[PATCH 0/12] PV on HVM Xen

[PATCH 0/12] PV on HVM Xen

[Stefano Stabellini]

Hi all,
the last update on the PV on HVM Xen series contains the following
changes:

- some variables and functions have been renamed following Jeremy's
suggestions, in particular:
s/init_shared_info/xen_hvm_init_shared_info
s/xen_platform_pci/xen_platform_pci_enabled
s/UNPLUG_/XEN_UNPLUG_

- the two platform_pci.h header files have been merged and the useless
intro has been removed;

- the xen platform pci product number and driver versions have been made
static;

- the description on the VIRQ_TIMER patch has been improved;

- a new patch to fix hpet behaviour has been introduced: hpet_disable is
called unconditionally on machine shutdown, and doesn't check whether
hpet has been actually enabled, causing trouble if it hasn't;

- the vector callback patch has been improved: we don't use the ipi
vector anymore but we allocate our own so that we can avoid useless and
expensive vlapic acks.


Meanwhile the vector callback patch for xen has been checked into
xen-unstable, so you don't need a separate patch for xen anymore to take
full advantage of this patch series.

A git tree is available here:

git://xenbits.xen.org/people/sstabellini/linux-pvhvm.git

branch name 2.6.34-pvhvm-v3.

Cheers,

Stefano

Re: [PATCH 0/12] PV on HVM Xen

[Jeremy Fitzhardinge]

On 06/03/2010 06:07 AM, Stefano Stabellini wrote:
> Hi all,
> the last update on the PV on HVM Xen series contains the following
> changes:
>
> - some variables and functions have been renamed following Jeremy's
> suggestions, in particular:
> s/init_shared_info/xen_hvm_init_shared_info
> s/xen_platform_pci/xen_platform_pci_enabled
> s/UNPLUG_/XEN_UNPLUG_
>
> - the two platform_pci.h header files have been merged and the useless
> intro has been removed;
>
> - the xen platform pci product number and driver versions have been made
> static;
>
> - the description on the VIRQ_TIMER patch has been improved;
>
> - a new patch to fix hpet behaviour has been introduced: hpet_disable is
> called unconditionally on machine shutdown, and doesn't check whether
> hpet has been actually enabled, causing trouble if it hasn't;
>
> - the vector callback patch has been improved: we don't use the ipi
> vector anymore but we allocate our own so that we can avoid useless and
> expensive vlapic acks.
>   

I think they're worse than that; I think they could end up potentially
acking a pending interrupt which hasn't been delivered yet.

>
> Meanwhile the vector callback patch for xen has been checked into
> xen-unstable, so you don't need a separate patch for xen anymore to take
> full advantage of this patch series.
>
> A git tree is available here:
>
> git://xenbits.xen.org/people/sstabellini/linux-pvhvm.git
>
> branch name 2.6.34-pvhvm-v3.
>   

Thanks,

    J

[PATCH 0/12] PV on HVM Xen

[Stefano Stabellini]

Hi all,
this is another update on the PV on HVM Xen series; a list of changes
compared to the previous version follows:

- the bug caused by modprobing xen frontend modules when xenbus is not
initialized has been fixed: now xenbus is always initialized during
postcore_initcall, before device drivers are initialized, so that we
can be sure that loading xen pv frontends will always happen afterwards,
no matter how the xen platform pci driver is built.
In order to do that xenbus_probe has been moved out of the xenbus
initialization and called it later on at device_initcall.

- a new HVMOP_pagetable_dying hypercall has been added to notify Xen
that a pagetable is going to be destroyed: this improves performances
significantly when running on shadow pagetables.
A patch is currently need on the Xen side for this to work.

Jeremy's comments have been addressed:

- xen_guest_init has been renamed xen_hvm_guest_init;

- init_hvm_time has been moved to arch/x86/xen/time.c;

Konrad's comments have been addressed:

- gnttab_max_nr_grant_frames has been renamed gnttab_max_grant_frames;

- few inaccurate comments have been rewritten;

- the preprocessor checks in platform-pci-unplug.c have been moved to
platform_pci.h;

- few other code style improvements, like using dev_err instead of
printk and strncmp instead of strcmp.



A git tree is available here:

git://xenbits.xen.org/people/sstabellini/linux-pvhvm.git

branch name 2.6.34-pvhvm-v4.

Cheers,

Stefano

[PATCH 0/12] PV on HVM Xen

[Stefano Stabellini]

Hi all,
this is another update on the PV on HVM Xen series; a list of changes
compared to the previous version follows:

- the bug caused by modprobing xen frontend modules when xenbus is not
initialized has been fixed: now xenbus is always initialized during
postcore_initcall, before device drivers are initialized, so that we
can be sure that loading xen pv frontends will always happen afterwards,
no matter how the xen platform pci driver is built.
In order to do that xenbus_probe has been moved out of the xenbus
initialization and called it later on at device_initcall.

- a new HVMOP_pagetable_dying hypercall has been added to notify Xen
that a pagetable is going to be destroyed: this improves performances
significantly when running on shadow pagetables.
A patch is currently need on the Xen side for this to work.

Jeremy's comments have been addressed:

- xen_guest_init has been renamed xen_hvm_guest_init;

- init_hvm_time has been moved to arch/x86/xen/time.c;

Konrad's comments have been addressed:

- gnttab_max_nr_grant_frames has been renamed gnttab_max_grant_frames;

- few inaccurate comments have been rewritten;

- the preprocessor checks in platform-pci-unplug.c have been moved to
platform_pci.h;

- few other code style improvements, like using dev_err instead of
printk and strncmp instead of strcmp.



