linux/arch/arm/mach-omap2/clock.c

1025 lines
26 KiB
C
Raw Normal View History

/*
* linux/arch/arm/mach-omap2/clock.c
*
* Copyright (C) 2005-2008 Texas Instruments, Inc.
* Copyright (C) 2004-2008 Nokia Corporation
*
* Contacts:
* Richard Woodruff <r-woodruff2@ti.com>
* Paul Walmsley
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2 as
* published by the Free Software Foundation.
*/
#undef DEBUG
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/device.h>
#include <linux/list.h>
#include <linux/errno.h>
#include <linux/delay.h>
#include <linux/clk.h>
#include <linux/io.h>
#include <linux/bitops.h>
#include <mach/clock.h>
#include <mach/clockdomain.h>
#include <mach/cpu.h>
#include <asm/div64.h>
#include <mach/sdrc.h>
#include "sdrc.h"
#include "clock.h"
#include "prm.h"
#include "prm-regbits-24xx.h"
#include "cm.h"
#include "cm-regbits-24xx.h"
#include "cm-regbits-34xx.h"
#define MAX_CLOCK_ENABLE_WAIT 100000
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
/* DPLL rate rounding: minimum DPLL multiplier, divider values */
#define DPLL_MIN_MULTIPLIER 1
#define DPLL_MIN_DIVIDER 1
/* Possible error results from _dpll_test_mult */
#define DPLL_MULT_UNDERFLOW -1
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
/*
* Scale factor to mitigate roundoff errors in DPLL rate rounding.
* The higher the scale factor, the greater the risk of arithmetic overflow,
* but the closer the rounded rate to the target rate. DPLL_SCALE_FACTOR
* must be a power of DPLL_SCALE_BASE.
*/
#define DPLL_SCALE_FACTOR 64
#define DPLL_SCALE_BASE 2
#define DPLL_ROUNDING_VAL ((DPLL_SCALE_BASE / 2) * \
(DPLL_SCALE_FACTOR / DPLL_SCALE_BASE))
/* DPLL valid Fint frequency band limits - from 34xx TRM Section 4.7.6.2 */
#define DPLL_FINT_BAND1_MIN 750000
#define DPLL_FINT_BAND1_MAX 2100000
#define DPLL_FINT_BAND2_MIN 7500000
#define DPLL_FINT_BAND2_MAX 21000000
/* _dpll_test_fint() return codes */
#define DPLL_FINT_UNDERFLOW -1
#define DPLL_FINT_INVALID -2
u8 cpu_mask;
/*-------------------------------------------------------------------------
* OMAP2/3 specific clock functions
*-------------------------------------------------------------------------*/
/**
* _omap2xxx_clk_commit - commit clock parent/rate changes in hardware
* @clk: struct clk *
*
* If @clk has the DELAYED_APP flag set, meaning that parent/rate changes
* don't take effect until the VALID_CONFIG bit is written, write the
* VALID_CONFIG bit and wait for the write to complete. No return value.
*/
static void _omap2xxx_clk_commit(struct clk *clk)
{
if (!cpu_is_omap24xx())
return;
if (!(clk->flags & DELAYED_APP))
return;
prm_write_mod_reg(OMAP24XX_VALID_CONFIG, OMAP24XX_GR_MOD,
OMAP24XX_PRCM_CLKCFG_CTRL_OFFSET);
/* OCP barrier */
prm_read_mod_reg(OMAP24XX_GR_MOD, OMAP24XX_PRCM_CLKCFG_CTRL_OFFSET);
}
/*
* _dpll_test_fint - test whether an Fint value is valid for the DPLL
* @clk: DPLL struct clk to test
* @n: divider value (N) to test
*
* Tests whether a particular divider @n will result in a valid DPLL
* internal clock frequency Fint. See the 34xx TRM 4.7.6.2 "DPLL Jitter
* Correction". Returns 0 if OK, -1 if the enclosing loop can terminate
* (assuming that it is counting N upwards), or -2 if the enclosing loop
* should skip to the next iteration (again assuming N is increasing).
