05.09.2020 23:41, Dmitry Osipenko пишет: > Apparently barrier() was intended to reduce possibility of racing > with the interrupt handler, but driver's code evolved significantly > and today's driver enables interrupt only when it waits for completion > notification. Hence barrier() has no good use anymore, let's remove it. > > Signed-off-by: Dmitry Osipenko <digetx@xxxxxxxxx> > --- > drivers/i2c/busses/i2c-tegra.c | 2 -- > 1 file changed, 2 deletions(-) > > diff --git a/drivers/i2c/busses/i2c-tegra.c b/drivers/i2c/busses/i2c-tegra.c > index 33d37a40fa83..f69587ca163b 100644 > --- a/drivers/i2c/busses/i2c-tegra.c > +++ b/drivers/i2c/busses/i2c-tegra.c > @@ -600,7 +600,6 @@ static int tegra_i2c_fill_tx_fifo(struct tegra_i2c_dev *i2c_dev) > i2c_dev->msg_buf_remaining = buf_remaining; > i2c_dev->msg_buf = buf + > words_to_transfer * BYTES_PER_FIFO_WORD; > - barrier(); > > i2c_writesl(i2c_dev, buf, I2C_TX_FIFO, words_to_transfer); > > @@ -624,7 +623,6 @@ static int tegra_i2c_fill_tx_fifo(struct tegra_i2c_dev *i2c_dev) > /* Again update before writing to FIFO to make sure isr sees. */ > i2c_dev->msg_buf_remaining = 0; > i2c_dev->msg_buf = NULL; > - barrier(); > > i2c_writel(i2c_dev, val, I2C_TX_FIFO); > } > It just caught my eye that there is actually a comment there saying that barrier() was intended to mitigate the racing with the ISR. Hence that comment is outdated now and needs to be removed. I'll correct it in v5.
