public inbox for drm-ai-reviews@public-inbox.freedesktop.org
 help / color / mirror / Atom feed
From: Claude Code Review Bot <claude-review@example.com>
To: dri-devel-reviews@example.com
Subject: Claude review: drm/msm/dp: rework HPD handling
Date: Mon, 25 May 2026 17:13:08 +1000	[thread overview]
Message-ID: <review-patch7-20260524-hpd-refactor-v6-7-cf3ab488dd7b@oss.qualcomm.com> (raw)
In-Reply-To: <20260524-hpd-refactor-v6-7-cf3ab488dd7b@oss.qualcomm.com>

Patch Review

This is the core patch (~600 lines changed). It drops the entire event queue/state machine and replaces it with a threaded IRQ handler that calls `drm_bridge_hpd_notify()` directly.

**Concern (potential bug):** The IRQ thread handler does not check `DP_DP_HPD_REPLUG_INT_MASK`:

```c
+	if (hpd_isr_status & DP_DP_HPD_UNPLUG_INT_MASK)
+		drm_bridge_hpd_notify(..., connector_status_disconnected);
+
+	if (hpd_isr_status & DP_DP_HPD_PLUG_INT_MASK)
+		drm_bridge_hpd_notify(..., connector_status_connected);
+
+	if (hpd_isr_status & DP_DP_IRQ_HPD_INT_MASK)
+		drm_bridge_hpd_notify(..., connector_status_connected);
```

The old code explicitly handled `DP_DP_HPD_REPLUG_INT_MASK` with an unplug+delayed plug sequence. The new code relies on `msm_dp_bridge_hpd_notify()` detecting replug via `msm_dp_aux_is_link_connected()` returning `ISR_HPD_REPLUG_COUNT`. But if the hardware sets only the REPLUG bit (without PLUG or UNPLUG), the IRQ thread will not call `drm_bridge_hpd_notify()` at all and the replug event is silently dropped. The old code previously queued explicit events for this. If hardware always sets PLUG alongside REPLUG this is fine, but that's a hardware assumption that should be documented.

**Observation:** The old replug handling had a 3-tick delay between unplug and plug events. The new `hpd_notify` handler does them back-to-back:

```c
+		if (hpd_link_status == ISR_HPD_REPLUG_COUNT) {
+			msm_dp_hpd_unplug_handle(dp);
+			msm_dp_hpd_plug_handle(dp);
```

This removes the debounce timing. If the display needs settling time between unplug and plug (which is common for DP), this could cause link training failures on replug events.

**Observation:** `msm_dp_snapshot()` loses its `event_mutex` protection. The function reads `dp->power_on` and accesses MMIO registers. Without locking, there's a TOCTOU race where power_on is true at the check but the device gets powered off before the register reads. This was pre-existing but the mutex removal makes it slightly worse.

---

---
Generated by Claude Code Patch Reviewer

  reply	other threads:[~2026-05-25  7:13 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-05-24 10:33 [PATCH v6 00/10] drm/msm/dp: Drop the HPD state machine Dmitry Baryshkov
2026-05-24 10:33 ` [PATCH v6 01/10] drm/msm/dp: fix HPD state status bit shift value Dmitry Baryshkov
2026-05-25  7:13   ` Claude review: " Claude Code Review Bot
2026-05-24 10:33 ` [PATCH v6 02/10] drm/msm/dp: Fix the ISR_* enum values Dmitry Baryshkov
2026-05-25  7:13   ` Claude review: " Claude Code Review Bot
2026-05-24 10:33 ` [PATCH v6 03/10] drm/msm/dp: Read DPCD and sink count in bridge detect() Dmitry Baryshkov
2026-05-25  7:13   ` Claude review: " Claude Code Review Bot
2026-05-24 10:33 ` [PATCH v6 04/10] drm/msm/dp: Move link training to atomic_enable() Dmitry Baryshkov
2026-05-25  7:13   ` Claude review: " Claude Code Review Bot
2026-05-24 10:33 ` [PATCH v6 05/10] drm/msm/dp: Drop EV_USER_NOTIFICATION Dmitry Baryshkov
2026-05-25  7:13   ` Claude review: " Claude Code Review Bot
2026-05-24 10:33 ` [PATCH v6 06/10] drm/msm/dp: drop event data Dmitry Baryshkov
2026-05-25  7:13   ` Claude review: " Claude Code Review Bot
2026-05-24 10:33 ` [PATCH v6 07/10] drm/msm/dp: rework HPD handling Dmitry Baryshkov
2026-05-25  7:13   ` Claude Code Review Bot [this message]
2026-05-24 10:33 ` [PATCH v6 08/10] drm/msm/dp: Add sink_count to debug logs Dmitry Baryshkov
2026-05-25  7:13   ` Claude review: " Claude Code Review Bot
2026-05-24 10:33 ` [PATCH v6 09/10] drm/msm/dp: turn link_ready into plugged Dmitry Baryshkov
2026-05-25  7:13   ` Claude review: " Claude Code Review Bot
2026-05-24 10:33 ` [PATCH v6 10/10] drm/msm/dp: clear EDID on display unplug Dmitry Baryshkov
2026-05-25  7:13   ` Claude review: " Claude Code Review Bot
2026-05-24 21:32 ` [PATCH v6 00/10] drm/msm/dp: Drop the HPD state machine Val Packett
2026-05-25  7:13 ` Claude review: " Claude Code Review Bot
  -- strict thread matches above, loose matches on Subject: below --
2026-03-14  1:09 [PATCH v5 00/10] " Dmitry Baryshkov
2026-03-14  1:09 ` [PATCH v5 07/10] drm/msm/dp: rework HPD handling Dmitry Baryshkov
2026-03-16  2:12   ` Claude review: " Claude Code Review Bot

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=review-patch7-20260524-hpd-refactor-v6-7-cf3ab488dd7b@oss.qualcomm.com \
    --to=claude-review@example.com \
    --cc=dri-devel-reviews@example.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox