public inbox for stable@vger.kernel.org
 help / color / mirror / Atom feed
From: Saikiran <bjsaikiran@gmail.com>
To: linux-media@vger.kernel.org
Cc: linux-arm-msm@vger.kernel.org, rfoss@kernel.org,
	todor.too@gmail.com, bryan.odonoghue@linaro.org, bod@kernel.org,
	vladimir.zapolskiy@linaro.org, hansg@kernel.org,
	sakari.ailus@linux.intel.com, mchehab@kernel.org,
	stable@vger.kernel.org
Subject: [PATCH v3 0/3] media: i2c: ov02c10: Fix brownouts and power sequence
Date: Mon, 26 Jan 2026 23:04:41 +0530	[thread overview]
Message-ID: <20260126173444.10228-1-bjsaikiran@gmail.com> (raw)

This series addresses stability issues with the OV02C10 sensor on Qualcomm
Snapdragon X Elite (X1E80100) platforms, specifically the Lenovo Yoga Slim 7x.

Note: This series supersedes all previous patch sets submitted by me regarding
OV02C10 / media stability and cleanup on X1E80100 (including the 'pipeline
lock' and 'brownout' series). Please disregard prior versions.

Problem 1: Brownouts during rapid cycling
On this platform, the RPMh-controlled regulators lack active discharge, taking
~2.3s to passively discharge. Rapid close/open cycles (e.g., WebRTC checks)
re-enable regulators while the rails are floating, causing the sensor to hang.
Previous attempts to manage this via open-loop delays in power_on were deemed
incorrect.

Problem 2: Incorrect Power Sequence
The driver was not strictly following the datasheet power-up timing (T1/T2),
potentially leading to race conditions between the reset pin and power rails.

Problem 3: Race condition on removal
The remove() function freed resources before powering off the device, causing
use-after-free errors if userspace (PipeWire) accessed controls during removal.

Solution in v3:
1. Implement Runtime PM Autosuspend (1000ms). This prevents the driver from
   cutting power during rapid user interactions, sidestepping the slow
   regulator discharge window entirely. (Patch 3)
2. Enforce strict datasheet power-on sequencing with ample delays to satisfy
   maintainer requirements for clean boot. (Patch 2)
3. Fix the remove() race condition by reordering cleanup. (Patch 1)

Changes in v3:
- Dropped the "always-on" regulator patch from v2.
- Added Runtime PM Autosuspend support (Patch 3).
- Added strict power-on sequencing with 10ms/20ms delays (Patch 2).
- Added fix for use-after-free in remove() (Patch 1).

Link: https://lore.kernel.org/linux-media/20260125171745.484806-1-bjsaikiran@gmail.com/T/#t [1]

Saikiran (3):
  media: i2c: ov02c10: Fix use-after-free in remove function
  media: i2c: ov02c10: Correct power-on sequence and timing
  media: i2c: ov02c10: Use runtime PM autosuspend to avoid brownouts

 drivers/media/i2c/ov02c10.c | 70 ++++++++++++++++++++++++++++++-------
 1 file changed, 58 insertions(+), 12 deletions(-)

             reply	other threads:[~2026-01-26 17:34 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-01-26 17:34 Saikiran [this message]
2026-01-26 17:34 ` [PATCH v3 1/3] media: i2c: ov02c10: Fix use-after-free in remove function Saikiran
2026-01-27 10:30   ` Hans de Goede
2026-01-26 17:34 ` [PATCH v3 2/3] media: i2c: ov02c10: Correct power-on sequence and timing Saikiran
2026-01-27 10:40   ` Hans de Goede
2026-01-27 10:47     ` Bryan O'Donoghue
2026-01-27 10:50     ` Hans de Goede
2026-01-26 17:34 ` [PATCH v3 3/3] media: i2c: ov02c10: Use runtime PM autosuspend to avoid brownouts Saikiran
2026-01-27  9:46   ` Bryan O'Donoghue
2026-01-27 10:43     ` Hans de Goede
2026-01-27 10:44     ` Hans de Goede
     [not found]     ` <CAAFDt1tsyvtAa84bFK2Hq5yG_F15SUUseBd5Xi-DB8GnUj7+7A@mail.gmail.com>
2026-01-27 10:50       ` Bryan O'Donoghue
     [not found]         ` <CAAFDt1vKn5ssoTQZduGKb5eOeN74P=FVk9f01go1d-JS71Zt0A@mail.gmail.com>
2026-01-27 11:06           ` Bryan O'Donoghue
2026-01-27 11:11             ` Bryan O'Donoghue
2026-01-27 16:20               ` Saikiran B

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=20260126173444.10228-1-bjsaikiran@gmail.com \
    --to=bjsaikiran@gmail.com \
    --cc=bod@kernel.org \
    --cc=bryan.odonoghue@linaro.org \
    --cc=hansg@kernel.org \
    --cc=linux-arm-msm@vger.kernel.org \
    --cc=linux-media@vger.kernel.org \
    --cc=mchehab@kernel.org \
    --cc=rfoss@kernel.org \
    --cc=sakari.ailus@linux.intel.com \
    --cc=stable@vger.kernel.org \
    --cc=todor.too@gmail.com \
    --cc=vladimir.zapolskiy@linaro.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