|
From: | raif at swiftdsl dot com dot au |
Subject: | [Bug crypto/28678] Misleading error message from gnu.java.security.Engine |
Date: | 12 Aug 2006 22:26:55 -0000 |
------- Comment #9 from raif at swiftdsl dot com dot au 2006-08-12 22:26 ------- i've done some tests with the RI (both 1.4 and 1.5) and the results are: * original exceptions (e.g. class not found) are _not_ swallowed; instead they are chained as cause(s) of exceptions that are raised later in other classes of the xxx.getInstance() invocation chain. * the exception for one provider (last one tried) appear when the getInstance() methods do not designate a specific provider. a new Mauve test gnu/testlet/gnu/java/security/TestOfPR28678.java has been added to highlight and validate the up-coming patch. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28678
[Prev in Thread] | Current Thread | [Next in Thread] |