[ARM] 3990/1: i.MX/MX1 more precise PLL decode

The future high resolution support inclusion utilizes
imx_decode_pll() in timer base frequency computation.
This use requires more precise computation without
discarding 10 bits by shifting left.

Signed-off-by: Pavel Pisa <pisa@cmp.felk.cvut.cz>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

diff --git a/arch/arm/mach-imx/generic.c b/arch/arm/mach-imx/generic.c
index 12ea58a..b5aa49d 100644 (file)
--- a/arch/arm/mach-imx/generic.c
+++ b/arch/arm/mach-imx/generic.c
@@ -104,6 +104,9 @@ EXPORT_SYMBOL(imx_gpio_mode);
  */
 static unsigned int imx_decode_pll(unsigned int pll)
 {
+       unsigned long long ll;
+       unsigned long quot;
+
        u32 mfi = (pll >> 10) & 0xf;
        u32 mfn = pll & 0x3ff;
        u32 mfd = (pll >> 16) & 0x3ff;
@@ -112,7 +115,11 @@ static unsigned int imx_decode_pll(unsigned int pll)
 
        mfi = mfi <= 5 ? 5 : mfi;
 
-       return (2 * (f_ref>>10) * ( (mfi<<10) + (mfn<<10) / (mfd+1) )) / (pd+1);
+       ll = 2 * (unsigned long long)f_ref * ( (mfi<<16) + (mfn<<16) / (mfd+1) );
+       quot = (pd+1) * (1<<16);
+       ll += quot / 2;
+       do_div(ll, quot);
+       return (unsigned int) ll;
 }
 
 unsigned int imx_get_system_clk(void)