[OE-core] [PATCH] openssl: fix for CVE-2010-5298

[Douglas Royds]

On 19/05/14 19:00, rongqing.li at windriver.com wrote:
> From: Yue Tao <Yue.Tao at windriver.com>
>
> Race condition in the ssl3_read_bytes function in s3_pkt.c in OpenSSL
> through 1.0.1g, when SSL_MODE_RELEASE_BUFFERS is enabled, allows remote
> attackers to inject data across sessions or cause a denial of service
> (use-after-free and parsing error) via an SSL connection in a
> multithreaded environment.
>
> http://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2010-5298
>
> Signed-off-by: Yue Tao <Yue.Tao at windriver.com>
> Signed-off-by: Roy Li <rongqing.li at windriver.com>
> ---
>   .../openssl-1.0.1e/openssl-CVE-2010-5298.patch     |   24 ++++++++++++++++++++
>   1 file changed, 24 insertions(+)
>   create mode 100644 meta/recipes-connectivity/openssl/openssl-1.0.1e/openssl-CVE-2010-5298.patch
>
> diff --git a/meta/recipes-connectivity/openssl/openssl-1.0.1e/openssl-CVE-2010-5298.patch b/meta/recipes-connectivity/openssl/openssl-1.0.1e/openssl-CVE-2010-5298.patch
> new file mode 100644
> index 0000000..417a774
> --- /dev/null
> +++ b/meta/recipes-connectivity/openssl/openssl-1.0.1e/openssl-CVE-2010-5298.patch
> @@ -0,0 +1,24 @@
> +openssl fix for CVE-2010-5298
> +
> +Upstream-Status: Backport
> +
> +Race condition in the ssl3_read_bytes function in s3_pkt.c in OpenSSL
> +through 1.0.1g, when SSL_MODE_RELEASE_BUFFERS is enabled, allows remote
> +attackers to inject data across sessions or cause a denial of service
> +(use-after-free and parsing error) via an SSL connection in a
> +multithreaded environment.
> +
> +http://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2010-5298
> +
> +Signed-off-by: Yue Tao <Yue.Tao at windriver.com>
> +--- a/ssl/s3_pkt.c
> ++++ b/ssl/s3_pkt.c
> +@@ -1013,7 +1013,7 @@ start:
> + 				{
> + 				s->rstate=SSL_ST_READ_HEADER;
> + 				rr->off=0;
> +-				if (s->mode & SSL_MODE_RELEASE_BUFFERS)
> ++				if (s->mode & SSL_MODE_RELEASE_BUFFERS && s->s3->rbuf.left == 0)
> + 					ssl3_release_read_buffer(s);
> + 				}
> + 			}

This patch file appears to be orphaned in openssl-1.0.1e/, when the 
current version is 1.0.1g.

Also, there is no file://openssl-CVE-2010-5298.patch entry in the SRC_URI.

Was this deliberate? Am I missing something?