| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <Z1hC-V8LZcXO_Qzo@kekkonen.localdomain>
Date: Tue, 10 Dec 2024 13:32:41 +0000
From: Sakari Ailus <sakari.ailus@...ux.intel.com>
To: Jai Luthra <jai.luthra@...asonboard.com>
Cc: Dave Stevenson <dave.stevenson@...pberrypi.com>,
Mauro Carvalho Chehab <mchehab@...nel.org>,
linux-media@...r.kernel.org, linux-kernel@...r.kernel.org,
Jacopo Mondi <jacopo.mondi@...asonboard.com>,
Laurent Pinchart <laurent.pinchart@...asonboard.com>,
Naushir Patuck <naush@...pberrypi.com>,
Vinay Varma <varmavinaym@...il.com>
Subject: Re: [PATCH v3 3/3] media: i2c: imx219: Scale the pixel rate for
analog binning
Hi Jai,
On Wed, Nov 27, 2024 at 03:07:14PM +0530, Jai Luthra wrote:
> Hi Sakari,
>
> On Nov 26, 2024 at 14:36:24 +0000, Sakari Ailus wrote:
> > Hi Jai,
> >
> > On Mon, Nov 25, 2024 at 08:36:27PM +0530, Jai Luthra wrote:
> > > When the analog binning mode is used for high framerate operation,
> > > the pixel rate is effectively doubled. Account for this when setting up
> > > the pixel clock rate, and applying the vblank and exposure controls.
> > >
> > > The previous logic only used analog binning for 8-bit modes, but normal
> > > binning limits the framerate on 10-bit 480p [1]. So with this patch we
> > > switch to using special binning (with 2x pixel rate) for all formats of
> > > 480p mode and 8-bit 1232p.
> > >
> > > [1]: https://github.com/raspberrypi/linux/issues/5493
> > >
> > > Co-developed-by: Naushir Patuck <naush@...pberrypi.com>
> > > Signed-off-by: Naushir Patuck <naush@...pberrypi.com>
> > > Co-developed-by: Vinay Varma <varmavinaym@...il.com>
> > > Signed-off-by: Vinay Varma <varmavinaym@...il.com>
> > > Signed-off-by: Jai Luthra <jai.luthra@...asonboard.com>
> > > ---
> > > drivers/media/i2c/imx219.c | 120 ++++++++++++++++++++++++++++-----------------
> > > 1 file changed, 76 insertions(+), 44 deletions(-)
> > >
> > > diff --git a/drivers/media/i2c/imx219.c b/drivers/media/i2c/imx219.c
> > > index 970e6362d0ae3a9078daf337155e83d637bc1ca1..39b85cdee58318b080c867afd68ca33d14d3eda7 100644
> > > --- a/drivers/media/i2c/imx219.c
> > > +++ b/drivers/media/i2c/imx219.c
> > > @@ -149,6 +149,12 @@
> > > #define IMX219_PIXEL_ARRAY_WIDTH 3280U
> > > #define IMX219_PIXEL_ARRAY_HEIGHT 2464U
> > >
> > > +enum binning_mode {
> > > + BINNING_NONE = IMX219_BINNING_NONE,
> > > + BINNING_X2 = IMX219_BINNING_X2,
> > > + BINNING_ANALOG_X2 = IMX219_BINNING_X2_ANALOG,
> > > +};
> > > +
> > > /* Mode : resolution and related config&values */
> > > struct imx219_mode {
> > > /* Frame width */
> > > @@ -337,6 +343,10 @@ struct imx219 {
> > >
> > > /* Two or Four lanes */
> > > u8 lanes;
> > > +
> > > + /* Binning mode */
> > > + enum binning_mode bin_h;
> > > + enum binning_mode bin_v;
> > > };
> > >
> > > static inline struct imx219 *to_imx219(struct v4l2_subdev *_sd)
> > > @@ -362,6 +372,36 @@ static u32 imx219_get_format_code(struct imx219 *imx219, u32 code)
> > > return imx219_mbus_formats[i];
> > > }
> > >
> > > +static u32 imx219_get_format_bpp(const struct v4l2_mbus_framefmt *format)
> > > +{
> > > + switch (format->code) {
> > > + case MEDIA_BUS_FMT_SRGGB8_1X8:
> > > + case MEDIA_BUS_FMT_SGRBG8_1X8:
> > > + case MEDIA_BUS_FMT_SGBRG8_1X8:
> > > + case MEDIA_BUS_FMT_SBGGR8_1X8:
> > > + return 8;
> > > +
> > > + case MEDIA_BUS_FMT_SRGGB10_1X10:
> > > + case MEDIA_BUS_FMT_SGRBG10_1X10:
> > > + case MEDIA_BUS_FMT_SGBRG10_1X10:
> > > + case MEDIA_BUS_FMT_SBGGR10_1X10:
> > > + default:
> > > + return 10;
> > > + }
> > > +}
> > > +
> > > +static int imx219_get_rate_factor(struct imx219 *imx219)
> > > +{
> > > + switch (imx219->bin_v) {
> > > + case BINNING_NONE:
> > > + case BINNING_X2:
> > > + return 1;
> > > + case BINNING_ANALOG_X2:
> > > + return 2;
> >
> > FWIW, what the CCS driver does is that it exposes different horizontal
> > blanking ranges for devices that use analogue binning. The rate is really
> > about reading pixels and with analogue binning the rate is the same, it's
> > just that fewer pixels are being (digitally) read (as they are binned). I
> > wonder if this would be a workable approach for this sensor, too. Of course
> > if the LLP behaves differently for this sensor, then we should probably
> > just accept that.
