Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Luca Ceresoli <luca.ceresoli@bootlin.com>
To:     philippe.schenker@toradex.com
Cc:     devicetree@vger.kernel.org, festevam@gmail.com,
        gregkh@linuxfoundation.org, jun.li@nxp.com, kernel@pengutronix.de,
        linux-arm-kernel@lists.infradead.org, linux-imx@nxp.com,
        linux-kernel@vger.kernel.org, linux-usb@vger.kernel.org,
        peter.chen@nxp.com, robh+dt@kernel.org, s.hauer@pengutronix.de,
        shawnguo@kernel.org,
        Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Subject: Re: [PATCH v2 2/3] usb: chipidea: imx: support disabling runtime-pm
Date:   Thu,  4 May 2023 18:23:12 +0200
Message-Id: <20230504162312.1506763-1-luca.ceresoli@bootlin.com>
In-Reply-To: <23672d66d229d3be4cc854ddf1462c3507f1c2fc.camel@toradex.com>
References: <23672d66d229d3be4cc854ddf1462c3507f1c2fc.camel@toradex.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

Hello Philippe,

I found this thread after several hours spent in debugging why USB host is
not detecting new devices on a custom board using the iMX6ULL Colibri
SoM.

My best workaround at the moment is:

  --- a/drivers/usb/chipidea/ci_hdrc_imx.c
  +++ b/drivers/usb/chipidea/ci_hdrc_imx.c
  @@ -56,7 +58,7 @@ static const struct ci_hdrc_imx_platform_flag imx6sx_usb_data = {
   };
 
   static const struct ci_hdrc_imx_platform_flag imx6ul_usb_data = {
  -       .flags = CI_HDRC_SUPPORTS_RUNTIME_PM |
  +       .flags = //CI_HDRC_SUPPORTS_RUNTIME_PM |
                  CI_HDRC_TURN_VBUS_EARLY_ON |
                  CI_HDRC_DISABLE_DEVICE_STREAMING,
   };

so it's equivalent to yours, but hard-coded.

I haven't found any follow-up patches from you, so I'm wondering whether
you have made any progress on this issue.

As I see it, a quirk in the driver would make sense. What needs to be
sorted out is how to enable it from device tree.

I think a DT boolean as you propose would do, but with a name describing
the hardware aspects of it ("chipidea,vbus-detection-is-broken"?) and not
the driver behaviour.

Otherwise looking for the "toradex,colibri-imx6ul" prefix in the top-level
/compatible strings seems doable. I don't know how acceptable this is from
a driver maintainer's point of view though.

Best regards,
Luca