Dr. Christian Kohlschütter comments

Results 100 comments of


                                            Dr. Christian Kohlschütter

MC_OVERRIDABLE_METHOD_CALL_IN_READ_OBJECT false positive in 4.8.4

With JNI, the Java security model is entirely defeated (you can access private members, raw memory, etc.). There are separate warnings that cover this scenario already. Reflection access has been...

MC_OVERRIDABLE_METHOD_CALL_IN_READ_OBJECT false positive in 4.8.4

@gtoison Correct, and that (possibility 2) is why I think having this warning in the general case is important, rather than to silence it altogether. I do see some paranoid...

MC_OVERRIDABLE_METHOD_CALL_IN_READ_OBJECT false positive in 4.8.4

Any update on this? I'm still seeing the false positive errors in 4.8.6.1

[MINSTALL-201] Regression in 3.1.2 when installing pom artifact

@slawekjaranowski MINSTALL-189 refers to a fix to 3.1.1; this is about 3.1.2. fwiw, the problem remains unresolved in 3.1.4.

passthrough_ll: "No file descriptors available" errors

Given that "passthrough_ll" is supposed to be a "proper" implementation, I think it should do the right thing, especially given that it is an example implementation to tell users of...

Use Opaque for locking instead of byte[], improve reuse of base64 keys

Hang on, a few more changes (low-hanging fruit) coming...

Use Opaque for locking instead of byte[], improve reuse of base64 keys

@kofemann Please review again, thanks!

Use Opaque for locking instead of byte[], improve reuse of base64 keys

one last improvement for this PR coming up in a sec...

Use Opaque for locking instead of byte[], improve reuse of base64 keys

I assume you're not reviewing right now — I will add two more changes before I log off.

[java] UseTryWithResources false positive with simple close call

Thanks for the tip, @oowekyala, I didn't know that! However, in order to have no errors, it would have to be ```java try (Closeable unused = source()) { // nothing...