The clock for UARTs 0 through 2 is UART012, the clock for UARTs 3 through
7 is UART34567.
For UART012, we stop the clock driver from changing the clock rate. This
is because the Synopsys UART driver simply sets the reference clock to 16x
the baud rate, but doesn't check if the actual rate is within the required
tolerance. The RZ/N1 clock divider can't provide this (we have to rely on
the UART's internal divider to set the correct clock rate), so you end up
with a clock rate that is way off what you wanted.
In addition, since the clock is shared between multiple UARTs, you don't
want the driver trying to change the clock rate as it may affect the other
UARTs (which may not have been configured yet, so you don't know what baud
rate they will use). Normally, the clock rate is set early on before Linux
to some very high rate that supports all of the clock rates you want.
This change stops the UART34567 clock rate from changing for the same
reasons.
Signed-off-by: Phil Edworthy <[email protected]>
---
drivers/clk/renesas/r9a06g032-clocks.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/clk/renesas/r9a06g032-clocks.c b/drivers/clk/renesas/r9a06g032-clocks.c
index a0b6ecd..6d2b568 100644
--- a/drivers/clk/renesas/r9a06g032-clocks.c
+++ b/drivers/clk/renesas/r9a06g032-clocks.c
@@ -539,7 +539,8 @@ r9a06g032_div_round_rate(struct clk_hw *hw,
* several uarts attached to this divider, and changing this impacts
* everyone.
*/
- if (clk->index == R9A06G032_DIV_UART) {
+ if (clk->index == R9A06G032_DIV_UART ||
+ clk->index == R9A06G032_DIV_P2_PG) {
pr_devel("%s div uart hack!\n", __func__);
return clk_get_rate(hw->clk);
}
--
2.7.4
Hi Phil,
On Fri, Aug 31, 2018 at 1:26 PM Phil Edworthy <[email protected]> wrote:
> The clock for UARTs 0 through 2 is UART012, the clock for UARTs 3 through
> 7 is UART34567.
> For UART012, we stop the clock driver from changing the clock rate. This
> is because the Synopsys UART driver simply sets the reference clock to 16x
> the baud rate, but doesn't check if the actual rate is within the required
> tolerance. The RZ/N1 clock divider can't provide this (we have to rely on
> the UART's internal divider to set the correct clock rate), so you end up
> with a clock rate that is way off what you wanted.
>
> In addition, since the clock is shared between multiple UARTs, you don't
> want the driver trying to change the clock rate as it may affect the other
> UARTs (which may not have been configured yet, so you don't know what baud
> rate they will use). Normally, the clock rate is set early on before Linux
> to some very high rate that supports all of the clock rates you want.
>
> This change stops the UART34567 clock rate from changing for the same
> reasons.
>
> Signed-off-by: Phil Edworthy <[email protected]>
Reviewed-by: Geert Uytterhoeven <[email protected]>
i.e. will queue in clk-renesas-for-v4.20.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]
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