[OpenAFS-devel] Issues with my precache patch (was: Documentation for fs precache)

Jeffrey Hutzelman jhutz@cmu.edu
Wed, 20 Oct 2010 18:04:13 -0400


--On Tuesday, October 19, 2010 02:21:18 PM -0400 Phillip Moore 
<w.phillip.moore@gmail.com> wrote:

> Well, if this simple patch to the current pioctl can be made without all
> that overhead, then someone else can fix this, unfortunately.  I thought
> this would be relatively simple, and it's a fix I don't really need
> personally.  Sorry to be selfish...
>
> The real question is: do I need a new pioctl for this?  I'm just modifying
> the existing code and changing it's behavior.

I think you can safely change the behavior of the existing pioctl in the 
way you describe, because your new behavior would be backward-compatible 
with existing callers.  What you cannot do is assume that every kernel 
module you encounter will support your change.  Those that do not will fail 
with EINVAL if presented with empty input, and will not return the current 
value in any case.

-- Jeff