Devicetree
 help / color / mirror / Atom feed
From: Dave Stevenson <dave.stevenson@raspberrypi.com>
To: Sakari Ailus <sakari.ailus@linux.intel.com>,
	 Mauro Carvalho Chehab <mchehab@kernel.org>,
	Rob Herring <robh@kernel.org>,
	 Krzysztof Kozlowski <krzk+dt@kernel.org>,
	 Conor Dooley <conor+dt@kernel.org>,
	 Richard Acayan <mailingradian@gmail.com>
Cc: linux-media@vger.kernel.org, linux-kernel@vger.kernel.org,
	 David Heidelberg <david@ixit.cz>,
	 Jacopo Mondi <jacopo.mondi@ideasonboard.com>,
	devicetree@vger.kernel.org,
	 Dave Stevenson <dave.stevenson@raspberrypi.com>,
	 Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com>
Subject: [PATCH v3 00/20] media/imx355: General code cleanups, and adding support for 2 lane operation
Date: Wed, 08 Jul 2026 15:48:36 +0100	[thread overview]
Message-ID: <20260708-media-imx355-v3-0-9df386a623d7@raspberrypi.com> (raw)

This series came about as we had a customer ask about using imx355,
so I had a look at what was going on.

There was significant duplication between the various mode register lists,
unused parameters around, and no support for advertising the crop
rectangles through get_selection. It also wasn't using V4L2_CCI so
had all the boilerplate for writing registers.
The customer also wanted to run the sensor over only 2 CSI2 data lanes
which required some rework of the clock setup so that the MIPI link
frequency could be run at 445MHz instead of the 360MHz used for 4 lane
mode.

V2 has some fairly substantial differences to V1 as I obtained a datasheet.
Whilst some things worked perfectly, the datasheet declared them as
prohibited. This was largely using dual PLL mode in 2 lane mode, and
LLP is restricted to specific values. Link frequencies are therefore now
derived from the PLL configuration, and pixel rate is computed from that.

Being able to refer against the datasheet, I've pulled an updated version
of David Heidelberg's patch for 24MHz clock support into this series.

I've tried to keep it that the earlier patches are largely cleanups or
minor improvements, and then we get adding 2 lane support, and converting
to using subdev state as they are larger patches.

---
Changes in v3:
- Fixed a number of issues picked up by sashiko-bot
  - Unintentionally dropped writing reg 0x030b, which is at the default value anyway.
  - Swapped len and val in intermediate patches adding imx355_write_reg before converting
    to cci_write
  - Initialise colorspace in _open for try_fmt.
  - IMX355_REG_TEST_PATTERN is a 16 bit register, but was defined as 8 bit.
  - Call pm_runtime_dont_use_autosuspend on remove / probe error
  - Missed subdev_lock in set_stream.
  - Not addressed the pre-existing issues of double power_off in _probe error path, nor
    potential for VBLANK update on mode change not calling s_ctrl if the new value happens
    to be the same and therefore exposure range being incorrect. I'll look at them as
    separate fixes.
- Variable definition for format and state in imx355_set_ctrl got pushed into the wrong patch 
- Collected Krzysztof's R-b tag for the binding patch.
- Link to v2: https://lore.kernel.org/r/20260707-media-imx355-v2-0-1683ec07b897@raspberrypi.com

Changes in v2:
- Collected Jacopo's R-b tags and implement his review comments.
- Add patches to switch to using the subdev state, clean up colorspace
  (sets RAW), use a define for the exposure offset.
- Added DT binding update to permit 2 CSI2 data lanes.
- Link to v1: https://lore.kernel.org/r/20260506-media-imx355-v1-0-660685030455@raspberrypi.com

---
Dave Stevenson (19):
      dt-bindings: media: imx355: Allow 2 CSI2 data lane output
      media: imx355: Remove duplicated registers from the mode tables
      media: imx355: Remove setting FRM_LENGTH_LINES in the mode regs
      media: imx355: Programmatically set the crop parameters for each mode
      media: imx355: Set register LINE_LENGTH_PCK programmatically
      media: imx355: Set binning mode registers programmatically
      media: imx355: Remove link_freq_index from each mode as ununsed
      media: imx355: pixel_rate never changes, so don't recompute
      media: imx355: Remove redundant fll_min, and implement fixed offset
      media: imx355: Add support for get_selection
      media: imx355: Use pm_runtime autosuspend_delay
      media: imx355: Convert to new CCI register access helpers
      media: imx355: Set the colorspace in the format
      media: imx355: Define the exposure offset, and use that define
      media: imx355: Use NULL ctrl_ops for HBLANK as it is a read-only control
      media: imx355: Compute link frequency from PLL setup
      media: imx355: Support 2 lane readout
      media: imx355: Switch to using the subdev state
      media: imx355: Remove storing cur_mode in the state

