RFR JDK-8206925,,Support the certificate_authorities extension
Alexey Bakhtin
alexey at azul.com
Wed May 13 10:00:06 UTC 2020
Hello Xuelei,
I’m not a reviewer but I have some comment which could be helpful for your implementation.
We’ve developed CA Extension in the OpenJSSE provider [1] and found an issue with a third party server implementations.
According to RFC-8446 specification [2] the maximum size of the CA extension is 2^16 bytes. The maximum TLS record size is 2^14 bytes. In case of handshake message is bigger then maximum TLS record size, it should be splitted into several records. In fact, some server implementations does not allow ClientHello message bigger than the Maximum TLS record size and aborts connection immediately with “illegal_parameter” fatal alert.
This issue can be easily reproduced on the client side:
1) put additional certificates into cacerts file, about 200 certs in total,
2) enable certificate_authorities extension in the ClientHello message
3) connect to https://www.google.com
[1] - https://github.com/openjsse/openjsse/blob/master/src/main/java/org/openjsse/sun/security/ssl/CertificateAuthorityExtension.java
[2] - https://tools.ietf.org/html/rfc8446#page-45
Thank you
Alexey
> On 13 May 2020, at 00:43, Xuelei Fan <xuelei.fan at oracle.com> wrote:
>
> Updated webrev: http://cr.openjdk.java.net/~xuelei/8206925/webrev.01/
>
> On 5/12/2020 12:40 PM, Sean Mullan wrote:
>> On 5/5/20 2:29 PM, Xuelei Fan wrote:
>>> Hi,
>>>
>>> Could I get the following update reviewed?
>>>
>>> RFE: https://bugs.openjdk.java.net/browse/JDK-8206925
>>> CSR: https://bugs.openjdk.java.net/browse/JDK-8244441
>> We have previously used the syntax "enable[Extension]" when naming system properties that enable optional extensions. Thus, it seems this name would be more consistent: "jdk.tls.client.enableCertificateAuthoritiesExtension"
>> However, it is a bit long, so maybe we could abbreviate it to CA: "jdk.tls.client.enableCAExtension"
> "enableCAExtension" looks fine, but it is not as instinctive as "indicateCertificateAuthorities".
>
> We used to use "enableXXExtension" because normally there is only one behavior for the extension. However, for the Certificate Authorities extension, it could be requested by server side to indicate client cert selection, or by client side to indicate server cert selection. It is not straightforward to know if "enableCAExtension" means accepting server request, or produce client request.
>
> It is not expected to use this extension regularly.
>
> Please let me know if you still prefer to use "enableCAExtension".
>
>> Also, it is a bit unfortunate that we have to have a system property to enable it. Can we not enable it based on whether the configured X509TrustManager.getAcceptedIssuers returns a non-empty list?
> We can do that on server side, but there are compatibility impact on client behavior if we did it in client side. See #2 in the "Specification" section.
>
>>> Release-note: https://bugs.openjdk.java.net/browse/JDK-8244460
>>> webrev: http://cr.openjdk.java.net/~xuelei/8206925/webrev.00/
>> * src/java.base/share/classes/sun/security/ssl/CertificateRequest.java
>> Missing copyright update.
>> * src/java.base/share/classes/sun/security/ssl/SSLExtension.java
>> 748 // Switch on certificate_authorities extention in ClientHello?
>> typo: s/extention/extension
>> Is the '?' at the end intentional or a typo?
> Updated.
>
>> * src/java.base/share/classes/sun/security/ssl/CertificateAuthoritiesExtension.java 70 if (!authorities.contains(encodedPrincipal)) {
>> 71 authorities.add(encodedPrincipal);
>> 72 }
>> Is it really necessary to remove duplicates? Seems kind of expensive to iterate over the list every single time for what should be a rare case.
> Good point! Update here, and revert the update for CertificateRequest.java as well.
>
>> 108 X500Principal[] getAuthorities() {
>> Here you know the size of the array up front so you could avoid using a List and populate the array directly.
> Good catch!
>
>> * test/jdk/sun/security/ssl/X509KeyManager/CertificateAuthorities.java
>> The test doesn't seem to do much, other than make sure you can make a connection if the extension is enabled. Can you test the scenario below where you can show that the extension addresses the issue where the certificate selected may not be the one the peer can accept?
> With this update, the certificate selected must be the one the peer can accept. It may be helpful to add test cases to make sure the connection is terminated if no certification can be selected. I will try to add more test cases.
>
> Thanks,
> Xuelei
>
>> --Sean
>>>
>>> The "certificate_authorities" extension is an optional extension introduced in TLS 1.3 and used to indicate the certificate authorities (CAs) which an endpoint supports and which SHOULD be used by the receiving endpoint to guide certificate selection.
>>>
>>> In TLS 1.2, this function is built in the CertificateRequest handshake massage.
>>>
>>> This function is supported in TLS 1.2 and prior versions. However, it is not implemented in the TLS 1.3 implementation. Without this function, the authentication certificate selected may be not the one the peer could accepted, when there are multiple certificates available.
>>>
>>> Thanks,
>>> Xuelei
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: Message signed with OpenPGP
URL: <https://mail.openjdk.org/pipermail/security-dev/attachments/20200513/b9e4e5e3/signature.asc>
More information about the security-dev
mailing list