diff mbox series

powerpc: Add KVM guest defconfig

Message ID 20181108050826.17796-1-sathnaga@linux.vnet.ibm.com (mailing list archive)
State Superseded
Headers show
Series powerpc: Add KVM guest defconfig | expand

Checks

Context Check Description
snowpatch_ozlabs/apply_patch success next/apply_patch Successfully applied
snowpatch_ozlabs/checkpatch success Test checkpatch on branch next
snowpatch_ozlabs/build-ppc64le success Test build-ppc64le on branch next
snowpatch_ozlabs/build-ppc64be success Test build-ppc64be on branch next
snowpatch_ozlabs/build-ppc64e success Test build-ppc64e on branch next
snowpatch_ozlabs/build-ppc32 success Test build-ppc32 on branch next

Commit Message

Satheesh Rajendran Nov. 8, 2018, 5:08 a.m. UTC 
From: Satheesh Rajendran <sathnaga@linux.vnet.ibm.com>

This patch adds new defconfig options for powerpc KVM guest
and guest.config with additional config symbols enabled,
which is to build kernel to boot without initramfs and can be used
as place holder for guest specific additional config symbols in future.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Satheesh Rajendran <sathnaga@linux.vnet.ibm.com>
---
 arch/powerpc/Makefile             |  8 ++++++++
 arch/powerpc/configs/guest.config | 14 ++++++++++++++
 2 files changed, 22 insertions(+)
 create mode 100644 arch/powerpc/configs/guest.config

Comments

Breno Leitao Nov. 8, 2018, 6:23 p.m. UTC | #1 
hi Satheesh,

On 11/08/2018 03:08 AM, sathnaga@linux.vnet.ibm.com wrote:
> --- /dev/null
> +++ b/arch/powerpc/configs/guest.config
> @@ -0,0 +1,14 @@
> +CONFIG_VIRTIO_BLK=y
> +CONFIG_VIRTIO_BLK_SCSI=y
> +CONFIG_SCSI_VIRTIO=y
> +CONFIG_VIRTIO_NET=y
> +CONFIG_NET_FAILOVER=y
> +CONFIG_VIRTIO_CONSOLE=y
> +CONFIG_VIRTIO=y
> +CONFIG_VIRTIO_PCI=y
> +CONFIG_KVM_GUEST=y
> +CONFIG_EPAPR_PARAVIRT=y
> +CONFIG_XFS_FS=y

Why a guest kernel needs to have XFS integrated in the core image? I am
wondering if it is a requirement from another CONFIG_ option.

If it is not a strict requirement from another config, I think we can keep it
as defined at ppc64_defconfig, which defines it as module (CONFIG_XFS_FS=m).

Thanks for this patch, very useful.
Breno
Satheesh Rajendran Nov. 9, 2018, 9:59 a.m. UTC | #2 
On Thu, Nov 08, 2018 at 04:23:07PM -0200, Breno Leitao wrote:
> hi Satheesh,
> 
> On 11/08/2018 03:08 AM, sathnaga@linux.vnet.ibm.com wrote:
> > --- /dev/null
> > +++ b/arch/powerpc/configs/guest.config
> > @@ -0,0 +1,14 @@
> > +CONFIG_VIRTIO_BLK=y
> > +CONFIG_VIRTIO_BLK_SCSI=y
> > +CONFIG_SCSI_VIRTIO=y
> > +CONFIG_VIRTIO_NET=y
> > +CONFIG_NET_FAILOVER=y
> > +CONFIG_VIRTIO_CONSOLE=y
> > +CONFIG_VIRTIO=y
> > +CONFIG_VIRTIO_PCI=y
> > +CONFIG_KVM_GUEST=y
> > +CONFIG_EPAPR_PARAVIRT=y
> > +CONFIG_XFS_FS=y
> 
> Why a guest kernel needs to have XFS integrated in the core image? I am
> wondering if it is a requirement from another CONFIG_ option.

Idea is to have a working config which would boot guest without initramfs,
other FS(like EXT4) is already integrated in the core image, 
thought this would be helpful for distributions, which default XFS as root disk.

Hope this should be fine?

> 
> If it is not a strict requirement from another config, I think we can keep it
> as defined at ppc64_defconfig, which defines it as module (CONFIG_XFS_FS=m).
> 
> Thanks for this patch, very useful.

Thanks Breno! :-)

-Satheesh.

