>-----Original Message----- >From: Mark Brown [mailto:broo...@kernel.org] >Sent: Tuesday, August 29, 2017 12:00 PM >To: Ryan Lee <ryans....@maximintegrated.com> >Cc: lgirdw...@gmail.com; pe...@perex.cz; ti...@suse.com; >kuninori.morimoto...@renesas.com; alsa-de...@alsa-project.org; linux- >ker...@vger.kernel.org; ryan.lee.ma...@gmail.com >Subject: Re: [PATCH 1/9] ASoC: max98927: Added support for DSP_A and DSP_B >format > >On Mon, Aug 28, 2017 at 04:30:53PM -0700, Ryan Lee wrote: > >> + if ((max98927->iface == SND_SOC_DAIFMT_DSP_A) || >> + (max98927->iface == SND_SOC_DAIFMT_DSP_B)) >> + return 0; >> + >> + /* BCLK configuration */ > >Why do we not configure the BCLK in DSP modes? That's unusual and seems >likely to break some systems that rely on exact clocking.
I put this code to avoid overwrite BCLK value in TDM mode. BCLK is being configured by ' max98927_dai_tdm_slot'. Now I added one more variable to check TDM mode instead of checking DAI_FMT. > >Also if we're selecting on format a switch statement is generally better to >make >any further special casing easier in future. OK. Thanks. But I kept if statement on the modified version because it only have two cases, tdm and non-tdm.
