2020-04-22 15:38:13

by Masahiro Yamada

[permalink] [raw]
Subject: [PATCH] mmc: sdhci-of-at91: make MMC_SDHCI_OF_AT91 depend on HAVE_CLK

If sdhci-of-at91.c is compiled without CONFIG_HAVE_CLK, the line

caps1 |= FIELD_PREP(SDHCI_CLOCK_MUL_MASK, clk_mul);

... emits "FIELD_PREP: value too large for the field" warning.

The compiler seems to decide clk_mul is constant (unsigned int)-1,
because clk_get_rate() returns 0 when CONFIG_HAVE_CLK is disabled.

Add HAVE_CLK to the depenency since this driver does not work without
the clock APIs anyway.

Link: https://lkml.org/lkml/2020/4/17/613
Fixes: linux-next ("mmc: sdhci: use FIELD_GET/PREP for capabilities bit masks")
Reported-by: Randy Dunlap <[email protected]>
Suggested-by: Adrian Hunter <[email protected]>
Signed-off-by: Masahiro Yamada <[email protected]>
---

Ulf,

I do not know how to fill the Fixes tag.
It is currently 8da1ff4f68a2 in linux-next, but I am not sure it is
stable. I just added 'linux-next'.

If you have a preferred way, please modify it.





drivers/mmc/host/Kconfig | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig
index 462b5352fea7..2aee844722d6 100644
--- a/drivers/mmc/host/Kconfig
+++ b/drivers/mmc/host/Kconfig
@@ -171,7 +171,7 @@ config MMC_SDHCI_OF_ASPEED
config MMC_SDHCI_OF_AT91
tristate "SDHCI OF support for the Atmel SDMMC controller"
depends on MMC_SDHCI_PLTFM
- depends on OF
+ depends on OF && HAVE_CLK
help
This selects the Atmel SDMMC driver

--
2.25.1


2020-04-22 16:11:24

by Ulf Hansson

[permalink] [raw]
Subject: Re: [PATCH] mmc: sdhci-of-at91: make MMC_SDHCI_OF_AT91 depend on HAVE_CLK

On Wed, 22 Apr 2020 at 17:34, Masahiro Yamada
<[email protected]> wrote:
>
> If sdhci-of-at91.c is compiled without CONFIG_HAVE_CLK, the line
>
> caps1 |= FIELD_PREP(SDHCI_CLOCK_MUL_MASK, clk_mul);
>
> ... emits "FIELD_PREP: value too large for the field" warning.
>
> The compiler seems to decide clk_mul is constant (unsigned int)-1,
> because clk_get_rate() returns 0 when CONFIG_HAVE_CLK is disabled.
>
> Add HAVE_CLK to the depenency since this driver does not work without
> the clock APIs anyway.
>
> Link: https://lkml.org/lkml/2020/4/17/613
> Fixes: linux-next ("mmc: sdhci: use FIELD_GET/PREP for capabilities bit masks")

I don't know about these, but I just decided to drop them.

> Reported-by: Randy Dunlap <[email protected]>
> Suggested-by: Adrian Hunter <[email protected]>
> Signed-off-by: Masahiro Yamada <[email protected]>

Applied for next, thanks!

Kind regards
Uffe

> ---
>
> Ulf,
>
> I do not know how to fill the Fixes tag.
> It is currently 8da1ff4f68a2 in linux-next, but I am not sure it is
> stable. I just added 'linux-next'.

...and there are no usefull commit hash to use from my tree either, as
I am occasionally rebasing my next branch.


>
> If you have a preferred way, please modify it.
>
>
>
>
>
> drivers/mmc/host/Kconfig | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig
> index 462b5352fea7..2aee844722d6 100644
> --- a/drivers/mmc/host/Kconfig
> +++ b/drivers/mmc/host/Kconfig
> @@ -171,7 +171,7 @@ config MMC_SDHCI_OF_ASPEED
> config MMC_SDHCI_OF_AT91
> tristate "SDHCI OF support for the Atmel SDMMC controller"
> depends on MMC_SDHCI_PLTFM
> - depends on OF
> + depends on OF && HAVE_CLK
> help
> This selects the Atmel SDMMC driver
>
> --
> 2.25.1
>

2020-04-22 16:38:03

by Randy Dunlap

[permalink] [raw]
Subject: Re: [PATCH] mmc: sdhci-of-at91: make MMC_SDHCI_OF_AT91 depend on HAVE_CLK

On 4/22/20 8:34 AM, Masahiro Yamada wrote:
> If sdhci-of-at91.c is compiled without CONFIG_HAVE_CLK, the line
>
> caps1 |= FIELD_PREP(SDHCI_CLOCK_MUL_MASK, clk_mul);
>
> ... emits "FIELD_PREP: value too large for the field" warning.
>
> The compiler seems to decide clk_mul is constant (unsigned int)-1,
> because clk_get_rate() returns 0 when CONFIG_HAVE_CLK is disabled.
>
> Add HAVE_CLK to the depenency since this driver does not work without
> the clock APIs anyway.
>
> Link: https://lkml.org/lkml/2020/4/17/613
> Fixes: linux-next ("mmc: sdhci: use FIELD_GET/PREP for capabilities bit masks")
> Reported-by: Randy Dunlap <[email protected]>
> Suggested-by: Adrian Hunter <[email protected]>
> Signed-off-by: Masahiro Yamada <[email protected]>

Acked-by: Randy Dunlap <[email protected]> # build-tested

Thanks.

> ---
>
> Ulf,
>
> I do not know how to fill the Fixes tag.
> It is currently 8da1ff4f68a2 in linux-next, but I am not sure it is
> stable. I just added 'linux-next'.
>
> If you have a preferred way, please modify it.
>
>
>
>
>
> drivers/mmc/host/Kconfig | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig
> index 462b5352fea7..2aee844722d6 100644
> --- a/drivers/mmc/host/Kconfig
> +++ b/drivers/mmc/host/Kconfig
> @@ -171,7 +171,7 @@ config MMC_SDHCI_OF_ASPEED
> config MMC_SDHCI_OF_AT91
> tristate "SDHCI OF support for the Atmel SDMMC controller"
> depends on MMC_SDHCI_PLTFM
> - depends on OF
> + depends on OF && HAVE_CLK
> help
> This selects the Atmel SDMMC driver
>
>


--
~Randy