Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1545918 > unrolled thread

RE: [PATCH 00/15] hyperv: more stuff to uapi + cleanup

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  RE: [PATCH 00/15] hyperv: more stuff to uapi + cleanup KY Srinivasan <kys@microsoft.com> - 2016-12-21 19:10 +0100
    RE: [PATCH 00/15] hyperv: more stuff to uapi + cleanup KY Srinivasan <kys@microsoft.com> - 2016-12-30 20:50 +0100

#1545918 — RE: [PATCH 00/15] hyperv: more stuff to uapi + cleanup

> -----Original Message-----
> From: Roman Kagan [mailto:rkagan@virtuozzo.com]
> Sent: Tuesday, December 20, 2016 7:56 AM
> To: Paolo Bonzini <pbonzini@redhat.com>; Radim Krčmář
> <rkrcmar@redhat.com>; KY Srinivasan <kys@microsoft.com>; Vitaly
> Kuznetsov <vkuznets@redhat.com>
> Cc: Thomas Gleixner <tglx@linutronix.de>; Ingo Molnar
> <mingo@redhat.com>; H. Peter Anvin <hpa@zytor.com>; x86@kernel.org;
> Haiyang Zhang <haiyangz@microsoft.com>; kvm@vger.kernel.org; linux-
> kernel@vger.kernel.org; devel@linuxdriverproject.org; Denis V . Lunev
> <den@openvz.org>; Roman Kagan <rkagan@virtuozzo.com>
> Subject: [PATCH 00/15] hyperv: more stuff to uapi + cleanup
> 
> Expose more Hyper-V-related definitions in the uapi header for
> consumption by userspace.
> 
> While doing so, get rid of a number of duplications between the KVM and
> the guest driver code.  Also a few other cleanups are made which are not
> strictly necessary for the main purpose of the series but appear
> reasonable to do at the same time.
> 
> The most controversial is the last patch which modifies the stuff
> already published in the uapi header, in the hope that no userspace
> applications have started relying on it; I'm ok dropping it if this is
> unacceptable.

Roman,

First, let me thank you. Broadly, this patch-set can be broken into
1. Moving existing definitions around - (to make it possible to share these
between Hyper-V guest drivers and KVM)
2. Cleanup of the existing code in the VMBUS driver (under drivers/hv).

To the extent possible, I want to take all non-KVM code through Greg's tree.
We can then modify the KVM code to use these common definitions. Currently, I too
am working on restructuring VMBUS driver code to fully isolate all x86 dependencies.
I can work with you on integration as I too am moving things around.

Regards,

K. Y 

> 
> Roman Kagan (15):
>   hyperv: consolidate TSC ref page definitions
>   hyperv: uapi-fy synic event flags definitions
>   hyperv: use standard bitops
>   hyperv: define VMBus message type
>   hyperv: GFP_ATOMIC -> GFP_KERNEL
>   hyperv: avoid unnecessary vmalloc
>   hyperv: dedup cpuid definitions
>   hyperv: dedup crash msr related definitions
>   hyperv: unify Hyper-V msr definitions
>   hyperv: uapi-fy PostMessage and SignalEvent hypercall structures
>   hyperv: uapi-fy monitored notification structures
>   hyperv: move VMBus connection ids to uapi
>   hyperv: move function close to its only callsite
>   hyperv_vmbus: drop unused definitions
>   hyperv: redefine hv_message without bitfields
> 
>  arch/x86/include/asm/kvm_host.h    |   2 +-
>  arch/x86/include/uapi/asm/hyperv.h | 101 +++++++---
>  drivers/hv/hyperv_vmbus.h          | 399 +------------------------------------
>  include/linux/hyperv.h             |  24 +--
>  arch/x86/kvm/hyperv.c              |  14 +-
>  drivers/hv/channel.c               |   8 +-
>  drivers/hv/channel_mgmt.c          |  30 +--
>  drivers/hv/connection.c            |  65 ++----
>  drivers/hv/hv.c                    | 300 +++++++++++++---------------
>  drivers/hv/vmbus_drv.c             |  67 +++----
>  10 files changed, 288 insertions(+), 722 deletions(-)
> 
> --
> 2.9.3

[toc] | [next] | [standalone]

#1548710

> -----Original Message-----
> From: Roman Kagan [mailto:rkagan@virtuozzo.com]
> Sent: Wednesday, December 28, 2016 8:57 AM
> To: KY Srinivasan <kys@microsoft.com>
> Cc: Paolo Bonzini <pbonzini@redhat.com>; Radim Krčmář
> <rkrcmar@redhat.com>; Vitaly Kuznetsov <vkuznets@redhat.com>; Thomas
> Gleixner <tglx@linutronix.de>; Ingo Molnar <mingo@redhat.com>; H. Peter
> Anvin <hpa@zytor.com>; x86@kernel.org; Haiyang Zhang
> <haiyangz@microsoft.com>; kvm@vger.kernel.org; linux-
> kernel@vger.kernel.org; devel@linuxdriverproject.org; Denis V . Lunev
> <den@openvz.org>
> Subject: Re: [PATCH 00/15] hyperv: more stuff to uapi + cleanup
> 
> [ Sorry for such a slow reply; flu and office relocation knocked me out
> for a while ]
> 
> On Wed, Dec 21, 2016 at 06:00:17PM +0000, KY Srinivasan wrote:
> > > -----Original Message-----
> > > From: Roman Kagan [mailto:rkagan@virtuozzo.com]
> > > Sent: Tuesday, December 20, 2016 7:56 AM
> > > Subject: [PATCH 00/15] hyperv: more stuff to uapi + cleanup
> > >
> > > Expose more Hyper-V-related definitions in the uapi header for
> > > consumption by userspace.
> > >
> > > While doing so, get rid of a number of duplications between the KVM and
> > > the guest driver code.  Also a few other cleanups are made which are not
> > > strictly necessary for the main purpose of the series but appear
> > > reasonable to do at the same time.
> > >
> > > The most controversial is the last patch which modifies the stuff
> > > already published in the uapi header, in the hope that no userspace
> > > applications have started relying on it; I'm ok dropping it if this is
> > > unacceptable.
> >
> > First, let me thank you. Broadly, this patch-set can be broken into
> > 1. Moving existing definitions around - (to make it possible to share these
> > between Hyper-V guest drivers and KVM)
> > 2. Cleanup of the existing code in the VMBUS driver (under drivers/hv).
> 
> Right.  Another significant part of the series is, where two sets of
> definitions exist for the same entity, consolidate on the one that looks
> more Linux-style, usually the one in the (currently) uapi header.
> 
> > To the extent possible, I want to take all non-KVM code through Greg's
> tree.
> > We can then modify the KVM code to use these common definitions.
> 
> Well, this patchset touches almost no KVM code, so we're fine here I
> think.
> 
> > Currently, I too am working on restructuring VMBUS driver code to
> > fully isolate all x86 dependencies.  I can work with you on
> > integration as I too am moving things around.
> 
> Great!  Do you want me to rebase on some your public tree?  (Once I
> split the patches as you requested, of course)?

I just sent out the patches that I have been working on. These are based on
Greg's char-misc tree. I do have a bunch of patches in flight that have not been
checked in yet. Sorry, I don't have a public tree I can point you to.
I will work with you and coordinate our work here.

Happy New Year!

K. Y 

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web