Hi Hans,
On Friday 21 Jul 2017 11:02:34 Hans Verkuil wrote:
> From: Hans Verkuil <hans.verkuil@xxxxxxxxx>
>
> This is now set by media_device_init.
>
> Drop the print of driver_version in the error message: the driver
> version is 1) not yet set at this point (the media_device_init call
> comes later AFAICS), and 2) irrelevant here, since it is the hw_revision
> that is important, not the driver version (which is identical to the
> kernel version anyway).
>
> Signed-off-by: Hans Verkuil <hans.verkuil@xxxxxxxxx>
> ---
> drivers/staging/media/atomisp/pci/atomisp2/atomisp_v4l2.c | 5 +----
> 1 file changed, 1 insertion(+), 4 deletions(-)
>
> diff --git a/drivers/staging/media/atomisp/pci/atomisp2/atomisp_v4l2.c
> b/drivers/staging/media/atomisp/pci/atomisp2/atomisp_v4l2.c index
> 2f49562377e6..29387c03fae9 100644
> --- a/drivers/staging/media/atomisp/pci/atomisp2/atomisp_v4l2.c
> +++ b/drivers/staging/media/atomisp/pci/atomisp2/atomisp_v4l2.c
> @@ -1100,8 +1100,7 @@ atomisp_load_firmware(struct atomisp_device *isp)
>
> if (!fw_path) {
> dev_err(isp->dev,
> - "Unsupported driver_version 0x%x, hw_revision 0x%x\n",
> - isp->media_dev.driver_version,
> + "Unsupported hw_revision 0x%x\n",
Nitpicking, you can now merge the first two lines.
> isp->media_dev.hw_revision);
> return NULL;
> }
> @@ -1249,8 +1248,6 @@ static int atomisp_pci_probe(struct pci_dev *dev,
> /* This is not a true PCI device on SoC, so the delay is not needed.
*/
> isp->pdev->d3_delay = 0;
>
> - isp->media_dev.driver_version = LINUX_VERSION_CODE;
> -
> switch (id->device & ATOMISP_PCI_DEVICE_SOC_MASK) {
> case ATOMISP_PCI_DEVICE_SOC_MRFLD:
> isp->media_dev.hw_revision =
--
Regards,
Laurent Pinchart
