Le Tue, Feb 15, 2022 at 03:07:56PM +0100, Johan Jonker a écrit : > Hi Heiko, > > On 2/11/22 12:59, Corentin Labbe wrote: > > Convert rockchip-crypto to yaml > > > > Signed-off-by: Corentin Labbe <clabbe@xxxxxxxxxxxx> > > --- > > Changes since v1: > > - fixed example > > - renamed to a new name > > - fixed some maxItems > > > > Change since v2: > > - Fixed maintainers section > > > > .../crypto/rockchip,rk3288-crypto.yaml | 66 +++++++++++++++++++ > > .../bindings/crypto/rockchip-crypto.txt | 28 -------- > > 2 files changed, 66 insertions(+), 28 deletions(-) > > create mode 100644 Documentation/devicetree/bindings/crypto/rockchip,rk3288-crypto.yaml > > rockchip,crypto.yaml Hello DT maintainer asked for rockchip,rk3288-crypto, so I will keep it. > > > delete mode 100644 Documentation/devicetree/bindings/crypto/rockchip-crypto.txt > > > > diff --git a/Documentation/devicetree/bindings/crypto/rockchip,rk3288-crypto.yaml b/Documentation/devicetree/bindings/crypto/rockchip,rk3288-crypto.yaml > > new file mode 100644 > > index 000000000000..2e1e9fa711c4 > > --- /dev/null > > +++ b/Documentation/devicetree/bindings/crypto/rockchip,rk3288-crypto.yaml > > @@ -0,0 +1,66 @@ > > +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) > > +%YAML 1.2 > > +--- > > +$id: http://devicetree.org/schemas/crypto/rockchip,rk3288-crypto.yaml# > > rockchip,crypto.yaml > > > +$schema: http://devicetree.org/meta-schemas/core.yaml# > > + > > +title: Rockchip Electronics And Security Accelerator > > + > > +maintainers: > > + - Heiko Stuebner <heiko@xxxxxxxxx> > > + > > +properties: > > + compatible: > > oneOf: > - const: rockchip,rk3288-crypto > - items: > - enum: > - rockchip,rk3228-crypto > - rockchip,rk3328-crypto > - rockchip,rk3368-crypto > - rockchip,rk3399-crypto > - const: rockchip,rk3288-crypto > > rk3288 was the first in line that had support, so we use that as fall > back string. I will dont add compatible which are not handled by the driver unless DT maintainer said I should. > > > + const: rockchip,rk3288-crypto > > + > > + reg: > > + maxItems: 1 > > + > > + interrupts: > > + maxItems: 1 > > + > > + clocks: > > + items: > > + - description: clock data > > + - description: clock data > > + - description: clock crypto accelerator > > > + - description: clock dma > > remove ??? > > > + > > + clock-names: > > + items: > > + - const: aclk > > + - const: hclk > > + - const: sclk > > > + - const: apb_pclk > > remove ??? > > Similar to the rk3568 pclk_xpcs discussion ACLK_DMAC1 belongs to the > dmac_bus_s node and should have been enabled by the DMA driver I think. > Could you advise if this is correct or should we remove parsing/enabling > ACLK_DMAC1 in rk3288_crypto.c in order to it easier > porting/adding/syncing nodes for other SoC types? > > Johan I tested on my rk3328-rock64 and I confirm apb_pclk is unnessecary. I will issue patch for fixing this. Regards