From owner-cvs-all Sun Nov 25 15:21:10 2001 Delivered-To: cvs-all@freebsd.org Received: from salmon.maths.tcd.ie (salmon.maths.tcd.ie [134.226.81.11]) by hub.freebsd.org (Postfix) with SMTP id C756837B41A; Sun, 25 Nov 2001 15:21:03 -0800 (PST) Received: from walton.maths.tcd.ie by salmon.maths.tcd.ie with SMTP id ; 25 Nov 2001 23:21:02 +0000 (GMT) To: Kris Kennaway Cc: cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/usr.bin/tail forward.c In-Reply-To: Your message of "Sun, 25 Nov 2001 14:45:40 PST." <20011125144540.A93369@xor.obsecurity.org> Date: Sun, 25 Nov 2001 23:21:02 +0000 From: Ian Dowse Message-ID: <200111252321.aa42920@salmon.maths.tcd.ie> Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG In message <20011125144540.A93369@xor.obsecurity.org>, Kris Kennaway writes: > >Hmm, this strikes me as a bit nasty...ideally you shouldn't have to >poll for any status changes in a kqueue world. Is there an easy way >kq could be improved to handle this? I guess tail could add a kevent for every component in the specified path including expanded symlinks (and resort to polling if any were on filesystems that don't support kqueue). I think that would just over-complicate tail for very little real gain. Note that tail -f doesn't use polling; it's only when you explicitly tell tail that the path might change by specifying "-F" that it gets used. I suppose for -F, the timeout could be increased to 5 or 10 seconds also because simple renames within the same directory are noticed immediately. Ian To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message