Il 28/03/22 00:39, Guillaume Ranquet ha scritto:

From: Markus Schneider-Pargmann <msp@baylibre.com>

This patch adds a DisplayPort driver for the Mediatek mt8195 SoC.

It supports the mt8195, the embedded DisplayPort units. It offers
DisplayPort 1.4 with up to 4 lanes.

The driver shares its iomap range with the mtk-dp-phy driver using
the regmap/syscon facility.

This driver is based on an initial version by
Jason-JH.Lin [off-list ref].

Signed-off-by: Markus Schneider-Pargmann <msp@baylibre.com>
Signed-off-by: Guillaume Ranquet <redacted>
Reported-by: kernel test robot <redacted>

Hello Guillaume,
as you know, there's some more work to do on this driver.

I will also mention here, not on the code, that at this point, your
mtk_dp_aux_transfer() function has something VERY similar to function
drm_dp_dpcd_access(), so I really believe that you can instead use
functions drm_dp_dpcd_read() and drm_dp_dpcd_write(), avoiding code
duplication around.

Please check drivers/gpu/drm/dp/drm_dp.c.

Actually, I think that the best would be DRM_MEDIATEK_DP_TX or DRM_MEDIATEK_DP...
...also, ordering is important, please!

Assuming that your TODO is about the maximum number of lanes,
{ .compatible = "mediatek,mt8195-edp-tx", .data = &some_const_structure },

+#define MTK_DP_MAX_LANES 4
+#define MTK_DP_MAX_LINK_RATE MTK_DP_LINKRATE_HBR3

..snip..

+
+static int mtk_dp_bulk_16bit_write(struct mtk_dp *mtk_dp, u32 offset, u8 *buf,
+				   size_t length)
+{
+	int i;
+	int ret = 0;
+	int num_regs = (length + 1) / 2;
+
+	/* 2 bytes per register */
+	for (i = 0; i < num_regs; i++) {
+		u32 val = buf[i * 2] |
+			  (i * 2 + 1 < length ? buf[i * 2 + 1] << 8 : 0);
+
+		ret = mtk_dp_write(mtk_dp, offset + i * 4, val);
+		if (ret)
+			goto out;

if (ret)
     return ret;

+	}
+
+out:

Remove this label.

+	return ret;
+}
+
+static unsigned long mtk_dp_sip_atf_call(unsigned int cmd, unsigned int para)
+{
+	struct arm_smccc_res res;
+
+	arm_smccc_smc(MTK_DP_SIP_CONTROL_AARCH32, cmd, para, 0, 0, 0, 0, 0,
+		      &res);
+
+	pr_debug("[DPTX]%s cmd 0x%x, p1 0x%x, ret 0x%lx-0x%lx", __func__, cmd,
+		 para, res.a0, res.a1);
+	return res.a1;
+}
+
+static int mtk_dp_msa_bypass_disable(struct mtk_dp *mtk_dp)
+{
+	const u16 bits_to_set =
+		BIT(HTOTAL_SEL_DP_ENC0_P0_SHIFT) |
+		BIT(VTOTAL_SEL_DP_ENC0_P0_SHIFT) |
+		BIT(HSTART_SEL_DP_ENC0_P0_SHIFT) |
+		BIT(VSTART_SEL_DP_ENC0_P0_SHIFT) |
+		BIT(HWIDTH_SEL_DP_ENC0_P0_SHIFT) |
+		BIT(VHEIGHT_SEL_DP_ENC0_P0_SHIFT) |
+		BIT(HSP_SEL_DP_ENC0_P0_SHIFT) | BIT(HSW_SEL_DP_ENC0_P0_SHIFT) |
+		BIT(VSP_SEL_DP_ENC0_P0_SHIFT) | BIT(VSW_SEL_DP_ENC0_P0_SHIFT);
+	return mtk_dp_update_bits(mtk_dp, MTK_DP_ENC0_P0_3030, bits_to_set,
+			   bits_to_set);
+}
+
+#define MTK_UPD_BITS_OR_OUT(mtk_dp, offset, val, mask, ret, label) \
+	do {\
+		ret = mtk_dp_update_bits(mtk_dp, offset, val, mask); \
+		if (ret) \
+			goto label; \
+	} while (0)

I'm sorry, no offense - but this macro is a bit ugly...

