On 5/3/2012 4:11 AM, Mike Turquette wrote:

quoted

I will take this patch in before sending my request to Arnd.  In your
patch 0/8 you ask me to take in the series, but it looks like not all of
the patches are going through me.  For instance patch 2/8 certainly
looks destined for Russell.

I think you only need to take this patch, i.e. 2/8.
Others will go through Arnd directly, keeping your branch as an dependency
branch for it.

quoted

After fixing up this patch can you resend a clean series which is only
headed for clk-next?

So, i would be sending this patch only after applying the last fixup
(hopefully :) ).

@Sascha: Please see if this patch is sufficient.

---
 drivers/clk/clk-fixed-factor.c |    9 ++++-----
 1 files changed, 4 insertions(+), 5 deletions(-)

diff --git a/drivers/clk/clk-fixed-factor.c b/drivers/clk/clk-fixed-factor.c
index 9afbdc4..8d25c2e 100644
--- a/drivers/clk/clk-fixed-factor.c
+++ b/drivers/clk/clk-fixed-factor.c

@@ -37,16 +37,15 @@ static long clk_factor_round_rate(struct clk_hw *hw, unsigned long rate,
 {
        struct clk_fixed_factor *fix = to_clk_fixed_factor(hw);
 
-       if (prate) {
+       if (__clk_get_flags(hw->clk) & CLK_SET_RATE_PARENT) {
                unsigned long best_parent;
+
                best_parent = (rate / fix->mult) * fix->div;
                *prate = __clk_round_rate(__clk_get_parent(hw->clk),
                                best_parent);
-               return (*prate / fix->div) * fix->mult;
-       } else {
-               return (__clk_get_rate(__clk_get_parent(hw->clk)) / fix->div) *
-                       fix->mult;
        }
+
+       return (*prate / fix->div) * fix->mult;
 }