[PATCH] i2c: stm32f7: truncate clock period instead of rounding it

[Guillermo Rodriguez Garcia]

Hi,

Thank you everyone.

A possible improvement would be to derive the clock period in
picoseconds rather than ns and do all the timing calculations based on
that.
This would reduce the rounding error (from max 1ns to max 1ps) and
help the solver find TIMINGR values closer to the target bus
frequency, while still keeping both freq and tLOW within spec.

I wanted to keep this patch minimal to address the spec violation
itself, and also to make it easier to backport to stable. The
suggested improvement is a bit more involved, but I wanted to mention
it in case you want to consider it.

Thanks,

Guillermo


El mié, 17 jun 2026 a las 10:38, Andi Shyti (<andi.shyti at kernel.org>) escribió:
>
> Hi Guillermo,
>
> On Thu, Jun 11, 2026 at 12:48:56PM +0200, Guillermo Rodríguez wrote:
> > stm32f7_i2c_compute_timing() derives the I2C clock source period
> > (i2cclk) with DIV_ROUND_CLOSEST, which may round it up. When the
> > period is overestimated, all timings computed from it (SCLDEL,
> > SDADEL, SCLL, SCLH) come out shorter on the wire than calculated,
> > and the resulting bus rate can exceed the requested speed, violating
> > the I2C specification minimums for tLOW and tHIGH.
> >
> > For example, with a 104.45 MHz clock source (e.g. PCLK1, the
> > reset-default I2C clock source on STM32MP1), i2cclk is rounded from
> > 9.574 ns up to 10 ns. Requesting a 400 kHz fast mode bus with
> > 72/27 ns rise/fall times and no analog/digital filters then produces
> > an actual bus rate of 415.6 kHz with tLOW = 1254 ns, violating both
> > the 400 kHz maximum rate and the 1300 ns tLOW minimum of the
> > specification.
> >
> > Truncate the period instead, so that it can only be underestimated.
> > The error then falls on the safe side: the programmed timings come
> > out slightly longer than computed and the bus runs marginally below
> > the target rate (375.3 kHz in the example above) while meeting the
> > specification.
> >
> > i2cbus is left rounded-to-closest: it is only used as the target of
> > the clk_error comparison and is never multiplied into the programmed
> > timings, so nearest rounding remains accurate there.
> >
> > Fixes: aeb068c57214 ("i2c: i2c-stm32f7: add driver")
> > Cc: stable at vger.kernel.org
> > Signed-off-by: Guillermo Rodríguez <guille.rodriguez at gmail.com>
>
> Merged to i2c/i2c-host.
>
> Thanks to Alain and Pierre-Yves for their review!
>
> Andi



--
Guillermo Rodriguez Garcia
guille.rodriguez at gmail.com