Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8;
 format=flowed
Content-Transfer-Encoding: 8bit
Date:   Fri, 20 Nov 2020 09:59:02 +0000
From:   Marc Zyngier <maz@kernel.org>
To:     Neil Armstrong <narmstrong@baylibre.com>
Cc:     Jerome Brunet <jbrunet@baylibre.com>,
        Guillaume Tucker <guillaume.tucker@collabora.com>,
        kernelci-results@groups.io, Kevin Hilman <khilman@baylibre.com>,
        linux-arm-kernel@lists.infradead.org,
        linux-amlogic@lists.infradead.org,
        Martin Blumenstingl <martin.blumenstingl@googlemail.com>,
        David Airlie <airlied@linux.ie>,
        Daniel Vetter <daniel@ffwll.ch>,
        dri-devel@lists.freedesktop.org, linux-kernel@vger.kernel.org
Subject: Re: next/master bisection: baseline.dmesg.emerg on meson-gxbb-p200
In-Reply-To: <93dcf61a-be80-8870-48f5-30e215cc56ed@baylibre.com>
References: <5fb5e094.1c69fb81.a2014.2e62@mx.google.com>
 <a0bec7c4-9bec-8858-4879-52f4688d9992@collabora.com>
 <630e00e83cdd07ee5a0eaba9d3479554@kernel.org>
 <3f54de27-0fef-c5a1-8991-0a0614c90667@baylibre.com>
 <c76273f5fe483766e6a7f509f82d928a@kernel.org>
 <f59922c6-69f5-c70e-b424-0659bf91a4fd@collabora.com>
 <1jr1op8bbc.fsf@starbuckisacylon.baylibre.com>
 <00a10c12a4eb2a9cdd9f50e88a293c3f@kernel.org>
 <93dcf61a-be80-8870-48f5-30e215cc56ed@baylibre.com>
User-Agent: Roundcube Webmail/1.4.9
Message-ID: <45c711c9d5fb5ef0e3b80e82b10f5998@kernel.org>
Precedence: bulk

On 2020-11-20 09:26, Neil Armstrong wrote:
> On 19/11/2020 19:35, Marc Zyngier wrote:
>>>>> diff --git a/drivers/gpu/drm/meson/meson_dw_hdmi.c 
>>>>> b/drivers/gpu/drm/meson/meson_dw_hdmi.c
>>>>> index 7f8eea494147..52af8ba94311 100644
>>>>> --- a/drivers/gpu/drm/meson/meson_dw_hdmi.c
>>>>> +++ b/drivers/gpu/drm/meson/meson_dw_hdmi.c
>>>>> @@ -146,6 +146,7 @@ struct meson_dw_hdmi {
>>>>>      struct reset_control *hdmitx_ctrl;
>>>>>      struct reset_control *hdmitx_phy;
>>>>>      struct clk *hdmi_pclk;
>>>>> +    struct clk *iahb_clk;
>>>>>      struct clk *venci_clk;
>>>>>      struct regulator *hdmi_supply;
>>>>>      u32 irq_stat;
>>>>> @@ -1033,6 +1034,13 @@ static int meson_dw_hdmi_bind(struct device 
>>>>> *dev, struct device *master,
>>>>>      }
>>>>>      clk_prepare_enable(meson_dw_hdmi->hdmi_pclk);
>>>>> 
>>>>> +    meson_dw_hdmi->iahb_clk = devm_clk_get(dev, "iahb");
>>>>> +    if (IS_ERR(meson_dw_hdmi->iahb_clk)) {
>>>>> +        dev_err(dev, "Unable to get iahb clk\n");
>>>>> +        return PTR_ERR(meson_dw_hdmi->iahb_clk);
>>>>> +    }
>>>>> +    clk_prepare_enable(meson_dw_hdmi->iahb_clk);
> 
> 
> On previous SoCs, iahb was directly the bus clock (clk81), and on 
> recent socs
> this clock is a gate.
> 
> The question is why is it disabled. Maybe a previous failed probe 
> disabled it
> in the dw-hdmi probe failure code and this clock is needed for
> meson_dw_hdmi_init(),
> so yeah this is the right fix.
> 
> Thanks.
> 
> Could you send a revert of b33340e33acdfe5ca6a5aa1244709575ae1e0432
> and then proper fix with clk_disable_unprepare added ?

Bah. I missed that email and sent a slightly different resolution.
Hopefully that'll be good enough.

Thanks,

         M.
-- 
Jazz is not dead. It just smells funny...