[PATCH v3 3/4] phy: rockchip-typec: Avoid magic numbers + add delays in aux calib
Douglas Anderson
dianders at chromium.org
Fri Sep 22 09:44:05 PDT 2017
NOTE: nothing is known to be fixed by this change, but it does enforce
some delays that are documented to be necessary. Possibly this could
fix some corner cases.
The function tcphy_dp_aux_calibration(), like most of the functions in
the type C PHY, is mostly undocumented and filled with mysterious,
hardcoded numbers.
Let's attempt to try to document some of these numbers and clean the
function up a little bit. Here's the actual cleanup that happened
here:
1. All magic numbers were replaced with bit definitions.
2. For registers that we modify multiple times I now keep track of the
value of the register rather than randomly doing a
read/modify/write or just hardcoding a new number based on knowing
what the old number was.
3. Delay 10 ms (vs 1 ms) after writing the calibration code. No idea
if this is important but it matches the example in the docs.
4. Whenever setting a "delayed" version of a signal always put an
explicit delay in the code. No known problems were seen without
this delay but it seems wise to have it. Whenever a delay of "at
least 100 ns" was specified I used a delay of 1 us.
5. Added comments to some of the bits of code.
6. Removed duplicate setting of TX_ANA_CTRL_REG_5 (to 0)
7. Moved setting of TX_ANA_CTRL_REG_3 to the same place it was in the
sample code. Note that TX_ANA_CTRL_REG_3 ought to be initted to 0
(and elsewhere we assume that we just got a reset), but it seems
fine to be explicit.
8. Treats the calibration code as a 7-bit two's complement number.
This isn't strictly required, but seems slightly cleaner. The docs
say "treat this as a two's complement number, but it should never
be negative". If we ever read the "adjustment" codes as documented
then perhaps the two's complement bit will matter more.
There are still a few weird / mysterious things around aux init and
this doesn't attempt to fix all of them. Mostly it's aimed at doing
changes that should be _very_ safe and add a lot of clarity. Things
specifically not done:
A) Resolve the fact that some registers are read/modify/write and
others are explicitly initted to a value. We always call
tcphy_dp_aux_calibration() right after resetting the PHY so it's
probably not critical, but it's a little weird that the code is
inconsistent.
B) Fully resolve the documented init sequence with the current one.
We still have a few mystery steps and we also leave out turning on
TXDA_DRV_LDO_BG_FB_EN and TXDA_DRV_LDO_BG_REF_EN, which is in the
sample code.
C) Clean things up to read all the bits of the calibration code. This
will hopefully come in a followup change.
This also doesn't attempt to document any of the other parts of the
PHY--just the aux init which is all I got docs for.
Reviewed-by: Chris Zhong <zyw at rock-chips.com>
Signed-off-by: Douglas Anderson <dianders at chromium.org>
---
Changes in v3: None
Changes in v2: None
drivers/phy/rockchip/phy-rockchip-typec.c | 202 ++++++++++++++++++++++++------
1 file changed, 163 insertions(+), 39 deletions(-)
diff --git a/drivers/phy/rockchip/phy-rockchip-typec.c b/drivers/phy/rockchip/phy-rockchip-typec.c
index b25c00432f9b..95f8f23676b4 100644
--- a/drivers/phy/rockchip/phy-rockchip-typec.c
+++ b/drivers/phy/rockchip/phy-rockchip-typec.c
@@ -102,9 +102,40 @@
#define CMN_PLL1_SS_CTRL1 (0xb8 << 2)
#define CMN_PLL1_SS_CTRL2 (0xb9 << 2)
#define CMN_RXCAL_OVRD (0xd1 << 2)
+
#define CMN_TXPUCAL_CTRL (0xe0 << 2)
#define CMN_TXPUCAL_OVRD (0xe1 << 2)
+#define CMN_TXPDCAL_CTRL (0xf0 << 2)
#define CMN_TXPDCAL_OVRD (0xf1 << 2)
+
+/* For CMN_TXPUCAL_CTRL, CMN_TXPDCAL_CTRL */
+#define CMN_TXPXCAL_START BIT(15)
+#define CMN_TXPXCAL_DONE BIT(14)
+#define CMN_TXPXCAL_NO_RESPONSE BIT(13)
+#define CMN_TXPXCAL_CURRENT_RESPONSE BIT(12)
+
+#define CMN_TXPU_ADJ_CTRL (0x108 << 2)
+#define CMN_TXPD_ADJ_CTRL (0x10c << 2)
+
+/*
+ * For CMN_TXPUCAL_CTRL, CMN_TXPDCAL_CTRL,
+ * CMN_TXPU_ADJ_CTRL, CMN_TXPDCAL_CTRL
+ *
+ * NOTE: some of these registers are documented to be 2's complement
+ * signed numbers, but then documented to be always positive. Weird.