> Breno
>
Breno Leitao Nov. 9, 2018, 12:30 p.m. UTC | #3 
hi Satheesh,

On 11/09/2018 07:59 AM, Satheesh Rajendran wrote:
> On Thu, Nov 08, 2018 at 04:23:07PM -0200, Breno Leitao wrote:
>> hi Satheesh,
>>
>> On 11/08/2018 03:08 AM, sathnaga@linux.vnet.ibm.com wrote:
>>> --- /dev/null
>>> +++ b/arch/powerpc/configs/guest.config
>>> @@ -0,0 +1,14 @@
>>> +CONFIG_VIRTIO_BLK=y
>>> +CONFIG_VIRTIO_BLK_SCSI=y
>>> +CONFIG_SCSI_VIRTIO=y
>>> +CONFIG_VIRTIO_NET=y
>>> +CONFIG_NET_FAILOVER=y
>>> +CONFIG_VIRTIO_CONSOLE=y
>>> +CONFIG_VIRTIO=y
>>> +CONFIG_VIRTIO_PCI=y
>>> +CONFIG_KVM_GUEST=y
>>> +CONFIG_EPAPR_PARAVIRT=y
>>> +CONFIG_XFS_FS=y
>>
>> Why a guest kernel needs to have XFS integrated in the core image? I am
>> wondering if it is a requirement from another CONFIG_ option.
> 
> Idea is to have a working config which would boot guest without initramfs,
> other FS(like EXT4) is already integrated in the core image, 
> thought this would be helpful for distributions, which default XFS as root disk.

Nice! That makes sense and it is a good idea.

Thanks!
Michael Ellerman Nov. 12, 2018, 12:24 p.m. UTC | #4 
Satheesh Rajendran <sathnaga@linux.vnet.ibm.com> writes:

> On Thu, Nov 08, 2018 at 04:23:07PM -0200, Breno Leitao wrote:
>> hi Satheesh,
>> 
>> On 11/08/2018 03:08 AM, sathnaga@linux.vnet.ibm.com wrote:
>> > --- /dev/null
>> > +++ b/arch/powerpc/configs/guest.config
>> > @@ -0,0 +1,14 @@
>> > +CONFIG_VIRTIO_BLK=y
>> > +CONFIG_VIRTIO_BLK_SCSI=y
>> > +CONFIG_SCSI_VIRTIO=y
>> > +CONFIG_VIRTIO_NET=y
>> > +CONFIG_NET_FAILOVER=y
>> > +CONFIG_VIRTIO_CONSOLE=y
>> > +CONFIG_VIRTIO=y
>> > +CONFIG_VIRTIO_PCI=y
>> > +CONFIG_KVM_GUEST=y
>> > +CONFIG_EPAPR_PARAVIRT=y
>> > +CONFIG_XFS_FS=y
>> 
>> Why a guest kernel needs to have XFS integrated in the core image? I am
>> wondering if it is a requirement from another CONFIG_ option.
>
> Idea is to have a working config which would boot guest without initramfs,
> other FS(like EXT4) is already integrated in the core image, 
> thought this would be helpful for distributions, which default XFS as root disk.

Maybe we should switch XFS_FS to Y in ppc64_defconfig ?

cheers
Satheesh Rajendran Nov. 13, 2018, 3:43 a.m. UTC | #5 
On Mon, Nov 12, 2018 at 11:24:08PM +1100, Michael Ellerman wrote:
> Satheesh Rajendran <sathnaga@linux.vnet.ibm.com> writes:
> 
> > On Thu, Nov 08, 2018 at 04:23:07PM -0200, Breno Leitao wrote:
> >> hi Satheesh,
> >> 
> >> On 11/08/2018 03:08 AM, sathnaga@linux.vnet.ibm.com wrote:
> >> > --- /dev/null
> >> > +++ b/arch/powerpc/configs/guest.config
> >> > @@ -0,0 +1,14 @@
> >> > +CONFIG_VIRTIO_BLK=y
> >> > +CONFIG_VIRTIO_BLK_SCSI=y
> >> > +CONFIG_SCSI_VIRTIO=y
> >> > +CONFIG_VIRTIO_NET=y
> >> > +CONFIG_NET_FAILOVER=y
> >> > +CONFIG_VIRTIO_CONSOLE=y
> >> > +CONFIG_VIRTIO=y
> >> > +CONFIG_VIRTIO_PCI=y
> >> > +CONFIG_KVM_GUEST=y
> >> > +CONFIG_EPAPR_PARAVIRT=y
> >> > +CONFIG_XFS_FS=y
> >> 
> >> Why a guest kernel needs to have XFS integrated in the core image? I am
> >> wondering if it is a requirement from another CONFIG_ option.
> >
> > Idea is to have a working config which would boot guest without initramfs,
> > other FS(like EXT4) is already integrated in the core image, 
> > thought this would be helpful for distributions, which default XFS as root disk.
> 
> Maybe we should switch XFS_FS to Y in ppc64_defconfig ?

