From patchwork Thu Jul 18 08:18:07 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Timo Aaltonen X-Patchwork-Id: 1133625 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=lists.ubuntu.com (client-ip=91.189.94.19; helo=huckleberry.canonical.com; envelope-from=kernel-team-bounces@lists.ubuntu.com; receiver=) Authentication-Results: ozlabs.org; dmarc=none (p=none dis=none) header.from=ubuntu.com Received: from huckleberry.canonical.com (huckleberry.canonical.com [91.189.94.19]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 45q6Tb68YWz9sNC; Thu, 18 Jul 2019 18:18:27 +1000 (AEST) Received: from localhost ([127.0.0.1] helo=huckleberry.canonical.com) by huckleberry.canonical.com with esmtp (Exim 4.86_2) (envelope-from ) id 1ho1cN-0002EX-15; Thu, 18 Jul 2019 08:18:23 +0000 Received: from youngberry.canonical.com ([91.189.89.112]) by huckleberry.canonical.com with esmtps (TLS1.0:DHE_RSA_AES_128_CBC_SHA1:128) (Exim 4.86_2) (envelope-from ) id 1ho1cE-0002BA-GH for kernel-team@lists.ubuntu.com; Thu, 18 Jul 2019 08:18:14 +0000 Received: from 1.general.tjaalton.uk.vpn ([10.172.193.56] helo=localhost.localdomain) by youngberry.canonical.com with esmtpsa (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.76) (envelope-from ) id 1ho1cE-0000Cx-11 for kernel-team@lists.ubuntu.com; Thu, 18 Jul 2019 08:18:14 +0000 From: Timo Aaltonen To: kernel-team@lists.ubuntu.com Subject: [E][PATCH 2/2] drm/i915: Enable hotplug retry Date: Thu, 18 Jul 2019 11:18:07 +0300 Message-Id: <20190718081809.329-3-tjaalton@ubuntu.com> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20190718081809.329-1-tjaalton@ubuntu.com> References: <20190718081809.329-1-tjaalton@ubuntu.com> MIME-Version: 1.0 X-BeenThere: kernel-team@lists.ubuntu.com X-Mailman-Version: 2.1.20 Precedence: list List-Id: Kernel team discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: kernel-team-bounces@lists.ubuntu.com Sender: "kernel-team" From: José Roberto de Souza BugLink: http://bugs.launchpad.net/bugs/1835001 Right now we are aware of two cases that needs another hotplug retry: - Unpowered type-c dongles - HDMI slow unplug Both have a complete explanation in the code to schedule another run of the hotplug handler. It could have more checks to just trigger the retry in those two specific cases but why would sink signal a long pulse if there is no change? Also the drawback of running the hotplug handler again is really low and that could fix another cases that we are not aware. Also retrying for old DP ports(non-DDI) to make it consistent and not cause CI failures if those systems are connected to chamelium boards that will be used to simulate the issues reported in here. v2: Also retrying for old DP ports(non-DDI)(Imre) v4: Renamed INTEL_HOTPLUG_NOCHANGE to INTEL_HOTPLUG_UNCHANGED to keep it consistent(Rodrigo) Tested-by: Timo Aaltonen Cc: Ville Syrjälä Cc: Imre Deak Cc: Jani Nikula Reviewed-by: Imre Deak Signed-off-by: José Roberto de Souza Link: https://patchwork.freedesktop.org/patch/msgid/20190712005343.24571-2-jose.souza@intel.com (backported from commit bb80c9255770fa1ed54e889a6bee628bdd0f6762 git://anongit.freedesktop.org/drm-intel) Signed-off-by: Timo Aaltonen --- drivers/gpu/drm/i915/intel_ddi.c | 21 +++++++++++++++++++++ drivers/gpu/drm/i915/intel_dp.c | 7 +++++++ drivers/gpu/drm/i915/intel_hdmi.c | 28 +++++++++++++++++++++++++++- 3 files changed, 55 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index 0b3e82da2c00..48b273a946c9 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -4037,6 +4037,7 @@ intel_ddi_hotplug(struct intel_encoder *encoder, struct