[PATCH v3 3/6] OMAP3+: use DPLL's round_rate when setting rate
From: paul@pwsan.com (Paul Walmsley)
Date: 2011-09-28 07:02:12
Also in:
linux-omap
Subsystem:
arm port, omap2+ support, the rest · Maintainers:
Russell King, Aaro Koskinen, Andreas Kemnade, Kevin Hilman, Roger Quadros, Tony Lindgren, Linus Torvalds
Hi, On Fri, 16 Sep 2011, Jon Hunter wrote:
From: Mike Turquette <redacted> omap3_noncore_dpll_set_rate uses omap2_dpll_round_rate explicitly. Instead use the struct clk pointer's round_rate function to allow for DPLL's with special needs. Also the rounded rate can differ from target rate, so to better reflect reality set clk->rate equal to the rounded rate when setting DPLL frequency. This avoids issues where the DPLL frequency is slightly different than what debugfs clock tree reports using the old target rate. An example of both of these needs is DPLL_ABE on OMAP4 which can have a 4x multiplier on top of the usual MN dividers depending on register settings. This requires a special round_rate function that might yield a rate different from the initial target. Signed-off-by: Mike Turquette <redacted> Signed-off-by: Jon Hunter <redacted>
The two separate changes in this patch have been separated out into two patches - both included below. Please let me know if you have any comments; otherwise, I'll queue for 3.2. - Paul From: Mike Turquette <redacted> Date: Wed, 28 Sep 2011 00:00:32 -0600 Subject: [PATCH] ARM: OMAP3+: dpll: use DPLL's round_rate when setting rate omap3_noncore_dpll_set_rate uses omap2_dpll_round_rate explicitly. Instead use the struct clk pointer's round_rate function to allow for DPLL's with special needs. An example of a clock that requires this is DPLL_ABE on OMAP4 which can have a 4x multiplier on top of the usual MN dividers depending on register settings. This requires a special round_rate function that might yield a rate different from the initial target. Signed-off-by: Mike Turquette <redacted> Signed-off-by: Jon Hunter <redacted> [paul at pwsan.com: split rate assignment portion into a separate patch] Signed-off-by: Paul Walmsley <paul@pwsan.com> --- arch/arm/mach-omap2/dpll3xxx.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/arch/arm/mach-omap2/dpll3xxx.c b/arch/arm/mach-omap2/dpll3xxx.c
index f77022b..6b0fa37 100644
--- a/arch/arm/mach-omap2/dpll3xxx.c
+++ b/arch/arm/mach-omap2/dpll3xxx.c@@ -455,7 +455,7 @@ int omap3_noncore_dpll_set_rate(struct clk *clk, unsigned long rate) new_parent = dd->clk_bypass; } else { if (dd->last_rounded_rate != rate) - omap2_dpll_round_rate(clk, rate); + clk->round_rate(clk, rate); if (dd->last_rounded_rate == 0) return -EINVAL;
--
1.7.6.3
--------------------
From: Mike Turquette <mturquette@ti.com>
Date: Wed, 28 Sep 2011 00:00:32 -0600
Subject: [PATCH] ARM: OMAP3+: dpll: assign clk rate from rounded rate during
rate set
The rounded rate can differ from target rate, so to better reflect
reality set clk->rate equal to the rounded rate when setting DPLL frequency.
This avoids issues where the DPLL frequency is slightly different than what
debugfs clock tree reports using the old target rate.
An example of a clock that requires this is DPLL_ABE on OMAP4 which
can have a 4x multiplier on top of the usual MN dividers depending on
register settings. This requires a special round_rate function that
might yield a rate different from the initial target.
Signed-off-by: Mike Turquette <mturquette@ti.com>
Signed-off-by: Jon Hunter <jon-hunter@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
---
arch/arm/mach-omap2/dpll3xxx.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/arch/arm/mach-omap2/dpll3xxx.c b/arch/arm/mach-omap2/dpll3xxx.c
index 6b0fa37..73a1595 100644
--- a/arch/arm/mach-omap2/dpll3xxx.c
+++ b/arch/arm/mach-omap2/dpll3xxx.c
@@ -455,7 +455,7 @@ int omap3_noncore_dpll_set_rate(struct clk *clk, unsigned long rate)
new_parent = dd->clk_bypass;
} else {
if (dd->last_rounded_rate != rate)
- clk->round_rate(clk, rate);
+ rate = clk->round_rate(clk, rate);
if (dd->last_rounded_rate == 0)
return -EINVAL;
--
1.7.6.3