Linux Virtualization - Apr 2020 - [PATCH V2] vhost: do not enable VHOST_MENU by default

On 2020/4/17 ??4:29, Michael S. Tsirkin wrote:
> On Fri, Apr 17, 2020 at 03:36:52PM +0800, Jason Wang wrote:
>> On 2020/4/17 ??2:33, Michael S. Tsirkin wrote:
>>> On Fri, Apr 17, 2020 at 11:12:14AM +0800, Jason Wang wrote:
>>>> On 2020/4/17 ??6:55, Michael S. Tsirkin wrote:
>>>>> On Wed, Apr 15, 2020 at 10:43:56AM +0800, Jason Wang wrote:
>>>>>> We try to keep the defconfig untouched after decoupling
CONFIG_VHOST
>>>>>> out of CONFIG_VIRTUALIZATION in commit 20c384f1ea1a
>>>>>> ("vhost: refine vhost and vringh kconfig") by
enabling VHOST_MENU by
>>>>>> default. Then the defconfigs can keep enabling
CONFIG_VHOST_NET
>>>>>> without the caring of CONFIG_VHOST.
>>>>>>
>>>>>> But this will leave a "CONFIG_VHOST_MENU=y"
in all defconfigs and even
>>>>>> for the ones that doesn't want vhost. So it
actually shifts the
>>>>>> burdens to the maintainers of all other to add
"CONFIG_VHOST_MENU is
>>>>>> not set". So this patch tries to enable
CONFIG_VHOST explicitly in
>>>>>> defconfigs that enables CONFIG_VHOST_NET and
CONFIG_VHOST_VSOCK.
>>>>>>
>>>>>> Acked-by: Christian Borntraeger<borntraeger at
de.ibm.com>  (s390)
>>>>>> Acked-by: Michael Ellerman<mpe at ellerman.id.au>
(powerpc)
>>>>>> Cc: Thomas Bogendoerfer<tsbogend at
alpha.franken.de>
>>>>>> Cc: Benjamin Herrenschmidt<benh at
kernel.crashing.org>
>>>>>> Cc: Paul Mackerras<paulus at samba.org>
>>>>>> Cc: Michael Ellerman<mpe at ellerman.id.au>
>>>>>> Cc: Heiko Carstens<heiko.carstens at de.ibm.com>
>>>>>> Cc: Vasily Gorbik<gor at linux.ibm.com>
>>>>>> Cc: Christian Borntraeger<borntraeger at
de.ibm.com>
>>>>>> Reported-by: Geert Uytterhoeven<geert at
linux-m68k.org>
>>>>>> Signed-off-by: Jason Wang<jasowang at redhat.com>
>>>>> I rebased this on top of OABI fix since that
>>>>> seems more orgent to fix.
>>>>> Pushed to my vhost branch pls take a look and
>>>>> if possible test.
>>>>> Thanks!
>>>> I test this patch by generating the defconfigs that wants
vhost_net or
>>>> vhost_vsock. All looks fine.
>>>>
>>>> But having CONFIG_VHOST_DPN=y may end up with the similar
situation that
>>>> this patch want to address.
>>>> Maybe we can let CONFIG_VHOST depends on !ARM || AEABI then add
another
>>>> menuconfig for VHOST_RING and do something similar?
>>>>
>>>> Thanks
>>> Sorry I don't understand. After this patch CONFIG_VHOST_DPN is
just
>>> an internal variable for the OABI fix. I kept it separate
>>> so it's easy to revert for 5.8. Yes we could squash it into
>>> VHOST directly but I don't see how that changes logic at all.
>>
>> Sorry for being unclear.
>>
>> I meant since it was enabled by default, "CONFIG_VHOST_DPN=y"
will be left
>> in the defconfigs.
> But who cares?

FYI, please see https://www.spinics.net/lists/kvm/msg212685.html

> That does not add any code, does it?

It doesn't.