*/
static int _dpll_test_fint(struct clk *clk, u8 n)
{
struct dpll_data *dd;
long fint;
int ret = 0;
dd = clk->dpll_data;
/* DPLL divider must result in a valid jitter correction val */
fint = clk->parent->rate / (n + 1);
if (fint < DPLL_FINT_BAND1_MIN) {
pr_debug("rejecting n=%d due to Fint failure, "
"lowering max_divider\n", n);
dd->max_divider = n;
ret = DPLL_FINT_UNDERFLOW;
} else if (fint > DPLL_FINT_BAND1_MAX &&
fint < DPLL_FINT_BAND2_MIN) {
pr_debug("rejecting n=%d due to Fint failure\n", n);
ret = DPLL_FINT_INVALID;
} else if (fint > DPLL_FINT_BAND2_MAX) {
pr_debug("rejecting n=%d due to Fint failure, "
"boosting min_divider\n", n);
dd->min_divider = n;
ret = DPLL_FINT_INVALID;
}
return ret;
}
/**
* omap2_init_clk_clkdm - look up a clockdomain name, store pointer in clk
* @clk: OMAP clock struct ptr to use
*
* Convert a clockdomain name stored in a struct clk 'clk' into a
* clockdomain pointer, and save it into the struct clk. Intended to be
* called during clk_register(). No return value.
*/
void omap2_init_clk_clkdm(struct clk *clk)
{
struct clockdomain *clkdm;
if (!clk->clkdm_name)
return;
clkdm = clkdm_lookup(clk->clkdm_name);
if (clkdm) {
pr_debug("clock: associated clk %s to clkdm %s\n",
clk->name, clk->clkdm_name);
clk->clkdm = clkdm;
} else {
pr_debug("clock: could not associate clk %s to "
"clkdm %s\n", clk->name, clk->clkdm_name);
}
}
/**
* omap2_init_clksel_parent - set a clksel clk's parent field from the hardware
* @clk: OMAP clock struct ptr to use
*
* Given a pointer to a source-selectable struct clk, read the hardware
* register and determine what its parent is currently set to. Update the
* clk->parent field with the appropriate clk ptr.
*/
void omap2_init_clksel_parent(struct clk *clk)
{
const struct clksel *clks;
const struct clksel_rate *clkr;
u32 r, found = 0;
if (!clk->clksel)
return;
r = __raw_readl(clk->clksel_reg) & clk->clksel_mask;
r >>= __ffs(clk->clksel_mask);
for (clks = clk->clksel; clks->parent && !found; clks++) {
for (clkr = clks->rates; clkr->div && !found; clkr++) {
if ((clkr->flags & cpu_mask) && (clkr->val == r)) {
if (clk->parent != clks->parent) {
pr_debug("clock: inited %s parent "
"to %s (was %s)\n",
clk->name, clks->parent->name,
((clk->parent) ?
clk->parent->name : "NULL"));
clk_reparent(clk, clks->parent);
};
found = 1;
}
}
}
if (!found)
printk(KERN_ERR "clock: init parent: could not find "
"regval %0x for clock %s\n", r, clk->name);
return;
}
/* Returns the DPLL rate */
u32 omap2_get_dpll_rate(struct clk *clk)
{
long long dpll_clk;
u32 dpll_mult, dpll_div, dpll;
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
struct dpll_data *dd;
dd = clk->dpll_data;
/* REVISIT: What do we return on error? */
if (!dd)
return 0;
dpll = __raw_readl(dd->mult_div1_reg);
dpll_mult = dpll & dd->mult_mask;
dpll_mult >>= __ffs(dd->mult_mask);
dpll_div = dpll & dd->div1_mask;
dpll_div >>= __ffs(dd->div1_mask);
dpll_clk = (long long)clk->parent->rate * dpll_mult;
do_div(dpll_clk, dpll_div + 1);
return dpll_clk;
}
/*
* Used for clocks that have the same value as the parent clock,
* divided by some factor
*/
void omap2_fixed_divisor_recalc(struct clk *clk)
{
WARN_ON(!clk->fixed_div);
clk->rate = clk->parent->rate / clk->fixed_div;
}
/**
* omap2_wait_clock_ready - wait for clock to enable
* @reg: physical address of clock IDLEST register
* @mask: value to mask against to determine if the clock is active
* @name: name of the clock (for printk)
*
* Returns 1 if the clock enabled in time, or 0 if it failed to enable
* in roughly MAX_CLOCK_ENABLE_WAIT microseconds.
*/
int omap2_wait_clock_ready(void __iomem *reg, u32 mask, const char *name)
{
int i = 0;
int ena = 0;
/*
* 24xx uses 0 to indicate not ready, and 1 to indicate ready.
