On Fri, Jun 22, 2018 at 11:27:55AM +0530, Sriram Periyasamy wrote:
On Thu, Jun 21, 2018 at 01:14:28PM +0200, Takashi Iwai wrote:
On Thu, 21 Jun 2018 12:45:34 +0200, Sriram Periyasamy wrote:
On Thu, Jun 21, 2018 at 10:28:33AM +0200, Takashi Iwai wrote:
On Thu, 21 Jun 2018 09:54:20 +0200, Sriram Periyasamy wrote:
On Wed, Jun 20, 2018 at 03:01:13PM +0200, Takashi Iwai wrote:
On Wed, 20 Jun 2018 13:08:21 +0200, Sriram Periyasamy wrote: > > During d3/d0 cycle, the connection selection index of all pins points > to the default value. This needs to be restored to ensure audio is > restored after d3/d0 cycle. > > So store the connection selection index and program it during jack > report event which gets invoked in cases like d3/d0 cycle, hot plug > detection when multiple displays are connected.
Hm, I thought *_CONNECT_SEL being cached and restored at resume already by regmap?
Yes, legacy HDA HDMI driver has this method whereas ASoC based HDAC HDMI driver doesn't use regmap. Hence we needed this solution.
If so, why not just use the standard snd_hdac_codec_write()?
Because snd_hdac_codec_write doesn't power up the device to send the verb across the link when the device is runtime suspended.
Then why hdac_hdmi_port_select_set() would work? It's using snd_hdac_codec_write(), too.
We will check and come back on this.
In case of MST, it will work since snd_hdac_read_parm_uncached() wakes up the device and put the device to runtime auto suspend. The auto suspend delay was helping this snd_hdac_codec_write() pass through.
In case of non-MST, hdac_hdmi_port_select_set() will simply return 0 by checking the MST capable port->pin->mst_capable. Hence the device is not runtime resumed and subsequent calls of snd_hdac_codec_write() will fail.
We will post v2 to address this.
Thanks, Sriram.