From: Pierre-Louis Bossart Subject: Re: [PATCH v5 8/8] ASoC: Intel: Skylake: add option to control HDAudio + DSP usage Date: Wed, 1 Aug 2018 07:50:06 -0700 Message-ID: References: <20180727230554.31027-1-pierre-louis.bossart@linux.intel.com> <20180727230554.31027-9-pierre-louis.bossart@linux.intel.com> <5f73f810-6fe5-bad9-81fb-fd587ef39e8c@linux.intel.com> <20180731162000.GJ5719@sirena.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; Format="flowed" Content-Transfer-Encoding: 7bit Return-path: Received: from mga03.intel.com (mga03.intel.com [134.134.136.65]) by alsa0.perex.cz (Postfix) with ESMTP id A8B062677AA for ; Wed, 1 Aug 2018 16:51:08 +0200 (CEST) In-Reply-To: <20180731162000.GJ5719@sirena.org.uk> Content-Language: en-US List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: alsa-devel-bounces@alsa-project.org Sender: alsa-devel-bounces@alsa-project.org To: Mark Brown Cc: Takashi Iwai , liam.r.girdwood@linux.intel.com, alsa-devel@alsa-project.org, vkoul@kernel.org List-Id: alsa-devel@alsa-project.org On 7/31/18 9:20 AM, Mark Brown wrote: > On Tue, Jul 31, 2018 at 09:06:54AM -0700, Pierre-Louis Bossart wrote: >> On 7/31/18 3:48 AM, Takashi Iwai wrote: >>> Pierre-Louis Bossart wrote: > >>>> +config SND_SOC_INTEL_SKYLAKE_FORCE_HDA_DSP >>>> + bool "Force HDAudio Codec + DSP support" >>>> + depends on SND_SOC_INTEL_SKYLAKE_HDA_DSP >>>> + help >>>> + If you have a Intel Skylake+ platform with the DSP enabled in >>>> + the BIOS and an HDAudio codec, then enable this option by saying Y >>>> + This option will ignore information from the BIOS and force the use >>>> + of the HDaudio codec, if present. >>>> + This is a debug option not recommended for distros. > >>> ... and this one is better to be a module option. >>> Distros tend to enable all possible options, and this may be set >>> unnecessarily. If any, this kconfig should be only toggling the >>> default option value. > >> Sorry, I don't get this one. this wouldn't change anything between built-in >> or module, it's just a yes/no answer to ignore ACPI stuff. the default is >> also no and there is a clear statement not to include it except for debug. >> We have a similar option for SOF to bypass all ACPI information and go >> straight to 'nocodec' mode. > > Debug options like these are really a lot more useful as something > runtime selectable with something like a module parameter as Takashi > suggests - an innatentive distro maintainer could easily enable it by > mistake when going through a bunch of new options in a new kernel > release and if someone needs it for debugging purposes in a distro > context it's a real pain to have to rebuild the kernel. Ah yes, I get it now and yes it's probably a better idea to have a kernel parameter. I saw multiple cases of distros such as Arch enabling the baytrail nocodec mode and it's not straightforward to reach all maintainers of all distros. Will fix, thanks for the feedback. > > It's also a bit confusing as the config text says: > >>>> + If you have a Intel Skylake+ platform with the DSP enabled in >>>> + the BIOS and an HDAudio codec, then enable this option by saying Y > > which sounds like people should turn it on, the bit about it being a > debug option is quite deep in the text. >