linux-kernel - Re: [PATCH 3/6] usb: bdc: driver may fail to get USB PHY

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

Message-ID: <1561095579.32589.3.camel@mhfsdcap03>
Date:   Fri, 21 Jun 2019 13:39:39 +0800
From:   Chunfeng Yun <chunfeng.yun@...iatek.com>
To:     Al Cooper <alcooperx@...il.com>
CC:     <linux-kernel@...r.kernel.org>, <devicetree@...r.kernel.org>,
        Felipe Balbi <balbi@...nel.org>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        <linux-usb@...r.kernel.org>, Luis Chamberlain <mcgrof@...nel.org>,
        "Mark Rutland" <mark.rutland@....com>,
        Rob Herring <robh+dt@...nel.org>
Subject: Re: [PATCH 3/6] usb: bdc: driver may fail to get USB PHY

On Thu, 2019-06-20 at 17:09 -0400, Al Cooper wrote:
> Initialization order is important for the USB PHY and the PHY clients.
> The init order is based on the build order of the drivers in the
> makefiles and the PHY drivers are built early to help with
> dependencies, but the new SCMI based clock subsystem has the side
> effect of making some additional drivers DEFER until the clock
> is ready. This is causing the USB PHY driver to defer which is causing
> some PHY clients to fail when they try to get the PHY. The fix is to have
> the client driver return DEFER when it's "get phy" routine returns DEFER.
> 
> Signed-off-by: Al Cooper <alcooperx@...il.com>
> ---
>  drivers/usb/gadget/udc/bdc/bdc_core.c | 6 +++++-
>  1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/usb/gadget/udc/bdc/bdc_core.c b/drivers/usb/gadget/udc/bdc/bdc_core.c
> index 11a43de6c1c6..c794890d785b 100644
> --- a/drivers/usb/gadget/udc/bdc/bdc_core.c
> +++ b/drivers/usb/gadget/udc/bdc/bdc_core.c
> @@ -543,9 +543,13 @@ static int bdc_probe(struct platform_device *pdev)
>  			dev, dev->of_node, phy_num);
>  		if (IS_ERR(bdc->phys[phy_num])) {
>  			ret = PTR_ERR(bdc->phys[phy_num]);
> +			if (ret == -EPROBE_DEFER) {
> +				dev_dbg(bdc->dev, "DEFER, waiting for PHY\n");
why not disable clock here? when re-probe, will enable clock again.
to me, no need check -EPROBE_DEFFER.
> +				return ret;
> +			}

>  			dev_err(bdc->dev,
>  				"BDC phy specified but not found:%d\n", ret);
> -			return ret;
> +			goto clk_cleanup;
>  		}
>  	}
>