On Tue, 2017-10-10 at 10:32 +0200, Matthias Brugger wrote:

On 10/10/2017 10:22 AM, Chaotian Jing wrote:

quoted

On Tue, 2017-10-10 at 10:09 +0200, Ulf Hansson wrote:

quoted

On 10 October 2017 at 09:35, Chaotian Jing [off-list ref] wrote:

quoted

On Tue, 2017-10-10 at 09:26 +0200, Ulf Hansson wrote:

quoted

[...]

quoted

+
+static const struct of_device_id msdc_of_ids[] = {
+       { .compatible = "mediatek,mt8135-mmc", .data = &mt8135_compat},
+       { .compatible = "mediatek,mt8173-mmc", .data = &mt8173_compat},
+       { .compatible = "mediatek,mt2701-mmc", .data = &mt2701_compat},
+       { .compatible = "mediatek,mt2712-mmc", .data = &mt2712_compat},
+       {}
+};
+MODULE_DEVICE_TABLE(of, msdc_of_ids);

As already stated in the other reply. These new compatible changes
needs to be discussed and acked before the driver starts using them.

In other words, make patch3 to precede this one.

but, then there will have a probe issue as mentioned in previous mail
list. that's why I separate the binding file changes to 2 patches.

That sounds seriously wrong. Aren't the bindings backwards compatible?

Kind regards
Uffe

As Matthias mentioned before:
"

quoted

quoted

NAK, this has to be:
You have to add the fallback compatible ("mediatek,mt8135-mmc") to

the binding

quoted

quoted

description as otherwise the driver does not get probed.

Regards,
Matthias

"
the original compatible in driver is "mediatek,mt8135-mmc", so that if
drop "mediatek,mt8135-mmc" in bindings file(User may refer it and drop
"mediatek,mt8135-mmc" in their DTS), the driver will not get probed.

If you change the driver as you did in this patch, then there is no probing problem.

Regards,
Matthias

Okay, will make patch3 to precede this one at next version.