> >
>
> IMX219 seems to be odd in this case, as the LLP doesn't change during
> analog binning. Shared some more details in this thread:
>
> https://lore.kernel.org/linux-media/20241125-imx219_fixes-v3-0-434fc0b541c8@ideasonboard.com/T/#m1da4206e91db12b8e377dc686935195fc5f4bb68
Ack.
>
> > > + }
> > > + return -EINVAL;
> > > +}
> > > +
> > > /* -----------------------------------------------------------------------------
> > > * Controls
> > > */
> > > @@ -373,10 +413,12 @@ static int imx219_set_ctrl(struct v4l2_ctrl *ctrl)
> > > struct i2c_client *client = v4l2_get_subdevdata(&imx219->sd);
> > > const struct v4l2_mbus_framefmt *format;
> > > struct v4l2_subdev_state *state;
> > > + int rate_factor;
> >
> > u32?
> >
>
> Fixed.
>
> > > int ret = 0;
> > >
> > > state = v4l2_subdev_get_locked_active_state(&imx219->sd);
> > > format = v4l2_subdev_state_get_format(state, 0);
> > > + rate_factor = imx219_get_rate_factor(imx219);
> > >
> > > if (ctrl->id == V4L2_CID_VBLANK) {
> > > int exposure_max, exposure_def;
> > > @@ -405,7 +447,7 @@ static int imx219_set_ctrl(struct v4l2_ctrl *ctrl)
> > > break;
> > > case V4L2_CID_EXPOSURE:
> > > cci_write(imx219->regmap, IMX219_REG_EXPOSURE,
> > > - ctrl->val, &ret);
> > > + ctrl->val / rate_factor, &ret);
> >
> > Isn't the exposure in lines? It shouldn't be affected by the rate change,
> > shouldn't it?
> >
>
> From the sensor datasheet the unit of FRAME_LENGTH register is updated
> to 2xLines when analog binning is used. And exposure and vertical
> blanking values are also in units of FRAME_LENGTH. This is also
> consistent with the behavior seen while testing.
>
> > > break;
> > > case V4L2_CID_DIGITAL_GAIN:
> > > cci_write(imx219->regmap, IMX219_REG_DIGITAL_GAIN,
> > > @@ -422,7 +464,7 @@ static int imx219_set_ctrl(struct v4l2_ctrl *ctrl)
> > > break;
> > > case V4L2_CID_VBLANK:
> > > cci_write(imx219->regmap, IMX219_REG_VTS,
> > > - format->height + ctrl->val, &ret);
> > > + (format->height + ctrl->val) / rate_factor, &ret);
> >
> > The same for vertical blanking.
