| Message ID | 1390505535-30229-1-git-send-email-paul.gortmaker@windriver.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Quoting Paul Gortmaker (2014-01-23 11:32:15) > Commit 3966c947f45911e093114371462687134d5e8d40 ("clk: max77686: > Refactor driver data handling") added a call to __clk_get_hw() > but this function is not exported (as __<name> is typically a > private/internal thing). > > Hence this driver fails to build modular, and has been causing > allmodconfig build breakage in arch outside of just ARM. Since > the only defconfig that uses it sets it as =y, lets just make > it non-modular for now. > > Mike says that changes are pending to export similar functionality > in the future[1], so at that point in time, it can be returned to > tristate if desired. Hi Paul, I've just sent off my 2nd pull request for 3.14 and it includes the export for __clk_get_hw. Regards, Mike > > [1] https://lkml.org/lkml/2014/1/20/21 > > Reported-by: SeongJae Park <sj38.park@gmail.com> > Cc: SeongJae Park <sj38.park@gmail.com> > Cc: Tomasz Figa <t.figa@samsung.com> > Cc: Kyungmin Park <kyungmin.park@samsung.com> > Cc: Mike Turquette <mturquette@linaro.org> > Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com> > > diff --git a/drivers/clk/Kconfig b/drivers/clk/Kconfig > index efb508a..009fb9b 100644 > --- a/drivers/clk/Kconfig > +++ b/drivers/clk/Kconfig > @@ -40,7 +40,7 @@ config COMMON_CLK_VERSATILE > - Versatile Express > > config COMMON_CLK_MAX77686 > - tristate "Clock driver for Maxim 77686 MFD" > + bool "Clock driver for Maxim 77686 MFD" > depends on MFD_MAX77686 > ---help--- > This driver supports Maxim 77686 crystal oscillator clock. > -- > 1.8.5.2 >
On 14-01-27 04:52 PM, Mike Turquette wrote: > Quoting Paul Gortmaker (2014-01-23 11:32:15) >> Commit 3966c947f45911e093114371462687134d5e8d40 ("clk: max77686: >> Refactor driver data handling") added a call to __clk_get_hw() >> but this function is not exported (as __<name> is typically a >> private/internal thing). >> >> Hence this driver fails to build modular, and has been causing >> allmodconfig build breakage in arch outside of just ARM. Since >> the only defconfig that uses it sets it as =y, lets just make >> it non-modular for now. >> >> Mike says that changes are pending to export similar functionality >> in the future[1], so at that point in time, it can be returned to >> tristate if desired. > > Hi Paul, > > I've just sent off my 2nd pull request for 3.14 and it includes the > export for __clk_get_hw. Sounds good ; I wasn't too partial to how it got fixed, but just wanting to see the ARCH != ARM build fallout go away. Thanks, Paul. -- > > Regards, > Mike > >> >> [1] https://lkml.org/lkml/2014/1/20/21 >> >> Reported-by: SeongJae Park <sj38.park@gmail.com> >> Cc: SeongJae Park <sj38.park@gmail.com> >> Cc: Tomasz Figa <t.figa@samsung.com> >> Cc: Kyungmin Park <kyungmin.park@samsung.com> >> Cc: Mike Turquette <mturquette@linaro.org> >> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com> >> >> diff --git a/drivers/clk/Kconfig b/drivers/clk/Kconfig >> index efb508a..009fb9b 100644 >> --- a/drivers/clk/Kconfig >> +++ b/drivers/clk/Kconfig >> @@ -40,7 +40,7 @@ config COMMON_CLK_VERSATILE >> - Versatile Express >> >> config COMMON_CLK_MAX77686 >> - tristate "Clock driver for Maxim 77686 MFD" >> + bool "Clock driver for Maxim 77686 MFD" >> depends on MFD_MAX77686 >> ---help--- >> This driver supports Maxim 77686 crystal oscillator clock. >> -- >> 1.8.5.2 >>
diff --git a/drivers/clk/Kconfig b/drivers/clk/Kconfig index efb508a..009fb9b 100644 --- a/drivers/clk/Kconfig +++ b/drivers/clk/Kconfig @@ -40,7 +40,7 @@ config COMMON_CLK_VERSATILE - Versatile Express config COMMON_CLK_MAX77686 - tristate "Clock driver for Maxim 77686 MFD" + bool "Clock driver for Maxim 77686 MFD" depends on MFD_MAX77686 ---help--- This driver supports Maxim 77686 crystal oscillator clock.
Commit 3966c947f45911e093114371462687134d5e8d40 ("clk: max77686: Refactor driver data handling") added a call to __clk_get_hw() but this function is not exported (as __<name> is typically a private/internal thing). Hence this driver fails to build modular, and has been causing allmodconfig build breakage in arch outside of just ARM. Since the only defconfig that uses it sets it as =y, lets just make it non-modular for now. Mike says that changes are pending to export similar functionality in the future[1], so at that point in time, it can be returned to tristate if desired. [1] https://lkml.org/lkml/2014/1/20/21 Reported-by: SeongJae Park <sj38.park@gmail.com> Cc: SeongJae Park <sj38.park@gmail.com> Cc: Tomasz Figa <t.figa@samsung.com> Cc: Kyungmin Park <kyungmin.park@samsung.com> Cc: Mike Turquette <mturquette@linaro.org> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>