* 34xx reverses this, just to keep us on our toes
*/
if (cpu_mask & (RATE_IN_242X | RATE_IN_243X))
ena = mask;
else if (cpu_mask & RATE_IN_343X)
ena = 0;
/* Wait for lock */
while (((__raw_readl(reg) & mask) != ena) &&
(i++ < MAX_CLOCK_ENABLE_WAIT)) {
udelay(1);
}
if (i < MAX_CLOCK_ENABLE_WAIT)
pr_debug("Clock %s stable after %d loops\n", name, i);
else
printk(KERN_ERR "Clock %s didn't enable in %d tries\n",
name, MAX_CLOCK_ENABLE_WAIT);
return (i < MAX_CLOCK_ENABLE_WAIT) ? 1 : 0;
};
/*
* Note: We don't need special code here for INVERT_ENABLE
* for the time being since INVERT_ENABLE only applies to clocks enabled by
* CM_CLKEN_PLL
*/
static void omap2_clk_wait_ready(struct clk *clk)
{
void __iomem *reg, *other_reg, *st_reg;
u32 bit;
/*
* REVISIT: This code is pretty ugly. It would be nice to generalize
* it and pull it into struct clk itself somehow.
*/
reg = clk->enable_reg;
/*
* Convert CM_ICLKEN* <-> CM_FCLKEN*. This conversion assumes
* it's just a matter of XORing the bits.
*/
other_reg = (void __iomem *)((u32)reg ^ (CM_FCLKEN ^ CM_ICLKEN));
/* Check if both functional and interface clocks
* are running. */
bit = 1 << clk->enable_bit;
if (!(__raw_readl(other_reg) & bit))
return;
st_reg = (void __iomem *)(((u32)other_reg & ~0xf0) | 0x20); /* CM_IDLEST* */
omap2_wait_clock_ready(st_reg, bit, clk->name);
}
static int omap2_dflt_clk_enable(struct clk *clk)
{
u32 v;
if (unlikely(clk->enable_reg == NULL)) {
printk(KERN_ERR "clock.c: Enable for %s without enable code\n",
clk->name);
return 0; /* REVISIT: -EINVAL */
}
v = __raw_readl(clk->enable_reg);
if (clk->flags & INVERT_ENABLE)
v &= ~(1 << clk->enable_bit);
else
v |= (1 << clk->enable_bit);
__raw_writel(v, clk->enable_reg);
wmb();
return 0;
}
static int omap2_dflt_clk_enable_wait(struct clk *clk)
{
int ret;
if (!clk->enable_reg) {
printk(KERN_ERR "clock.c: Enable for %s without enable code\n",
clk->name);
return 0; /* REVISIT: -EINVAL */
}
ret = omap2_dflt_clk_enable(clk);
if (ret == 0)
omap2_clk_wait_ready(clk);
return ret;
}
static void omap2_dflt_clk_disable(struct clk *clk)
{
u32 v;
if (!clk->enable_reg) {
/*
* 'Independent' here refers to a clock which is not
* controlled by its parent.
*/
printk(KERN_ERR "clock: clk_disable called on independent "
"clock %s which has no enable_reg\n", clk->name);
return;
}
v = __raw_readl(clk->enable_reg);
if (clk->flags & INVERT_ENABLE)
v |= (1 << clk->enable_bit);
else
v &= ~(1 << clk->enable_bit);
__raw_writel(v, clk->enable_reg);
[ARM] OMAP2/3 clock: don't use a barrier after clk_disable() clk_disable() previously used an ARM barrier, wmb(), to try to ensure that the hardware write completed before continuing. There are some problems with this approach. The first problem is that wmb() only ensures that the write leaves the ARM -- not that it actually reaches the endpoint device. In this case, the endpoint device - either the PRM, CM, or SCM - is three interconnects away from the ARM, and the final interconnect is low-speed. And the OCP interconnects will post the write, who knows how long that will take to complete. So the wmb() is not really what we want. Worse, the wmb() is indiscriminate; it will cause the ARM to flush any other unrelated buffered writes and wait for the local interconnect to acknowledge them - potentially very expensive. This first problem could be fixed by doing a readback of the same PRM/CM/SCM register. Since these devices use a single OCP thread, this will cause the MPU to wait for the write to complete. But the primary problem is a conceptual one: clk_disable() should not need any kind of barrier. clk_enable() needs one since device driver code must not access a device until its clocks are known to be enabled. But clk_disable() has no such restriction. Since blocking the MPU on a PRM/CM/SCM write can be a very high-latency operation - several hundred MPU cycles - it's worth avoiding this barrier if possible. linux-omap source commit is f4aacad2c0ed1055622d5c1e910befece24ef0e2. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2009-01-29 03:35:01 +08:00
/* No OCP barrier needed here since it is a disable operation */
}
const struct clkops clkops_omap2_dflt_wait = {
.enable = omap2_dflt_clk_enable_wait,
.disable = omap2_dflt_clk_disable,
};
const struct clkops clkops_omap2_dflt = {
.enable = omap2_dflt_clk_enable,
.disable = omap2_dflt_clk_disable,
};
/* Enables clock without considering parent dependencies or use count
* REVISIT: Maybe change this to use clk->enable like on omap1?
