Hi Judy,
On Thu, Sep 08, 2022 at 03:11:42PM +0000, Judy Hsiao wrote:
Use regmap_read_poll_timeout to poll I2S_CLR. It also fixes the 'rockchip-i2s ff070000.i2s; fail to clear' and return -EBUSY when the read of I2S_CLR exceeds the retry limit.
Fixes: 0ff9f8b9f592 ("ASoC: rockchip: i2s: Fix error code when fail to read I2S_CLR") Signed-off-by: Judy Hsiao judyhsiao@chromium.org
Thanks for the patch!
sound/soc/rockchip/rockchip_i2s.c | 39 +++++++++++++------------------ 1 file changed, 16 insertions(+), 23 deletions(-)
diff --git a/sound/soc/rockchip/rockchip_i2s.c b/sound/soc/rockchip/rockchip_i2s.c index f5f3540a9e18..02905eec615e 100644 --- a/sound/soc/rockchip/rockchip_i2s.c +++ b/sound/soc/rockchip/rockchip_i2s.c @@ -163,18 +163,14 @@ static int rockchip_snd_txctrl(struct rk_i2s_dev *i2s, int on) I2S_CLR_TXC | I2S_CLR_RXC); if (ret < 0) goto end;
regmap_read(i2s->regmap, I2S_CLR, &val);
/* Should wait for clear operation to finish */
while (val) {
regmap_read(i2s->regmap, I2S_CLR, &val);
retry--;
I think you're leaving |retry| as a "set but unused" variable now. Presumably some compile-testing bots will eventually complain at you about that.
if (!retry) {
dev_warn(i2s->dev, "fail to clear\n");
ret = -EBUSY;
break;
}
}
ret = regmap_read_poll_timeout(i2s->regmap,
I2S_CLR,
val,
val != 0,
20,
200);
if (ret < 0)
dev_warn(i2s->dev, "fail to clear\n");
Now would be a good time to print the return code, I think:
dev_warn(i2s->dev, "fail to clear: %d\n", ret);
}
} end: @@ -226,17 +222,14 @@ static int rockchip_snd_rxctrl(struct rk_i2s_dev *i2s, int on) I2S_CLR_TXC | I2S_CLR_RXC); if (ret < 0) goto end;
regmap_read(i2s->regmap, I2S_CLR, &val);
/* Should wait for clear operation to finish */
while (val) {
regmap_read(i2s->regmap, I2S_CLR, &val);
retry--;
if (!retry) {
Same.
dev_warn(i2s->dev, "fail to clear\n");
ret = -EBUSY;
break;
}
}
ret = regmap_read_poll_timeout(i2s->regmap,
I2S_CLR,
val,
val != 0,
20,
200);
if (ret < 0)
dev_warn(i2s->dev, "fail to clear\n");
Same.
}
} end:
Otherwise, this is definitely a good change, since we have no timing guarantee on the existing retry loop, which will surely make it flaky. So, with those fixups:
Reviewed-by: Brian Norris briannorris@chromium.org