From: Bjorn Andersson <bjorn.andersson@linaro.org>
To: Kuogee Hsieh <khsieh@codeaurora.org>
Cc: robdclark@gmail.com, sean@poorly.run, swboyd@chromium.org,
vkoul@kernel.org, agross@kernel.org, robh+dt@kernel.org,
devicetree@vger.kernel.org, abhinavk@codeaurora.org,
aravindh@codeaurora.org, freedreno@lists.freedesktop.org,
linux-arm-msm@vger.kernel.org, linux-kernel@vger.kernel.org,
kishon@ti.com, p.zabel@pengutronix.de
Subject: Re: [PATCH] phy: qcom-qmp: add support for voltage and pre emphesis swing
Date: Wed, 25 Aug 2021 13:57:52 -0700 [thread overview]
Message-ID: <YSau0HrLMZOAu2Nx@ripper> (raw)
In-Reply-To: <1629847775-16767-1-git-send-email-khsieh@codeaurora.org>
On Tue 24 Aug 16:29 PDT 2021, Kuogee Hsieh wrote:
> Add voltage and pre emphesis swing tables so that voltage and
> pre emphsis swing level can be configured base on link rate.
>
I think it would be nice if $subject, or at least the commit message
mentioned that this relates to the DisplayPort part of the QMP driver.
Also the commit message states that this allows someone/something to
configure the properties based on link rate. But it doesn't state why
this is needed.
> Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
> ---
> drivers/phy/qualcomm/phy-qcom-qmp.c | 95 ++++++++++++++++++++++++++++++++-----
> 1 file changed, 82 insertions(+), 13 deletions(-)
>
> diff --git a/drivers/phy/qualcomm/phy-qcom-qmp.c b/drivers/phy/qualcomm/phy-qcom-qmp.c
> index 31036aa..52bab6e 100644
> --- a/drivers/phy/qualcomm/phy-qcom-qmp.c
> +++ b/drivers/phy/qualcomm/phy-qcom-qmp.c
> @@ -1916,7 +1916,7 @@ static const struct qmp_phy_init_tbl qmp_v4_dp_tx_tbl[] = {
> QMP_PHY_INIT_CFG(QSERDES_V4_TX_RES_CODE_LANE_OFFSET_RX, 0x11),
> QMP_PHY_INIT_CFG(QSERDES_V4_TX_TX_BAND, 0x4),
> QMP_PHY_INIT_CFG(QSERDES_V4_TX_TX_POL_INV, 0x0a),
> - QMP_PHY_INIT_CFG(QSERDES_V4_TX_TX_DRV_LVL, 0x2a),
> + QMP_PHY_INIT_CFG(QSERDES_V4_TX_TX_DRV_LVL, 0x22),
Why is this initial value changed in order to make the swing and
emphasis configurable?
> QMP_PHY_INIT_CFG(QSERDES_V4_TX_TX_EMP_POST1_LVL, 0x20),
> };
>
> @@ -3727,6 +3727,81 @@ static int qcom_qmp_v3_dp_phy_calibrate(struct qmp_phy *qphy)
>
> return 0;
> }
> +/*
> + * 0x20 deducted from tables
> + *
> + * swing_value |= DP_PHY_TXn_TX_DRV_LVL_MUX_EN;
> + * pre_emphasis_value |= DP_PHY_TXn_TX_EMP_POST1_LVL_MUX_EN;
How about rewriting this as something like
"The values in these tables are given without MUX_EN (0x20) bit set"
?
