SWEET32 not being flagged when AES-CBC is used (False negative) #1296
Comments
|
Hello @nrathaus, This is not a false negative. SWEET32 is about the use of 64-bit block ciphers, such as 3DES (see https://sweet32.info/). In fact, the main recommendation on https://sweet32.info is to use AES. Depending on the versions of SSL/TLS that your server supports, there are vulnerabilities associated with the use of CBC ciphers (including AES-CBC), such as POODLE, BEAST, and LUCKY13. However, SWEET32 isn't about the use of CBC or AES. Can you tell of what the Cloudflare advisory said? I tried to search for this, but the only possible "hits" I found were Cloudflare support pages, which were only available to their customers. |
|
Hi
I am seeing a difference between SSLLabs findings of
websso.nsd.co.jp (for example) and testssl.sh
SSLlabs says it's vulnerable -
https://www.ssllabs.com/ssltest/analyze.html?d=websso.nsd.co.jp
- while testssl.sh does not flag it
The only possible reason I see is because of the AES-CBC but maybe my guess
here is incorrect
On Mon, 15 Jul 2019 at 16:35, David Cooper ***@***.***> wrote:
Hello @nrathaus <https://github.com/nrathaus>,
This is not a false negative. SWEET32 is about the use of 64-bit block
ciphers, such as 3DES (see https://sweet32.info/). In fact, the main
recommendation on https://sweet32.info is to use AES.
Depending on the versions of SSL/TLS that your server supports, there are
vulnerabilities associated with the use of CBC ciphers (including AES-CBC),
such as POODLE, BEAST, and LUCKY13. However, SWEET32 isn't about the use of
CBC or AES.
Can you tell of what the Cloudflare advisory said? I tried to search for
this, but the only possible "hits" I found were Cloudflare support pages,
which were only available to their customers.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#1296>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAPQE3RZP5FQGBA7IBTTMJDP7R4JJANCNFSM4IDNVAFQ>
.
--
Thanks,
Noam Rathaus
|
I took a look at https://www.ssllabs.com/ssltest/analyze.html?d=websso.nsd.co.jp and saw no mention of SWEET32 at all. It does mark the AES-CBC ciphers as "WEAK," but that is because of various POODLE attacks, not SWEET32 (see https://blog.qualys.com/technology/2019/04/22/zombie-poodle-and-goldendoodle-vulnerabilities). The only mention I could find about SWEET32 from SSL Labs was at https://blog.qualys.com/ssllabs/2017/01/18/ssl-labs-grading-changes-january-2017, where it mentions SWEET32 in relation to imposing a penalty for the use of 3DES. |
|
I made a mixup sorry
I titles this incorrectly
Please close it - the issue is not with SWEET32 rather with POODLE Oracle
On Mon, 15 Jul 2019 at 18:17, David Cooper ***@***.***> wrote:
I am seeing a difference between SSLLabs findings of
websso.nsd.co.jp (for example) and testssl.sh
SSLlabs says it's vulnerable -
https://www.ssllabs.com/ssltest/analyze.html?d=websso.nsd.co.jp
- while testssl.sh does not flag it
I took a look at
https://www.ssllabs.com/ssltest/analyze.html?d=websso.nsd.co.jp and saw
no mention of SWEET32 at all. It does mark the AES-CBC ciphers as "WEAK,"
but that is because of various POODLE attacks, not SWEET32 (see
https://blog.qualys.com/technology/2019/04/22/zombie-poodle-and-goldendoodle-vulnerabilities
).
The only mention I could find about SWEET32 from SSL Labs was at
https://blog.qualys.com/ssllabs/2017/01/18/ssl-labs-grading-changes-january-2017,
where it mentions SWEET32 in relation to imposing a penalty for the use of
3DES.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#1296>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAPQE3QFZOTZP3D7G6AQIUDP7SIIHANCNFSM4IDNVAFQ>
.
--
Thanks,
Noam Rathaus
|
|
Noam: Next time you should be able to close it yourself too as you opened it. |
Please make sure that you provide enough information so that we understand what your issue is about.
uname -a
Linux cloud2-LSS4SL-c5 3.16.0-4-amd64 Heartbleed for STARTTLS #1 SMP Debian 3.16.39-1 (2016-12-30) x86_64 GNU/Linux
testssl version from the banner: testssl.sh -b 2>/dev/null | head -4 | tail -2
testssl.sh 3.0rc4 from https://testssl.sh/dev/
git log | head -1 (if running from git repo)
openssl version used by testssl.sh: testssl.sh -b 2>/dev/null | awk -F':' '/openssl/ { print $2}'
./bin/openssl.Linux.x86_64
steps to reproduce: testssl.sh or docker command line, if possible incl. host
what exactly was happening, output is needed
SWEET32 (CVE-2016-2183, CVE-2016-6329) not vulnerable (OK)
what did you expect instead?
SWEET32 (CVE-2016-2183, CVE-2016-6329) vulnerable
A target having AES-CBC enabled, for example:
Will not get listed as vulnerable to SWEET32, even though it has AES_128_CBC listed as being used, based on the advisory of SWEET32 released by cloudflare this configuration is also vulnerable.
I believe
local sweet32_ciphers
Should be adjusted to include them
I can provide two domains that are not being flagged even though they should - but discreetly not in a public issue ticket
The text was updated successfully, but these errors were encountered: