Hi Ricardo,
On 09.03.2023 09:55, Ricardo Cañuelo wrote:
> On lun 09-01-2023 23:00:33, Marek Szyprowski wrote:
>> devm_regulator_get_enable_optional() function returns 0 on success, so
>> use it for the check if function succeded instead of the -ENODEV value.
>>
>> Fixes: 429e87063661 ("drm/meson: dw-hdmi: Use devm_regulator_*get_enable*()")
>> Signed-off-by: Marek Szyprowski <m.szyprowski@xxxxxxxxxxx>
>> ---
>> drivers/gpu/drm/meson/meson_dw_hdmi.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/gpu/drm/meson/meson_dw_hdmi.c b/drivers/gpu/drm/meson/meson_dw_hdmi.c
>> index 7642f740272b..534621a13a34 100644
>> --- a/drivers/gpu/drm/meson/meson_dw_hdmi.c
>> +++ b/drivers/gpu/drm/meson/meson_dw_hdmi.c
>> @@ -718,7 +718,7 @@ static int meson_dw_hdmi_bind(struct device *dev, struct device *master,
>> dw_plat_data = &meson_dw_hdmi->dw_plat_data;
>>
>> ret = devm_regulator_get_enable_optional(dev, "hdmi");
>> - if (ret != -ENODEV)
>> + if (ret < 0)
>> return ret;
>>
>> meson_dw_hdmi->hdmitx_apb = devm_reset_control_get_exclusive(dev,
>> --
>> 2.39.0
> This patch seems to have caused (or uncovered) a regression in the
> bootrr.deferred-probe-empty test for meson-g12a-sei510. Apparently, this
> causes a driver to stay in the deferred probe list after boot.
Indeed my fix causes regression on the platform, which doesn't have
"hdmi" regulator. Thanks for pointing that. I was really convinced that
devm_regulator_get_enable_*optional*() returns 0 if requested regulator
is not present (as "_optional" suffix suggests). Unfortunately it
returns -ENODEV, what breaks meson_dw_hdmi driver. I will send a fix in
a few minutes.
Best regards
--
Marek Szyprowski, PhD
Samsung R&D Institute Poland