*/
static int _omap2_clk_enable(struct clk *clk)
{
return clk->ops->enable(clk);
}
/* Disables clock without considering parent dependencies or use count */
static void _omap2_clk_disable(struct clk *clk)
{
clk->ops->disable(clk);
}
void omap2_clk_disable(struct clk *clk)
{
if (clk->usecount > 0 && !(--clk->usecount)) {
_omap2_clk_disable(clk);
if (clk->parent)
omap2_clk_disable(clk->parent);
if (clk->clkdm)
omap2_clkdm_clk_disable(clk->clkdm, clk);
}
}
int omap2_clk_enable(struct clk *clk)
{
int ret = 0;
if (clk->usecount++ == 0) {
if (clk->parent)
ret = omap2_clk_enable(clk->parent);
if (ret != 0) {
clk->usecount--;
return ret;
}
if (clk->clkdm)
omap2_clkdm_clk_enable(clk->clkdm, clk);
ret = _omap2_clk_enable(clk);
if (ret != 0) {
if (clk->clkdm)
omap2_clkdm_clk_disable(clk->clkdm, clk);
if (clk->parent) {
omap2_clk_disable(clk->parent);
clk->usecount--;
}
}
}
return ret;
}
/*
* Used for clocks that are part of CLKSEL_xyz governed clocks.
* REVISIT: Maybe change to use clk->enable() functions like on omap1?
*/
void omap2_clksel_recalc(struct clk *clk)
{
u32 div = 0;
pr_debug("clock: recalc'ing clksel clk %s\n", clk->name);
div = omap2_clksel_get_divisor(clk);
if (div == 0)
return;
if (clk->rate == (clk->parent->rate / div))
return;
clk->rate = clk->parent->rate / div;
pr_debug("clock: new clock rate is %ld (div %d)\n", clk->rate, div);
}
/**
* omap2_get_clksel_by_parent - return clksel struct for a given clk & parent
* @clk: OMAP struct clk ptr to inspect
* @src_clk: OMAP struct clk ptr of the parent clk to search for
*
* Scan the struct clksel array associated with the clock to find
* the element associated with the supplied parent clock address.
* Returns a pointer to the struct clksel on success or NULL on error.
*/
static const struct clksel *omap2_get_clksel_by_parent(struct clk *clk,
struct clk *src_clk)
{
const struct clksel *clks;
if (!clk->clksel)
return NULL;
for (clks = clk->clksel; clks->parent; clks++) {
if (clks->parent == src_clk)
break; /* Found the requested parent */
}
if (!clks->parent) {
printk(KERN_ERR "clock: Could not find parent clock %s in "
"clksel array of clock %s\n", src_clk->name,
clk->name);
return NULL;
}
return clks;
}
/**
* omap2_clksel_round_rate_div - find divisor for the given clock and rate
* @clk: OMAP struct clk to use
* @target_rate: desired clock rate
* @new_div: ptr to where we should store the divisor
*
* Finds 'best' divider value in an array based on the source and target
* rates. The divider array must be sorted with smallest divider first.
* Note that this will not work for clocks which are part of CONFIG_PARTICIPANT,
* they are only settable as part of virtual_prcm set.
*
* Returns the rounded clock rate or returns 0xffffffff on error.
