On Mon, Aug 03, 2020 at 06:57:05PM -0400, Tom Rini wrote:
> On Fri, Jul 17, 2020 at 09:03:17PM -0600, Simon Glass wrote:
>
> > Drop use of the preprocessor where possible.
> >
> > Signed-off-by: Simon Glass <s...@chromium.org>
> [snip]
> > + if (IS_ENABLED(CONFIG_DM_VIDEO)) {
> > + /*
> > + * If the console setting is not in environment variables then
> > + * console_init_r() will not be calling iomux_doenv() (which
> > + * calls search_device()). So we will not dynamically add
> > + * devices by calling stdio_probe_device().
> > + *
> > + * So just probe all video devices now so that whichever one is
> > + * required will be available.
> > + */
> > + struct udevice *vdev;
> > + int ret;
> > +
> > + if (!IS_ENABLED(CONFIG_SYS_CONSOLE_IS_IN_ENV)) {
> > + for (ret = uclass_first_device(UCLASS_VIDEO, &vdev);
> > + vdev;
> > + ret = uclass_next_device(&vdev))
> > + ;
> > + if (ret)
> > + printf("%s: Video device failed (ret=%d)\n",
> > + __func__, ret);
> > + }
> > + if (IS_ENABLED(CONFIG_SPLASH_SCREEN) &&
> > + IS_ENABLED(CONFIG_CMD_BMP))
> > + splash_display();
>
> We can't do this yet because CONFIG_SPLASH_SCREEN isn't migrated so the
> test fails and we drop the bmp logo. I got this run-time tested and
> confirmed on colibri imx6 by my colleague Matt Porter. So I'm holding
> off on _this_ patch (and then the style clean up patch) until I can push
> SPLASH_SCREEN migration through.This is also an issue for "stdio: Tidy up use of CONFIG_SYS_DEVICE_NULLDEV" as well, so I'm going to put aside the whole series until I can convert SPLASH_SCREEN. -- Tom
signature.asc
Description: PGP signature
