On Fri, 2020-05-01 at 12:44 +0300, Dan Carpenter wrote: > [External] > > Hello Dragos Bogdan, > > The patch 3a6af93dd66e: "iio: adc: ad7476: Add IIO_CHAN_INFO_RAW for > AD7091R" from Mar 11, 2020, leads to the following static checker > warning: > > drivers/iio/adc/ad7476.c:312 ad7476_probe() > warn: this array is non-NULL. 'st->chip_info->convst_channel' > > drivers/iio/adc/ad7476.c > 288 ret = devm_add_action_or_reset(&spi->dev, ad7476_reg_disable, > 289 st); > 290 if (ret) > 291 return ret; > 292 > 293 st->convst_gpio = devm_gpiod_get_optional(&spi->dev, > 294 "adi,conversion- > start", > 295 GPIOD_OUT_LOW); > 296 if (IS_ERR(st->convst_gpio)) > 297 return PTR_ERR(st->convst_gpio); > 298 > 299 spi_set_drvdata(spi, indio_dev); > 300 > 301 st->spi = spi; > 302 > 303 /* Establish that the iio_dev is a child of the spi device */ > 304 indio_dev->dev.parent = &spi->dev; > 305 indio_dev->dev.of_node = spi->dev.of_node; > 306 indio_dev->name = spi_get_device_id(spi)->name; > 307 indio_dev->modes = INDIO_DIRECT_MODE; > 308 indio_dev->channels = st->chip_info->channel; > 309 indio_dev->num_channels = 2; > 310 indio_dev->info = &ad7476_info; > 311 > 312 if (st->convst_gpio && st->chip_info->convst_channel) > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > This is an array, not a pointer and it it's in the middle of the struct > so it can't possibly be NULL. I think there's already a patch applied for this. https://git.kernel.org/pub/scm/linux/kernel/git/jic23/iio.git/commit/?h=testing&id=4d22ab7446c69f7ebbc8c8a486ebd9aba68823c0 > > 313 indio_dev->channels = st->chip_info->convst_channel; > 314 /* Setup default message */ > 315 > 316 st->xfer.rx_buf = &st->data; > 317 st->xfer.len = st->chip_info->channel[0].scan_type.storagebits > / 8; > 318 > 319 spi_message_init(&st->msg); > 320 spi_message_add_tail(&st->xfer, &st->msg); > 321 > > regards, > dan carpenter
