[PATCH v2 03/10] clk: berlin: add driver for BG2x audio/video PLL
Sebastian Hesselbarth
sebastian.hesselbarth at gmail.com
Wed May 14 13:15:14 PDT 2014
This is a driver for the dual-VCO PLL with 8 channels each found on
Marvell Berlin2 SoCs. While both VCOs share the same register set,
sometimes registers shifts for one of the VCOs is a bit off. Nothing
serious that should require a separate driver, so deal with both VCOs
in a single driver instead.
Signed-off-by: Alexandre Belloni <alexandre.belloni at free-electrons.com>
Signed-off-by: Sebastian Hesselbarth <sebastian.hesselbarth at gmail.com>
---
Changelog:
v1->v2:
- dropped usage of postponed of_clk_create_name()
Cc: Mike Turquette <mturquette at linaro.org>
Cc: Alexandre Belloni <alexandre.belloni at free-electrons.com>
Cc: Jisheng Zhang <jszhang at marvell.com>
Cc: linux-arm-kernel at lists.infradead.org
Cc: linux-kernel at vger.kernel.org
---
drivers/clk/Makefile | 1 +
drivers/clk/berlin/Makefile | 1 +
drivers/clk/berlin/berlin2-avpll.c | 368 +++++++++++++++++++++++++++++++++++++
3 files changed, 370 insertions(+)
create mode 100644 drivers/clk/berlin/Makefile
create mode 100644 drivers/clk/berlin/berlin2-avpll.c
diff --git a/drivers/clk/Makefile b/drivers/clk/Makefile
index 5f8a28735c96..f651b2ace915 100644
--- a/drivers/clk/Makefile
+++ b/drivers/clk/Makefile
@@ -31,6 +31,7 @@ obj-$(CONFIG_COMMON_CLK_WM831X) += clk-wm831x.o
obj-$(CONFIG_COMMON_CLK_XGENE) += clk-xgene.o
obj-$(CONFIG_COMMON_CLK_AT91) += at91/
obj-$(CONFIG_ARCH_BCM_MOBILE) += bcm/
+obj-$(CONFIG_ARCH_BERLIN) += berlin/
obj-$(CONFIG_ARCH_HI3xxx) += hisilicon/
obj-$(CONFIG_ARCH_HIP04) += hisilicon/
obj-$(CONFIG_COMMON_CLK_KEYSTONE) += keystone/
diff --git a/drivers/clk/berlin/Makefile b/drivers/clk/berlin/Makefile
new file mode 100644
index 000000000000..5905733fc7c7
--- /dev/null
+++ b/drivers/clk/berlin/Makefile
@@ -0,0 +1 @@
+obj-y += berlin2-avpll.o
diff --git a/drivers/clk/berlin/berlin2-avpll.c b/drivers/clk/berlin/berlin2-avpll.c
new file mode 100644
index 000000000000..a06cde9193f6
--- /dev/null
+++ b/drivers/clk/berlin/berlin2-avpll.c
@@ -0,0 +1,368 @@
+/*
+ * Copyright (c) 2014 Marvell Technology Group Ltd.
+ *
+ * Sebastian Hesselbarth <sebastian.hesselbarth at gmail.com>
+ * Alexandre Belloni <alexandre.belloni at free-electrons.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program. If not, see <http://www.gnu.org/licenses/>.
+ */
+#include <linux/clk-provider.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/slab.h>
+
+/*
+ * Berlin2 AVPLL comprises two PLLs (VCOs) with 8 channels each,
+ * channel 8 is the odd-one-out and does not provide mul/div.
+ *
+ * Unfortunately, its registers are just numbered from 0-63. To
+ * get in at least some kind of structure, we split both VCOs and
+ * each of the channels into separate clock drivers.
+ *
+ * Also, here and there the VCO registers are a bit different with
+ * respect to bit shifts. Make sure to add a comment for those.
