RE: [PATCH ] Staging: hv: Hyper-V driver cleanup
From: KY Srinivasan
Date: Thu Feb 24 2011 - 20:07:09 EST
> -----Original Message-----
> From: Greg KH [mailto:gregkh@xxxxxxx]
> Sent: Thursday, February 24, 2011 7:32 PM
> To: KY Srinivasan
> Cc: linux-kernel@xxxxxxxxxxxxxxx; devel@xxxxxxxxxxxxxxxxxxxxxx;
> virtualization@xxxxxxxxxxxxxx; Haiyang Zhang; Hank Janssen
> Subject: Re: [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@xxxxxxx]
> > > Sent: Thursday, February 24, 2011 6:46 PM
> > > To: KY Srinivasan
> > > Cc: linux-kernel@xxxxxxxxxxxxxxx; devel@xxxxxxxxxxxxxxxxxxxxxx;
> > > virtualization@xxxxxxxxxxxxxx; 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 hyperv_driver abstraction. Would breaking this patch up into two patches,
> > one dealing with the device abstraction consolidation and the other dealing
> > the consolidation of driver abstractions satisfy your concern. Even if I partition
> > 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.)
If it is ok with you, I will do two patches - one for dealing with consolidating device structure
and the other for consolidating the driver abstractions.
> > > 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.
While the hv code is from the tip of the tree; the kernel I am running is a little dated.
> greg k-h
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/