From: Claudiu Beznea <[email protected]>
Hardware user manual of RZ/G2L (r01uh0914ej0130-rzg2l-rzg2lc.pdf,
chapter 7.4.7 Procedure for Switching Clocks by the Dynamic Switching
Frequency Selectors) specifies that we need to check CPG_PL2SDHI_DSEL for
SD clock switching status.
Fixes: eaff33646f4cb ("clk: renesas: rzg2l: Add SDHI clk mux support")
Signed-off-by: Claudiu Beznea <[email protected]>
---
drivers/clk/renesas/rzg2l-cpg.c | 14 +++++++++-----
1 file changed, 9 insertions(+), 5 deletions(-)
diff --git a/drivers/clk/renesas/rzg2l-cpg.c b/drivers/clk/renesas/rzg2l-cpg.c
index 47f488387f33..70d1c28ba088 100644
--- a/drivers/clk/renesas/rzg2l-cpg.c
+++ b/drivers/clk/renesas/rzg2l-cpg.c
@@ -188,7 +188,8 @@ static int rzg2l_cpg_sd_clk_mux_set_parent(struct clk_hw *hw, u8 index)
u32 off = GET_REG_OFFSET(hwdata->conf);
u32 shift = GET_SHIFT(hwdata->conf);
const u32 clk_src_266 = 2;
- u32 bitmask;
+ u32 msk, val, bitmask;
+ int ret;
/*
* As per the HW manual, we should not directly switch from 533 MHz to
@@ -203,9 +204,6 @@ static int rzg2l_cpg_sd_clk_mux_set_parent(struct clk_hw *hw, u8 index)
*/
bitmask = (GENMASK(GET_WIDTH(hwdata->conf) - 1, 0) << shift) << 16;
if (index != clk_src_266) {
- u32 msk, val;
- int ret;
-
writel(bitmask | ((clk_src_266 + 1) << shift), priv->base + off);
msk = off ? CPG_CLKSTATUS_SELSDHI1_STS : CPG_CLKSTATUS_SELSDHI0_STS;
@@ -221,7 +219,13 @@ static int rzg2l_cpg_sd_clk_mux_set_parent(struct clk_hw *hw, u8 index)
writel(bitmask | ((index + 1) << shift), priv->base + off);
- return 0;
+ ret = readl_poll_timeout(priv->base + CPG_CLKSTATUS, val,
+ !(val & msk), 100,
+ CPG_SDHI_CLK_SWITCH_STATUS_TIMEOUT_US);
+ if (ret)
+ dev_err(priv->dev, "failed to switch clk source\n");
+
+ return ret;
}
static u8 rzg2l_cpg_sd_clk_mux_get_parent(struct clk_hw *hw)
--
2.39.2
Hi Claudiu,
On Tue, Sep 12, 2023 at 6:52 AM Claudiu <[email protected]> wrote:
> From: Claudiu Beznea <[email protected]>
>
> Hardware user manual of RZ/G2L (r01uh0914ej0130-rzg2l-rzg2lc.pdf,
> chapter 7.4.7 Procedure for Switching Clocks by the Dynamic Switching
> Frequency Selectors) specifies that we need to check CPG_PL2SDHI_DSEL for
> SD clock switching status.
>
> Fixes: eaff33646f4cb ("clk: renesas: rzg2l: Add SDHI clk mux support")
> Signed-off-by: Claudiu Beznea <[email protected]>
Thanks for your patch!
> --- a/drivers/clk/renesas/rzg2l-cpg.c
> +++ b/drivers/clk/renesas/rzg2l-cpg.c
> @@ -188,7 +188,8 @@ static int rzg2l_cpg_sd_clk_mux_set_parent(struct clk_hw *hw, u8 index)
> u32 off = GET_REG_OFFSET(hwdata->conf);
> u32 shift = GET_SHIFT(hwdata->conf);
> const u32 clk_src_266 = 2;
> - u32 bitmask;
> + u32 msk, val, bitmask;
> + int ret;
>
> /*
> * As per the HW manual, we should not directly switch from 533 MHz to
> @@ -203,9 +204,6 @@ static int rzg2l_cpg_sd_clk_mux_set_parent(struct clk_hw *hw, u8 index)
> */
> bitmask = (GENMASK(GET_WIDTH(hwdata->conf) - 1, 0) << shift) << 16;
> if (index != clk_src_266) {
> - u32 msk, val;
> - int ret;
> -
> writel(bitmask | ((clk_src_266 + 1) << shift), priv->base + off);
>
> msk = off ? CPG_CLKSTATUS_SELSDHI1_STS : CPG_CLKSTATUS_SELSDHI0_STS;
> @@ -221,7 +219,13 @@ static int rzg2l_cpg_sd_clk_mux_set_parent(struct clk_hw *hw, u8 index)
>
> writel(bitmask | ((index + 1) << shift), priv->base + off);
>
> - return 0;
> + ret = readl_poll_timeout(priv->base + CPG_CLKSTATUS, val,
> + !(val & msk), 100,
"msk" may be uninitialized.
