On Tue, Nov 23, 2021 at 05:53:06PM +0000, Mark Brown wrote:
On Tue, Nov 23, 2021 at 04:31:40PM +0000, Lucas Tanure wrote:
--- a/sound/soc/codecs/Makefile +++ b/sound/soc/codecs/Makefile @@ -54,8 +54,8 @@ snd-soc-cs35l33-objs := cs35l33.o snd-soc-cs35l34-objs := cs35l34.o snd-soc-cs35l35-objs := cs35l35.o snd-soc-cs35l36-objs := cs35l36.o -snd-soc-cs35l41-spi-objs := cs35l41-spi.o cs35l41.o cs35l41-tables.o -snd-soc-cs35l41-i2c-objs := cs35l41-i2c.o cs35l41.o cs35l41-tables.o +snd-soc-cs35l41-spi-objs := cs35l41-spi.o cs35l41.o cs35l41-lib.o +snd-soc-cs35l41-i2c-objs := cs35l41-i2c.o cs35l41.o cs35l41-lib.o snd-soc-cs42l42-objs := cs42l42.o snd-soc-cs42l51-objs := cs42l51.o snd-soc-cs42l51-i2c-objs := cs42l51-i2c.o
As was already called out on the HDA part of this patch I was expecting this to be more like how we handle sharing the PXA2xx driver code between the ASoC and AC'97 implementations - making it a library rather than linking the object files directly into both drivers. I'm not sure that the current approach would DTRT if one or both of the drivers is built in.
Yeah looking closer the Prince driver itself is kinda weird here since it is building separate modules for SPI and I2C but including cs25l41.o into both, rather than the more normal 3 modules approach.
I will have a look at doing a patch to fix that up, separately from Lucas's series here.
Thanks, Charles