| Message ID | 20191130150958.5AE2668BE1@verein.lst.de (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | drm/bridge: analogix-anx6345: Avoid duplicate -supply suffix | expand |
On Sat, Nov 30, 2019 at 04:09:58PM +0100, Torsten Duwe wrote:
> IMHO that commit message should have ended up somewhere under Documentation/.
This is documented already in the generic regulator bindings and
has been since they were added, they specify that supplies should
always have the suffix -supply. The commit you reference was
part of the initial merge of the device tree support for the
regulator API. Including the -supply suffix in the name used
internally in Linux would create problems for all other firmware
interfaces that map names differently.
On Sat, Nov 30, 2019 at 04:09:58PM +0100, Torsten Duwe wrote: > of_get_regulator() will unconditionally add "-supply" to form the > property name. This is documented in commit 69511a452e6dc ("map consumer > regulator based on device tree"). Sorry, I meant to also say: Reviewed-by: Mark Brown <broonie@kernel.org>
--- a/drivers/gpu/drm/bridge/analogix/analogix-anx6345.c +++ b/drivers/gpu/drm/bridge/analogix/analogix-anx6345.c @@ -712,14 +709,14 @@ static int anx6345_i2c_probe(struct i2c_client *client, DRM_DEBUG("No panel found\n"); /* 1.2V digital core power regulator */ - anx6345->dvdd12 = devm_regulator_get(dev, "dvdd12-supply"); + anx6345->dvdd12 = devm_regulator_get(dev, "dvdd12"); if (IS_ERR(anx6345->dvdd12)) { DRM_ERROR("dvdd12-supply not found\n"); return PTR_ERR(anx6345->dvdd12); } /* 2.5V digital core power regulator */ - anx6345->dvdd25 = devm_regulator_get(dev, "dvdd25-supply"); + anx6345->dvdd25 = devm_regulator_get(dev, "dvdd25"); if (IS_ERR(anx6345->dvdd25)) { DRM_ERROR("dvdd25-supply not found\n"); return PTR_ERR(anx6345->dvdd25);
of_get_regulator() will unconditionally add "-supply" to form the property name. This is documented in commit 69511a452e6dc ("map consumer regulator based on device tree"). Signed-off-by: Torsten Duwe <duwe@suse.de> --- IMHO that commit message should have ended up somewhere under Documentation/.