Yes, it is the same issue. Sorry I didn't see the previous report -- I
didn't imagine an issue like this, with such a simple fix, could have
been reported back in 2006 without the fix ever having been implemented,
so I didn't bother searching back that far!

The fix in that report contains the same one-line fix as I proposed,
plus additional changes to ensure that the code won't break if in
the future additional changes are made to allow a second possible
exit-out-of-the-loop-with-d-being-non-null code path.

All that's strictly necessary to solve the current problem is the
one line I proposed, but the other proposed changes would make the
safer against future potential breakage if the code gets rearranged
later.