linux-kernel - Re: [PATCH v3 4/4] media: tc358746: add Toshiba TC358746 Parallel to CSI-2 bridge driver

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

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:   Fri, 23 Sep 2022 16:38:11 +0200
From:   Marco Felsch <m.felsch@...gutronix.de>
To:     Sakari Ailus <sakari.ailus@...ux.intel.com>
Cc:     mchehab@...nel.org, robh+dt@...nel.org,
        krzysztof.kozlowski+dt@...aro.org, kishon@...com, vkoul@...nel.org,
        laurent.pinchart@...asonboard.com, jacopo@...ndi.org,
        hverkuil@...all.nl, linux-phy@...ts.infradead.org,
        linux-kernel@...r.kernel.org, devicetree@...r.kernel.org,
        linux-media@...r.kernel.org, kernel@...gutronix.de
Subject: Re: [PATCH v3 4/4] media: tc358746: add Toshiba TC358746 Parallel to
 CSI-2 bridge driver

Hi Sakari,

On 22-09-23, Sakari Ailus wrote:
> Hi Marco,
> 
> Thanks for the update. A few small things below.
> 
> On Thu, Sep 22, 2022 at 03:48:43PM +0200, Marco Felsch wrote:
> ...
> > +static int tc358746_apply_misc_config(struct tc358746 *tc358746)
> > +{
> > +	const struct v4l2_mbus_framefmt *mbusfmt;
> > +	struct v4l2_subdev *sd = &tc358746->sd;
> > +	struct v4l2_subdev_state *sink_state;
> > +	const struct tc358746_format *fmt;
> > +	struct device *dev = sd->dev;
> > +	u32 val;
> > +	int err;
> > +
> > +	sink_state = v4l2_subdev_lock_and_get_active_state(sd);
> > +	mbusfmt = v4l2_subdev_get_pad_format(sd, sink_state, TC358746_SINK);
> > +	v4l2_subdev_unlock_state(sink_state);
> > +
> > +	fmt = tc358746_get_format_by_code(TC358746_SINK, mbusfmt->code);
> 
> You can no longer access mbusfmt here as you've already unlocked subdev
> state.

You're right, Laurent also mentioned this but this function gets called
during .s_stream(on) so the format shouldn't be changed that late.
However since you're the 2nd with the thought, I will fix it in the v4.

> > +
> > +	/* Self defined CSI user data type id's are not supported yet */
> > +	val = PDFMT(fmt->pdformat);
> > +	dev_dbg(dev, "DATAFMT: 0x%x\n", val);
> > +	err = tc358746_write(tc358746, DATAFMT_REG, val);
> > +	if (err)
> > +		return err;
> > +
> > +	val = PDATAF(fmt->pdataf);
> > +	dev_dbg(dev, "CONFCTL[PDATAF]: 0x%x\n", fmt->pdataf);
> > +	err = tc358746_update_bits(tc358746, CONFCTL_REG, PDATAF_MASK, val);
> > +	if (err)
> > +		return err;
> > +
> > +	val = tc358746->vb_size / 32;
> > +	dev_dbg(dev, "FIFOCTL: %u (0x%x)\n", val, val);
> > +	err = tc358746_write(tc358746, FIFOCTL_REG, val);
> > +	if (err)
> > +		return err;
> > +
> > +	/* Total number of bytes for each line/width */
> > +	val = mbusfmt->width * fmt->bpp / 8;
> > +	dev_dbg(dev, "WORDCNT: %u (0x%x)\n", val, val);
> > +	return tc358746_write(tc358746, WORDCNT_REG, val);
> > +}
> 
> ...
> 
> > +static int tc358746_s_stream(struct v4l2_subdev *sd, int enable)
> > +{
> > +	struct tc358746 *tc358746 = to_tc358746(sd);
> > +	struct v4l2_subdev *src;
> > +	int err;
> > +
> > +	dev_dbg(sd->dev, "%sable\n", enable ? "en" : "dis");
> > +
> > +	src = tc358746_get_remote_sd(&tc358746->pads[TC358746_SINK]);
> > +	if (!src)
> > +		return -EPIPE;
> > +
> > +	if (enable) {
> > +		err = pm_runtime_resume_and_get(sd->dev);
> > +		if (err)
> > +			return err;
> > +
> > +		err = tc358746_apply_dphy_config(tc358746);
> > +		if (err)
> > +			goto err_out;
> > +
> > +		err = tc358746_apply_misc_config(tc358746);
> > +		if (err)
> > +			goto err_out;
> > +
> > +		err = tc358746_enable_csi_lanes(tc358746, 1);
> > +		if (err)
> > +			goto err_out;
> > +
> > +		err = tc358746_enable_csi_module(tc358746, 1);
> > +		if (err)
> > +			goto err_out;
> > +
> > +		err = tc358746_enable_parallel_port(tc358746, 1);
> > +		if (err)
> > +			goto err_out;
> > +
> > +		err = v4l2_subdev_call(src, video, s_stream, 1);
> > +		if (err)
> > +			goto err_out;
> > +
> > +		return err;
> 
> This can be return 0.

Sure.

Regards,
  Marco

> 
> > +
> > +err_out:
> > +		pm_runtime_mark_last_busy(sd->dev);
> > +		pm_runtime_put_sync_autosuspend(sd->dev);
> > +
> > +		return err;
> > +	}
> 
> -- 
> Kind regards,
> 
> Sakari Ailus
>