On Wed, 5 Jun 2024 20:08:10 GMT, Robert Toyonaga <d...@openjdk.org> wrote:
> ### Summary > This change ensures we don't get undefined behavior when > calling[`isspace`](https://pubs.opengroup.org/onlinepubs/007904975/functions/isspace.html). > `isspace` accepts an `int` argument that "the application shall ensure is a > character representable as an unsigned char or equal to the value of the > macro EOF.". > > Previously, there was no checking of the values passed to `isspace`. I've > replaced direct calls with a new wrapper `os::is_space` that performs a range > check and prevents the possibility of undefined behavior from happening. For > instances outside of Hotspot, I've added casts to `unsigned char`. > > **Testing** > - Added a new test in `test/hotspot/gtest/runtime/test_os.cpp` to check > `os::is_space` is working correctly. > - tier1 This pull request has now been integrated. Changeset: cc64aeac Author: Robert Toyonaga <rtoyo...@redhat.com> Committer: Thomas Stuefe <stu...@openjdk.org> URL: https://git.openjdk.org/jdk/commit/cc64aeac47917f20a6d70e9796f0de9aa165ce62 Stats: 17 lines in 8 files changed: 0 ins; 0 del; 17 mod 8332400: isspace argument should be a valid unsigned char Reviewed-by: dholmes, amitkumar, stuefe, jwaters ------------- PR: https://git.openjdk.org/jdk/pull/19567
