialized between files processed in the same invocation, allowing an out-of-bounds read in the LZH decoder.

Adapted for gzip 1.14:
- Refreshed NEWS and THANKS hunks to match 1.14 release context.

Reference: https://nvd.nist.gov/vuln/detail/CVE-2026-41992

---

Changes in v2:
- Include full upstream patch (NEWS, THANKS, unlzh.c) instead of only unlzh.c
- Preserve original UTF-8 encoding in Subject and commit message
- Refreshed NEWS and THANKS hunks for target gzip version

 .../gzip/gzip-1.14/CVE-2026-41992.patch       | 64 +++++++++++++++++++
 meta/recipes-extended/gzip/gzip_1.14.bb       |  1 +
 2 files changed, 65 insertions(+)
 create mode 100644 meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch
meta/recipes-extended/gzip/gzip-1.14/CVE-2026-41992.patch new file +64 -0
new file mode 100644
index 0000000000..f55c89978d
@@ -0,0 +1,64 @@
+From 63dbf6b3b9e6e781df1a6a64e609b10e23969681 Mon Sep 17 00:00:00 2001
+From: Paul Eggert <[email protected]>
+Date: Wed, 15 Apr 2026 12:00:17 -0700
+Subject: =?UTF-8?q?gzip:=20don=E2=80=99t=20mishandle=20.lzh=20after=20.Z?=
+MIME-Version: 1.0
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: 8bit
+
+Problem reported by Michał Majchrowicz.
+* unlzh.c (read_c_len): Clear left and right when n == 0.
+
+CVE: CVE-2026-41992
+Upstream-Status: Backport [https://cgit.git.savannah.gnu.org/cgit/gzip.git/commit/?id=63dbf6b3b9e6e781df1a6a64e609b10e23969681]
+Signed-off-by: Jaipaul Cheernam <[email protected]>
+---
+ NEWS    | 4 ++++
+ THANKS  | 1 +
+ unlzh.c | 6 ++++++
+ 3 files changed, 11 insertions(+)
+
+diff --git a/NEWS b/NEWS
+index 6388227..8fb8918 100644
+--- a/NEWS
++++ b/NEWS
+@@ -4,6 +4,10 @@ GNU gzip NEWS                                    -*- outline -*-
+ 
+ ** Bug fixes
+ 
++  A buffer overflow has been fixed when decompressing an .lzh file
++  after decompressing a .Z file.
++  [bug present since the beginning]
++
+   'gzip -d' no longer omits the last partial output buffer when the
+   input ends unexpectedly on an IBM Z platform.
+   [bug introduced in gzip-1.11]
+diff --git a/THANKS b/THANKS
+index 4e545d9..a7d25e4 100644
+--- a/THANKS
++++ b/THANKS
+@@ -186,6 +186,7 @@ Jamie Lokier            [email protected]
+ Richard Lloyd           [email protected]
+ David J. MacKenzie	[email protected]
+ John R MacMillan        [email protected]
++Michał Majchrowicz	[email protected]
+ Ron Male                [email protected]
+ Jakub Martisko		[email protected]
+ Don R. Maszle           [email protected]
+diff --git a/unlzh.c b/unlzh.c
+index 3320196..a6cf109 100644
+--- a/unlzh.c
++++ b/unlzh.c
+@@ -232,6 +232,12 @@ read_c_len ()
+         c = getbits(CBIT);
+         for (i = 0; i < NC; i++) c_len[i] = 0;
+         for (i = 0; i < 4096; i++) c_table[i] = c;
++
++        /* Needed in case LEFT and RIGHT are reused from a previous
++           LZW decompression.  It may be overkill to clear all of both
++           arrays, but nobody has had time to analyze this carefully.  */
++        memzero(left, (2 * NC - 1) * sizeof *left);
++        memzero(right, (2 * NC - 1) * sizeof *left);
+     } else {
+         i = 0;
+         while (i < n) {
meta/recipes-extended/gzip/gzip_1.14.bb +1 -0
index c7837cdae0..99174a58c5 100644
@@ -6,6 +6,7 @@ LICENSE = "GPL-3.0-or-later"
 
 SRC_URI = "${GNU_MIRROR}/gzip/${BP}.tar.gz \
            file://run-ptest \
+           file://CVE-2026-41992.patch \
            "
 SRC_URI:append:class-target = " file://wrong-path-fix.patch"
 
Signature · 1 line
2.34.1