*/
u32 omap2_clksel_round_rate_div(struct clk *clk, unsigned long target_rate,
u32 *new_div)
{
unsigned long test_rate;
const struct clksel *clks;
const struct clksel_rate *clkr;
u32 last_div = 0;
printk(KERN_INFO "clock: clksel_round_rate_div: %s target_rate %ld\n",
clk->name, target_rate);
*new_div = 1;
clks = omap2_get_clksel_by_parent(clk, clk->parent);
if (!clks)
return ~0;
for (clkr = clks->rates; clkr->div; clkr++) {
if (!(clkr->flags & cpu_mask))
continue;
/* Sanity check */
if (clkr->div <= last_div)
printk(KERN_ERR "clock: clksel_rate table not sorted "
"for clock %s", clk->name);
last_div = clkr->div;
test_rate = clk->parent->rate / clkr->div;
if (test_rate <= target_rate)
break; /* found it */
}
if (!clkr->div) {
printk(KERN_ERR "clock: Could not find divisor for target "
"rate %ld for clock %s parent %s\n", target_rate,
clk->name, clk->parent->name);
return ~0;
}
*new_div = clkr->div;
printk(KERN_INFO "clock: new_div = %d, new_rate = %ld\n", *new_div,
(clk->parent->rate / clkr->div));
return (clk->parent->rate / clkr->div);
}
/**
* omap2_clksel_round_rate - find rounded rate for the given clock and rate
* @clk: OMAP struct clk to use
* @target_rate: desired clock rate
*
* Compatibility wrapper for OMAP clock framework
* Finds best target rate based on the source clock and possible dividers.
* rates. The divider array must be sorted with smallest divider first.
* Note that this will not work for clocks which are part of CONFIG_PARTICIPANT,
* they are only settable as part of virtual_prcm set.
*
* Returns the rounded clock rate or returns 0xffffffff on error.
*/
long omap2_clksel_round_rate(struct clk *clk, unsigned long target_rate)
{
u32 new_div;
return omap2_clksel_round_rate_div(clk, target_rate, &new_div);
}
/* Given a clock and a rate apply a clock specific rounding function */
long omap2_clk_round_rate(struct clk *clk, unsigned long rate)
{
if (clk->round_rate)
return clk->round_rate(clk, rate);
if (clk->flags & RATE_FIXED)
printk(KERN_ERR "clock: generic omap2_clk_round_rate called "
"on fixed-rate clock %s\n", clk->name);
return clk->rate;
}
/**
* omap2_clksel_to_divisor() - turn clksel field value into integer divider
* @clk: OMAP struct clk to use
* @field_val: register field value to find
*
* Given a struct clk of a rate-selectable clksel clock, and a register field
* value to search for, find the corresponding clock divisor. The register
* field value should be pre-masked and shifted down so the LSB is at bit 0
* before calling. Returns 0 on error
*/
u32 omap2_clksel_to_divisor(struct clk *clk, u32 field_val)
{
const struct clksel *clks;
const struct clksel_rate *clkr;
clks = omap2_get_clksel_by_parent(clk, clk->parent);
if (!clks)
return 0;
for (clkr = clks->rates; clkr->div; clkr++) {
if ((clkr->flags & cpu_mask) && (clkr->val == field_val))
break;
}
if (!clkr->div) {
printk(KERN_ERR "clock: Could not find fieldval %d for "
"clock %s parent %s\n", field_val, clk->name,
clk->parent->name);
return 0;
}
return clkr->div;
}
/**
* omap2_divisor_to_clksel() - turn clksel integer divisor into a field value
* @clk: OMAP struct clk to use
* @div: integer divisor to search for
*
* Given a struct clk of a rate-selectable clksel clock, and a clock divisor,
* find the corresponding register field value. The return register value is
* the value before left-shifting. Returns 0xffffffff on error
*/
u32 omap2_divisor_to_clksel(struct clk *clk, u32 div)
{
const struct clksel *clks;
const struct clksel_rate *clkr;
/* should never happen */
WARN_ON(div == 0);
clks = omap2_get_clksel_by_parent(clk, clk->parent);
if (!clks)
return 0;
for (clkr = clks->rates; clkr->div; clkr++) {
if ((clkr->flags & cpu_mask) && (clkr->div == div))
break;
}
if (!clkr->div) {
printk(KERN_ERR "clock: Could not find divisor %d for "
"clock %s parent %s\n", div, clk->name,
clk->parent->name);
return 0;
}
return clkr->val;
}
/**
* omap2_clksel_get_divisor - get current divider applied to parent clock.
* @clk: OMAP struct clk to use.
*
* Returns the integer divisor upon success or 0 on error.
