On Thu, 23 Sep 2021 14:32:01 GMT, Weijun Wang <wei...@openjdk.org> wrote:

> This code change removes weak etypes from the default list so it's safer to 
> enable one of them. See the corresponding CSR at 
> https://bugs.openjdk.java.net/browse/JDK-8274207 for more explanation. BTW, 
> please review the CSR as well.

src/java.security.jgss/share/classes/sun/security/krb5/internal/crypto/EType.java
 line 101:

> 99:         if (allowWeakCrypto) {
> 100:             result[num++] = EncryptedData.ETYPE_DES3_CBC_HMAC_SHA1_KD;
> 101:             result[num++] = EncryptedData.ETYPE_ARCFOUR_HMAC;

MIT still has arcfour-hmac-md5 in the enabled list - do you think there is any 
reason (compatibility) we should do the same? Note that I think it is better 
that we don't though. See "permitted_enctypes" at 
https://web.mit.edu/Kerberos/krb5-1.19/doc/admin/conf_files/krb5_conf.html#libdefaults.

src/java.security.jgss/share/classes/sun/security/krb5/internal/crypto/EType.java
 line 242:

> 240:     // used in Config
> 241:     public static int[] getBuiltInDefaults() {
> 242:         return defaultETypes;

It might be safer to return a clone here since it is mutable. The previous code 
always returned a new array. This array gets passed back to calling code via 
Etype.getDefaults(), returning a clone would prevent the configured value from 
being accidentally modified.

-------------

PR: https://git.openjdk.java.net/jdk/pull/5654

Reply via email to