On Fri Mar 1, 2024 at 10:48 PM EET, Stefan Berger wrote:
>
>
> On 3/1/24 15:26, Jarkko Sakkinen wrote:
> > On Thu Feb 29, 2024 at 4:57 PM EET, Stefan Berger wrote:
> >>
> >>
> >> On 2/29/24 04:11, Lukas Wunner wrote:
> >>> On Fri, Feb 23, 2024 at 03:41:40PM -0500, Stefan Berger wrote:
> >>>> +static inline void ecc_digits_from_bytes(const u8 *in, unsigned int nbytes,
> >>>> + u64 *out, unsigned int ndigits)
> >>>> +{
> >>>> + unsigned int sz = ndigits << ECC_DIGITS_TO_BYTES_SHIFT;
> >>>> + u8 tmp[ECC_MAX_DIGITS << ECC_DIGITS_TO_BYTES_SHIFT];
> >>>> + unsigned int o = sz - nbytes;
> >>>> +
> >>>> + memset(tmp, 0, o);
> >>>> + memcpy(&tmp[o], in, nbytes);
> >>>> + ecc_swap_digits(tmp, out, ndigits);
> >>>> +}
> >>>
> >>> Copying the whole key into tmp seems inefficient. You only need
> >>> special handling for the first few bytes of "in" (6 bytes in the
> >>> P521 case) and could use ecc_swap_digits() to convert the rest
> >>> of "in" directly to "out" without using tmp.
> >>>
> >>> So it would be sufficient to allocate the first digit on the stack,
> >>> memset + memcpy, then convert that to native byte order into "in[0]"
> >>> and use ecc_swap_digits() for the rest.
> >>>
> >>> And the special handling would be conditional on "!o", so is skipped
> >>> for existing curves.
> >>
> >> Thanks. It looks like this now:
> >>
> >> static inline void ecc_digits_from_bytes(const u8 *in, unsigned int nbytes,
> >> u64 *out, unsigned int ndigits)
> >> {
> >> unsigned int o = nbytes & 7;
> >> u64 msd = 0;
> >> size_t i;
> >>
> >> if (o == 0) {
> >> ecc_swap_digits(in, out, ndigits);
> >> } else {
> >> for (i = 0; i < o; i++)
> >> msd = (msd << 8) | in[i];
> >> out[ndigits - 1] = msd;
> >> ecc_swap_digits(&in[o], out, ndigits - 1);
> >
> > This would be more stream-lined IMHO:
> >
> > unsigned int o = nbytes & 7;
> > unsigned int n = ndigits;
> > u64 msd = 0;
> > size_t i;
> >
> > if (o != 0) {
> > for (i = 0; i < o; i++)
> > msd = (msd << 8) | in[i];
> >
> > out[--n] = msd;
> > }
> >
> > ecc_swap_digits(in, out, n);
>
> You forgot to advance 'in'.
yeah, pointing out that two call sites is unnecessary complexity, that's
all
BR, Jarkko
