All of lore.kernel.org
 help / color / mirror / Atom feed
From: Wayne Lin <Wayne.Lin@amd.com>
To: <amd-gfx@lists.freedesktop.org>
Cc: <alexander.deucher@amd.com>, <Harry.Wentland@amd.com>,
	<nicholas.kazlauskas@amd.com>, <Rodrigo.Siqueira@amd.com>,
	<wayne.lin@amd.com>, <stylon.wang@amd.com>, <jude.shih@amd.com>,
	<jimmy.kizito@amd.com>, <meenakshikumar.somasundaram@amd.com>,
	Jimmy Kizito <Jimmy.Kizito@amd.com>, Jun Lei <Jun.Lei@amd.com>,
	Wayne Lin <Wayne.Lin@amd.com>
Subject: [PATCH v2 13/23] drm/amd/display: Implement end of training for hop in DPIA display path
Date: Tue, 5 Oct 2021 15:51:55 +0800	[thread overview]
Message-ID: <20211005075205.3467938-14-Wayne.Lin@amd.com> (raw)
In-Reply-To: <20211005075205.3467938-1-Wayne.Lin@amd.com>

From: Jimmy Kizito <Jimmy.Kizito@amd.com>

[Why & How]
Clear training pattern sequence for hop in display path once clock
recovery and equalization phases of DP tunnel link training completed.

Reviewed-by: Jun Lei <Jun.Lei@amd.com>
Acked-by: Wayne Lin <Wayne.Lin@amd.com>
Acked-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: Jimmy Kizito <Jimmy.Kizito@amd.com>
---
 .../gpu/drm/amd/display/dc/core/dc_link_dp.c  | 10 ++-
 .../drm/amd/display/dc/core/dc_link_dpia.c    | 77 ++++++++++++++++++-
 2 files changed, 82 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
index 7f6fd0a3bf18..bfba1d2c6a18 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
@@ -2391,14 +2391,20 @@ bool perform_link_training_with_retries(
 			dc_link_dp_perform_link_training_skip_aux(link, &current_setting);
 			return true;
 		} else {
-			if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA)
+			/** @todo Consolidate USB4 DP and DPx.x training. */
+			if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA) {
 				status = dc_link_dpia_perform_link_training(link,
 									    &current_setting,
 									    skip_video_pattern);
-			else
+
+				/* Transmit idle pattern once training successful. */
+				if (status == LINK_TRAINING_SUCCESS)
+					dp_set_hw_test_pattern(link, DP_TEST_PATTERN_VIDEO_MODE, NULL, 0);
+			} else {
 				status = dc_link_dp_perform_link_training(link,
 									  &current_setting,
 									  skip_video_pattern);
+			}
 
 			if (status == LINK_TRAINING_SUCCESS)
 				return true;
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dpia.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dpia.c
index fa7539916c77..4b1ad057bd1f 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dpia.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dpia.c
@@ -730,13 +730,84 @@ static enum link_training_result dpia_training_eq_phase(struct dc_link *link,
 }
 
 /* End training of specified hop in display path. */
+static enum dc_status dpcd_clear_lt_pattern(struct dc_link *link, uint32_t hop)
+{
+	union dpcd_training_pattern dpcd_pattern = { {0} };
+	uint32_t dpcd_tps_offset = DP_TRAINING_PATTERN_SET;
+	enum dc_status status;
+
+	if (hop != DPRX)
+		dpcd_tps_offset = DP_TRAINING_PATTERN_SET_PHY_REPEATER1 +
+			((DP_REPEATER_CONFIGURATION_AND_STATUS_SIZE) * (hop - 1));
+
+	status = core_link_write_dpcd(link,
+			DP_TRAINING_PATTERN_SET,
+			&dpcd_pattern.raw,
+			sizeof(dpcd_pattern.raw));
+
+	return status;
+}
+
+/* End training of specified hop in display path.
+ *
+ * In transparent LTTPR mode:
+ * - driver clears training pattern for the specified hop in DPCD.
+ * In non-transparent LTTPR mode:
+ * - in addition to clearing training pattern, driver issues USB4 tunneling
+ * (SET_CONFIG) messages to notify DPOA when training is done for first hop
+ * (DPTX-to-DPIA) and last hop (DPRX).
+ *
+ * @param link DPIA link being trained.
+ * @param hop Hop in display path. DPRX = 0.
+ */
 static enum link_training_result dpia_training_end(struct dc_link *link,
 		uint32_t hop)
 {
-	enum link_training_result result;
+	enum link_training_result result = LINK_TRAINING_SUCCESS;
+	uint8_t repeater_cnt = 0; /* Number of hops/repeaters in display path. */
+	enum dc_status status;
 