I think I understand why you have introduced it, but in my opinion this decreases
human readability a lot, I was even about to point out multiple functions that
you had unused labels before checking this macro, as that was totally unexpected...

In my opinion, this should be open-coded everywhere... yes it makes the file a
bit fatter in terms of amount of text, but eh... it's life :)))

+

....snip....

+
+static int mtk_dp_set_color_format(struct mtk_dp *mtk_dp,
+				   enum mtk_dp_color_format color_format)
+{
+	u32 val;
+	int ret;
+
+	mtk_dp->info.format = color_format;
+
+	/* Update MISC0 */
+	ret = mtk_dp_update_bits(mtk_dp, MTK_DP_ENC0_P0_3034,
+				 color_format << DP_TEST_COLOR_FORMAT_SHIFT,
+			   DP_TEST_COLOR_FORMAT_MASK);
+
+	if (ret)
+		return ret;
+
+	switch (color_format) {
+	case MTK_DP_COLOR_FORMAT_YUV_422:
+		val = PIXEL_ENCODE_FORMAT_DP_ENC0_P0_YCBCR422;
+		break;
+	case MTK_DP_COLOR_FORMAT_YUV_420:
+		val = PIXEL_ENCODE_FORMAT_DP_ENC0_P0_YCBCR420;
+		break;
+	case MTK_DP_COLOR_FORMAT_YONLY:
+	case MTK_DP_COLOR_FORMAT_RAW:
+	case MTK_DP_COLOR_FORMAT_RESERVED:
+	case MTK_DP_COLOR_FORMAT_UNKNOWN:
+		drm_warn(mtk_dp->drm_dev, "Unsupported color format: %d\n",
+			 color_format);
+		fallthrough;

Uhm, are you sure that you should fallthrough?
If we get a color format that we don't understand, the best thing that would
happen is that we'd get scrambled colors... which is something unwanted.

The best idea is to return -EINVAL here, hence refuse to go on with an invalid
color format.

+	case MTK_DP_COLOR_FORMAT_RGB_444:
+	case MTK_DP_COLOR_FORMAT_YUV_444:
+		val = PIXEL_ENCODE_FORMAT_DP_ENC0_P0_RGB;
+		break;
+	}
+
+	return mtk_dp_update_bits(mtk_dp, MTK_DP_ENC0_P0_303C, val,
+			   PIXEL_ENCODE_FORMAT_DP_ENC0_P0_MASK);
+}
+

...snip...

+
+static void mtk_dp_initialize_hpd_detect_settings(struct mtk_dp *mtk_dp)
+{
+	// Debounce threshold

I think I already told you to fix comments style.

+	mtk_dp_update_bits(mtk_dp, MTK_DP_TRANS_P0_3410,
+			   8 << HPD_DEB_THD_DP_TRANS_P0_SHIFT,
+			   HPD_DEB_THD_DP_TRANS_P0_MASK);
+	mtk_dp_update_bits(mtk_dp, MTK_DP_TRANS_P0_3410,
+			   (HPD_INT_THD_DP_TRANS_P0_LOWER_500US |
+			    HPD_INT_THD_DP_TRANS_P0_UPPER_1100US)
+				   << HPD_INT_THD_DP_TRANS_P0_SHIFT,
+			   HPD_INT_THD_DP_TRANS_P0_MASK);
+
+	// Connect threshold 1.5ms + 5 x 0.1ms = 2ms
+	// Disconnect threshold 1.5ms + 5 x 0.1ms = 2ms
+	mtk_dp_update_bits(mtk_dp, MTK_DP_TRANS_P0_3410,
+			   (5 << HPD_DISC_THD_DP_TRANS_P0_SHIFT) |
+				   (5 << HPD_CONN_THD_DP_TRANS_P0_SHIFT),
+			   HPD_DISC_THD_DP_TRANS_P0_MASK |
+				   HPD_CONN_THD_DP_TRANS_P0_MASK);
+	mtk_dp_update_bits(mtk_dp, MTK_DP_TRANS_P0_3430,
+			   HPD_INT_THD_ECO_DP_TRANS_P0_HIGH_BOUND_EXT,
+			   HPD_INT_THD_ECO_DP_TRANS_P0_MASK);
+}
+

..snip..

