From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark Fasheh Date: Wed Sep 13 11:17:41 2006 Subject: [Ocfs2-devel] cluster/heartbeat.c::compute_max_sectors is invalid. In-Reply-To: <20060913104442.6d8a541b@mathieu.toulouse> References: <44FFEF06.5030001@seanodes.com> <20060908182833.GF8792@ca-server1.us.oracle.com> <20060913104442.6d8a541b@mathieu.toulouse> Message-ID: <20060913181737.GH8792@ca-server1.us.oracle.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: ocfs2-devel@oss.oracle.com On Wed, Sep 13, 2006 at 10:44:42AM +0200, Mathieu Avila wrote: > A proposal patch (against the 2.6.17.11 vanilla kernel) for this is > attached. It works for me, and shouldn't break any existing device. Ok, that patch looks good - I've put it in my 2.6.19 queue. > And i still think that it is a bad idea to try figuring out how many > BIOs are needed, because it really depends on the underlying > architecture of the device. Although it might work on most hardware, > some strange one might break one day... Well, I'm pretty sure some other software uses that approach, but that might just mean other software might have similar problems :) So yeah. I don't reallly have any qualms changing ocfs2 heartbeat to a more "natural" usage of the bio API. Thanks! --Mark -- Mark Fasheh Senior Software Developer, Oracle mark.fasheh@oracle.com