A git tree is available here:

git://xenbits.xen.org/people/sstabellini/linux-pvhvm.git

branch name 2.6.34-pvhvm-v4.

Cheers,

Stefano

Re: [PATCH 0/12] PV on HVM Xen

[Stefano Stabellini]

As usual I mistyped the topic, the series is made up of 13 patches now
:)

On Mon, 21 Jun 2010, Stefano Stabellini wrote:
> Hi all,
> this is another update on the PV on HVM Xen series; a list of changes
> compared to the previous version follows:
> 
> - the bug caused by modprobing xen frontend modules when xenbus is not
> initialized has been fixed: now xenbus is always initialized during
> postcore_initcall, before device drivers are initialized, so that we
> can be sure that loading xen pv frontends will always happen afterwards,
> no matter how the xen platform pci driver is built.
> In order to do that xenbus_probe has been moved out of the xenbus
> initialization and called it later on at device_initcall.
> 
> - a new HVMOP_pagetable_dying hypercall has been added to notify Xen
> that a pagetable is going to be destroyed: this improves performances
> significantly when running on shadow pagetables.
> A patch is currently need on the Xen side for this to work.
> 
> Jeremy's comments have been addressed:
> 
> - xen_guest_init has been renamed xen_hvm_guest_init;
> 
> - init_hvm_time has been moved to arch/x86/xen/time.c;
> 
> Konrad's comments have been addressed:
> 
> - gnttab_max_nr_grant_frames has been renamed gnttab_max_grant_frames;
> 
> - few inaccurate comments have been rewritten;
> 
> - the preprocessor checks in platform-pci-unplug.c have been moved to
> platform_pci.h;
> 
> - few other code style improvements, like using dev_err instead of
> printk and strncmp instead of strcmp.
> 
> 
> 
> A git tree is available here:
> 
> git://xenbits.xen.org/people/sstabellini/linux-pvhvm.git
> 
> branch name 2.6.34-pvhvm-v4.
> 
> Cheers,
> 
> Stefano
> 

Re: [PATCH 0/12] PV on HVM Xen

[Stefano Stabellini]

As usual I mistyped the topic, the series is made up of 13 patches now
:)

On Mon, 21 Jun 2010, Stefano Stabellini wrote:
> Hi all,
> this is another update on the PV on HVM Xen series; a list of changes
> compared to the previous version follows:
> 
> - the bug caused by modprobing xen frontend modules when xenbus is not
> initialized has been fixed: now xenbus is always initialized during
> postcore_initcall, before device drivers are initialized, so that we
> can be sure that loading xen pv frontends will always happen afterwards,
> no matter how the xen platform pci driver is built.
> In order to do that xenbus_probe has been moved out of the xenbus
> initialization and called it later on at device_initcall.
> 
> - a new HVMOP_pagetable_dying hypercall has been added to notify Xen
> that a pagetable is going to be destroyed: this improves performances
> significantly when running on shadow pagetables.
> A patch is currently need on the Xen side for this to work.
> 
> Jeremy's comments have been addressed:
> 
> - xen_guest_init has been renamed xen_hvm_guest_init;
> 
> - init_hvm_time has been moved to arch/x86/xen/time.c;
> 
> Konrad's comments have been addressed:
> 
> - gnttab_max_nr_grant_frames has been renamed gnttab_max_grant_frames;
> 
> - few inaccurate comments have been rewritten;
> 
> - the preprocessor checks in platform-pci-unplug.c have been moved to
> platform_pci.h;
> 
> - few other code style improvements, like using dev_err instead of
> printk and strncmp instead of strcmp.
> 
> 
> 
> A git tree is available here:
> 
> git://xenbits.xen.org/people/sstabellini/linux-pvhvm.git
> 
> branch name 2.6.34-pvhvm-v4.
> 
> Cheers,
> 
> Stefano
> 

Re: [PATCH 0/12] PV on HVM Xen

[Konrad Rzeszutek Wilk]

On Mon, Jun 21, 2010 at 05:12:20PM +0100, Stefano Stabellini wrote:
> Hi all,
> this is another update on the PV on HVM Xen series; a list of changes
> compared to the previous version follows:

Can you use the [v4] in the main patch (this is the fourth version of
these patches, right?)? I am getting confused as to which version I am
reviewing and if I've missed one or what not.

And in this first intro email state:

Since v7 the following things have changed:
 - Konrad made some comments.
 - etc.

Re: [PATCH 0/12] PV on HVM Xen

[Stefano Stabellini]

On Wed, 30 Jun 2010, Konrad Rzeszutek Wilk wrote:
> On Mon, Jun 21, 2010 at 05:12:20PM +0100, Stefano Stabellini wrote:
> > Hi all,
> > this is another update on the PV on HVM Xen series; a list of changes
> > compared to the previous version follows:
> 
> Can you use the [v4] in the main patch (this is the fourth version of
> these patches, right?)? I am getting confused as to which version I am
> reviewing and if I've missed one or what not.
> 
> And in this first intro email state:
> 
> Since v7 the following things have changed:
>  - Konrad made some comments.
>  - etc.
> 

Good point, I'll do that from now on.
I am pretty sure that you are reviewing the latest version I sent.