diff options
author | Uwe Kleine-König <u.kleine-koenig@pengutronix.de> | 2011-04-04 14:06:12 (GMT) |
---|---|---|
committer | Sascha Hauer <s.hauer@pengutronix.de> | 2011-04-12 07:48:02 (GMT) |
commit | f61b9fc27e5b61dbc330696f040cc66ba1dbcbaa (patch) | |
tree | 96ed9c645dd14f70548601c9717875d651bb27df /kernel | |
parent | 32a90b6e65792260d6212ac52e8f5be140b6f5be (diff) | |
download | linux-fsl-qoriq-f61b9fc27e5b61dbc330696f040cc66ba1dbcbaa.tar.xz |
ARM: mxs/clock-mx28: fix up name##_set_rate
For the lcdif clock get_rate looks as follows:
read div from HW_CLKCTRL_DIS_LCDIF.DIV
return clk_get_rate(clk->parent) / div
with clk->parent being ref_pix_clk on my system.
ref_pix_clk's rate depends on HW_CLKCTRL_FRAC1.PIXFRAC.
The set_rate function for lcdif does:
parent_rate = clk_get_rate(clk->parent);
based on that calculate frac and div such that
parent_rate * 18 / frac / div is near the requested rate.
HW_CLKCTRL_FRAC1.PIXFRAC is updated with frac
HW_CLKCTRL_DIS_LCDIF.DIV is updated with div
For this calculation to be correct parent_rate needs to be
initialized not with the clock rate of lcdif's parent (i.e. ref_pix) but
that of its grandparent (i.e. ref_pix' parent == pll0_clk).
The obvious downside of this patch is that now set_rate(lcdif) changes
its parent's rate, too. Still this is better than a wrong rate.
Acked-by: Shawn Guo <shawn.guo@freescale.com>
LAKML-Reference: 20110225084950.GA13684@S2101-09.ap.freescale.net
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Diffstat (limited to 'kernel')
0 files changed, 0 insertions, 0 deletions