From: Karthik Nayak <karthik.188@gmail.com>
Date: Fri, 04 Sep 2015 12:35:36 GMT
Subject: Re: [PATCH v15 07/13] ref-filter: add support for %(contents:lines=X)
Message-ID: <CAOLa=ZT8uAWp=janWUX5iBKZZ7cj-j2Z7OynYxhJ+v4THZJZXw@mail.gmail.com>
In-Reply-To: <xmqqpp1zlb79.fsf@gitster.mtv.corp.google.com>

On Thu, Sep 3, 2015 at 9:57 PM, Junio C Hamano <gitster@pobox.com> wrote:
> Eric Sunshine <sunshine@sunshineco.com> writes:
>
>> Also, if 'buf' is indeed unconditionally NUL-terminated, then the (sp
>> <= buf + size) check is wasted code since the result of strstr() will
>> always be either NULL or pointing somewhere within the NUL-terminated
>> string.
>
> A caller can give a buf that is NUL terminated but specify that the
> only early part of the buffer to be used by giving you a shorter
> size, no?  In such a case, strstr() is safe in the sense that it is
> guaranteed not to go on forever, but you need to verify the location
> of the string it found is within the bounds.

That was the idea behind this, but will stick to Matthieu's suggestion.

-- 
Regards,
Karthik Nayak