[meta-core,1/1] Signed-off-by: Manoj Saun <manojsingh.saun@windriver.com>

-> libksba: fix CVE-2022-3515

-> libksba: integer overflow may lead to remote code execution.

-> Reference:
-> https://www.gnupg.org/blog/20221017-pepe-left-the-ksba.html

-> Upstream fixes:
-> https://git.gnupg.org/cgi-bin/gitweb.cgi?p=libksba.git;a=commit;h=4b7d9cd4a018898d7714ce06f3faf2626c14582b
---
 ...e-overflow-directly-in-the-TLV-parse.patch | 42 +++++++++++++++++++
 meta/recipes-support/libksba/libksba_1.6.2.bb |  3 +-
 2 files changed, 44 insertions(+), 1 deletion(-)
 create mode 100644 meta/recipes-support/libksba/libksba/0001-Detect-a-possible-overflow-directly-in-the-TLV-parse.patch

Manoj,

Welcome to Yocto/oe-core.

I'm not sure what happened with your subject line but it's not
what we expect for patches on the oe-core list.

Please read:
https://www.openembedded.org/wiki/How_to_submit_a_patch_to_OpenEmbedded
and several of the commits to this list and then we can  work on this 
together offline.


On 2022-12-07 07:30, Manoj Saun via lists.openembedded.org wrote:
> -> libksba: fix CVE-2022-3515

This should be your subject line:

libksba: fix CVE-2022-3515

>
> -> libksba: integer overflow may lead to remote code execution.
Drop the "-> " prefix entirely, it's not our style.
>
> -> Reference:
> -> https://www.gnupg.org/blog/20221017-pepe-left-the-ksba.html
>
> -> Upstream fixes:
> -> https://git.gnupg.org/cgi-bin/gitweb.cgi?p=libksba.git;a=commit;h=4b7d9cd4a018898d7714ce06f3faf2626c14582b
> ---
>   ...e-overflow-directly-in-the-TLV-parse.patch | 42 +++++++++++++++++++
>   meta/recipes-support/libksba/libksba_1.6.2.bb |  3 +-
>   2 files changed, 44 insertions(+), 1 deletion(-)
>   create mode 100644 meta/recipes-support/libksba/libksba/0001-Detect-a-possible-overflow-directly-in-the-TLV-parse.patch
>
> diff --git a/meta/recipes-support/libksba/libksba/0001-Detect-a-possible-overflow-directly-in-the-TLV-parse.patch b/meta/recipes-support/libksba/libksba/0001-Detect-a-possible-overflow-directly-in-the-TLV-parse.patch
> new file mode 100644
> index 0000000000..e2cb842a4d
> --- /dev/null
> +++ b/meta/recipes-support/libksba/libksba/0001-Detect-a-possible-overflow-directly-in-the-TLV-parse.patch
> @@ -0,0 +1,42 @@
> +From 4b7d9cd4a018898d7714ce06f3faf2626c14582b Mon Sep 17 00:00:00 2001
> +From: Werner Koch <wk@gnupg.org>
> +Date: Wed, 5 Oct 2022 14:19:06 +0200
> +Subject: [PATCH] Detect a possible overflow directly in the TLV parser.
> +
> +* src/ber-help.c (_ksba_ber_read_tl): Check for overflow of a commonly
> +used sum.

As per the document above, we insert:

+CVE: CVE-2022-3515
+Upstream-Status: Backport
+Signed-off-by: First Lastname<First.Lastname@example.com>

> +--
> +
> +It is quite common to have checks like
> +
> +    if (ti.nhdr + ti.length >= DIM(tmpbuf))
> +       return gpg_error (GPG_ERR_TOO_LARGE);
> +
> +This patch detects possible integer overflows immmediately when
> +creating the TI object.
> +
> +Reported-by: ZDI-CAN-18927, ZDI-CAN-18928, ZDI-CAN-18929
> +---
> + src/ber-help.c | 6 ++++++
> + 1 file changed, 6 insertions(+)
> +
> +diff --git a/src/ber-help.c b/src/ber-help.c
> +index 81c31ed..56efb6a 100644
> +--- a/src/ber-help.c
> ++++ b/src/ber-help.c
> +@@ -182,6 +182,12 @@ _ksba_ber_read_tl (ksba_reader_t reader, struct tag_info *ti)
> +       ti->length = len;
> +     }
> +
> ++  if (ti->length > ti->nhdr && (ti->nhdr + ti->length) < ti->length)
> ++    {
> ++      ti->err_string = "header+length would overflow";
> ++      return gpg_error (GPG_ERR_EOVERFLOW);
> ++    }
> ++
> +   /* Without this kludge some example certs can't be parsed */
> +   if (ti->class == CLASS_UNIVERSAL && !ti->tag)
> +     ti->length = 0;
> +--
> +2.34.1
> +
> diff --git a/meta/recipes-support/libksba/libksba_1.6.2.bb b/meta/recipes-support/libksba/libksba_1.6.2.bb
> index f6ecb9aec4..c25c23ef0f 100644
> --- a/meta/recipes-support/libksba/libksba_1.6.2.bb
> +++ b/meta/recipes-support/libksba/libksba_1.6.2.bb
> @@ -22,7 +22,8 @@ inherit autotools binconfig-disabled pkgconfig texinfo
>   
>   UPSTREAM_CHECK_URI = "https://gnupg.org/download/index.html"
>   SRC_URI = "${GNUPG_MIRROR}/${BPN}/${BPN}-${PV}.tar.bz2 \
> -           file://ksba-add-pkgconfig-support.patch"
> +           file://ksba-add-pkgconfig-support.patch \
> +           file://0001-Detect-a-possible-overflow-directly-in-the-TLV-parse.patch"
The patch above is fine.

You find that if you keep the trailing close quote on a new line,

like this:

  UPSTREAM_CHECK_URI = "https://gnupg.org/download/index.html"
  SRC_URI = "${GNUPG_MIRROR}/${BPN}/${BPN}-${PV}.tar.bz2 \
-           file://ksba-add-pkgconfig-support.patch"
+           file://ksba-add-pkgconfig-support.patch \
+           file://0001-Detect-a-possible-overflow-directly-in-the-TLV-parse.patch \
+ "

then the next time a patch is added, the diff will be one line smaller and just a bit
easier for readers to parse. ;-)

../Randy

>   
>   SRC_URI[sha256sum] = "fce01ccac59812bddadffacff017dac2e4762bdb6ebc6ffe06f6ed4f6192c971"
>   
>
> -=-=-=-=-=-=-=-=-=-=-=-
> Links: You receive all messages sent to this group.
> View/Reply Online (#174351): https://lists.openembedded.org/g/openembedded-core/message/174351
> Mute This Topic: https://lists.openembedded.org/mt/95514199/3616765
> Group Owner: openembedded-core+owner@lists.openembedded.org
> Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [randy.macleod@windriver.com]
> -=-=-=-=-=-=-=-=-=-=-=-
>