RFR: 8346059: [ASAN] awt_LoadLibrary.c reported compile warning ignoring return value of function by clang17

[Magnus Ihse Bursie]

On Fri, 13 Dec 2024 04:29:09 GMT, SendaoYan <syan at openjdk.org> wrote:

>>> better might be to just copy dlinfo.dli_fname into buf
>> 
>> "Something would be very, very broken if this were to fail.", So I think `perror` maybe better, it will print which file and the reason get realpath fail.
>
> Should I combine 3 similar issues([JDK-8346104](https://bugs.openjdk.org/browse/JDK-8346104)/[JDK-8346103](https://bugs.openjdk.org/browse/JDK-8346103)/[JDK-8346059](https://bugs.openjdk.org/browse/JDK-8346059)) to 1 issue

> `perror` will print the given message accrording the previous `errno` info to stderr, so this change will not change the original behaviour.

What do you mean? The original behavior would not write anything to stderr under any circumstance.

Now since the file name comes from a loaded library, I can't possibly imagine anything (short of a provoked situation where someone deliberatly removes the library after starting the JVM or something like that) that would cause realpath to fail. So this is basically just about keeping the compiler quiet.

Had we *really* expected that this could happen, this would not have been an accepted solution. Then you'd had to raise an exception, perhaps, and see if/how the operation could be retried, what consequences a failure would have, etc.

But as I said, in this particular instance, it might be okay.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/22701#discussion_r1883520144