author | Sekhar Nori <nsekhar@ti.com> | |
Thu, 22 Oct 2009 09:42:16 +0000 (15:12 +0530) | ||
committer | Kevin Hilman <khilman@deeprootsystems.com> | |
Wed, 25 Nov 2009 18:21:37 +0000 (10:21 -0800) | ||
commit | 13d5e27a4482e43bea9073706033c84cd873b5ca | |
tree | 2b66f5238dce2563741bf0524b2d731a7dc10ded | tree | snapshot (tar.xz tar.gz zip) |
parent | 5aeb15aafc0c378d4ae2aff7b5d5c686793b8a0e | commit | diff |
davinci: DA850/OMAP-L138: avoid using separate initcall for initializing regulator
Using a device_initcall() for initializing the voltage regulator
on DA850 is not such a good idea because it gets called for all
platforms - even those who do not have a regulator implemented.
This leads to a big fat warning message during boot-up when
regulator cannot be found.
Instead, tie initialization of voltage regulator to cpufreq init.
Define a platform specific init call which in case of DA850 gets
used for initializing the regulator. On other future platforms it
can be used for other purposes.
Signed-off-by: Sekhar Nori <nsekhar@ti.com>
Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com>
Using a device_initcall() for initializing the voltage regulator
on DA850 is not such a good idea because it gets called for all
platforms - even those who do not have a regulator implemented.
This leads to a big fat warning message during boot-up when
regulator cannot be found.
Instead, tie initialization of voltage regulator to cpufreq init.
Define a platform specific init call which in case of DA850 gets
used for initializing the regulator. On other future platforms it
can be used for other purposes.
Signed-off-by: Sekhar Nori <nsekhar@ti.com>
Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com>
arch/arm/mach-davinci/cpufreq.c | diff | blob | history | |
arch/arm/mach-davinci/da850.c | diff | blob | history | |
arch/arm/mach-davinci/include/mach/cpufreq.h | diff | blob | history |