+ * In such a case, using CMN_CALIB_CODE_POS() avoids the unnecessary
+ * sign extension.
+ */
+#define CMN_CALIB_CODE_WIDTH 7
+#define CMN_CALIB_CODE_OFFSET 0
+#define CMN_CALIB_CODE_MASK GENMASK(CMN_CALIB_CODE_WIDTH, 0)
+#define CMN_CALIB_CODE(x) \
+ sign_extend32((x) >> CMN_CALIB_CODE_OFFSET, CMN_CALIB_CODE_WIDTH)
+
+#define CMN_CALIB_CODE_POS_MASK GENMASK(CMN_CALIB_CODE_WIDTH - 1, 0)
+#define CMN_CALIB_CODE_POS(x) \
+ (((x) >> CMN_CALIB_CODE_OFFSET) & CMN_CALIB_CODE_POS_MASK)
+
#define CMN_DIAG_PLL0_FBH_OVRD (0x1c0 << 2)
#define CMN_DIAG_PLL0_FBL_OVRD (0x1c1 << 2)
#define CMN_DIAG_PLL0_OVRD (0x1c2 << 2)
@@ -138,6 +169,15 @@
#define TX_TXCC_MGNFS_MULT_101(n) ((0x4055 | ((n) << 9)) << 2)
#define TX_TXCC_MGNFS_MULT_110(n) ((0x4056 | ((n) << 9)) << 2)
#define TX_TXCC_MGNFS_MULT_111(n) ((0x4057 | ((n) << 9)) << 2)
+#define TX_TXCC_MGNLS_MULT_000(n) ((0x4058 | ((n) << 9)) << 2)
+#define TX_TXCC_MGNLS_MULT_001(n) ((0x4059 | ((n) << 9)) << 2)
+#define TX_TXCC_MGNLS_MULT_010(n) ((0x405a | ((n) << 9)) << 2)
+#define TX_TXCC_MGNLS_MULT_011(n) ((0x405b | ((n) << 9)) << 2)
+#define TX_TXCC_MGNLS_MULT_100(n) ((0x405c | ((n) << 9)) << 2)
+#define TX_TXCC_MGNLS_MULT_101(n) ((0x405d | ((n) << 9)) << 2)
+#define TX_TXCC_MGNLS_MULT_110(n) ((0x405e | ((n) << 9)) << 2)
+#define TX_TXCC_MGNLS_MULT_111(n) ((0x405f | ((n) << 9)) << 2)
+
#define XCVR_DIAG_PLLDRC_CTRL(n) ((0x40e0 | ((n) << 9)) << 2)
#define XCVR_DIAG_BIDI_CTRL(n) ((0x40e8 | ((n) << 9)) << 2)
#define XCVR_DIAG_LANE_FCM_EN_MGN(n) ((0x40f2 | ((n) << 9)) << 2)
@@ -150,10 +190,63 @@
#define TX_RCVDET_ST_TMR(n) ((0x4123 | ((n) << 9)) << 2)
#define TX_DIAG_TX_DRV(n) ((0x41e1 | ((n) << 9)) << 2)
#define TX_DIAG_BGREF_PREDRV_DELAY (0x41e7 << 2)
+
+/* Use this for "n" in macros like "_MULT_XXX" to target the aux channel */
+#define AUX_CH_LANE 8
+
#define TX_ANA_CTRL_REG_1 (0x5020 << 2)
+
+#define TXDA_DP_AUX_EN BIT(15)
+#define AUXDA_SE_EN BIT(14)
+#define TXDA_CAL_LATCH_EN BIT(13)
+#define AUXDA_POLARITY BIT(12)
+#define TXDA_DRV_POWER_ISOLATION_EN BIT(11)
+#define TXDA_DRV_POWER_EN_PH_2_N BIT(10)
+#define TXDA_DRV_POWER_EN_PH_1_N BIT(9)
+#define TXDA_BGREF_EN BIT(8)
+#define TXDA_DRV_LDO_EN BIT(7)
+#define TXDA_DECAP_EN_DEL BIT(6)
+#define TXDA_DECAP_EN BIT(5)
+#define TXDA_UPHY_SUPPLY_EN_DEL BIT(4)
