Re: [PATCH v2] KVM: arm64: fix compile error because of shift overflow

From: Marc Zyngier
Date: Tue Aug 09 2022 - 09:54:04 EST


On Tue, 09 Aug 2022 14:51:27 +0100,
Yang Yingliang <yangyingliang@xxxxxxxxxx> wrote:
>
> Using GENMASK() to generate the masks of device type and device id, it makes
> code unambiguous, also it can fix the following fix compile error because of
> shift overflow when using low verison gcc(mine version is 7.5):
>
> In function ‘kvm_vm_ioctl_set_device_addr.isra.38’,
> inlined from ‘kvm_arch_vm_ioctl’ at arch/arm64/kvm/arm.c:1454:10:
> ././include/linux/compiler_types.h:354:38: error: call to ‘__compiletime_assert_599’ \
> declared with attribute error: FIELD_GET: mask is not constant
> _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
>
> Fixes: 9f968c9266aa ("KVM: arm64: vgic-v2: Add helper for legacy dist/cpuif base address setting")
> Signed-off-by: Yang Yingliang <yangyingliang@xxxxxxxxxx>
> ---
> v2:
> Using GENMASK() to generate the masks.
> ---
> arch/arm64/include/uapi/asm/kvm.h | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/arch/arm64/include/uapi/asm/kvm.h b/arch/arm64/include/uapi/asm/kvm.h
> index 3bb134355874..5e7dfaf76ec1 100644
> --- a/arch/arm64/include/uapi/asm/kvm.h
> +++ b/arch/arm64/include/uapi/asm/kvm.h
> @@ -75,9 +75,9 @@ struct kvm_regs {
>
> /* KVM_ARM_SET_DEVICE_ADDR ioctl id encoding */
> #define KVM_ARM_DEVICE_TYPE_SHIFT 0
> -#define KVM_ARM_DEVICE_TYPE_MASK (0xffff << KVM_ARM_DEVICE_TYPE_SHIFT)
> +#define KVM_ARM_DEVICE_TYPE_MASK GENMASK(15, KVM_ARM_DEVICE_TYPE_SHIFT)
> #define KVM_ARM_DEVICE_ID_SHIFT 16
> -#define KVM_ARM_DEVICE_ID_MASK (0xffff << KVM_ARM_DEVICE_ID_SHIFT)
> +#define KVM_ARM_DEVICE_ID_MASK GENMASK(31, KVM_ARM_DEVICE_ID_SHIFT)
>
> /* Supported device IDs */
> #define KVM_ARM_DEVICE_VGIC_V2 0

This is marginally better, but what I was expecting is something like:

#define KVM_ARM_DEVICE_ID_MASK GENMASK(KVM_ARM_DEVICE_ID_SHIFT + 15, \
KVM_ARM_DEVICE_ID_SHIFT)

which I find a bit more readable.

Thanks,

M.

--
Without deviation from the norm, progress is not possible.