From: Junio C Hamano Date: Thu, 20 Oct 2011 01:30:45 GMT Subject: Re: [PATCH 3/6] revert: fix buffer overflow in insn sheet parser Message-ID: <7v8vogbgai.fsf@alter.siamese.dyndns.org> In-Reply-To: <1319058208-17923-4-git-send-email-artagnon@gmail.com> Ramkumar Ramachandra writes: > Check that the commit name argument to a "pick" or "revert" action in > '.git/sequencer/todo' is not too long, to avoid overflowing an > on-stack buffer. This fixes a regression introduced by 5a5d80f4 > (revert: Introduce --continue to continue the operation, 2011-08-04). Given that this function is going to be fixed properly so that it does not even need to use the "on-stack buffer", is this really necessary?