> + CPG_SDHI_CLK_SWITCH_STATUS_TIMEOUT_US);
> + if (ret)
> + dev_err(priv->dev, "failed to switch clk source\n");
> +
> + return ret;
This is now (supposed to be) doing the same thing twice, once using
clk_src_266, and then again with the wanted index, so why not introduce
a small helper? That would have avoided the uninitialized variable, too.
I know you're rewriting this code in "[PATCH 18/37] clk: renesas:
rzg2l: refactor sd mux driver", but even after that, you always do
a register write before calling rzg2l_cpg_wait_clk_update_done(),
so it may still be a net win.
> }
>
> static u8 rzg2l_cpg_sd_clk_mux_get_parent(struct clk_hw *hw)
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
Hi, Geert,
On 14.09.2023 14:42, Geert Uytterhoeven wrote:
> Hi Claudiu,
>
> On Tue, Sep 12, 2023 at 6:52 AM Claudiu <[email protected]> wrote:
>> From: Claudiu Beznea <[email protected]>
>>
>> Hardware user manual of RZ/G2L (r01uh0914ej0130-rzg2l-rzg2lc.pdf,
>> chapter 7.4.7 Procedure for Switching Clocks by the Dynamic Switching
>> Frequency Selectors) specifies that we need to check CPG_PL2SDHI_DSEL for
>> SD clock switching status.
>>
>> Fixes: eaff33646f4cb ("clk: renesas: rzg2l: Add SDHI clk mux support")
>> Signed-off-by: Claudiu Beznea <[email protected]>
>
> Thanks for your patch!
>
>> --- a/drivers/clk/renesas/rzg2l-cpg.c
>> +++ b/drivers/clk/renesas/rzg2l-cpg.c
>> @@ -188,7 +188,8 @@ static int rzg2l_cpg_sd_clk_mux_set_parent(struct clk_hw *hw, u8 index)
>> u32 off = GET_REG_OFFSET(hwdata->conf);
>> u32 shift = GET_SHIFT(hwdata->conf);
>> const u32 clk_src_266 = 2;
>> - u32 bitmask;
>> + u32 msk, val, bitmask;
>> + int ret;
>>
>> /*
>> * As per the HW manual, we should not directly switch from 533 MHz to
>> @@ -203,9 +204,6 @@ static int rzg2l_cpg_sd_clk_mux_set_parent(struct clk_hw *hw, u8 index)
>> */
>> bitmask = (GENMASK(GET_WIDTH(hwdata->conf) - 1, 0) << shift) << 16;
>> if (index != clk_src_266) {
>> - u32 msk, val;
>> - int ret;
>> -
>> writel(bitmask | ((clk_src_266 + 1) << shift), priv->base + off);
>>
>> msk = off ? CPG_CLKSTATUS_SELSDHI1_STS : CPG_CLKSTATUS_SELSDHI0_STS;
>> @@ -221,7 +219,13 @@ static int rzg2l_cpg_sd_clk_mux_set_parent(struct clk_hw *hw, u8 index)
>>
>> writel(bitmask | ((index + 1) << shift), priv->base + off);
>>
>> - return 0;
>> + ret = readl_poll_timeout(priv->base + CPG_CLKSTATUS, val,
>> + !(val & msk), 100,
>
> "msk" may be uninitialized.
Indeed! I'll update it in next version.
>
>> + CPG_SDHI_CLK_SWITCH_STATUS_TIMEOUT_US);
>> + if (ret)
>> + dev_err(priv->dev, "failed to switch clk source\n");
>> +
>> + return ret;
>
> This is now (supposed to be) doing the same thing twice, once using
> clk_src_266, and then again with the wanted index, so why not introduce
> a small helper? That would have avoided the uninitialized variable, too.
Initially I thought about it but I found it too much for this stage as it
is only about the readl_poll_timeout() and the debug message. I may keep
the debug message in a local variable if you think worth it (but FMPOV it
the code will look a bit... unusual). Moreover, as the code is rewritten in
patch "[PATCH 18/37] clk: renesas:
rzg2l: refactor sd mux driver" I thought it doesn't worth introducing a new
helper in this patch.
Thank you,
Claudiu Beznea
>
> I know you're rewriting this code in "[PATCH 18/37] clk: renesas:
> rzg2l: refactor sd mux driver", but even after that, you always do
> a register write before calling rzg2l_cpg_wait_clk_update_done(),
> so it may still be a net win.
>
>> }
>>
>> static u8 rzg2l_cpg_sd_clk_mux_get_parent(struct clk_hw *hw)
>
> Gr{oetje,eeting}s,
>
> Geert
>