+
+static int mtk_dp_power_disable(struct mtk_dp *mtk_dp)
+{
+	int ret;
+
+	ret = mtk_dp_write(mtk_dp, MTK_DP_TOP_PWR_STATE, 0);
+
+	if (ret)
+		goto out;
+
+	usleep_range(10, 200);
+	ret = mtk_dp_write(mtk_dp, MTK_DP_0034,
+			   DA_CKM_CKTX0_EN_FORCE_EN | DA_CKM_BIAS_LPF_EN_FORCE_VAL |
+		     DA_CKM_BIAS_EN_FORCE_VAL |
+		     DA_XTP_GLB_LDO_EN_FORCE_VAL |
+		     DA_XTP_GLB_AVD10_ON_FORCE_VAL);
+
+	if (ret)
+		goto out;
+
+	/* Disable RX */
+	ret = mtk_dp_write(mtk_dp, MTK_DP_1040, 0);
+
+	if (ret)
+		goto out;
+
+	ret = mtk_dp_write(mtk_dp, MTK_DP_TOP_MEM_PD,
+			   0x550 | BIT(FUSE_SEL_SHIFT) | BIT(MEM_ISO_EN_SHIFT));
+
+out:

You are using a goto instead of a return, but you're not reverting any change,
so this goto usage is not justified - hence, wrong.

Please simply return ret when anything fails, and return 0 at the end of this
function.

	.......
	(blahblah function code)

	ret = something()
	if (ret)
		return ret;

	return 0;
}

+	return ret;
+}
+

...snip...

+
+static int mtk_dp_dt_parse(struct mtk_dp *mtk_dp,
+			   struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	int ret = 0;
+	void __iomem *base;
+
+	base = devm_platform_ioremap_resource(pdev, 0);
+	if (IS_ERR(base))
+		return PTR_ERR(base);
+
+	mtk_dp->regs = syscon_node_to_regmap(dev->of_node);
+	if (IS_ERR(mtk_dp->regs))
+		return PTR_ERR(mtk_dp->regs);
+
+	//TODO: optional clock?

Well, if it's optional, you should use devm_clk_get_optional(), meaning
that......

+	mtk_dp->dp_tx_clk = devm_clk_get(dev, "faxi");
+	if (IS_ERR(mtk_dp->dp_tx_clk)) {
+		ret = PTR_ERR(mtk_dp->dp_tx_clk);
+		dev_info(dev, "Failed to get dptx clock: %d\n", ret);
+		mtk_dp->dp_tx_clk = NULL;

...I shouldn't see this snippet of code, as it should be a bit different... :)

+	}
+
+	return 0;
+}
+

..snip..

+static void mtk_dp_bridge_atomic_disable(struct drm_bridge *bridge,
+					 struct drm_bridge_state *old_state)
+{
+	struct mtk_dp *mtk_dp = mtk_dp_from_bridge(bridge);
+
+	mtk_dp_video_mute(mtk_dp, true);
+	mtk_dp->state = MTK_DP_STATE_IDLE;
+	mtk_dp->train_state = MTK_DP_TRAIN_STATE_STARTUP;
+
+	mtk_dp->enabled = false;

+	msleep(100);

100 milliseconds is an eternity, why are we sleeping for *so long* here?

Please, either add a comment that fully explains the resons for that, or
remove the sleep entirely: to my eyes, it doesn't look like this sleep is
really needed for anything important because here you are disabling the
bridge and *powering off* the IP entirely.

+	mtk_dp_poweroff(mtk_dp);
+}
+

..snip..

+
+static enum drm_mode_status
+mtk_dp_bridge_mode_valid(struct drm_bridge *bridge,
+			 const struct drm_display_info *info,
+			 const struct drm_display_mode *mode)
+{
+	struct mtk_dp *mtk_dp = mtk_dp_from_bridge(bridge);
+	u32 rx_linkrate = (u32)mtk_dp->train_info.link_rate * 27000;
+	u32 bpp = info->color_formats & DRM_COLOR_FORMAT_YCBCR422 ? 16 : 24;
+
+	if (rx_linkrate * mtk_dp->train_info.lane_count < mode->clock * bpp / 8)
+		return MODE_CLOCK_HIGH;
+
+	if (mode->clock > 600000)
+		return MODE_CLOCK_HIGH;
+
+	//TODO: explain magic number: 62?

Please fix this TODO for v10.

+	if ((mode->clock * 1000) / (mode->htotal * mode->vtotal) > 62)
+		return MODE_CLOCK_HIGH;
+
+	return MODE_OK;
+}
+