> +*/
> +static const u8 qmp_dp_v4_pre_emphasis_hbr3_hbr2[4][4] = {
> + /* p0 p1 p2 p3 */
> + { 0x00, 0x0c, 0x15, 0x1b }, /* s0 */
> + { 0x02, 0x0e, 0x16, 0xff }, /* s1 */
> + { 0x02, 0x11, 0xff, 0xff }, /* s2 */
> + { 0x04, 0xff, 0xff, 0xff } /* s3 */
> +};
> +
> +static const u8 qmp_dp_v4_voltage_swing_hbr3_hbr2[4][4] = {
> + /* p0 p1 p2 p3 */
> + { 0x02, 0x12, 0x16, 0x1a }, /* s0 */
> + { 0x09, 0x19, 0x1f, 0xff }, /* s1 */
> + { 0x10, 0x1f, 0xff, 0xff }, /* s2 */
> + { 0x1f, 0xff, 0xff, 0xff } /* s3 */
> +};
> +
> +static const u8 qmp_dp_v4_pre_emphasis_hbr_rbr[4][4] = {
> + /* p0 p1 p2 p3 */
> + { 0x00, 0x0e, 0x15, 0x1b }, /* s0 */
> + { 0x00, 0x0e, 0x15, 0xff }, /* s1 */
> + { 0x00, 0x0e, 0xff, 0xff }, /* s2 */
> + { 0x04, 0xff, 0xff, 0xff } /* s3 */
> +};
> +
> +static const u8 qmp_dp_v4_voltage_swing_hbr_rbr[4][4] = {
> + /* p0 p1 p2 p3 */
> + { 0x08, 0x0f, 0x16, 0x1f }, /* s0 */
> + { 0x11, 0x1e, 0x1f, 0xff }, /* s1 */
> + { 0x16, 0x1f, 0xff, 0xff }, /* s2 */
> + { 0x1f, 0xff, 0xff, 0xff } /* s3 */
> +};
> +
> +static int qcom_qmp_v4_phy_configure_dp_swing(struct qmp_phy *qphy,
> + unsigned int drv_lvl_reg, unsigned int emp_post_reg)
> +{
> + const struct phy_configure_opts_dp *dp_opts = &qphy->dp_opts;
> + unsigned int v_level = 0, p_level = 0;
> + u8 voltage_swing_cfg, pre_emphasis_cfg;
The "_cfg" suffix on these variables doesn't really add any value.
Frankly, calling them "voltage" (or "swing") and "emphasis" seems just
as expressive, but easier to read.
> + int i;
> +
> + for (i = 0; i < dp_opts->lanes; i++) {
> + v_level = max(v_level, dp_opts->voltage[i]);
> + p_level = max(p_level, dp_opts->pre[i]);
> + }
> +
> +
> + if (dp_opts->link_rate <= 2700) {
> + voltage_swing_cfg = qmp_dp_v4_voltage_swing_hbr_rbr[v_level][p_level];
> + pre_emphasis_cfg = qmp_dp_v4_pre_emphasis_hbr_rbr[v_level][p_level];
> + } else {
> + voltage_swing_cfg = qmp_dp_v4_voltage_swing_hbr3_hbr2[v_level][p_level];
> + pre_emphasis_cfg = qmp_dp_v4_pre_emphasis_hbr3_hbr2[v_level][p_level];
> + }
> +
> + /* TODO: Move check to config check */
> + if (voltage_swing_cfg == 0xFF && pre_emphasis_cfg == 0xFF)
Why is this && and not || ?
> + return -EINVAL;
> +
> + /* Enable MUX to use Cursor values from these registers */
> + voltage_swing_cfg |= DP_PHY_TXn_TX_DRV_LVL_MUX_EN;
> + pre_emphasis_cfg |= DP_PHY_TXn_TX_EMP_POST1_LVL_MUX_EN;
> +
> + writel(voltage_swing_cfg, qphy->tx + drv_lvl_reg);
> + writel(pre_emphasis_cfg, qphy->tx + emp_post_reg);
> + writel(voltage_swing_cfg, qphy->tx2 + drv_lvl_reg);
> + writel(pre_emphasis_cfg, qphy->tx2 + emp_post_reg);
> +
This function is called once, so why is drv_lvl_reg and emp_post_reg
variables passed to the function, rather than just using the defines
directly?