Thanks

>
>> This requires the arch maintainers to add
>> "CONFIG_VHOST_VDPN is not set". (Geert complains about this)
>>
>> Thanks
>>
>>

On Fri, Apr 17, 2020 at 04:39:49PM +0800, Jason Wang
wrote:
> 
> On 2020/4/17 ??4:29, Michael S. Tsirkin wrote:
> > On Fri, Apr 17, 2020 at 03:36:52PM +0800, Jason Wang wrote:
> > > On 2020/4/17 ??2:33, Michael S. Tsirkin wrote:
> > > > On Fri, Apr 17, 2020 at 11:12:14AM +0800, Jason Wang wrote:
> > > > > On 2020/4/17 ??6:55, Michael S. Tsirkin wrote:
> > > > > > On Wed, Apr 15, 2020 at 10:43:56AM +0800, Jason
Wang wrote:
> > > > > > > We try to keep the defconfig untouched after
decoupling CONFIG_VHOST
> > > > > > > out of CONFIG_VIRTUALIZATION in commit
20c384f1ea1a
> > > > > > > ("vhost: refine vhost and vringh
kconfig") by enabling VHOST_MENU by
> > > > > > > default. Then the defconfigs can keep
enabling CONFIG_VHOST_NET
> > > > > > > without the caring of CONFIG_VHOST.
> > > > > > > 
> > > > > > > But this will leave a
"CONFIG_VHOST_MENU=y" in all defconfigs and even
> > > > > > > for the ones that doesn't want vhost. So
it actually shifts the
> > > > > > > burdens to the maintainers of all other to
add "CONFIG_VHOST_MENU is
> > > > > > > not set". So this patch tries to enable
CONFIG_VHOST explicitly in
> > > > > > > defconfigs that enables CONFIG_VHOST_NET and
CONFIG_VHOST_VSOCK.
> > > > > > > 
> > > > > > > Acked-by: Christian
Borntraeger<borntraeger at de.ibm.com>  (s390)
> > > > > > > Acked-by: Michael Ellerman<mpe at
ellerman.id.au>  (powerpc)
> > > > > > > Cc: Thomas Bogendoerfer<tsbogend at
alpha.franken.de>
> > > > > > > Cc: Benjamin Herrenschmidt<benh at
kernel.crashing.org>
> > > > > > > Cc: Paul Mackerras<paulus at samba.org>
> > > > > > > Cc: Michael Ellerman<mpe at
ellerman.id.au>
> > > > > > > Cc: Heiko Carstens<heiko.carstens at
de.ibm.com>
> > > > > > > Cc: Vasily Gorbik<gor at linux.ibm.com>
> > > > > > > Cc: Christian Borntraeger<borntraeger at
de.ibm.com>
> > > > > > > Reported-by: Geert Uytterhoeven<geert at
linux-m68k.org>
> > > > > > > Signed-off-by: Jason Wang<jasowang at
redhat.com>
> > > > > > I rebased this on top of OABI fix since that
> > > > > > seems more orgent to fix.
> > > > > > Pushed to my vhost branch pls take a look and
> > > > > > if possible test.
> > > > > > Thanks!
> > > > > I test this patch by generating the defconfigs that
wants vhost_net or
> > > > > vhost_vsock. All looks fine.
> > > > > 
> > > > > But having CONFIG_VHOST_DPN=y may end up with the
similar situation that
> > > > > this patch want to address.
> > > > > Maybe we can let CONFIG_VHOST depends on !ARM || AEABI
then add another
> > > > > menuconfig for VHOST_RING and do something similar?
> > > > > 
> > > > > Thanks
> > > > Sorry I don't understand. After this patch
CONFIG_VHOST_DPN is just
> > > > an internal variable for the OABI fix. I kept it separate
> > > > so it's easy to revert for 5.8. Yes we could squash it
into
> > > > VHOST directly but I don't see how that changes logic at
all.
> > > 
> > > Sorry for being unclear.
> > > 
> > > I meant since it was enabled by default,
"CONFIG_VHOST_DPN=y" will be left
> > > in the defconfigs.
> > But who cares?
> 
> 
> FYI, please see https://www.spinics.net/lists/kvm/msg212685.html
The complaint was not about the symbol IIUC.  It was that we caused
everyone to build vhost unless they manually disabled it.
> 
> > That does not add any code, does it?
> 
> 
> It doesn't.
> 
> Thanks
> 
> 
> > 
> > > This requires the arch maintainers to add
> > > "CONFIG_VHOST_VDPN is not set". (Geert complains about
this)
> > > 
> > > Thanks
> > > 
> > >