David Heidelberg (1):
      media: i2c: imx355: Add support for 24 MHz external clock

 .../devicetree/bindings/media/i2c/sony,imx355.yaml |   11 +-
 drivers/media/i2c/Kconfig                          |    1 +
 drivers/media/i2c/imx355.c                         | 1518 +++++++-------------
 3 files changed, 498 insertions(+), 1032 deletions(-)
---
base-commit: 254f49634ee16a731174d2ae34bc50bd5f45e731
change-id: 20260506-media-imx355-a30f5be3deda

Best regards,
-- 
Dave Stevenson <dave.stevenson@raspberrypi.com>


             reply	other threads:[~2026-07-08 14:49 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-07-08 14:48 Dave Stevenson [this message]
2026-07-08 14:48 ` [PATCH v3 01/20] media: i2c: imx355: Add support for 24 MHz external clock Dave Stevenson
2026-07-08 14:48 ` [PATCH v3 02/20] dt-bindings: media: imx355: Allow 2 CSI2 data lane output Dave Stevenson
2026-07-08 14:48 ` [PATCH v3 03/20] media: imx355: Remove duplicated registers from the mode tables Dave Stevenson
2026-07-08 15:00   ` sashiko-bot
2026-07-08 14:48 ` [PATCH v3 04/20] media: imx355: Remove setting FRM_LENGTH_LINES in the mode regs Dave Stevenson
2026-07-08 14:48 ` [PATCH v3 05/20] media: imx355: Programmatically set the crop parameters for each mode Dave Stevenson
2026-07-08 14:48 ` [PATCH v3 06/20] media: imx355: Set register LINE_LENGTH_PCK programmatically Dave Stevenson
2026-07-08 14:48 ` [PATCH v3 07/20] media: imx355: Set binning mode registers programmatically Dave Stevenson
2026-07-08 14:48 ` [PATCH v3 08/20] media: imx355: Remove link_freq_index from each mode as ununsed Dave Stevenson
2026-07-08 14:48 ` [PATCH v3 09/20] media: imx355: pixel_rate never changes, so don't recompute Dave Stevenson
2026-07-08 15:01   ` sashiko-bot
2026-07-08 14:48 ` [PATCH v3 10/20] media: imx355: Remove redundant fll_min, and implement fixed offset Dave Stevenson
2026-07-08 15:05   ` sashiko-bot
2026-07-08 14:48 ` [PATCH v3 11/20] media: imx355: Add support for get_selection Dave Stevenson
2026-07-08 15:11   ` sashiko-bot
2026-07-08 14:48 ` [PATCH v3 12/20] media: imx355: Use pm_runtime autosuspend_delay Dave Stevenson
2026-07-08 15:06   ` sashiko-bot
2026-07-08 14:48 ` [PATCH v3 13/20] media: imx355: Convert to new CCI register access helpers Dave Stevenson
2026-07-08 15:08   ` sashiko-bot
2026-07-08 14:48 ` [PATCH v3 14/20] media: imx355: Set the colorspace in the format Dave Stevenson
2026-07-08 15:01   ` sashiko-bot
2026-07-08 14:48 ` [PATCH v3 15/20] media: imx355: Define the exposure offset, and use that define Dave Stevenson
2026-07-08 15:07   ` sashiko-bot
2026-07-08 14:48 ` [PATCH v3 16/20] media: imx355: Use NULL ctrl_ops for HBLANK as it is a read-only control Dave Stevenson
2026-07-08 14:48 ` [PATCH v3 17/20] media: imx355: Compute link frequency from PLL setup Dave Stevenson
2026-07-08 14:48 ` [PATCH v3 18/20] media: imx355: Support 2 lane readout Dave Stevenson
2026-07-08 14:48 ` [PATCH v3 19/20] media: imx355: Switch to using the subdev state Dave Stevenson
2026-07-08 15:10   ` sashiko-bot
2026-07-08 14:48 ` [PATCH v3 20/20] media: imx355: Remove storing cur_mode in the state Dave Stevenson
2026-07-08 15:17   ` sashiko-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=20260708-media-imx355-v3-0-9df386a623d7@raspberrypi.com \
    --to=dave.stevenson@raspberrypi.com \
    --cc=conor+dt@kernel.org \
    --cc=david@ixit.cz \
    --cc=devicetree@vger.kernel.org \
    --cc=jacopo.mondi@ideasonboard.com \
    --cc=krzk+dt@kernel.org \
    --cc=krzysztof.kozlowski@oss.qualcomm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-media@vger.kernel.org \
    --cc=mailingradian@gmail.com \
    --cc=mchehab@kernel.org \
    --cc=robh@kernel.org \
    --cc=sakari.ailus@linux.intel.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