-	/** @todo Fail until implemented. */
-	result = LINK_TRAINING_ABORT;
+	if (link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) {
+		repeater_cnt = dp_convert_to_count(link->dpcd_caps.lttpr_caps.phy_repeater_cnt);
+
+		if (hop == repeater_cnt) { /* DPTX-to-DPIA */
+			/* Send SET_CONFIG(SET_TRAINING:0xff) to notify DPOA that
+			 * DPTX-to-DPIA hop trained. No DPCD write needed for first hop.
+			 */
+			status = core_link_send_set_config(link,
+					DPIA_SET_CFG_SET_TRAINING,
+					DPIA_TS_UFP_DONE);
+			if (status != DC_OK)
+				result = LINK_TRAINING_ABORT;
+		} else { /* DPOA-to-x */
+			/* Write 0x0 to TRAINING_PATTERN_SET */
+			status = dpcd_clear_lt_pattern(link, hop);
+			if (status != DC_OK)
+				result = LINK_TRAINING_ABORT;
+		}
+
+		/* Notify DPOA that non-transparent link training of DPRX done. */
+		if (hop == DPRX && result != LINK_TRAINING_ABORT) {
+			status = core_link_send_set_config(link,
+					DPIA_SET_CFG_SET_TRAINING,
+					DPIA_TS_DPRX_DONE);
+			if (status != DC_OK)
+				result = LINK_TRAINING_ABORT;
+		}
+
+	} else { /* non-LTTPR or transparent LTTPR. */
+		/* Write 0x0 to TRAINING_PATTERN_SET */
+		status = dpcd_clear_lt_pattern(link, hop);
+		if (status != DC_OK)
+			result = LINK_TRAINING_ABORT;
+	}
+
+	DC_LOG_HW_LINK_TRAINING("%s\n DPIA(%d) end\n - hop(%d)\n - result(%d)\n - LTTPR mode(%d)\n",
+				__func__,
+				link->link_id.enum_id - ENUM_ID_1,
+				hop,
+				result,
+				link->lttpr_mode);
 
 	return result;
 }
-- 
2.25.1


  parent reply	other threads:[~2021-10-05  7:58 UTC|newest]