+ */
+#define NUM_VCOS 2
+#define NUM_CHANNELS 8
+
+#define AVPLL_CTRL(x) ((x) * 0x4)
+/* Second VCO starts at AVPLL_CTRL31 */
+#define VCO_OFFSET AVPLL_CTRL(31)
+
+#define VCO_CTRL0 AVPLL_CTRL(0)
+/* VCO_B has an additional shift of 4 for its VCO_CTRL0 reg */
+#define VCO_RESET BIT(0)
+#define VCO_POWERUP BIT(1)
+#define VCO_INTERPOL_SHIFT 2
+#define VCO_INTERPOL_MASK (0xf << VCO_INTERPOL_SHIFT)
+#define VCO_REG1V45_SEL_SHIFT 6
+#define VCO_REG1V45_SEL(x) ((x) << VCO_REG1V45_SEL_SHIFT)
+#define VCO_REG1V45_SEL_1V40 VCO_REG1V45_SEL(0)
+#define VCO_REG1V45_SEL_1V45 VCO_REG1V45_SEL(1)
+#define VCO_REG1V45_SEL_1V50 VCO_REG1V45_SEL(2)
+#define VCO_REG1V45_SEL_1V55 VCO_REG1V45_SEL(3)
+#define VCO_REG1V45_SEL_MASK VCO_REG1V45_SEL(3)
+#define VCO_REG0V9_SEL_SHIFT 8
+#define VCO_REG0V9_SEL_MASK (0xf << VCO_REG0V9_SEL_SHIFT)
+#define VCO_VTHCAL_SHIFT 12
+#define VCO_VTHCAL(x) ((x) << VCO_VTHCAL_SHIFT)
+#define VCO_VTHCAL_0V90 VCO_VTHCAL(0)
+#define VCO_VTHCAL_0V95 VCO_VTHCAL(1)
+#define VCO_VTHCAL_1V00 VCO_VTHCAL(2)
+#define VCO_VTHCAL_1V05 VCO_VTHCAL(3)
+#define VCO_VTHCAL_MASK VCO_VTHCAL(3)
+#define VCO_KVCOEXT_SHIFT 14
+#define VCO_KVCOEXT_MASK (0x3 << VCO_KVCOEXT_SHIFT)
+#define VCO_KVCOEXT_ENABLE BIT(17)
+#define VCO_V2IEXT_SHIFT 18
+#define VCO_V2IEXT_MASK (0xf << VCO_V2IEXT_SHIFT)
+#define VCO_V2IEXT_ENABLE BIT(22)
+#define VCO_SPEED_SHIFT 23
+#define VCO_SPEED(x) ((x) << VCO_SPEED_SHIFT)
+#define VCO_SPEED_1G08_1G21 VCO_SPEED(0)
+#define VCO_SPEED_1G21_1G40 VCO_SPEED(1)
+#define VCO_SPEED_1G40_1G61 VCO_SPEED(2)
+#define VCO_SPEED_1G61_1G86 VCO_SPEED(3)
+#define VCO_SPEED_1G86_2G00 VCO_SPEED(4)
+#define VCO_SPEED_2G00_2G22 VCO_SPEED(5)
+#define VCO_SPEED_2G22 VCO_SPEED(6)
+#define VCO_SPEED_MASK VCO_SPEED(0x7)
+#define VCO_CLKDET_ENABLE BIT(26)
+#define VCO_CTRL1 AVPLL_CTRL(1)
+#define VCO_REFDIV_SHIFT 0
+#define VCO_REFDIV(x) ((x) << VCO_REFDIV_SHIFT)
+#define VCO_REFDIV_1 VCO_REFDIV(0)
+#define VCO_REFDIV_2 VCO_REFDIV(1)
+#define VCO_REFDIV_4 VCO_REFDIV(2)
+#define VCO_REFDIV_3 VCO_REFDIV(3)
+#define VCO_REFDIV_MASK VCO_REFDIV(0x3f)
+#define VCO_FBDIV_SHIFT 6
+#define VCO_FBDIV(x) ((x) << VCO_FBDIV_SHIFT)
+#define VCO_FBDIV_MASK VCO_FBDIV(0xff)
+#define VCO_ICP_SHIFT 14
+/* PLL Charge Pump Current = 10uA * (x + 1) */
+#define VCO_ICP(x) ((x) << VCO_ICP_SHIFT)
+#define VCO_ICP_MASK VCO_ICP(0xf)
+#define VCO_LOAD_CAP BIT(18)
+#define VCO_CALIBRATION_START BIT(19)
+#define VCO_FREQOFFSETn(x) AVPLL_CTRL(3 + (x))
+#define VCO_FREQOFFSET_MASK 0x7ffff
+#define VCO_CTRL11 AVPLL_CTRL(11)
+#define VCO_CTRL12 AVPLL_CTRL(12)
+#define VCO_CTRL13 AVPLL_CTRL(13)
+#define VCO_CTRL14 AVPLL_CTRL(14)
+#define VCO_CTRL15 AVPLL_CTRL(15)
+#define VCO_SYNC1n(x) AVPLL_CTRL(15 + (x))
+#define VCO_SYNC1_MASK 0x1ffff
+#define VCO_SYNC2n(x) AVPLL_CTRL(23 + (x))
+#define VCO_SYNC2_MASK 0x1ffff
+#define VCO_CTRL30 AVPLL_CTRL(30)
+#define VCO_DPLL_CH1_ENABLE BIT(17)
+
+struct avpll_vco;
+struct avpll;
+
+struct avpll_channel {
+ struct clk_hw hw;
+ struct avpll_vco *vco;
+ u8 index;
+};
+
