In the discussions here: https://lore.kernel.org/all/CAMuHMdXM34nNz1nfowNqDvdsdg+d69Bo3_ufs6fbcq65iYd5-A@xxxxxxxxxxxxxx/T/#m3449390080c5a0c8f59b8f8286a87e31e093b98b Marc Zyngier points out that the ls-extirq driver should have never used the "interrupt-map" keyword for mapping its channels to GIC interrupts. This change modifies the device trees for the LS1046A to use a driver-specific OF property for that purpose. The change is intended to target stable kernels, to accelerate the conversion and ultimate removal of the improperly used "interrupt-map" property. Note that "interrupt-map-mask" isn't needed now (nor was it needed before, it was just there to keep an apparent compatibility in form with the reserved "interrupt-map" property). So delete it. Fixes: 7968344126e5 ("arm64: dts: ls1046a: add DT node for external interrupt lines") Signed-off-by: Vladimir Oltean <vladimir.oltean@xxxxxxx> --- arch/arm64/boot/dts/freescale/fsl-ls1046a.dtsi | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/arm64/boot/dts/freescale/fsl-ls1046a.dtsi b/arch/arm64/boot/dts/freescale/fsl-ls1046a.dtsi index 687fea6d8afa..c735469010b7 100644 --- a/arch/arm64/boot/dts/freescale/fsl-ls1046a.dtsi +++ b/arch/arm64/boot/dts/freescale/fsl-ls1046a.dtsi @@ -328,7 +328,7 @@ extirq: interrupt-controller@1ac { #address-cells = <0>; interrupt-controller; reg = <0x1ac 4>; - interrupt-map = + fsl,extirq-map = <0 0 &gic GIC_SPI 131 IRQ_TYPE_LEVEL_HIGH>, <1 0 &gic GIC_SPI 132 IRQ_TYPE_LEVEL_HIGH>, <2 0 &gic GIC_SPI 133 IRQ_TYPE_LEVEL_HIGH>, @@ -341,7 +341,6 @@ extirq: interrupt-controller@1ac { <9 0 &gic GIC_SPI 149 IRQ_TYPE_LEVEL_HIGH>, <10 0 &gic GIC_SPI 150 IRQ_TYPE_LEVEL_HIGH>, <11 0 &gic GIC_SPI 151 IRQ_TYPE_LEVEL_HIGH>; - interrupt-map-mask = <0xffffffff 0x0>; }; }; -- 2.25.1
