linux-kernel - [PATCH 1/2] drm/i2c: tda998x: adjust CTS

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Date:   Thu, 21 Feb 2019 13:18:13 -0500
From:   Sven Van Asbroeck <thesven73@...il.com>
To:     Russell King <linux@...linux.org.uk>,
        David Airlie <airlied@...ux.ie>,
        Daniel Vetter <daniel@...ll.ch>
Cc:     dri-devel@...ts.freedesktop.org, linux-kernel@...r.kernel.org,
        Peter Rosin <peda@...ntia.se>
Subject: [PATCH 1/2] drm/i2c: tda998x: adjust CTS_N audio pre-divider calculation

My cpu dai driving the tda998x in master mode outputs
SNDRV_PCM_FORMAT_S24_LE, i2s left justified, two channels:

	[SNDRV_PCM_FORMAT_S24_LE] = {
		.width = 24, .phys = 32, .le = 1, .signd = 1,
		.silence = {},
	},

This format has a sample width of 24 bits, but a physical
size of 32 bits per channel. The redundant bits are padded
with zeros. This gives a total frame width of 64 bits.
According to the tda19988 datasheet, this format is supported:

"Audio samples with a precision better than 24-bit are
truncated to 24-bit. [...] If the input clock has a
frequency of 64fs and is left justified or Philips,
the audio word is truncated to 24-bit format and other
bits padded with zeros."
(tda19988 product datasheet Rev. 3 - 21 July 2011)

However, supplying this format to the chip results in distorted
audio. I noticed that the audio problem disappears when the
CTS_N pre-divider is calculated from the physical width, _not_
the sample width.

This patch adjusts the CTS_N calculation so that the audio
distortion goes away.

Caveats:
- I am only capable of generating audio with a frame width of
  64fs. So I cannot test if 32fs or 48fs audio formats still
  work. Such as S16_LE or S20_LE.
- I have no access to a datasheet which accurately describes
  the CTS_N register. So I cannot check if my assumption is
  correct.

Tested with an imx6 ssi.

Cc: Peter Rosin <peda@...ntia.se>
Signed-off-by: Sven Van Asbroeck <TheSven73@...il.com>
---
 drivers/gpu/drm/i2c/tda998x_drv.c | 4 ++--
 include/drm/i2c/tda998x.h         | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
index a7c39f39793f..ba9f55176e1b 100644
--- a/drivers/gpu/drm/i2c/tda998x_drv.c
+++ b/drivers/gpu/drm/i2c/tda998x_drv.c
@@ -893,7 +893,7 @@ tda998x_configure_audio(struct tda998x_priv *priv,
 		reg_write(priv, REG_MUX_AP, MUX_AP_SELECT_I2S);
 		clksel_aip = AIP_CLKSEL_AIP_I2S;
 		clksel_fs = AIP_CLKSEL_FS_ACLK;
-		switch (params->sample_width) {
+		switch (params->physical_width) {
 		case 16:
 			cts_n = CTS_N_M(3) | CTS_N_K(1);
 			break;
@@ -982,7 +982,7 @@ static int tda998x_audio_hw_params(struct device *dev, void *data,
 	struct tda998x_priv *priv = dev_get_drvdata(dev);
 	int i, ret;
 	struct tda998x_audio_params audio = {
-		.sample_width = params->sample_width,
+		.physical_width = params->physical_width,
 		.sample_rate = params->sample_rate,
 		.cea = params->cea,
 	};
diff --git a/include/drm/i2c/tda998x.h b/include/drm/i2c/tda998x.h
index 3cb25ccbe5e6..7e9fddcc4ddd 100644
--- a/include/drm/i2c/tda998x.h
+++ b/include/drm/i2c/tda998x.h
@@ -14,7 +14,7 @@ enum {
 struct tda998x_audio_params {
 	u8 config;
 	u8 format;
-	unsigned sample_width;
+	unsigned int physical_width;
 	unsigned sample_rate;
 	struct hdmi_audio_infoframe cea;
 	u8 status[5];
-- 
2.17.1