drivers: video: dcmi: use devicetree compatibles for platform-detection

[josuah]

Summary

In order to allow some platform-specific HAL features to be detected, DCMI uses #if defined(CONFIG_SOC_SERIES_STM32F7X) || defined(CONFIG_SOC_SERIES_STM32H7X).
A suggestion from #92018 (comment) was to use the devicetree instead, with a different compatible for each variant.

Describe the solution you'd like

This can be migrated to the same strategy as i.e. i2s, and use platform-specific compatibles:

diff --git a/dts/arm/st/h7/stm32h7.dtsi b/dts/arm/st/h7/stm32h7.dtsi
index 5cbf3c9b090..6c05239127d 100644
--- a/dts/arm/st/h7/stm32h7.dtsi
+++ b/dts/arm/st/h7/stm32h7.dtsi
@@ -1081,7 +1081,7 @@
                };

                dcmi: dcmi@48020000 {
-                       compatible = "st,stm32-dcmi";
+                       compatible = "st,stm32h7-dcmi", "st,stm32-dcmi";
                        reg = <0x48020000 0x400>;
                        interrupts = <78 0>;
                        interrupt-names = "dcmi";
diff --git a/dts/arm/st/l4/stm32l4p5.dtsi b/dts/arm/st/l4/stm32l4p5.dtsi
index 9b0aca624ed..907674731f9 100644
--- a/dts/arm/st/l4/stm32l4p5.dtsi
+++ b/dts/arm/st/l4/stm32l4p5.dtsi
@@ -336,7 +336,7 @@
                };

                dcmi: dcmi@50050000 {
-                       compatible = "st,stm32-dcmi";
+                       compatible = "st,stm32l4-dcmi", "st,stm32-dcmi";
                        reg = <0x50050000 0x400>;
                        interrupts = <85 0>;
                        interrupt-names = "dcmi";

Then in the dcmi driver, do like on I2S:

#if DT_HAS_COMPAT_STATUS_OKAY(st_stm32h7_i2s)
                        (void *)LL_SPI_DMA_GetRxRegAddr(cfg->i2s),
#else
                        (void *)LL_SPI_DMA_GetRegAddr(cfg->i2s),
#endif

Alternatives

Use CONFIG_SOC_SERIES_STM32H7X like now, which is also used by drivers.

Additional Context

No preference on my side! Only making sure that the preferred way is the one in use.

[avolmat-st]

Approach taken in PR #92018 sounds a good approach to solve this issue so I think this issue can be closed now since PR #92018 has been merged.