Reported by coccinelle, there is a do_div call that does
64-by-32 divisions even in 64bit platforms, this patch will
move it to div64_ul macro that will decide the correct
division function for the platform underneath. The output
the warning follows:

arch/x86/kvm/lapic.c:1948:1-7: WARNING: do_div() does a 64-by-32 division, please consider using div64_ul instead.

Signed-off-by: José Pekkarinen <jose.pekkarinen@foxhound.fi>
---
 arch/x86/kvm/lapic.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index 3e977dbbf993..0b90c6ad5091 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -1945,7 +1945,7 @@ static void start_sw_tscdeadline(struct kvm_lapic *apic)
 	guest_tsc = kvm_read_l1_tsc(vcpu, rdtsc());
 
 	ns = (tscdeadline - guest_tsc) * 1000000ULL;
-	do_div(ns, this_tsc_khz);
+	div64_ul(ns, this_tsc_khz);
 
 	if (likely(tscdeadline > guest_tsc) &&
 	    likely(ns > apic->lapic_timer.timer_advance_ns)) {
-- 
2.39.2

From: José Pekkarinen
> Sent: 29 October 2023 09:39
> 
> Reported by coccinelle, there is a do_div call that does
> 64-by-32 divisions even in 64bit platforms, this patch will
> move it to div64_ul macro that will decide the correct
> division function for the platform underneath. The output
> the warning follows:
> 
> arch/x86/kvm/lapic.c:1948:1-7: WARNING: do_div() does a 64-by-32 division, please consider using
> div64_ul instead.

That is about the worst message from the coccinelle scripts.
It really ought to ask you to check the domain on the values.

> 
> Signed-off-by: José Pekkarinen <jose.pekkarinen@foxhound.fi>
> ---
>  arch/x86/kvm/lapic.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
> index 3e977dbbf993..0b90c6ad5091 100644
> --- a/arch/x86/kvm/lapic.c
> +++ b/arch/x86/kvm/lapic.c
> @@ -1945,7 +1945,7 @@ static void start_sw_tscdeadline(struct kvm_lapic *apic)
>  	guest_tsc = kvm_read_l1_tsc(vcpu, rdtsc());
> 
>  	ns = (tscdeadline - guest_tsc) * 1000000ULL;
> -	do_div(ns, this_tsc_khz);
> +	div64_ul(ns, this_tsc_khz);

Did you test this?
Hint - you didn't.

	David

> 
>  	if (likely(tscdeadline > guest_tsc) &&
>  	    likely(ns > apic->lapic_timer.timer_advance_ns)) {
> --
> 2.39.2

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)

On Sun, Oct 29, 2023, José Pekkarinen wrote:
> Reported by coccinelle, there is a do_div call that does
> 64-by-32 divisions even in 64bit platforms, this patch will
> move it to div64_ul macro that will decide the correct
> division function for the platform underneath. The output
> the warning follows:
> 
> arch/x86/kvm/lapic.c:1948:1-7: WARNING: do_div() does a 64-by-32 division, please consider using div64_ul instead.
> 
> Signed-off-by: José Pekkarinen <jose.pekkarinen@foxhound.fi>
> ---
>  arch/x86/kvm/lapic.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
> index 3e977dbbf993..0b90c6ad5091 100644
> --- a/arch/x86/kvm/lapic.c
> +++ b/arch/x86/kvm/lapic.c
> @@ -1945,7 +1945,7 @@ static void start_sw_tscdeadline(struct kvm_lapic *apic)
>  	guest_tsc = kvm_read_l1_tsc(vcpu, rdtsc());
>  
>  	ns = (tscdeadline - guest_tsc) * 1000000ULL;
> -	do_div(ns, this_tsc_khz);
> +	div64_ul(ns, this_tsc_khz);

Well this is silly, virtual_tsc_khz is a u32.

	unsigned long this_tsc_khz = vcpu->arch.virtual_tsc_khz;

struct kvm_vcpu_arch { 

	...

	u32 virtual_tsc_khz;

}

I assume this will make coccinelle happy?

diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index 245b20973cae..31e9c84b8791 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -1932,7 +1932,7 @@ static void start_sw_tscdeadline(struct kvm_lapic *apic)
        u64 ns = 0;
        ktime_t expire;
        struct kvm_vcpu *vcpu = apic->vcpu;
-       unsigned long this_tsc_khz = vcpu->arch.virtual_tsc_khz;
+       u32 this_tsc_khz = vcpu->arch.virtual_tsc_khz;
        unsigned long flags;
        ktime_t now;