| Message ID | 20201217142842.v3.3.I07afdedcc49655c5d26880f8df9170aac5792378@changeid (mailing list archive) |
|---|---|
| State | Accepted |
| Commit | 3d7d916f9bc98ce88272b3e4405c7c685afbfcd6 |
| Headers | show |
| Series | [v3,1/4] spi: spi-geni-qcom: Fix geni_spi_isr() NULL dereference in timeout case | expand |
Quoting Douglas Anderson (2020-12-17 14:29:13) > If we get a timeout sending then this happens: > > spi_transfer_one_message() > ->transfer_one() AKA spi_geni_transfer_one() > setup_fifo_xfer() > mas->cur_xfer = non-NULL > spi_transfer_wait() => TIMES OUT > if (msg->status != -EINPROGRESS) > goto out > if (ret != 0 ...) > spi_set_cs() > ->set_cs AKA spi_geni_set_cs() > # mas->cur_xfer is non-NULL > > The above happens _before_ the SPI core calls ->handle_err() AKA > handle_fifo_timeout(). > > Unfortunately that won't work so well on geni. If we got a timeout > transferring then it's likely that our interrupt handler is blocked, > but we need that same interrupt handler to run and the command channel > to be unblocked in order to adjust the chip select. Trying to set the > chip select doesn't crash us but ends up confusing our state machine > and leads to messages like: Premature done. rx_rem = 32 bpw8 > > Let's just drop the chip select request in this case. We can detect > the case because cur_xfer is non-NULL--it would have been set to NULL > in the interrupt handler if the previous transfer had finished. Sure, > we might leave the chip select in the wrong state but it's likely it > was going to fail anyway and this avoids getting the driver even more > confused about what it's doing. > > The SPI core in general assumes that setting chip select is a simple > operation that doesn't fail. Yet another reason to just reconfigure > the chip select line as GPIOs. > > Signed-off-by: Douglas Anderson <dianders@chromium.org> > --- Reviewed-by: Stephen Boyd <swboyd@chromium.org>
diff --git a/drivers/spi/spi-geni-qcom.c b/drivers/spi/spi-geni-qcom.c index cf3db40ae5ba..b3ba092db489 100644 --- a/drivers/spi/spi-geni-qcom.c +++ b/drivers/spi/spi-geni-qcom.c @@ -207,9 +207,14 @@ static void spi_geni_set_cs(struct spi_device *slv, bool set_flag) goto exit; } - mas->cs_flag = set_flag; - spin_lock_irq(&mas->lock); + if (mas->cur_xfer) { + dev_err(mas->dev, "Can't set CS when prev xfer running\n"); + spin_unlock_irq(&mas->lock); + goto exit; + } + + mas->cs_flag = set_flag; reinit_completion(&mas->cs_done); if (set_flag) geni_se_setup_m_cmd(se, SPI_CS_ASSERT, 0);
If we get a timeout sending then this happens: spi_transfer_one_message() ->transfer_one() AKA spi_geni_transfer_one() setup_fifo_xfer() mas->cur_xfer = non-NULL spi_transfer_wait() => TIMES OUT if (msg->status != -EINPROGRESS) goto out if (ret != 0 ...) spi_set_cs() ->set_cs AKA spi_geni_set_cs() # mas->cur_xfer is non-NULL The above happens _before_ the SPI core calls ->handle_err() AKA handle_fifo_timeout(). Unfortunately that won't work so well on geni. If we got a timeout transferring then it's likely that our interrupt handler is blocked, but we need that same interrupt handler to run and the command channel to be unblocked in order to adjust the chip select. Trying to set the chip select doesn't crash us but ends up confusing our state machine and leads to messages like: Premature done. rx_rem = 32 bpw8 Let's just drop the chip select request in this case. We can detect the case because cur_xfer is non-NULL--it would have been set to NULL in the interrupt handler if the previous transfer had finished. Sure, we might leave the chip select in the wrong state but it's likely it was going to fail anyway and this avoids getting the driver even more confused about what it's doing. The SPI core in general assumes that setting chip select is a simple operation that doesn't fail. Yet another reason to just reconfigure the chip select line as GPIOs. Signed-off-by: Douglas Anderson <dianders@chromium.org> --- Changes in v3: - xfter => xfer in error message. - More obvious in commit message how this happens Changes in v2: - ("spi: spi-geni-qcom: Don't try to set CS if an xfer is pending") new for v2. drivers/spi/spi-geni-qcom.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)