Hi Biju,
On Sun, 16 Mar 2025 at 18:02, Biju Das <biju.das.jz@xxxxxxxxxxxxxx> wrote:
> > From: Geert Uytterhoeven <geert@xxxxxxxxxxxxxx>
> > On Sat, 15 Mar 2025 at 18:31, Biju Das <biju.das.jz@xxxxxxxxxxxxxx> wrote:
> > > > From: Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> On Thu, 6 Mar 2025
> > > > at 13:43, Biju Das <biju.das.jz@xxxxxxxxxxxxxx> wrote:
> > > > > R-Car Gen3 and Gen4 has some differences in the shift bits. Add a
> > > > > shift table to handle these differences. After this drop the
> > > > > unused functions reg_gen4() and is_gen4().
> > > > >
> > > > > Signed-off-by: Biju Das <biju.das.jz@xxxxxxxxxxxxxx>
> > > >
> > > > Reviewed-by: Geert Uytterhoeven <geert+renesas@xxxxxxxxx>
> > > >
> > > > A suggestion for improvement below.
> > > >
> > > > > --- a/drivers/net/can/rcar/rcar_canfd.c
> > > > > +++ b/drivers/net/can/rcar/rcar_canfd.c
> > > > > @@ -90,11 +90,13 @@
> > > > > /* RSCFDnCFDGAFLCFG0 / RSCFDnGAFLCFG0 */ #define
> > > > > RCANFD_GAFLCFG_SETRNC(gpriv, n, x) \
> > > > > (((x) & (gpriv)->info->mask_table[RCANFD_RNC_MASK]) << \
> > > > > - (reg_gen4(gpriv, 16, 24) - ((n) & 1) * reg_gen4(gpriv, 16, 8)))
> > > > > + ((gpriv)->info->shift_table[RCANFD_FIRST_RNC_SH] - ((n) & 1) * \
> > > > > + (gpriv)->info->shift_table[RCANFD_SECOND_RNC_SH]))
> > > >
> > > > Both shifts are dictated by the field width:
> > > > - R-Car Gen4 packs 2 values in a 32-bit word, using a field width
> > > > of 16 bits,
> > > > - R-Car Gen3 packs up to 4 values in a 32-bit word, using a field
> > > > width of 8 bits.
> > > > The odd split in two shifts is due to the MSB-first numbering of the
> > > > fields. Hence I think it would be better to use a field width
> > > > parameter, and derive the proper shift value from that (like you
> > > > have already done for ICU ;-)
> > > >
> > > > Note that the formula will have to be reworked anyway to support the
> > > > third channel on R-Car V3H_2, as "n & "1 will no longer fly here...
> > >
> > > Agreed. Will introduce rnc_field_width variable in struct
> > > rcar_canfd_hw_info and Simplify the calculation using the formula (32 - (n + 1) * rnc_field_width).
> > > Which will work for third channel on R-Car V3H_2 as well.
> >
> > Note that you still need to mask n, as n can be up to 7 (on R-Car V4H), while the register holds only
> > 2 or 4 values.
>
> OK, will add separate patch for fixing
>
> -#define RCANFD_GAFLCFG(ch) (0x009c + (0x04 * ((ch) / 2)))
> +#define RCANFD_GAFLCFG(ch, n) (0x009c + (0x04 * ((ch) / (n))))
>
> Where n is gpriv->info->rnc_per_reg.
I think you better pass gpriv instead of n, i.e. RCANFD_GAFLCFG(gpriv, ch),
for consistency with other macros.
> and SETRNC is simplified to
>
> #define RCANFD_GAFLCFG_SETRNC(gpriv, n, x) \
> (((x) & ((gpriv)->info->num_supported_rules - 1)) << \
> - (reg_gen4(gpriv, 16, 24) - ((n) & 1) * reg_gen4(gpriv, 16, 8)))
> + (32 - (((n) % (gpriv)->info->rnc_per_reg + 1) * (gpriv)->info->rnc_field_width)))
FTR, gpriv->info->rnc_per_reg == 32 / gpriv->info->rnc_field_width.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
