Hi Eugeniu
We've been reviewing this patch in the context of Renesas-Yocto-v3.21.0 BSP integration, where it is contained as commit [1].
OK, now, you are using BSP.
diff --git a/sound/soc/sh/rcar/ssi.c b/sound/soc/sh/rcar/ssi.c index f5afab6..44bda21 100644 --- a/sound/soc/sh/rcar/ssi.c +++ b/sound/soc/sh/rcar/ssi.c @@ -303,6 +303,8 @@ static int rsnd_ssi_master_clk_start(struct rsnd_mod *mod, if (rsnd_runtime_is_tdm_split(io)) chan = rsnd_io_converted_chan(io);
- chan = rsnd_channel_normalization(chan);
Since the "chan" value is already normalized by calling: => rsnd_ssi_master_clk_start() => chan = rsnd_runtime_channel_for_ssi(io) => rsnd_runtime_channel_for_ssi_with_params() => rsnd_channel_normalization()
I was wondering if it is really required to call rsnd_channel_normalization() second time in ssi.c for fixing the issue described in this patch?
Please compare BSP and upstream patch.
BSP https://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas-bsp.git/commit...
upstream https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i...
I don't know detail of BSP side, but I guess it doesn't have TDM Split support (?). Because of that, it doesn't have (A) code.
int rsnd_ssi_master_clk_start(xxx) { ... int chan = rsnd_runtime_channel_for_ssi(io); ... if (rsnd_runtime_is_tdm_split(io)) (A) chan = rsnd_io_converted_chan(io);
chan = rsnd_channel_normalization(chan); ... }
Thank you for your help !! Best regards --- Kuninori Morimoto