Thread overview: 39+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-10-05  7:51 [PATCH v2 00/23] USB4 DP tunneling Wayne Lin
2021-10-05  7:51 ` [PATCH v2 01/23] drm/amd/display: Update link encoder object creation Wayne Lin
2021-10-05 14:15   ` Harry Wentland
2021-10-05 15:13     ` Harry Wentland
2021-10-05  7:51 ` [PATCH v2 02/23] drm/amd/display: USB4 DPIA enumeration and AUX Tunneling Wayne Lin
2021-10-05  7:51 ` [PATCH v2 03/23] drm/amd/display: Support for DMUB HPD and HPD RX interrupt handling Wayne Lin
2021-10-05  7:51 ` [PATCH v2 04/23] drm/amd/display: Support USB4 dynamic link encoder selection Wayne Lin
2021-10-05 15:46   ` Harry Wentland
2021-10-05  7:51 ` [PATCH v2 05/23] drm/amd/display: Support USB4 for display endpoint control path Wayne Lin
2021-10-05 15:50   ` Harry Wentland
2021-10-05  7:51 ` [PATCH v2 06/23] drm/amd/display: Support DP tunneling when DPRX detection Wayne Lin
2021-10-05 15:57   ` Harry Wentland
2021-10-05  7:51 ` [PATCH v2 07/23] drm/amd/display: Update training parameters for DPIA links Wayne Lin
2021-10-05 15:58   ` Harry Wentland
2021-10-05  7:51 ` [PATCH v2 08/23] drm/amd/display: Support USB4 when DP link training Wayne Lin
2021-10-05 15:59   ` Harry Wentland
2021-10-05  7:51 ` [PATCH v2 09/23] drm/amd/display: Implement DPIA training loop Wayne Lin
2021-10-05  7:51 ` [PATCH v2 10/23] drm/amd/display: Implement DPIA link configuration Wayne Lin
2021-10-05  7:51 ` [PATCH v2 11/23] drm/amd/display: Implement DPIA clock recovery phase Wayne Lin
2021-10-07 10:00   ` Mike Lothian
2021-10-08  8:32     ` Lin, Wayne
2021-10-05  7:51 ` [PATCH v2 12/23] drm/amd/display: Implement DPIA equalisation phase Wayne Lin
2021-10-05  7:51 ` Wayne Lin [this message]
2021-10-05  7:51 ` [PATCH v2 14/23] drm/amd/display: Support for SET_CONFIG processing with DMUB Wayne Lin
2021-10-05  7:51 ` [PATCH v2 15/23] drm/amd/display: Read USB4 DP tunneling data from DPCD Wayne Lin
2021-10-05  7:51 ` [PATCH v2 16/23] drm/amd/display: Add dpia debug options Wayne Lin
2021-10-05  7:51 ` [PATCH v2 17/23] drm/amd/display: Support for SET_CONFIG processing with DMUB Wayne Lin
2021-10-05  7:52 ` [PATCH v2 18/23] drm/amd/display: Fix DIG_HPD_SELECT for USB4 display endpoints Wayne Lin
2021-10-05  7:52 ` [PATCH v2 19/23] drm/amd/display: Add debug flags for USB4 DP link training Wayne Lin
2021-10-05 17:10   ` Harry Wentland
2021-10-06 10:14     ` Lin, Wayne
2021-10-06 14:02       ` Harry Wentland
2021-10-07  9:22         ` Lin, Wayne
2021-10-05  7:52 ` [PATCH v2 20/23] drm/amd/display: Fix for access for ddc pin and aux engine Wayne Lin
2021-10-05  7:52 ` [PATCH v2 21/23] drm/amd/display: Deadlock/HPD Status/Crash Bug Fix Wayne Lin
2021-10-05  7:52 ` [PATCH v2 22/23] drm/amd/display: Fix USB4 Aux via DMUB terminate unexpectedly Wayne Lin
2021-10-05  7:52 ` [PATCH v2 23/23] drm/amd/display: USB4 bring up set correct address Wayne Lin
2021-10-05 17:13 ` [PATCH v2 00/23] USB4 DP tunneling Harry Wentland
2021-10-06 10:14   ` Lin, Wayne

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=20211005075205.3467938-14-Wayne.Lin@amd.com \
    --to=wayne.lin@amd.com \
    --cc=Harry.Wentland@amd.com \
    --cc=Jun.Lei@amd.com \
    --cc=Rodrigo.Siqueira@amd.com \
    --cc=alexander.deucher@amd.com \
    --cc=amd-gfx@lists.freedesktop.org \
    --cc=jimmy.kizito@amd.com \
    --cc=jude.shih@amd.com \
    --cc=meenakshikumar.somasundaram@amd.com \
    --cc=nicholas.kazlauskas@amd.com \
    --cc=stylon.wang@amd.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.