From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark Brown Subject: Re: [PATCH] ASoC: wm_adsp: Dump DSP_SCRATCH1 on DSP shutdown Date: Thu, 28 May 2015 20:23:35 +0100 Message-ID: <20150528192335.GD21577@sirena.org.uk> References: <1432820714-31850-1-git-send-email-rf@opensource.wolfsonmicro.com> <20150528135751.GV21577@sirena.org.uk> <20150528142431.GA19558@opensource.wolfsonmicro.com> <20150528143724.GY21577@sirena.org.uk> <20150528161313.GB19558@opensource.wolfsonmicro.com> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha256; protocol="application/pgp-signature"; boundary="6KchevI7DAOLL/Ks" Return-path: Content-Disposition: inline In-Reply-To: <20150528161313.GB19558@opensource.wolfsonmicro.com> Sender: linux-kernel-owner@vger.kernel.org To: Richard Fitzgerald Cc: patches@opensource.wolfsonmicro.com, alsa-devel@alsa-project.org, linux-kernel@vger.kernel.org List-Id: alsa-devel@alsa-project.org --6KchevI7DAOLL/Ks Content-Type: text/plain; charset=us-ascii Content-Disposition: inline On Thu, May 28, 2015 at 05:13:13PM +0100, Richard Fitzgerald wrote: > On Thu, May 28, 2015 at 03:37:24PM +0100, Mark Brown wrote: > > The default is that dev_dbg() is available in dmesg but not on the > > console which usually seems like a reasonable balance for this sort of > > thing - it's there if people want it but not included in the default > > logging. > What you're describing seems to be the default for dev_info(). > For me I don't get any dev_dbg() output in the dmesg log unless the source > file has a #define DEBUG at the top. Interesting... the dev_dbg default appears to have been changed at some point, possibly with the addition of dynamic debug (which I tend to have on in my configurations and is common for distros and so on). The info default is overridden by a lot of distros. In any case, it's still the way we generally do debug - doing this on every DSP stop does seem quite chatty from a system point of view (and yes, some of the existing logging in the driver does seem a bit enthusiastic here too in retrospect, IIRC at the time request_firmware() was chatty too but that has been fixed). Users will typically have the option to get this via ftrace as well (we have trace points for register read), and you might also want to consider the option of exporting the value via a sysfs file so people can grab the latest value whenever they like. --6KchevI7DAOLL/Ks Content-Type: application/pgp-signature; name="signature.asc" Content-Description: Digital signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBCAAGBQJVZ2s2AAoJECTWi3JdVIfQkoYH/A95ztJTbS8GVPOSLrW8HCfQ rigChVehhG56OfjJrulUv+EpwLYvgaGNaSm3ImTzkleA4dFc2Vlwchh3zrjufMdD X1bJ/opWfObQC09hSXkHnltzaTndNs4r10Fx39uAW9DLtwOvoR5ctw0lb3QG/ZVO cU1oeMSLYRWpVJUqSwPCOCcjGeiavutk6UbMFXlxiJv+CmZIJF7sN4k4i5mbMSns PXadhDpvKeZ3/I2ibEuNw1VEaOZ6fGPhe/HNOdEyXJX/htg2TNNW+eS6eEgc0+hV 3cKZ7uyRDXzbc2px/YqrfVhFRK4M2J1KnDnS0jIUwiFUO0FWTN7HJxm+xicQO90= =AEB7 -----END PGP SIGNATURE----- --6KchevI7DAOLL/Ks--