+#define TXDA_UPHY_SUPPLY_EN BIT(3)
+#define TXDA_LOW_LEAKAGE_EN BIT(2)
+#define TXDA_DRV_IDLE_LOWI_EN BIT(1)
+#define TXDA_DRV_CMN_MODE_EN BIT(0)
+
#define TX_ANA_CTRL_REG_2 (0x5021 << 2)
+
+#define AUXDA_DEBOUNCING_CLK BIT(15)
+#define TXDA_LPBK_RECOVERED_CLK_EN BIT(14)
+#define TXDA_LPBK_ISI_GEN_EN BIT(13)
+#define TXDA_LPBK_SERIAL_EN BIT(12)
+#define TXDA_LPBK_LINE_EN BIT(11)
+#define TXDA_DRV_LDO_REDC_SINKIQ BIT(10)
+#define XCVR_DECAP_EN_DEL BIT(9)
+#define XCVR_DECAP_EN BIT(8)
+#define TXDA_MPHY_ENABLE_HS_NT BIT(7)
+#define TXDA_MPHY_SA_MODE BIT(6)
+#define TXDA_DRV_LDO_RBYR_FB_EN BIT(5)
+#define TXDA_DRV_RST_PULL_DOWN BIT(4)
+#define TXDA_DRV_LDO_BG_FB_EN BIT(3)
+#define TXDA_DRV_LDO_BG_REF_EN BIT(2)
+#define TXDA_DRV_PREDRV_EN_DEL BIT(1)
+#define TXDA_DRV_PREDRV_EN BIT(0)
+
#define TXDA_COEFF_CALC_CTRL (0x5022 << 2)
+
+#define TX_HIGH_Z BIT(6)
+#define TX_VMARGIN_OFFSET 3
+#define TX_VMARGIN_MASK 0x7
+#define LOW_POWER_SWING_EN BIT(2)
+#define TX_FCM_DRV_MAIN_EN BIT(1)
+#define TX_FCM_FULL_MARGIN BIT(0)
+
#define TX_DIG_CTRL_REG_2 (0x5024 << 2)
+
+#define TX_HIGH_Z_TM_EN BIT(15)
+#define TX_RESCAL_CODE_OFFSET 0
+#define TX_RESCAL_CODE_MASK 0x3f
+
#define TXDA_CYA_AUXDA_CYA (0x5025 << 2)
#define TX_ANA_CTRL_REG_3 (0x5026 << 2)
#define TX_ANA_CTRL_REG_4 (0x5027 << 2)
@@ -456,54 +549,63 @@ static void tcphy_dp_aux_set_flip(struct rockchip_typec_phy *tcphy)
*/
tx_ana_ctrl_reg_1 = readl(tcphy->base + TX_ANA_CTRL_REG_1);
if (!tcphy->flip)
- tx_ana_ctrl_reg_1 |= BIT(12);
+ tx_ana_ctrl_reg_1 |= AUXDA_POLARITY;
else
- tx_ana_ctrl_reg_1 &= ~BIT(12);
+ tx_ana_ctrl_reg_1 &= ~AUXDA_POLARITY;
writel(tx_ana_ctrl_reg_1, tcphy->base + TX_ANA_CTRL_REG_1);
}
static void tcphy_dp_aux_calibration(struct rockchip_typec_phy *tcphy)
{
+ u16 val;
u16 tx_ana_ctrl_reg_1;
- u16 rdata, rdata2, val;
+ u16 tx_ana_ctrl_reg_2;
+ s32 pu_calib_code;
/* disable txda_cal_latch_en for rewrite the calibration values */
tx_ana_ctrl_reg_1 = readl(tcphy->base + TX_ANA_CTRL_REG_1);
- tx_ana_ctrl_reg_1 &= ~BIT(13);
+ tx_ana_ctrl_reg_1 &= ~TXDA_CAL_LATCH_EN;
writel(tx_ana_ctrl_reg_1, tcphy->base + TX_ANA_CTRL_REG_1);
/*
- * read a resistor calibration code from CMN_TXPUCAL_CTRL[6:0] and
- * write it to TX_DIG_CTRL_REG_2[6:0], and delay 1ms to make sure it
- * works.