Sure, makes sense, will send it for ppc64_defconfig instead. 
Inaddition, Have few more symbols to be enabled for cgroups,
memhotplug,numa balancing.
I guess these symbols can also go to ppc64_defconfig itself?.

i.e,

CONFIG_CGROUP_SCHED=y
CONFIG_MEMCG=y
CONFIG_CGROUP_SCHED=y
CONFIG_CGROUP_FREEZER=y
CONFIG_CGROUP_DEVICE=y
CONFIG_CGROUP_CPUACCT=y
CONFIG_CGROUP_PERF=y
CONFIG_MEMORY_HOTPLUG=y
CONFIG_MEMORY_HOTREMOVE=y
CONFIG_NUMA_BALANCING=y

Thanks!
-Satheesh.
> 
> cheers
>
Michael Ellerman Nov. 13, 2018, 11:07 a.m. UTC | #6 
Satheesh Rajendran <sathnaga@linux.vnet.ibm.com> writes:
> On Mon, Nov 12, 2018 at 11:24:08PM +1100, Michael Ellerman wrote:
>> Satheesh Rajendran <sathnaga@linux.vnet.ibm.com> writes:
>> 
>> > On Thu, Nov 08, 2018 at 04:23:07PM -0200, Breno Leitao wrote:
>> >> hi Satheesh,
>> >> 
>> >> On 11/08/2018 03:08 AM, sathnaga@linux.vnet.ibm.com wrote:
>> >> > --- /dev/null
>> >> > +++ b/arch/powerpc/configs/guest.config
>> >> > @@ -0,0 +1,14 @@
>> >> > +CONFIG_VIRTIO_BLK=y
>> >> > +CONFIG_VIRTIO_BLK_SCSI=y
>> >> > +CONFIG_SCSI_VIRTIO=y
>> >> > +CONFIG_VIRTIO_NET=y
>> >> > +CONFIG_NET_FAILOVER=y
>> >> > +CONFIG_VIRTIO_CONSOLE=y
>> >> > +CONFIG_VIRTIO=y
>> >> > +CONFIG_VIRTIO_PCI=y
>> >> > +CONFIG_KVM_GUEST=y
>> >> > +CONFIG_EPAPR_PARAVIRT=y
>> >> > +CONFIG_XFS_FS=y
>> >> 
>> >> Why a guest kernel needs to have XFS integrated in the core image? I am
>> >> wondering if it is a requirement from another CONFIG_ option.
>> >
>> > Idea is to have a working config which would boot guest without initramfs,
>> > other FS(like EXT4) is already integrated in the core image, 
>> > thought this would be helpful for distributions, which default XFS as root disk.
>> 
>> Maybe we should switch XFS_FS to Y in ppc64_defconfig ?
>
> Sure, makes sense, will send it for ppc64_defconfig instead. 
> Inaddition, Have few more symbols to be enabled for cgroups,
> memhotplug,numa balancing.
> I guess these symbols can also go to ppc64_defconfig itself?.
>
> i.e,
>
> CONFIG_CGROUP_SCHED=y
> CONFIG_MEMCG=y
> CONFIG_CGROUP_SCHED=y
> CONFIG_CGROUP_FREEZER=y
> CONFIG_CGROUP_DEVICE=y
> CONFIG_CGROUP_CPUACCT=y
> CONFIG_CGROUP_PERF=y
> CONFIG_MEMORY_HOTPLUG=y
> CONFIG_MEMORY_HOTREMOVE=y
> CONFIG_NUMA_BALANCING=y

Yeah I don't see why not, they're probably more or less required by
modern distros.

cheers
diff mbox series

Patch

diff --git a/arch/powerpc/Makefile b/arch/powerpc/Makefile
index 8a2ce14d68d0..0bff8bd82ed5 100644
--- a/arch/powerpc/Makefile
+++ b/arch/powerpc/Makefile
@@ -317,6 +317,14 @@  PHONY += ppc64le_defconfig
 ppc64le_defconfig:
 	$(call merge_into_defconfig,ppc64_defconfig,le)
 
+PHONY += ppc64le_guest_defconfig
+ppc64le_guest_defconfig:
+	$(call merge_into_defconfig,ppc64_defconfig,le guest)
+
+PHONY += ppc64_guest_defconfig
+ppc64_guest_defconfig:
+	$(call merge_into_defconfig,ppc64_defconfig,be guest)
+
 PHONY += powernv_be_defconfig
 powernv_be_defconfig:
 	$(call merge_into_defconfig,powernv_defconfig,be)
diff --git a/arch/powerpc/configs/guest.config b/arch/powerpc/configs/guest.config
new file mode 100644
index 000000000000..a5eb8b904058
--- /dev/null
+++ b/arch/powerpc/configs/guest.config
@@ -0,0 +1,14 @@ 
+CONFIG_VIRTIO_BLK=y
+CONFIG_VIRTIO_BLK_SCSI=y
+CONFIG_SCSI_VIRTIO=y
+CONFIG_VIRTIO_NET=y
+CONFIG_NET_FAILOVER=y
+CONFIG_VIRTIO_CONSOLE=y
+CONFIG_VIRTIO=y
+CONFIG_VIRTIO_PCI=y
+CONFIG_KVM_GUEST=y
+CONFIG_EPAPR_PARAVIRT=y
+CONFIG_XFS_FS=y
+CONFIG_VIRTIO_BALLOON=y
+CONFIG_VHOST_NET=y
+CONFIG_VHOST=y