RFR: JDK-8230744 Several classes throw OutOfMemoryError without message

[Martin Buchholz]

I continue to think JDK library code should avoid exposing VM
implementation limits to users, i.e. exception detail messages should
not include MAX_ARRAY_SIZE

I think our grow code has been drifting to being inconsistent and
incoherent over the years.

In the code below, the detail message looks wrong; this exception is
thrown when minCapacity wraps around, not when between SAFE_BOUND and
UNSAFE_BOUND.

         int SAFE_BOUND = MAX_ARRAY_SIZE >> coder;
         int UNSAFE_BOUND = Integer.MAX_VALUE >> coder;
         if (UNSAFE_BOUND - minCapacity < 0) { // overflow
-            throw new OutOfMemoryError();
+            throw new
+            OutOfMemoryError("StringBuilder buffer exceeds maximum size: " +
+                             MAX_ARRAY_SIZE);

On Mon, Jun 1, 2020 at 12:19 PM Jim Laskey <james.laskey at oracle.com> wrote:
>
> To be honest I should (will) file another bug against the use of math exact here. I don't think the tests actually catch what they think they test. Looks good in theory but more often than not, the pos array above that code will blow up before you get there.
>
>
>
>
> > On Jun 1, 2020, at 4:09 PM, Paul Sandoz <paul.sandoz at oracle.com> wrote:
> >
> > +1
> >
> > I quibble about the “ignored” variable name for the caught ArithmeticException.
> >
> > Paul.
> >
> >> On Jun 1, 2020, at 5:48 AM, Jim Laskey <james.laskey at oracle.com> wrote:
> >>
> >> Cleans up every case of OutOfMemoryError without a message. Some logic changes in String::repeat and ByteArrayChannel:: hugeCapacity.
> >>
> >> Thank you.
> >>
> >> Cheers,
> >>
> >> -- Jim
> >>
> >> webrev: http://cr.openjdk.java.net/~jlaskey/8230744/webrev-00/index.html <http://cr.openjdk.java.net/~jlaskey/8230744/webrev-00/index.html>
> >> jbs: https://bugs.openjdk.java.net/browse/JDK-8230744 <https://bugs.openjdk.java.net/browse/JDK-8230744>
> >>
> >>
> >
>