On 2020/4/17 ??4:46, Michael S. Tsirkin wrote:
> On Fri, Apr 17, 2020 at 04:39:49PM +0800, Jason Wang wrote:
>> On 2020/4/17 ??4:29, Michael S. Tsirkin wrote:
>>> On Fri, Apr 17, 2020 at 03:36:52PM +0800, Jason Wang wrote:
>>>> On 2020/4/17 ??2:33, Michael S. Tsirkin wrote:
>>>>> On Fri, Apr 17, 2020 at 11:12:14AM +0800, Jason Wang wrote:
>>>>>> On 2020/4/17 ??6:55, Michael S. Tsirkin wrote:
>>>>>>> On Wed, Apr 15, 2020 at 10:43:56AM +0800, Jason
Wang wrote:
>>>>>>>> We try to keep the defconfig untouched after
decoupling CONFIG_VHOST
>>>>>>>> out of CONFIG_VIRTUALIZATION in commit
20c384f1ea1a
>>>>>>>> ("vhost: refine vhost and vringh
kconfig") by enabling VHOST_MENU by
>>>>>>>> default. Then the defconfigs can keep enabling
CONFIG_VHOST_NET
>>>>>>>> without the caring of CONFIG_VHOST.
>>>>>>>>
>>>>>>>> But this will leave a
"CONFIG_VHOST_MENU=y" in all defconfigs and even
>>>>>>>> for the ones that doesn't want vhost. So it
actually shifts the
>>>>>>>> burdens to the maintainers of all other to add
"CONFIG_VHOST_MENU is
>>>>>>>> not set". So this patch tries to enable
CONFIG_VHOST explicitly in
>>>>>>>> defconfigs that enables CONFIG_VHOST_NET and
CONFIG_VHOST_VSOCK.
>>>>>>>>
>>>>>>>> Acked-by: Christian Borntraeger<borntraeger
at de.ibm.com>   (s390)
>>>>>>>> Acked-by: Michael Ellerman<mpe at
ellerman.id.au>   (powerpc)
>>>>>>>> Cc: Thomas Bogendoerfer<tsbogend at
alpha.franken.de>
>>>>>>>> Cc: Benjamin Herrenschmidt<benh at
kernel.crashing.org>
>>>>>>>> Cc: Paul Mackerras<paulus at samba.org>
>>>>>>>> Cc: Michael Ellerman<mpe at
ellerman.id.au>
>>>>>>>> Cc: Heiko Carstens<heiko.carstens at
de.ibm.com>
>>>>>>>> Cc: Vasily Gorbik<gor at linux.ibm.com>
>>>>>>>> Cc: Christian Borntraeger<borntraeger at
de.ibm.com>
>>>>>>>> Reported-by: Geert Uytterhoeven<geert at
linux-m68k.org>
>>>>>>>> Signed-off-by: Jason Wang<jasowang at
redhat.com>
>>>>>>> I rebased this on top of OABI fix since that
>>>>>>> seems more orgent to fix.
>>>>>>> Pushed to my vhost branch pls take a look and
>>>>>>> if possible test.
>>>>>>> Thanks!
>>>>>> I test this patch by generating the defconfigs that
wants vhost_net or
>>>>>> vhost_vsock. All looks fine.
>>>>>>
>>>>>> But having CONFIG_VHOST_DPN=y may end up with the
similar situation that
>>>>>> this patch want to address.
>>>>>> Maybe we can let CONFIG_VHOST depends on !ARM || AEABI
then add another
>>>>>> menuconfig for VHOST_RING and do something similar?
>>>>>>
>>>>>> Thanks
>>>>> Sorry I don't understand. After this patch
CONFIG_VHOST_DPN is just
>>>>> an internal variable for the OABI fix. I kept it separate
>>>>> so it's easy to revert for 5.8. Yes we could squash it
into
>>>>> VHOST directly but I don't see how that changes logic
at all.
>>>> Sorry for being unclear.
>>>>
>>>> I meant since it was enabled by default,
"CONFIG_VHOST_DPN=y" will be left
>>>> in the defconfigs.
>>> But who cares?
>> FYI, please seehttps://www.spinics.net/lists/kvm/msg212685.html
> The complaint was not about the symbol IIUC.  It was that we caused
> everyone to build vhost unless they manually disabled it.

There could be some misunderstanding here. I thought it's somehow 
similar: a CONFIG_VHOST_MENU=y will be left in the defconfigs even if 
CONFIG_VHOST is not set.

Thanks

>