<!DOCTYPE html><html><head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
  </head>
  <body>
    <p>Hi Archie,<br>
      my 0.02$ below:<br>
    </p>
    <div class="moz-cite-prefix">On 07/05/2025 19:26, Archie Cobbs
      wrote:<br>
    </div>
    <blockquote type="cite" cite="mid:CANSoFxsZEAzsp5HF4J47-0Ygcd8vbjDbNjU+H80mZMvGCu=b1Q@mail.gmail.com">
      
      <div dir="ltr">
        <div dir="ltr">
          <div dir="ltr">
            <div dir="ltr">
              <div dir="ltr">
                <div dir="ltr">
                  <div dir="ltr">
                    <div dir="ltr">
                      <div dir="ltr">
                        <div>I'd like to restart this discussion of <a href="https://bugs.openjdk.org/browse/JDK-8344159" target="_blank" moz-do-not-send="true">JDK-8344159</a>  -
                          "Add lint warnings for unnecessary warning
                          suppression"</div>
                      </div>
                      <div dir="ltr">
                        <div><br>
                          <div>Note, this is likely not destined for JDK
                            25. I'm bringing it up now simply because
                            some refactoring that arose from the
                            previous discussion back in November (which
                            is <a href="https://mail.openjdk.org/pipermail/compiler-dev/2024-November/028573.html" target="_blank" moz-do-not-send="true">here</a>)
                            is now mostly done.
                            <div><br>
                            </div>
                            <div>I think a fair summary of the current
                              state is: Having the compiler emit
                              warnings for unnecessary warning
                              suppression is considered a useful idea in
                              principle, but it's unclear what is the
                              best way to configure/enable that
                              functionality in the compiler.</div>
                          </div>
                        </div>
                        <div><br>
                        </div>
                        <div>In particular, there are subtle corner
                          cases that could be problematic if these
                          warnings were enabled by <span style="font-family:monospace">-Xlint:all</span>,
                          especially for <span style="font-family:monospace">-Werror</span>
                          builds.</div>
                      </div>
                    </div>
                  </div>
                </div>
              </div>
            </div>
          </div>
        </div>
      </div>
    </blockquote>
    <p>IMHO, this is a problem present with the addition of every new
      lint warning. It's not specific to this one. Of course, some new
      lint warning operate on brand new language features, so the
      potential to compromise existing compilaton units is low. But I'd
      expect that warnings such as "serial" would have ran into similar
      corner cases (e.g. previously ok code started to report new
      warnings, and to fail with Werror).</p>
    <p>So I think this seems ok, and not enough to cause a tweak to the
      treatment of lint:all.</p>
    <p>I'll add here that a more robust fix for Werror issues is to
      enable developers to be more specific in the warnings they'd like
      to promote to errors (as discussed in the past). I'll CC Kevin who
      has always been a fan of this feature ;-)<br>
    </p>
    <blockquote type="cite" cite="mid:CANSoFxsZEAzsp5HF4J47-0Ygcd8vbjDbNjU+H80mZMvGCu=b1Q@mail.gmail.com">
      <div dir="ltr">
        <div dir="ltr">
          <div dir="ltr">
            <div dir="ltr">
              <div dir="ltr">
                <div dir="ltr">
                  <div dir="ltr">
                    <div dir="ltr">
                      <div dir="ltr">
                        <div><br>
                        </div>
                        <div>Also complicating things, there are two
                          different ways to suppress warnings, <span style="font-family:monospace">@SuppressWarnings("foo")</span>
                          and <span style="font-family:monospace">-Xlint:-foo</span>
                          and for similar "corner case" reasons they
                          need to be configurable separately.</div>
                      </div>
                    </div>
                  </div>
                </div>
              </div>
            </div>
          </div>
        </div>
      </div>
    </blockquote>
    <p>IMHO, the main value of this new lint warning is to detect the
      use of `@SuppressWarnings` that are redundant. Sure there's the
      command-line flavor as well, but that renders any source code
      analysis dependent on the flags passed to javac. In other words,
      detecting redundant `@SuppressWarnings` has a more "timeless"
      nature that does not apply to detecting redundant command-line
      suppressions. The former tells you something about _your code_, in
      a way that is independent from which flags you use to compile it.
      Which is why I think we should only worry about that.</p>
    <p>Maurizio</p>
    <br>
  </body>
</html>