*/
u32 omap2_clksel_get_divisor(struct clk *clk)
{
u32 v;
if (!clk->clksel_mask)
return 0;
v = __raw_readl(clk->clksel_reg) & clk->clksel_mask;
v >>= __ffs(clk->clksel_mask);
return omap2_clksel_to_divisor(clk, v);
}
int omap2_clksel_set_rate(struct clk *clk, unsigned long rate)
{
u32 v, field_val, validrate, new_div = 0;
if (!clk->clksel_mask)
return -EINVAL;
validrate = omap2_clksel_round_rate_div(clk, rate, &new_div);
if (validrate != rate)
return -EINVAL;
field_val = omap2_divisor_to_clksel(clk, new_div);
if (field_val == ~0)
return -EINVAL;
v = __raw_readl(clk->clksel_reg);
v &= ~clk->clksel_mask;
v |= field_val << __ffs(clk->clksel_mask);
__raw_writel(v, clk->clksel_reg);
wmb();
clk->rate = clk->parent->rate / new_div;
_omap2xxx_clk_commit(clk);
return 0;
}
/* Set the clock rate for a clock source */
int omap2_clk_set_rate(struct clk *clk, unsigned long rate)
{
int ret = -EINVAL;
pr_debug("clock: set_rate for clock %s to rate %ld\n", clk->name, rate);
/* CONFIG_PARTICIPANT clocks are changed only in sets via the
rate table mechanism, driven by mpu_speed */
if (clk->flags & CONFIG_PARTICIPANT)
return -EINVAL;
/* dpll_ck, core_ck, virt_prcm_set; plus all clksel clocks */
if (clk->set_rate)
ret = clk->set_rate(clk, rate);
return ret;
}
/*
* Converts encoded control register address into a full address
* On error, the return value (parent_div) will be 0.
*/
static u32 _omap2_clksel_get_src_field(struct clk *src_clk, struct clk *clk,
u32 *field_val)
{
const struct clksel *clks;
const struct clksel_rate *clkr;
clks = omap2_get_clksel_by_parent(clk, src_clk);
if (!clks)
return 0;
for (clkr = clks->rates; clkr->div; clkr++) {
if (clkr->flags & (cpu_mask | DEFAULT_RATE))
break; /* Found the default rate for this platform */
}
if (!clkr->div) {
printk(KERN_ERR "clock: Could not find default rate for "
"clock %s parent %s\n", clk->name,
src_clk->parent->name);
return 0;
}
/* Should never happen. Add a clksel mask to the struct clk. */
WARN_ON(clk->clksel_mask == 0);
*field_val = clkr->val;
return clkr->div;
}
int omap2_clk_set_parent(struct clk *clk, struct clk *new_parent)
{
u32 field_val, v, parent_div;
if (clk->flags & CONFIG_PARTICIPANT)
return -EINVAL;
if (!clk->clksel)
return -EINVAL;
parent_div = _omap2_clksel_get_src_field(new_parent, clk, &field_val);
if (!parent_div)
return -EINVAL;
if (clk->usecount > 0)
_omap2_clk_disable(clk);
/* Set new source value (previous dividers if any in effect) */
v = __raw_readl(clk->clksel_reg);
v &= ~clk->clksel_mask;
v |= field_val << __ffs(clk->clksel_mask);
__raw_writel(v, clk->clksel_reg);
wmb();
_omap2xxx_clk_commit(clk);
if (clk->usecount > 0)
_omap2_clk_enable(clk);
clk_reparent(clk, new_parent);
/* CLKSEL clocks follow their parents' rates, divided by a divisor */
clk->rate = new_parent->rate;
if (parent_div > 0)
clk->rate /= parent_div;
pr_debug("clock: set parent of %s to %s (new rate %ld)\n",
clk->name, clk->parent->name, clk->rate);
return 0;
}
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
/* DPLL rate rounding code */
/**
* omap2_dpll_set_rate_tolerance: set the error tolerance during rate rounding
* @clk: struct clk * of the DPLL
* @tolerance: maximum rate error tolerance
*
* Set the maximum DPLL rate error tolerance for the rate rounding
* algorithm. The rate tolerance is an attempt to balance DPLL power
* saving (the least divider value "n") vs. rate fidelity (the least
* difference between the desired DPLL target rate and the rounded
* rate out of the algorithm). So, increasing the tolerance is likely
* to decrease DPLL power consumption and increase DPLL rate error.
