diff options
| author | Karl Mehltretter <kmehltretter@gmail.com> | 2026-10-03 21:07:38 +0200 |
|---|---|---|
| committer | Mark Brown <broonie@kernel.org> | 2026-10-05 15:04:20 +0200 |
| commit | 58e7bca5cf264e910ad9199cfc40246e0a2834f4 (patch) | |
| tree | 9cfdf15360a560166e14fc5bf5e907ce7007bb4d /tools/lib/python | |
| parent | 9a773d9d21fa909e8babbe11e93c811cf1a6d22d (diff) | |
ASoC: codecs: es8375: use devm_clk_get_optional() for mclk
es8375_read_device_properities() gets "mclk" with devm_clk_get() and
then tests the result for NULL to print "assuming static mclk". A
missing clock makes devm_clk_get() return -ENOENT, so the probe fails
before that test can match.
The binding does not require clocks, its example has none, and
everest,mclk-src lets the codec run from SCLK instead of MCLK. Use
devm_clk_get_optional(), as es8311, es8316, es8323, es8326 and es8389
do for the same clock.
Found with an experimental clang warning about NULL tests on functions
that return error pointers. Tested only in QEMU with i2c-stub in place
of the codec, not on real hardware.
Fixes: de2b3119f9f7 ("ASoC: codecs: add support for ES8375")
Assisted-by: LLM
Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>
Link: https://patch.msgid.link/20261003190738.77564-1-kmehltretter@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Diffstat (limited to 'tools/lib/python')
0 files changed, 0 insertions, 0 deletions
