clk: tegra: make tegra_clocks_apply_init_table() arch_initcall

tegra_clocks_apply_init_table() needs to be called after the udelay
loop has been calibrated (see commit
441f199a37cfd66c5dd8dd45490bd3ea6971117d ("clk: tegra: defer
application of init table") for why that is).  On existing Tegra SoCs
this was done by calling tegra_clocks_apply_init_table() from
tegra_dt_init(). To make this also work on ARM64, we need to change
this into an initcall. tegra_dt_init() is called from
customize_machine which is an arch_initcall. Therefore this should
also work on existing 32bit Tegra SoCs.

Tested on Tegra20 (ventana), Tegra30 (beaverboard), Tegra124 (jetson TK1) and
Tegra132.

Signed-off-by: Peter De Schrijver <pdeschrijver@nvidia.com>
[paul@pwsan.com: tweaked the commit message]
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Paul Walmsley <pwalmsley@nvidia.com>
Cc: Thierry Reding <treding@nvidia.com>
Cc: Prashant Gaikwad <pgaikwad@nvidia.com>
Cc: Mike Turquette <mturquette@linaro.org>
Cc: Stephen Boyd <sboyd@codeaurora.org>
Cc: Stephen Warren <swarren@wwwdotorg.org>
Cc: Thierry Reding <thierry.reding@gmail.com>
Cc: Alexandre Courbot <gnurou@gmail.com>

1 files changed, 0 insertions, 2 deletions

diff --git a/include/linux/clk/tegra.h b/include/linux/clk/tegra.h
index 3ca9fca827a2..19c4208f4752 100644
--- a/include/linux/clk/tegra.h
+++ b/include/linux/clk/tegra.h
@@ -120,6 +120,4 @@ static inline void tegra_cpu_clock_resume(void)
 }
 #endif
 
-void tegra_clocks_apply_init_table(void);
-
 #endif /* __LINUX_CLK_TEGRA_H_ */