On 15/07/21 04:21PM, Yoshitaka Ikeda wrote: > Fix below division by zero warning: > - Added an if statement because buswidth can be zero, resulting in division by zero. > - The modified code was based on another driver (atmel-quadspi). > > [ 0.795337] Division by zero in kernel. > : > [ 0.834051] [<807fd40c>] (__div0) from [<804e1acc>] (Ldiv0+0x8/0x10) > [ 0.839097] [<805f0710>] (cqspi_exec_mem_op) from [<805edb4c>] (spi_mem_exec_op+0x3b0/0x3f8) Please add a fixes tag. Fixes: 7512eaf54190 ("spi: cadence-quadspi: Fix dummy cycle calculation when buswidth > 1") > > Signed-off-by: Yoshitaka Ikeda <ikeda@xxxxxxxxxxxx> > --- > v2: > - Fix commit message. > v3: > - repost. > > drivers/spi/spi-cadence-quadspi.c | 10 ++++++---- > 1 file changed, 6 insertions(+), 4 deletions(-) > > diff --git a/drivers/spi/spi-cadence-quadspi.c b/drivers/spi/spi-cadence-quadspi.c > index 7a00346ff9b9..13d1f0ce618e 100644 > --- a/drivers/spi/spi-cadence-quadspi.c > +++ b/drivers/spi/spi-cadence-quadspi.c > @@ -307,11 +307,13 @@ static unsigned int cqspi_calc_rdreg(struct cqspi_flash_pdata *f_pdata) > > static unsigned int cqspi_calc_dummy(const struct spi_mem_op *op, bool dtr) > { > - unsigned int dummy_clk; > + unsigned int dummy_clk = 0; I think you can just do: if (!op->dummy.nbytes) return 0; and leave the rest same. I don't think we should have to check for buswidth here, even though it is the one causing divide-by-zero. Any op with positive dummy nbytes but with 0 buswidth is invalid. That should be rejected by the SPI MEM core or the supports_op(), so it should never even get here. > > - dummy_clk = op->dummy.nbytes * (8 / op->dummy.buswidth); > - if (dtr) > - dummy_clk /= 2; > + if (op->dummy.buswidth && op->dummy.nbytes) { > + dummy_clk = op->dummy.nbytes * (8 / op->dummy.buswidth); > + if (dtr) > + dummy_clk /= 2; > + } > > return dummy_clk; > } > -- > 2.32.0 -- Regards, Pratyush Yadav Texas Instruments Inc.