lkml.org 
[lkml]   [2011]   [Feb]   [24]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    Date
    From
    SubjectRe: [PATCH ] Staging: hv: Hyper-V driver cleanup
    On Fri, Feb 25, 2011 at 12:24:57AM +0000, KY Srinivasan wrote:
    >
    >
    > > -----Original Message-----
    > > From: Greg KH [mailto:gregkh@suse.de]
    > > Sent: Thursday, February 24, 2011 6:46 PM
    > > To: KY Srinivasan
    > > Cc: linux-kernel@vger.kernel.org; devel@linuxdriverproject.org;
    > > virtualization@lists.osdl.org; Haiyang Zhang; Hank Janssen
    > > Subject: Re: [PATCH ] Staging: hv: Hyper-V driver cleanup
    > >
    > > On Thu, Feb 24, 2011 at 03:20:58PM -0800, K. Y. Srinivasan wrote:
    > > > This patch cleans up (a lot of the) naming issues that
    > > > various reviewers have noted. It also gets rid of
    > > > some unnecessary layering in the code.
    > >
    > > Whenever you have a patch description that says "It also..." you know
    > > you need to break this up into smaller, logical pieces.
    >
    > The name change was related to the layering issue. For instance I combined the
    > Vm_device and hv_device abstractions to build the hyperv_device abstraction.
    > Likewise, I combined the driver_context and the hv_driver abstractions to build the
    > the hyperv_driver abstraction. Would breaking this patch up into two patches,
    > one dealing with the device abstraction consolidation and the other dealing with
    > the consolidation of driver abstractions satisfy your concern. Even if I partition this
    > patch along these lines, it will still be a large set of patches; since these changes
    > are pervasive.

    pervasive patches are fine, just remember, "each patch can only do one
    thing". It sounds like you want to do at least 2 patches here, if not
    a lot more. Look at my past patches when I combined things and removed
    a whole layer for how to do this in a very incremental, piece-by-piece
    fashion (i.e, move one field over at a time until the structure is gone,
    and then remove it entirely.)

    > > There is no 2.6.38 kernel yet, so I find this very hard to believe :)
    >
    > My mistake; I did not specify the full output of uname -a on the box
    > that I tested this code. This box is running the LINUX-NEXT kernel :
    > 2.6.38-rc1-0.2-default.

    linux-next should be farther along than -rc1 as -rc6 is currently out.

    confused,

    greg k-h


    \
     
     \ /
      Last update: 2011-02-25 01:39    [W:0.022 / U:0.264 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site