+struct avpll_vco {
+ struct clk_hw hw;
+ struct avpll_channel channel[NUM_CHANNELS];
+ struct clk *clks[NUM_CHANNELS];
+ void __iomem *base;
+ struct avpll *pll;
+ u8 index;
+};
+
+struct avpll {
+ struct avpll_vco vco[NUM_VCOS];
+ void __iomem *base;
+};
+
+#define to_avpll_vco(hw) container_of(hw, struct avpll_vco, hw)
+#define to_avpll_channel(hw) container_of(hw, struct avpll_channel, hw)
+
+static u8 vco_refdiv[] = { 1, 2, 4, 3 };
+
+static unsigned long
+avpll_vco_recalc_rate(struct clk_hw *hw, unsigned long parent_rate)
+{
+ struct avpll_vco *vco = to_avpll_vco(hw);
+ u32 reg, refdiv;
+ u64 freq;
+
+ /* AVPLL VCO frequency: Fvco = (Fref / refdiv) * FBdiv */
+ reg = readl_relaxed(vco->base + VCO_CTRL1);
+ refdiv = (reg & VCO_REFDIV_MASK) >> VCO_REFDIV_SHIFT;
+ refdiv = vco_refdiv[refdiv];
+ freq = (reg & VCO_FBDIV_MASK) >> VCO_FBDIV_SHIFT;
+ freq *= parent_rate;
+ do_div(freq, refdiv);
+
+ return (unsigned long)freq;
+}
+
+static const struct clk_ops avpll_vco_ops = {
+ .recalc_rate = avpll_vco_recalc_rate,
+};
+
+static const u8 div_hdmi[] = { 1, 2, 4, 6 };
+static const u8 div_av1[] = { 1, 2, 5, 5 };
+
+static unsigned long
+avpll_channel_recalc_rate(struct clk_hw *hw, unsigned long parent_rate)
+{
+ struct avpll_channel *ch = to_avpll_channel(hw);
+ struct avpll_vco *vco = ch->vco;
+ u32 reg, divider = 1;
+ u64 freq = parent_rate;
+
+ reg = readl_relaxed(vco->base + VCO_CTRL30);
+ if ((reg & (VCO_DPLL_CH1_ENABLE << ch->index)) == 0)
+ goto skip_div;
+
+ /*
+ * Fch = (Fref * sync2) /
+ * (sync1 * div_hdmi * div_av1 * div_av2 * div_av3)
+ */
+
+ reg = readl_relaxed(vco->base + VCO_SYNC1n(ch->index));
+ /* SYNC1 for Channel 1 is shifted by 4 bits */
+ if (ch->index == 0)
+ reg >>= 4;
+ divider = reg & VCO_SYNC1_MASK;
+
+ reg = readl_relaxed(vco->base + VCO_SYNC2n(ch->index));
+ freq *= reg & VCO_SYNC2_MASK;
+
+ /* Channel 8 has no dividers */
+ if (ch->index == 7)
+ goto skip_div;
+
+ /*
+ * HDMI divider start at VCO_CTRL11, bit 7; each 3 bits wide but
+ * only 0-3 are valid values.
+ */
+ reg = readl_relaxed(vco->base + VCO_CTRL11) >> 7;
+ divider *= div_hdmi[(reg >> (ch->index * 3)) & 0x3];
+
+ /*
+ * AV1 divider start at VCO_CTRL11, bit 28; each 3 bits wide but
+ * only 0-3 are valid values.
+ */
+ if (ch->index == 0) {
+ reg = readl_relaxed(vco->base + VCO_CTRL11);
+ reg >>= 28;
+ } else {
+ reg = readl_relaxed(vco->base + VCO_CTRL12);
+ reg >>= (ch->index-1) * 3;
+ }
+ divider *= div_av1[reg & 0x3];
+
+ /*
+ * AV2 divider start at VCO_CTRL12, bit 18; each 7 bits wide,
+ * zero is not a valid value.
+ */
+ if (ch->index < 2) {
+ reg = readl_relaxed(vco->base + VCO_CTRL12);
+ reg >>= 18 + (ch->index * 7);
+ } else if (ch->index < 7) {
+ reg = readl_relaxed(vco->base + VCO_CTRL13);
+ reg >>= (ch->index - 2) * 7;
+ } else {
+ reg = readl_relaxed(vco->base + VCO_CTRL14);
+ }
+
+ reg &= 0x7f;
+ if (reg)
+ divider *= reg;
+
+ /*
+ * AV3 divider start at VCO_CTRL14, bit 7; each 4 bits wide,
+ * only 0, 1 are valid values. AV3=1 divides by 1/2, AV3=0 is bypass.