+ * read a resistor calibration code from CMN_TXPUCAL_CTRL[5:0] and
+ * write it to TX_DIG_CTRL_REG_2[5:0].
*/
- rdata = readl(tcphy->base + TX_DIG_CTRL_REG_2);
- rdata = rdata & 0xffc0;
-
- rdata2 = readl(tcphy->base + CMN_TXPUCAL_CTRL);
- rdata2 = rdata2 & 0x3f;
+ val = readl(tcphy->base + CMN_TXPUCAL_CTRL);
+ pu_calib_code = CMN_CALIB_CODE_POS(val);
- val = rdata | rdata2;
+ /* write the calibration, then delay 10 ms as sample in docs */
+ val = readl(tcphy->base + TX_DIG_CTRL_REG_2);
+ val &= ~(TX_RESCAL_CODE_MASK << TX_RESCAL_CODE_OFFSET);
+ val |= pu_calib_code << TX_RESCAL_CODE_OFFSET;
writel(val, tcphy->base + TX_DIG_CTRL_REG_2);
- usleep_range(1000, 1050);
+ usleep_range(10000, 10050);
/*
* Enable signal for latch that sample and holds calibration values.
* Activate this signal for 1 clock cycle to sample new calibration
* values.
*/
- tx_ana_ctrl_reg_1 |= BIT(13);
+ tx_ana_ctrl_reg_1 |= TXDA_CAL_LATCH_EN;
writel(tx_ana_ctrl_reg_1, tcphy->base + TX_ANA_CTRL_REG_1);
usleep_range(150, 200);
/* set TX Voltage Level and TX Deemphasis to 0 */
writel(0, tcphy->base + PHY_DP_TX_CTL);
+
/* re-enable decap */
- writel(0x100, tcphy->base + TX_ANA_CTRL_REG_2);
- writel(0x300, tcphy->base + TX_ANA_CTRL_REG_2);
- tx_ana_ctrl_reg_1 |= BIT(3);
+ tx_ana_ctrl_reg_2 = XCVR_DECAP_EN;
+ writel(tx_ana_ctrl_reg_2, tcphy->base + TX_ANA_CTRL_REG_2);
+ udelay(1);
+ tx_ana_ctrl_reg_2 |= XCVR_DECAP_EN_DEL;
+ writel(tx_ana_ctrl_reg_2, tcphy->base + TX_ANA_CTRL_REG_2);
+
+ writel(0, tcphy->base + TX_ANA_CTRL_REG_3);
+
+ tx_ana_ctrl_reg_1 |= TXDA_UPHY_SUPPLY_EN;
writel(tx_ana_ctrl_reg_1, tcphy->base + TX_ANA_CTRL_REG_1);
- tx_ana_ctrl_reg_1 |= BIT(4);
+ udelay(1);
+ tx_ana_ctrl_reg_1 |= TXDA_UPHY_SUPPLY_EN_DEL;
writel(tx_ana_ctrl_reg_1, tcphy->base + TX_ANA_CTRL_REG_1);
writel(0, tcphy->base + TX_ANA_CTRL_REG_5);
@@ -515,44 +617,66 @@ static void tcphy_dp_aux_calibration(struct rockchip_typec_phy *tcphy)
writel(0x1001, tcphy->base + TX_ANA_CTRL_REG_4);
/* re-enables Bandgap reference for LDO */
- tx_ana_ctrl_reg_1 |= BIT(7);
+ tx_ana_ctrl_reg_1 |= TXDA_DRV_LDO_EN;
writel(tx_ana_ctrl_reg_1, tcphy->base + TX_ANA_CTRL_REG_1);
- tx_ana_ctrl_reg_1 |= BIT(8);
+ udelay(5);
+ tx_ana_ctrl_reg_1 |= TXDA_BGREF_EN;
writel(tx_ana_ctrl_reg_1, tcphy->base + TX_ANA_CTRL_REG_1);
/*
* re-enables the transmitter pre-driver, driver data selection MUX,
* and receiver detect circuits.
