All of lore.kernel.org
 help / color / mirror / Atom feed
From: Shawn Guo <shawnguo@kernel.org>
To: Sean Paul <seanpaul@chromium.org>
Cc: Xin Zhou <zhou.xin8@zte.com.cn>,
	Daniel Vetter <daniel.vetter@ffwll.ch>,
	Baoyou Xie <xie.baoyou@zte.com.cn>,
	dri-devel@lists.freedesktop.org, Jun Nie <jun.nie@linaro.org>
Subject: Re: [PATCH 4/4] drm: zte: add VGA driver support
Date: Wed, 5 Apr 2017 22:08:21 +0800	[thread overview]
Message-ID: <20170405140815.GG3394@dragon> (raw)
In-Reply-To: <20170404174648.4khdo24dlmrfswuu@art_vandelay>

On Tue, Apr 04, 2017 at 01:46:48PM -0400, Sean Paul wrote:
> > +static int zx_vga_register(struct drm_device *drm, struct zx_vga *vga)
> > +{
> > +	struct drm_encoder *encoder = &vga->encoder;
> > +	struct drm_connector *connector = &vga->connector;
> > +
> > +	encoder->possible_crtcs = VOU_CRTC_MASK;
> > +
> > +	drm_encoder_init(drm, encoder, &zx_vga_encoder_funcs,
> > +			 DRM_MODE_ENCODER_DAC, NULL);
> > +	drm_encoder_helper_add(encoder, &zx_vga_encoder_helper_funcs);
> > +
> > +	vga->connector.polled = DRM_CONNECTOR_POLL_HPD;
> > +
> > +	drm_connector_init(drm, connector, &zx_vga_connector_funcs,
> > +			   DRM_MODE_CONNECTOR_VGA);
> > +	drm_connector_helper_add(connector, &zx_vga_connector_helper_funcs);
> > +
> > +	drm_mode_connector_attach_encoder(connector, encoder);
> 
> You're missing a couple of return value checks in this function.

Okay.  I will add return checks in the new version.

> 
> > +
> > +	return 0;
> > +}

<snip>

> > +static irqreturn_t zx_vga_irq_handler(int irq, void *dev_id)
> > +{
> > +	struct zx_vga *vga = dev_id;
> > +	u32 status;
> > +
> > +	status = zx_readl(vga->mmio + VGA_I2C_STATUS);
> > +
> > +	/* Clear interrupt status */
> > +	zx_writel_mask(vga->mmio + VGA_I2C_STATUS, VGA_CLEAR_IRQ,
> > +		       VGA_CLEAR_IRQ);
> > +
> > +	if (status & VGA_DEVICE_CONNECTED) {
> > +		/*
> > +		 * We should ideally set up VGA_DETECT_SEL_HAS_DEVICE bit here
> > +		 * for unplug detection, but doing so will stop DDC bus from
> > +		 * reading EDID later on. It looks like a HW limitation, and we
> > +		 * work around it by defering the bit setup to .get_modes hook
> > +		 * after EDID read succeeds.
> 
> get_modes resets VGA_AUTO_DETECT_SEL before and after reading edid, so is it
> really necessary to defer setting it here?

Okay, the comment in the code seems confusing.  Since we need to reset
VGA_AUTO_DETECT_SEL bits to switch DDC bus from device detecting to EDID
reading, VGA_DETECT_SEL_HAS_DEVICE will need to be set after reading
EDID anyway.  I will improve the comment.

> 
> > +		 */
> > +		vga->connected = true;
> > +		return IRQ_WAKE_THREAD;
> > +	}
> > +
> > +	if (status & VGA_DEVICE_DISCONNECTED) {
> > +		zx_writel(vga->mmio + VGA_AUTO_DETECT_SEL,
> > +			  VGA_DETECT_SEL_NO_DEVICE);
> 
> This races with get_modes, you should serialize access to this register.

Good catch.  I will add lock for it.

Shawn

> 
> > +		vga->connected = false;
> > +		return IRQ_WAKE_THREAD;
> > +	}
> > +
> > +	if (status & VGA_TRANS_DONE) {
> > +		complete(&vga->complete);
> > +		return IRQ_HANDLED;
> > +	}
> > +
> > +	return IRQ_NONE;
> > +}
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

      reply	other threads:[~2017-04-05 14:08 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-03-21 12:15 [PATCH 0/4] Add ZTE VGA driver support Shawn Guo
2017-03-21 12:15 ` [PATCH 1/4] drm: zte: do not enable clock auto-gating by default Shawn Guo
2017-04-04 17:48   ` Sean Paul
2017-03-21 12:15 ` [PATCH 2/4] drm: zte: move CSC register definitions into a common header Shawn Guo
2017-04-04 17:47   ` Sean Paul
2017-03-21 12:15 ` [PATCH 3/4] dt: add bindings for ZTE VGA device Shawn Guo
2017-03-21 12:15 ` [PATCH 4/4] drm: zte: add VGA driver support Shawn Guo
2017-04-03  9:23   ` Daniel Vetter
2017-04-04 13:47     ` Shawn Guo
2017-04-04 17:32       ` Sean Paul
2017-04-05  6:08         ` Shawn Guo
2017-04-05  6:25           ` Daniel Vetter
2017-04-04 17:46   ` Sean Paul
2017-04-05 14:08     ` Shawn Guo [this message]

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=20170405140815.GG3394@dragon \
    --to=shawnguo@kernel.org \
    --cc=daniel.vetter@ffwll.ch \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=jun.nie@linaro.org \
    --cc=seanpaul@chromium.org \
    --cc=xie.baoyou@zte.com.cn \
    --cc=zhou.xin8@zte.com.cn \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.