RFR: 8368365: ASAN errors should produce hs-err files and core dumps [v4]

[Thomas Stuefe]

On Wed, 24 Sep 2025 23:35:25 GMT, Dean Long <dlong at openjdk.org> wrote:

> Actually, what I was trying to get at with my question was whether we need the callback if abort_on_error=1 is set. If abort_on_error=1 is set, it will call abort(), which should send a SIGABRT and then cause an hs_err file to get generated.

We don't catch SIGABRT. Starting to do so would be a more invasive change (and add some complexities, since we ourselves use abort(3) to generate cores). It would also interfere with user SIGABRT handlers, possibly requiring them to start using libjsig etc.

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

PR Comment: https://git.openjdk.org/jdk/pull/27446#issuecomment-3332305635