clk: si5351: Fix clkout rate computation.
PATCH v2 1/2 clk: si5351: Fix clkout rate computation.
Rate was incorrectly computed because we read from wrong divider register.
--- drivers/clk/clk-si5351.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/clk/clk-si5351.c +1 -1
@@ -932,7 +932,7 @@ static unsigned long si5351_clkout_recalc_rate(struct clk_hw *hw,unsigned char reg;unsigned char rdiv; -if (hwdata->num >5 ) +if (hwdata->num <=5 ) reg =si5351_msynth_params_address (hwdata->num) +2 ;else reg =SI5351_CLK6_7_OUTPUT_DIVIDER ;
Signature · 1 line
1.7.9.5
PATCH v2 2/2 clk: si5351: Set initial clkout rate when defined in platform data.
---
Changes v1->v2:
- update commit message and comment to refer to platform data not DT
drivers/clk/clk-si5351.c | 10 ++++++++++ 1 file changed, 10 insertions(+)
drivers/clk/clk-si5351.c +10 -0
@@ -1477,6 +1477,16 @@ static int si5351_i2c_probe(struct i2c_client *client,return -EINVAL ; } drvdata->onecell.clks[n] = clk; + +/* set initial clkout rate */ +if (pdata->clkout[n].rate !=0 ) { +int ret; + ret =clk_set_rate (clk, pdata->clkout[n].rate); +if (ret !=0 ) { +dev_err (&client->dev,"Cannot set rate : %d\n" , + ret); + } + } } ret = of_clk_add_provider(client->dev.of_node, of_clk_src_onecell_get,
Signature · 1 line
1.7.9.5
Re: [PATCH v2 2/2] clk: si5351: Set initial clkout rate when defined in platform data.
Quoted 4 lines · Marek Belisko
Signed-off-by Marek Belisko<[email protected]>---
Changesv1->v2:
- update commit message and comment to refer to platform data not DT
Marek,
except that the commit msg itself is very short (there is no!) you can add my
Next time sending patches for clk-si5351, please ensure you are using get_maintainers.pl and add me to the list of people to Cc also.
Sebastian
Quoted 24 lines
drivers/clk/clk-si5351.c | 10 ++++++++++ 1 file changed, 10 insertions(+)drivers/clk/clk-si5351.c +10
-0@@ -1477,6 +1477,16 @@ static int si5351_i2c_probe(struct i2c_client *client,return -EINVAL ; } drvdata->onecell.clks[n] = clk; + +/* set initial clkout rate */ +if (pdata->clkout[n].rate !=0 ) { +int ret; + ret =clk_set_rate (clk, pdata->clkout[n].rate); +if (ret !=0 ) { +dev_err (&client->dev,"Cannot set rate : %d\n" , + ret); + } + } } ret = of_clk_add_provider(client->dev.of_node, of_clk_src_onecell_get,
PATCH v2 1/2 clk: si5351: Fix clkout rate computation.
In reply to Marek BeliskoQuoting Marek Belisko (2013-05-02 22:53:22)
Quoted 4 lines
Rate was incorrectly computed because we read from wrong divider register.
Signed-off-by Marek Belisko <[email protected]>
Do either of these patches need to go into clk-fixes? If so please be explicit about it in the changelog, the section under the three dashes.
Regards,
Mike
Quoted 19 lines
--- drivers/clk/clk-si5351.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)drivers/clk/clk-si5351.c +1
-1@@ -932,7 +932,7 @@ static unsigned long si5351_clkout_recalc_rate(struct clk_hw *hw,unsigned char reg;unsigned char rdiv; -if (hwdata->num >5 ) +if (hwdata->num <=5 ) reg =si5351_msynth_params_address (hwdata->num) +2 ;else reg =SI5351_CLK6_7_OUTPUT_DIVIDER ;Signature · 1 line
1.7.9.5
Quoted 8 lines · Mike Turquette
Quoting Marek Belisko (2013-05-02 22:53:22)
Quoted 4 lines
Rate was incorrectly computed because we read from wrong divider register.
Signed-off-by Marek Belisko <[email protected]>Do either of these patches need to go into clk-fixes? If so please be explicit about it in the changelog, the section under the three dashes.
Yes I believe they should go to clk-fixes (probably also to linux-stable).
Quoted 16 lines
Regards,
MikeQuoted 12 lines
--- drivers/clk/clk-si5351.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)drivers/clk/clk-si5351.c +0
-0@@ -932,7 +932,7 @@ static unsigned long si5351_clkout_recalc_rate(struct clk_hw *hw,unsigned char reg;unsigned char rdiv;
Quoted 8 lines
Quoted 7 lines
drivers/clk/clk-si5351.c +1
-2-if (hwdata->num >5 ) +if (hwdata->num <=5 ) reg =si5351_msynth_params_address (hwdata->num) +2 ;else reg =SI5351_CLK6_7_OUTPUT_DIVIDER ; --1.7.9.5
Regards,
Marek
Signature · 16 lines
Marek Belisko Software Developer StreamUnlimited Engineering GmbH Gutheil Schodergasse 8-12 A-1100 Vienna, Austria Office: +421 267200087 e-mail: marek.belisko at streamunlimited.com http://www.streamunlimited.com Meet us at: IFA - Berlin, 6-11 September CEDIA - Denver, 25-28 September
PATCH v2 1/2 clk: si5351: Fix clkout rate computation.
Quoting Marek Belisko (2013-05-16 00:53:26)
Quoted 10 lines · Mike Turquette
Quoted 8 lines
Quoting Marek Belisko (2013-05-02 22:53:22)
Quoted 4 lines
Rate was incorrectly computed because we read from wrong divider register.
Signed-off-by Marek Belisko <[email protected]>Do either of these patches need to go into clk-fixes? If so please be explicit about it in the changelog, the section under the three dashes.
Yes I believe they should go to clk-fixes (probably also to linux-stable).
Pulled both into clk-fixes. I added the missing changelog to patch #2 because I am such a nice guy.
Regards,
Mike
Quoted 45 lines
Quoted 16 lines
Regards,
MikeQuoted 12 lines
--- drivers/clk/clk-si5351.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)drivers/clk/clk-si5351.c +0
-0@@ -932,7 +932,7 @@ static unsigned long si5351_clkout_recalc_rate(struct clk_hw *hw,unsigned char reg;unsigned char rdiv;Quoted 8 lines
Quoted 7 lines
drivers/clk/clk-si5351.c +1
-2-if (hwdata->num >5 ) +if (hwdata->num <=5 ) reg =si5351_msynth_params_address (hwdata->num) +2 ;else reg =SI5351_CLK6_7_OUTPUT_DIVIDER ; --1.7.9.5
Regards,
Marek
Signature · 16 lines
Marek Belisko Software Developer StreamUnlimited Engineering GmbH Gutheil Schodergasse 8-12 A-1100 Vienna, Austria Office: +421 267200087 e-mail: marek.belisko at streamunlimited.com http://www.streamunlimited.com Meet us at: IFA - Berlin, 6-11 September CEDIA - Denver, 25-28 September