Re: [PATCH] virt: acrn: Fix vCPU removing code build error

[Shuo A Liu]

Hi Greg,

On Fri 12.Feb'21 at  8:52:33 +0100, Greg Kroah-Hartman wrote:
> On Fri, Feb 12, 2021 at 12:57:24PM +0800, shuo.a.liu@xxxxxxxxx wrote:
> > From: Shuo Liu <shuo.a.liu@xxxxxxxxx>
> >
> > vCPU removing code depends on CONFIG_HOTPLUG_CPU as it uses remove_cpu()
> > and add_cpu(). Make the vCPU removing interface building with
> > CONFIG_HOTPLUG_CPU.
> >
> > ../drivers/virt/acrn/hsm.c: In function ‘remove_cpu_store’:
> > ../drivers/virt/acrn/hsm.c:389:3: error: implicit declaration of function ‘remove_cpu’; [-Werror=implicit-function-declaration]
> >    remove_cpu(cpu);
> >
> > ../drivers/virt/acrn/hsm.c:402:2: error: implicit declaration of function ‘add_cpu’; [-Werror=implicit-function-declaration]
> >    add_cpu(cpu);
> >
> > Fixes: 279dcf693ac7 ("virt: acrn: Introduce an interface for Service VM to control vCPU")
> > Reported-by: Randy Dunlap <rdunlap@xxxxxxxxxxxxx>
> > Acked-by: Randy Dunlap <rdunlap@xxxxxxxxxxxxx> # build-tested
> > Signed-off-by: Shuo Liu <shuo.a.liu@xxxxxxxxx>
> > ---
> >  drivers/virt/acrn/hsm.c | 4 ++++
> >  1 file changed, 4 insertions(+)
> >
> > diff --git a/drivers/virt/acrn/hsm.c b/drivers/virt/acrn/hsm.c
> > index 1f6b7c54a1a4..e340788aacdf 100644
> > --- a/drivers/virt/acrn/hsm.c
> > +++ b/drivers/virt/acrn/hsm.c
> > @@ -372,6 +372,7 @@ static int acrn_dev_release(struct inode *inode, struct file *filp)
> >  	return 0;
> >  }
> >
> > +#ifdef CONFIG_HOTPLUG_CPU
> >  static ssize_t remove_cpu_store(struct device *dev,
> >  				struct device_attribute *attr,
> >  				const char *buf, size_t count)
> > @@ -403,9 +404,12 @@ static ssize_t remove_cpu_store(struct device *dev,
> >  	return ret;
> >  }
> >  static DEVICE_ATTR_WO(remove_cpu);
> > +#endif
> >
> >  static struct attribute *acrn_attrs[] = {
> > +#ifdef CONFIG_HOTPLUG_CPU
> >  	&dev_attr_remove_cpu.attr,
> > +#endif
> >  	NULL
> >  };
>
> Shouldn't the real solution for this be that remove_cpu() and add_cpu()
> have function prototypes for when this is not enabled in the kernel
> build?

Something like this in linux/cpu.h?

diff --git a/include/linux/cpu.h b/include/linux/cpu.h
index 3aaa0687e8df..94a578a96202 100644
--- a/include/linux/cpu.h
+++ b/include/linux/cpu.h
@@ -108,6 +108,8 @@ static inline void cpu_maps_update_done(void)
{
}

+static inline int add_cpu(unsigned int cpu) { return 0;}
+
#endif /* CONFIG_SMP */
extern struct bus_type cpu_subsys;

@@ -137,6 +139,7 @@ static inline int  cpus_read_trylock(void) { return
true; }
static inline void lockdep_assert_cpus_held(void) { }
static inline void cpu_hotplug_disable(void) { }
static inline void cpu_hotplug_enable(void) { }
+static inline int remove_cpu(unsigned int cpu) { return -EPERM; }
static inline void smp_shutdown_nonboot_cpus(unsigned int primary_cpu)
{ }
#endif /* !CONFIG_HOTPLUG_CPU */


> Putting #ifdef in .c files like this is not a good idea at all.
>
> Then, at runtime, you can determine if you need to create this sysfs
> file or not, as you do not want to expose it to userspace if the kernel
> can not handle it, right?

Right. I don't want to expose the sysfs to userspace if the kernel built
w/o CONFIG_HOTPLUG_CPU. But how to implement that if #ifdef is not used?
misc_register() creates sysfs with .groups (acrn_attr_groups)
unconditionally, then userspace can see the interface even it doesn't
work.

Thanks
shuo