* Returns -EINVAL if provided a null clock ptr or a clk that is not a
* DPLL; or 0 upon success.
*/
int omap2_dpll_set_rate_tolerance(struct clk *clk, unsigned int tolerance)
{
if (!clk || !clk->dpll_data)
return -EINVAL;
clk->dpll_data->rate_tolerance = tolerance;
return 0;
}
static unsigned long _dpll_compute_new_rate(unsigned long parent_rate,
unsigned int m, unsigned int n)
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
{
unsigned long long num;
num = (unsigned long long)parent_rate * m;
do_div(num, n);
return num;
}
/*
* _dpll_test_mult - test a DPLL multiplier value
* @m: pointer to the DPLL m (multiplier) value under test
* @n: current DPLL n (divider) value under test
* @new_rate: pointer to storage for the resulting rounded rate
* @target_rate: the desired DPLL rate
* @parent_rate: the DPLL's parent clock rate
*
* This code tests a DPLL multiplier value, ensuring that the
* resulting rate will not be higher than the target_rate, and that
* the multiplier value itself is valid for the DPLL. Initially, the
* integer pointed to by the m argument should be prescaled by
* multiplying by DPLL_SCALE_FACTOR. The code will replace this with
* a non-scaled m upon return. This non-scaled m will result in a
* new_rate as close as possible to target_rate (but not greater than
* target_rate) given the current (parent_rate, n, prescaled m)
* triple. Returns DPLL_MULT_UNDERFLOW in the event that the
* non-scaled m attempted to underflow, which can allow the calling
* function to bail out early; or 0 upon success.
*/
static int _dpll_test_mult(int *m, int n, unsigned long *new_rate,
unsigned long target_rate,
unsigned long parent_rate)
{
int r = 0, carry = 0;
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
/* Unscale m and round if necessary */
if (*m % DPLL_SCALE_FACTOR >= DPLL_ROUNDING_VAL)
carry = 1;
*m = (*m / DPLL_SCALE_FACTOR) + carry;
/*
* The new rate must be <= the target rate to avoid programming
* a rate that is impossible for the hardware to handle
*/
*new_rate = _dpll_compute_new_rate(parent_rate, *m, n);
if (*new_rate > target_rate) {
(*m)--;
*new_rate = 0;
}
/* Guard against m underflow */
if (*m < DPLL_MIN_MULTIPLIER) {
*m = DPLL_MIN_MULTIPLIER;
*new_rate = 0;
r = DPLL_MULT_UNDERFLOW;
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
}
if (*new_rate == 0)
*new_rate = _dpll_compute_new_rate(parent_rate, *m, n);
return r;
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
}
/**
* omap2_dpll_round_rate - round a target rate for an OMAP DPLL
* @clk: struct clk * for a DPLL
* @target_rate: desired DPLL clock rate
*
* Given a DPLL, a desired target rate, and a rate tolerance, round
* the target rate to a possible, programmable rate for this DPLL.
* Rate tolerance is assumed to be set by the caller before this
* function is called. Attempts to select the minimum possible n
* within the tolerance to reduce power consumption. Stores the
* computed (m, n) in the DPLL's dpll_data structure so set_rate()
* will not need to call this (expensive) function again. Returns ~0
* if the target rate cannot be rounded, either because the rate is
* too low or because the rate tolerance is set too tightly; or the
* rounded rate upon success.
