Re: [PATCH 1/2] i2c: tegra: Add logic to support different register offsets

[Kartik Rajput]

On 24/10/25 20:43, Jon Hunter wrote:
> On 01/10/2025 16:36, Kartik Rajput wrote:
> > Tegra410 use different offsets for existing I2C registers, update
> > the logic to use appropriate offsets per SoC.
> >
> > Signed-off-by: Kartik Rajput <kkartik@xxxxxxxxxx>
> > ---
> >   drivers/i2c/busses/i2c-tegra.c | 499 ++++++++++++++++++++++-----------
> >   1 file changed, 334 insertions(+), 165 deletions(-)
> >
> > diff --git a/drivers/i2c/busses/i2c-tegra.c b/drivers/i2c/busses/i2c-tegra.c
> > index 038809264526..1e26d67cbd30 100644
> > --- a/drivers/i2c/busses/i2c-tegra.c
> > +++ b/drivers/i2c/busses/i2c-tegra.c
>
> ...
>
> > +static const struct tegra_i2c_regs tegra20_i2c_regs_vi = {
> > +    .cnfg = 0x0c00 + (0x000 << 2),
> > +    .status = 0x0c00 + (0x01c << 2),
> > +    .sl_cnfg = 0x0c00 + (0x020 << 2),
> > +    .sl_addr1 = 0x0c00 + (0x02c << 2),
> > +    .sl_addr2 = 0x0c00 + (0x030 << 2),
> > +    .tlow_sext = 0x0c00 + (0x034 << 2),
> > +    .tx_fifo = 0x0c00 + (0x050 << 2),
> > +    .rx_fifo = 0x0c00 + (0x054 << 2),
> > +    .packet_transfer_status = 0x0c00 + (0x058 << 2),
> > +    .fifo_control = 0x0c00 + (0x05c << 2),
> > +    .fifo_status = 0x0c00 + (0x060 << 2),
> > +    .int_mask = 0x0c00 + (0x064 << 2),
> > +    .int_status = 0x0c00 + (0x068 << 2),
> > +    .clk_divisor = 0x0c00 + (0x06c << 2),
> > +    .bus_clear_cnfg = 0x0c00 + (0x084 << 2),
> > +    .bus_clear_status = 0x0c00 + (0x088 << 2),
> > +    .config_load = 0x0c00 + (0x08c << 2),
> > +    .clken_override = 0x0c00 + (0x090 << 2),
> > +    .interface_timing_0 = 0x0c00 + (0x094 << 2),
> > +    .interface_timing_1 = 0x0c00 + (0x098 << 2),
> > +    .hs_interface_timing_0 = 0x0c00 + (0x09c << 2),
> > +    .hs_interface_timing_1 = 0x0c00 + (0x0a0 << 2),
> > +    .master_reset_cntrl = 0x0c00 + (0x0a8 << 2),
> > +    .mst_fifo_control = 0x0c00 + (0x0b4 << 2),
> > +    .mst_fifo_status = 0x0c00 + (0x0b8 << 2),
> > +    .sw_mutex = 0x0c00 + (0x0ec << 2),
>
> Why do we define all the above with '<< 2'? Seems odd.
>
> Jon

Ack. I will update this with the calculated values.

Thanks,
Kartik