+ */
+ if (ch->index < 6) {
+ reg = readl_relaxed(vco->base + VCO_CTRL14);
+ reg >>= 7 + (ch->index * 4);
+ } else {
+ reg = readl_relaxed(vco->base + VCO_CTRL15);
+ }
+
+ if (reg & 0x1)
+ freq *= 2;
+
+skip_div:
+ do_div(freq, divider);
+ return (unsigned long)freq;
+}
+
+static const struct clk_ops avpll_channel_ops = {
+ .recalc_rate = avpll_channel_recalc_rate,
+};
+
+/*
+ * Another nice quirk: on production SoC revisions, AVPLL_B outputs
+ * are scrambled. Use a translation table to deal with it.
+ */
+static const u8 avpllb_channel_map[] = { 0, 6, 5, 4, 3, 2, 1, 7 };
+
+struct clk *avpll_src_get(struct of_phandle_args *clkspec, void *data)
+{
+ struct avpll *pll = data;
+ unsigned int vco = clkspec->args[0];
+ unsigned int ch = clkspec->args[1] - 1;
+
+ if (vco > NUM_VCOS || ch > NUM_CHANNELS) {
+ pr_err("%s: invalid clock index %d.%d\n",
+ __func__, vco, ch);
+ return ERR_PTR(-EINVAL);
+ }
+
+ if (vco == 1)
+ ch = avpllb_channel_map[ch];
+
+ return pll->vco[vco].clks[ch];
+}
+
+static void __init avpll_of_setup(struct device_node *np)
+{
+ struct avpll *pll;
+ struct clk_init_data init;
+ struct clk *refclk;
+ const char *refclk_name;
+ int nvco, nch;
+
+ pll = kzalloc(sizeof(*pll), GFP_KERNEL);
+ if (!pll)
+ return;
+
+ pll->base = of_iomap(np, 0);
+ if (!pll->base) {
+ pr_err("%s: Unable to map pll register\n", np->full_name);
+ kfree(pll);
+ return;
+ }
+
+ refclk = of_clk_get(np, 0);
+ if (IS_ERR(refclk)) {
+ pr_err("%s: Missing reference clock\n", np->full_name);
+ goto avpll_fail;
+ }
+ refclk_name = __clk_get_name(refclk);
+ clk_put(refclk);
+
+ for (nvco = 0; nvco < NUM_VCOS; nvco++) {
+ struct avpll_vco *vco = &pll->vco[nvco];
+ const char *vco_name =
+ kasprintf(GFP_KERNEL, "%s.vco%d", np->name, nvco);
+ struct clk *clk;
+
+ init.name = vco_name;
+ init.ops = &avpll_vco_ops;
+ init.parent_names = &refclk_name;
+ init.num_parents = 1;
+ init.flags = 0;
+
+ vco->hw.init = &init;
+ vco->base = pll->base + (nvco * VCO_OFFSET);
+ vco->pll = pll;
+ vco->index = nvco;
+
+ clk = clk_register(NULL, &vco->hw);
+ kfree(vco_name);
+ if (IS_ERR(clk)) {
+ pr_err("%s: Unable to register vco %d\n",
+ np->full_name, nvco);
+ goto avpll_fail;
+ }
+ vco_name = __clk_get_name(clk);
+
+ for (nch = 0; nch < NUM_CHANNELS; nch++) {
+ struct avpll_channel *ch = &vco->channel[nch];
+ char *ch_name = kasprintf(GFP_KERNEL,
+ "%s.%d", vco_name, nch);
+
+ init.name = ch_name;
+ init.ops = &avpll_channel_ops;
+ init.parent_names = &vco_name;
+ init.num_parents = 1;
+ init.flags = CLK_SET_RATE_PARENT;
+
+ ch->hw.init = &init;
+ ch->vco = vco;
+ ch->index = nch;
+
+ clk = clk_register(NULL, &ch->hw);
+ kfree(ch_name);
+ if (IS_ERR(clk)) {
+ pr_err("%s: Unable to register channel %d.%d\n",
+ np->full_name, nvco, nch);
+ goto avpll_fail;
+ }
+ vco->clks[nch] = clk;
+ }
+ }
+
+ of_clk_add_provider(np, avpll_src_get, pll);
+ return;
+
+avpll_fail:
+ iounmap(pll->base);
+ kfree(pll);
+}
+CLK_OF_DECLARE(berlin2_avpll, "marvell,berlin2-avpll", avpll_of_setup);
--
1.9.1
More information about the linux-arm-kernel
mailing list