*/
long omap2_dpll_round_rate(struct clk *clk, unsigned long target_rate)
{
int m, n, r, e, scaled_max_m;
unsigned long scaled_rt_rp, new_rate;
int min_e = -1, min_e_m = -1, min_e_n = -1;
struct dpll_data *dd;
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
if (!clk || !clk->dpll_data)
return ~0;
dd = clk->dpll_data;
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
pr_debug("clock: starting DPLL round_rate for clock %s, target rate "
"%ld\n", clk->name, target_rate);
scaled_rt_rp = target_rate / (clk->parent->rate / DPLL_SCALE_FACTOR);
scaled_max_m = dd->max_multiplier * DPLL_SCALE_FACTOR;
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
dd->last_rounded_rate = 0;
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
for (n = dd->min_divider; n <= dd->max_divider; n++) {
/* Is the (input clk, divider) pair valid for the DPLL? */
r = _dpll_test_fint(clk, n);
if (r == DPLL_FINT_UNDERFLOW)
break;
else if (r == DPLL_FINT_INVALID)
continue;
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
/* Compute the scaled DPLL multiplier, based on the divider */
m = scaled_rt_rp * n;
/*
* Since we're counting n up, a m overflow means we
* can bail out completely (since as n increases in
* the next iteration, there's no way that m can
* increase beyond the current m)
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
*/
if (m > scaled_max_m)
break;
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
r = _dpll_test_mult(&m, n, &new_rate, target_rate,
clk->parent->rate);
/* m can't be set low enough for this n - try with a larger n */
if (r == DPLL_MULT_UNDERFLOW)
continue;
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
e = target_rate - new_rate;
pr_debug("clock: n = %d: m = %d: rate error is %d "
"(new_rate = %ld)\n", n, m, e, new_rate);
if (min_e == -1 ||
min_e >= (int)(abs(e) - dd->rate_tolerance)) {
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
min_e = e;
min_e_m = m;
min_e_n = n;
pr_debug("clock: found new least error %d\n", min_e);
/* We found good settings -- bail out now */
if (min_e <= dd->rate_tolerance)
break;
}
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
}
if (min_e < 0) {
pr_debug("clock: error: target rate or tolerance too low\n");
return ~0;
}
dd->last_rounded_m = min_e_m;
dd->last_rounded_n = min_e_n;
dd->last_rounded_rate = _dpll_compute_new_rate(clk->parent->rate,
min_e_m, min_e_n);
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
pr_debug("clock: final least error: e = %d, m = %d, n = %d\n",
min_e, min_e_m, min_e_n);
pr_debug("clock: final rate: %ld (target rate: %ld)\n",
dd->last_rounded_rate, target_rate);
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
return dd->last_rounded_rate;
ARM: OMAP2: Clock: New OMAP2/3 DPLL rate rounding algorithm This patch adds a new rate rounding algorithm for DPLL clocks on the OMAP2/3 architecture. For a desired DPLL target rate, there may be several multiplier/divider (M, N) values which will generate a sufficiently close rate. Lower N values result in greater power economy. However, lower N values can cause the difference between the rounded rate and the target rate ("rate error") to be larger than it would be with a higher N. This can cause downstream devices to run more slowly than they otherwise would. This DPLL rate rounding algorithm: - attempts to find the lowest possible N (DPLL divider) to reach the target_rate (since, according to Richard Woodruff <r-woodruff@ti.com>, lower N values save more power than higher N values). - allows developers to set an upper bound on the error between the rounded rate and the desired target rate ("rate tolerance"), so an appropriate balance between rate fidelity and power savings can be set. This maximum rate error tolerance is set via omap2_set_dpll_rate_tolerance(). - never returns a rounded rate higher than the target rate. The rate rounding algorithm caches the last rounded M, N, and rate computation to avoid rounding the rate twice for each clk_set_rate() call. (This patch does not yet implement set_rate for DPLLs; that follows in a future patch.) The algorithm trades execution speed for rate accuracy. It will find the (M, N) set that results in the least rate error, within a specified rate tolerance. It does this by evaluating each divider setting - on OMAP3, this involves 128 steps. Another approach to DPLL rate rounding would be to bail out as soon as a valid rate is found within the rate tolerance, which would trade rate accuracy for execution speed. Alternate implementations welcome. This code is not yet used by the OMAP24XX DPLL clock, since it is currently defined as a composite clock, fusing the DPLL M,N and the M2 output divider. This patch also renames the existing OMAP24xx DPLL programming functions to highlight that they program both the DPLL and the DPLL's output multiplier. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
2008-07-03 17:24:46 +08:00
}
/*-------------------------------------------------------------------------
* Omap2 clock reset and init functions
*-------------------------------------------------------------------------*/
#ifdef CONFIG_OMAP_RESET_CLOCKS
void omap2_clk_disable_unused(struct clk *clk)
{
u32 regval32, v;
v = (clk->flags & INVERT_ENABLE) ? (1 << clk->enable_bit) : 0;
regval32 = __raw_readl(clk->enable_reg);
if ((regval32 & (1 << clk->enable_bit)) == v)
return;
printk(KERN_INFO "Disabling unused clock \"%s\"\n", clk->name);
if (cpu_is_omap34xx()) {
omap2_clk_enable(clk);
omap2_clk_disable(clk);
} else
_omap2_clk_disable(clk);
}
#endif