intel_connector *connector, bool irq_received) { + struct intel_digital_port *dig_port = enc_to_dig_port(&encoder->base); struct drm_modeset_acquire_ctx ctx; enum intel_hotplug_state state; int ret; @@ -4063,6 +4064,26 @@ intel_ddi_hotplug(struct intel_encoder *encoder, drm_modeset_acquire_fini(&ctx); WARN(ret, "Acquiring modeset locks failed with %i\n", ret); + /* + * Unpowered type-c dongles can take some time to boot and be + * responsible, so here giving some time to those dongles to power up + * and then retrying the probe. + * + * On many platforms the HDMI live state signal is known to be + * unreliable, so we can't use it to detect if a sink is connected or + * not. Instead we detect if it's connected based on whether we can + * read the EDID or not. That in turn has a problem during disconnect, + * since the HPD interrupt may be raised before the DDC lines get + * disconnected (due to how the required length of DDC vs. HPD + * connector pins are specified) and so we'll still be able to get a + * valid EDID. To solve this schedule another detection cycle if this + * time around we didn't detect any change in the sink's connection + * status. + */ + if (state == INTEL_HOTPLUG_UNCHANGED && irq_received && + !dig_port->dp.is_mst) + state = INTEL_HOTPLUG_RETRY; + return state; } diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 699b5f3dbd8c..be230d5dce92 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -4738,6 +4738,13 @@ intel_dp_hotplug(struct intel_encoder *encoder, drm_modeset_acquire_fini(&ctx); WARN(ret, "Acquiring modeset locks failed with %i\n", ret); + /* + * Keeping it consistent with intel_ddi_hotplug() and + * intel_hdmi_hotplug(). + */ + if (state == INTEL_HOTPLUG_UNCHANGED && irq_received) + state = INTEL_HOTPLUG_RETRY; + return state; } diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c index 34be2cfd0ec8..4a37c53c027e 100644 --- a/drivers/gpu/drm/i915/intel_hdmi.c +++ b/drivers/gpu/drm/i915/intel_hdmi.c @@ -3026,6 +3026,32 @@ void intel_hdmi_init_connector(struct intel_digital_port *intel_dig_port, DRM_DEBUG_KMS("CEC notifier get failed\n"); } +static enum intel_hotplug_state +intel_hdmi_hotplug(struct intel_encoder *encoder, + struct intel_connector *connector, bool irq_received) +{ + enum intel_hotplug_state state; + + state = intel_encoder_hotplug(encoder, connector, irq_received); + + /* + * On many platforms the HDMI live state signal is known to be + * unreliable, so we can't use it to detect if a sink is connected or + * not. Instead we detect if it's connected based on whether we can + * read the EDID or not. That in turn has a problem during disconnect, + * since the HPD interrupt may be raised before the DDC lines get + * disconnected (due to how the required length of DDC vs. HPD + * connector pins are specified) and so we'll still be able to get a + * valid EDID. To solve this schedule another detection cycle if this + * time around we didn't detect any change in the sink's connection + * status. + */ + if (state == INTEL_HOTPLUG_UNCHANGED && irq_received) + state = INTEL_HOTPLUG_RETRY; + + return state; +} + void intel_hdmi_init(struct drm_i915_private *dev_priv, i915_reg_t hdmi_reg, enum port port) { @@ -3049,7 +3075,7 @@ void intel_hdmi_init(struct drm_i915_private *dev_priv, &intel_hdmi_enc_funcs, DRM_MODE_ENCODER_TMDS, "HDMI %c", port_name(port)); - intel_encoder->hotplug = intel_encoder_hotplug; + intel_encoder->hotplug = intel_hdmi_hotplug; intel_encoder->compute_config = intel_hdmi_compute_config; if (HAS_PCH_SPLIT(dev_priv)) { intel_encoder->disable = pch_disable_hdmi;