Re: [PATCH RESEND] pwm: th1520: Remove requirement for mul_u64_u64_div_u64_roundup

[Michal Wilczynski]

On 5/25/26 15:11, Maurice Hieronymus via B4 Relay wrote:
> From: Maurice Hieronymus <mhi@xxxxxxxxxxx>
> 
> The cycle register is always u32, so cycles_to_ns() can take a u32
> instead of a u64. With that narrowing, cycles * NSEC_PER_SEC is at most
> u32::MAX * 1e9 (~4.3e18), which fits in u64 without overflow. The
> saturating arithmetic is therefore no longer needed, and the ceiling
> division can use Rust's u64::div_ceil() directly instead of the
> open-coded numerator/denominator form.
> 
> This also drops the TODO referring to a future
> mul_u64_u64_div_u64_roundup kernel helper, which is no longer required.
> 
> Signed-off-by: Maurice Hieronymus <mhi@xxxxxxxxxxx>

Hi Maurice,

Thanks for sending this.

I agree that this optimization makes perfect sense when looking at the
TH1520 driver in isolation. However, since this is currently the only
Rust PWM driver in the tree, it serves as the reference implementation
for future drivers.

We need cycles_to_ns to remain a generic blueprint. Many PWM controllers
have 64-bit registers, and in the C API, it is standard practice to use
helpers like DIV64_U64_ROUND_UP regardless of the register size (e.g.
axi_pwmgen). Resolving the TODO by assuming 32-bit registers removes
that generic capability.

At the time this was merged, the equivalent Rust math helpers didn't
exist yet, which is why the open-coded saturating math is there as a
placeholder.

To move forward, we should fix this by implementing the proper generic
64-bit math helper in the Rust abstractions, rather than narrowing the
types here. Are you open to looking into adding the Rust equivalent for
mul_u64_u64_div_u64_roundup instead?

Best regards,
-- 
Michal Wilczynski <m.wilczynski@xxxxxxxxxxx>