*/
- writel(0x301, tcphy->base + TX_ANA_CTRL_REG_2);
- writel(0x303, tcphy->base + TX_ANA_CTRL_REG_2);
+ tx_ana_ctrl_reg_2 |= TXDA_DRV_PREDRV_EN;
+ writel(tx_ana_ctrl_reg_2, tcphy->base + TX_ANA_CTRL_REG_2);
+ udelay(1);
+ tx_ana_ctrl_reg_2 |= TXDA_DRV_PREDRV_EN_DEL;
+ writel(tx_ana_ctrl_reg_2, tcphy->base + TX_ANA_CTRL_REG_2);
/*
- * Do some magic undocumented stuff, some of which appears to
- * undo the "re-enables Bandgap reference for LDO" above.
+ * Do all the undocumented magic:
+ * - Turn on TXDA_DP_AUX_EN, whatever that is, even though sample
+ * never shows this going on.
+ * - Turn on TXDA_DECAP_EN (and TXDA_DECAP_EN_DEL) even though
+ * docs say for aux it's always 0.
+ * - Turn off the LDO and BGREF, which we just spent time turning
+ * on above (???).
+ *
+ * Without this magic, things seem worse.
*/
- tx_ana_ctrl_reg_1 |= BIT(15);
- tx_ana_ctrl_reg_1 &= ~BIT(8);
- tx_ana_ctrl_reg_1 &= ~BIT(7);
- tx_ana_ctrl_reg_1 |= BIT(6);
- tx_ana_ctrl_reg_1 |= BIT(5);
+ tx_ana_ctrl_reg_1 |= TXDA_DP_AUX_EN;
+ tx_ana_ctrl_reg_1 |= TXDA_DECAP_EN;
+ tx_ana_ctrl_reg_1 &= ~TXDA_DRV_LDO_EN;
+ tx_ana_ctrl_reg_1 &= ~TXDA_BGREF_EN;
+ writel(tx_ana_ctrl_reg_1, tcphy->base + TX_ANA_CTRL_REG_1);
+ udelay(1);
+ tx_ana_ctrl_reg_1 |= TXDA_DECAP_EN_DEL;
writel(tx_ana_ctrl_reg_1, tcphy->base + TX_ANA_CTRL_REG_1);
-
- writel(0, tcphy->base + TX_ANA_CTRL_REG_3);
- writel(0, tcphy->base + TX_ANA_CTRL_REG_4);
- writel(0, tcphy->base + TX_ANA_CTRL_REG_5);
/*
- * Controls low_power_swing_en, don't set the voltage swing of the
- * driver to 400mv. The values below are peak to peak (differential)
- * values.
+ * Undo the work we did to set the LDO voltage.
+ * This doesn't seem to help nor hurt, but it kinda goes with the
+ * undocumented magic above.
*/
+ writel(0, tcphy->base + TX_ANA_CTRL_REG_4);
+
+ /* Don't set voltage swing to 400 mV peak to peak (differential) */
writel(0, tcphy->base + TXDA_COEFF_CALC_CTRL);
+
+ /* Init TXDA_CYA_AUXDA_CYA for unknown magic reasons */
writel(0, tcphy->base + TXDA_CYA_AUXDA_CYA);
- /* Controls tx_high_z_tm_en */
+ /*
+ * More undocumented magic, presumably the goal of which is to
+ * make the "auxda_source_aux_oen" be ignored and instead to decide
+ * about "high impedance state" based on what software puts in the
+ * register TXDA_COEFF_CALC_CTRL (see TX_HIGH_Z). Since we only
+ * program that register once and we don't set the bit TX_HIGH_Z,
+ * presumably the goal here is that we should never put the analog
+ * driver in high impedance state.
+ */
val = readl(tcphy->base + TX_DIG_CTRL_REG_2);
- val |= BIT(15);
+ val |= TX_HIGH_Z_TM_EN;
writel(val, tcphy->base + TX_DIG_CTRL_REG_2);
}
--
2.14.1.821.g8fa685d3b7-goog
More information about the linux-arm-kernel
mailing list