Regards,
Bjorn
> + return 0;
> +}
>
> static void qcom_qmp_v4_phy_dp_aux_init(struct qmp_phy *qphy)
> {
> @@ -3757,14 +3832,7 @@ static void qcom_qmp_v4_phy_dp_aux_init(struct qmp_phy *qphy)
>
> static void qcom_qmp_v4_phy_configure_dp_tx(struct qmp_phy *qphy)
> {
> - /* Program default values before writing proper values */
> - writel(0x27, qphy->tx + QSERDES_V4_TX_TX_DRV_LVL);
> - writel(0x27, qphy->tx2 + QSERDES_V4_TX_TX_DRV_LVL);
> -
> - writel(0x20, qphy->tx + QSERDES_V4_TX_TX_EMP_POST1_LVL);
> - writel(0x20, qphy->tx2 + QSERDES_V4_TX_TX_EMP_POST1_LVL);
> -
> - qcom_qmp_phy_configure_dp_swing(qphy,
> + qcom_qmp_v4_phy_configure_dp_swing(qphy,
> QSERDES_V4_TX_TX_DRV_LVL,
> QSERDES_V4_TX_TX_EMP_POST1_LVL);
> }
> @@ -3885,6 +3953,9 @@ static int qcom_qmp_v4_phy_configure_dp_phy(struct qmp_phy *qphy)
> writel(drvr1_en, qphy->tx2 + QSERDES_V4_TX_HIGHZ_DRVR_EN);
> writel(bias1_en, qphy->tx2 + QSERDES_V4_TX_TRANSCEIVER_BIAS_EN);
>
> + writel(0x0a, qphy->tx + QSERDES_V4_TX_TX_POL_INV);
> + writel(0x0a, qphy->tx2 + QSERDES_V4_TX_TX_POL_INV);
> +
> writel(0x18, qphy->pcs + QSERDES_DP_PHY_CFG);
> udelay(2000);
> writel(0x19, qphy->pcs + QSERDES_DP_PHY_CFG);
> @@ -3896,11 +3967,9 @@ static int qcom_qmp_v4_phy_configure_dp_phy(struct qmp_phy *qphy)
> 10000))
> return -ETIMEDOUT;
>
> - writel(0x0a, qphy->tx + QSERDES_V4_TX_TX_POL_INV);
> - writel(0x0a, qphy->tx2 + QSERDES_V4_TX_TX_POL_INV);
>
> - writel(0x27, qphy->tx + QSERDES_V4_TX_TX_DRV_LVL);
> - writel(0x27, qphy->tx2 + QSERDES_V4_TX_TX_DRV_LVL);
> + writel(0x22, qphy->tx + QSERDES_V4_TX_TX_DRV_LVL);
> + writel(0x22, qphy->tx2 + QSERDES_V4_TX_TX_DRV_LVL);
>
> writel(0x20, qphy->tx + QSERDES_V4_TX_TX_EMP_POST1_LVL);
> writel(0x20, qphy->tx2 + QSERDES_V4_TX_TX_EMP_POST1_LVL);
> --
> The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
> a Linux Foundation Collaborative Project
>
prev parent reply other threads:[~2021-08-25 20:56 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-08-24 23:29 [PATCH] phy: qcom-qmp: add support for voltage and pre emphesis swing Kuogee Hsieh
2021-08-25 17:49 ` Stephen Boyd
2021-08-27 20:00 ` khsieh
2021-08-25 20:57 ` Bjorn Andersson [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=YSau0HrLMZOAu2Nx@ripper \
--to=bjorn.andersson@linaro.org \
--cc=abhinavk@codeaurora.org \
--cc=agross@kernel.org \
--cc=aravindh@codeaurora.org \
--cc=devicetree@vger.kernel.org \
--cc=freedreno@lists.freedesktop.org \
--cc=khsieh@codeaurora.org \
--cc=kishon@ti.com \
--cc=linux-arm-msm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=p.zabel@pengutronix.de \
--cc=robdclark@gmail.com \
--cc=robh+dt@kernel.org \
--cc=sean@poorly.run \
--cc=swboyd@chromium.org \
--cc=vkoul@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).