Hi Stephen,
On 3/23/2018 12:13 AM, Stephen Boyd wrote:
> Quoting Manu Gautam (2018-03-22 01:50:41)
>> QMP PHY for USB mode requires pipe_clk for calibration and PLL lock
>> to take place. This lock is output from PHY to GCC clock_ctl and then
> s/lock/clock/
Yes, will fix typo.
>> fed back to QMP PHY and is output from PHY only after PHY is reset
>> and initialized, hence it can't be enabled too early in initialization
>> sequence.
>>
>> Signed-off-by: Manu Gautam <mgau...@codeaurora.org>
>> ---
>> drivers/phy/qualcomm/phy-qcom-qmp.c | 18 ++++++++++++++++++
>> 1 file changed, 18 insertions(+)
>>
>> diff --git a/drivers/phy/qualcomm/phy-qcom-qmp.c
>> b/drivers/phy/qualcomm/phy-qcom-qmp.c
>> index 6470c5d..73aa282 100644
>> --- a/drivers/phy/qualcomm/phy-qcom-qmp.c
>> +++ b/drivers/phy/qualcomm/phy-qcom-qmp.c
>> @@ -797,8 +797,13 @@ static int qcom_qmp_phy_poweron(struct phy *phy)
>> {
>> struct qmp_phy *qphy = phy_get_drvdata(phy);
>> struct qcom_qmp *qmp = qphy->qmp;
>> + const struct qmp_phy_cfg *cfg = qmp->cfg;
>> int ret;
>>
>> + /* Not needed for USB3 PHY as pipe_clk is enabled from phy_init */
>> + if (cfg->type == PHY_TYPE_USB3)
>> + return 0;
> Would be nice to not even assign phy_poweron when the PHY is USB3 type.
Yes, that should be better.
>
>> +
>> ret = clk_prepare_enable(qphy->pipe_clk);
>> if (ret)
>> dev_err(qmp->dev, "pipe_clk enable failed, err=%d\n", ret);
>> @@ -1008,6 +1013,19 @@ static int qcom_qmp_phy_init(struct phy *phy)
>> status = pcs + cfg->regs[QPHY_PCS_READY_STATUS];
>> mask = cfg->mask_pcs_ready;
>>
>> + /* USB3 PHY requires pipe_clk for PLL lock and calibration */
>> + if (cfg->type == PHY_TYPE_USB3) {
>> + ret = clk_prepare_enable(qphy->pipe_clk);
>> + if (ret)
>> + dev_err(qmp->dev, "pipe_clk enable err=%d\n", ret);
>> + /*
>> + * Ignore this error as pipe_clk might take some time to get
>> + * enabled. In any case following check for PHY PLL lock
>> would
>> + * timeout below if there is a fatal error and clock is not
>> fed
>> + * to PHY
>> + */
>> +
> This is odd. If clk_enable() fails then we don't keep parent clks (or
> the clk that's being operated on) up refcounted and enabled. How is that
> going to work?
I agree. My intent was to ignore only halt_check errors from clock driver.
In any case this condition (to ignore clk_enable error) should probably
be handled in clock driver. And I see that clock driver is already using
halt_check as BRANCH_HALT_DELAY, so I don't need to worry about that
in phy driver.
--
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project