> >
> > > break;
> > > case V4L2_CID_HBLANK:
> > > cci_write(imx219->regmap, IMX219_REG_HTS,
> > > @@ -463,7 +505,8 @@ static const struct v4l2_ctrl_ops imx219_ctrl_ops = {
> > >
> > > static unsigned long imx219_get_pixel_rate(struct imx219 *imx219)
> > > {
> > > - return (imx219->lanes == 2) ? IMX219_PIXEL_RATE : IMX219_PIXEL_RATE_4LANE;
> > > + return ((imx219->lanes == 2) ? IMX219_PIXEL_RATE :
> > > + IMX219_PIXEL_RATE_4LANE) * imx219_get_rate_factor(imx219);
> > > }
> > >
> > > /* Initialize control handlers */
> > > @@ -592,29 +635,12 @@ static int imx219_set_framefmt(struct imx219 *imx219,
> > > {
> > > const struct v4l2_mbus_framefmt *format;
> > > const struct v4l2_rect *crop;
> > > - unsigned int bpp;
> > > - u64 bin_h, bin_v;
> > > + u32 bpp;
> > > int ret = 0;
> > >
> > > format = v4l2_subdev_state_get_format(state, 0);
> > > crop = v4l2_subdev_state_get_crop(state, 0);
> > > -
> > > - switch (format->code) {
> > > - case MEDIA_BUS_FMT_SRGGB8_1X8:
> > > - case MEDIA_BUS_FMT_SGRBG8_1X8:
> > > - case MEDIA_BUS_FMT_SGBRG8_1X8:
> > > - case MEDIA_BUS_FMT_SBGGR8_1X8:
> > > - bpp = 8;
> > > - break;
> > > -
> > > - case MEDIA_BUS_FMT_SRGGB10_1X10:
> > > - case MEDIA_BUS_FMT_SGRBG10_1X10:
> > > - case MEDIA_BUS_FMT_SGBRG10_1X10:
> > > - case MEDIA_BUS_FMT_SBGGR10_1X10:
> > > - default:
> > > - bpp = 10;
> > > - break;
> > > - }
> > > + bpp = imx219_get_format_bpp(format);
> > >
> > > cci_write(imx219->regmap, IMX219_REG_X_ADD_STA_A,
> > > crop->left - IMX219_PIXEL_ARRAY_LEFT, &ret);
> > > @@ -625,28 +651,8 @@ static int imx219_set_framefmt(struct imx219 *imx219,
> > > cci_write(imx219->regmap, IMX219_REG_Y_ADD_END_A,
> > > crop->top - IMX219_PIXEL_ARRAY_TOP + crop->height - 1, &ret);
> > >
> > > - switch (crop->width / format->width) {
> > > - case 1:
> > > - default:
> > > - bin_h = IMX219_BINNING_NONE;
> > > - break;
> > > - case 2:
> > > - bin_h = bpp == 8 ? IMX219_BINNING_X2_ANALOG : IMX219_BINNING_X2;
> > > - break;
> > > - }
> > > -
> > > - switch (crop->height / format->height) {
> > > - case 1:
> > > - default:
> > > - bin_v = IMX219_BINNING_NONE;
> > > - break;
> > > - case 2:
> > > - bin_v = bpp == 8 ? IMX219_BINNING_X2_ANALOG : IMX219_BINNING_X2;
> > > - break;
> > > - }
> > > -
> > > - cci_write(imx219->regmap, IMX219_REG_BINNING_MODE_H, bin_h, &ret);
> > > - cci_write(imx219->regmap, IMX219_REG_BINNING_MODE_V, bin_v, &ret);
> > > + cci_write(imx219->regmap, IMX219_REG_BINNING_MODE_H, imx219->bin_h, &ret);
> > > + cci_write(imx219->regmap, IMX219_REG_BINNING_MODE_V, imx219->bin_v, &ret);
> >
> > Please run:
> >
> > $ ./scripts/checkpatch.pl --strict --max-line-length=80
> >
>
> Oops, fixed in next revision.
>
> > >
> > > cci_write(imx219->regmap, IMX219_REG_X_OUTPUT_SIZE,
> > > format->width, &ret);
> > > @@ -851,6 +857,27 @@ static int imx219_set_pad_format(struct v4l2_subdev *sd,
> > > int exposure_max;
> > > int exposure_def;
> > > int hblank;
> > > + int pixel_rate;
> > > + u32 bpp = imx219_get_format_bpp(format);
> > > + enum binning_mode binning = BINNING_NONE;
> > > +
> > > + /*
> > > + * For 8-bit formats, analog horizontal binning is required,
> > > + * else the output image is garbage.
> > > + * For 10-bit formats, analog horizontal binning is optional,
> > > + * but still useful as it doubles the effective framerate.
> > > + * We can only use it with width <= 1624, as for higher values
> > > + * there are blocky artefacts.
> >
> > This comment would benefit from rewrapping.
> >
>
> Fixed.
>
> > > + *
> > > + * Vertical binning should match the horizontal binning mode.
> > > + */
> > > + if (bin_h == 2 && (format->width <= 1624 || bpp == 8))
> > > + binning = BINNING_ANALOG_X2;
> > > + else
> > > + binning = BINNING_X2;
> > > +
> > > + imx219->bin_h = (bin_h == 2) ? binning : BINNING_NONE;
> > > + imx219->bin_v = (bin_v == 2) ? binning : BINNING_NONE;
> >
> > It'd be also nice to move the state information to sub-device state.
> >
>
> I'm not sure I follow, do you mean the framework should store the
> binning mode, similar to how crop rectangle and interval are stored in
> v4l2_subdev_state?
Yes, please. This is done to the CCS driver as part of the metadata set
(which we can hopefully merge in not too distant future)
<URL:https://git.retiisi.eu/?p=~sailus/linux.git;a=shortlog;h=refs/heads/metadata>.
--
Kind regards,
Sakari Ailus
Powered by blists - more mailing lists