linux-kernel - Re: [PATCH v2] arm64: dts: qcom: x1e80100: Add crypto engine

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <Z7yJt4ARAM39F9b6@linaro.org>
Date: Mon, 24 Feb 2025 16:01:11 +0100
From: Stephan Gerhold <stephan.gerhold@...aro.org>
To: Abel Vesa <abel.vesa@...aro.org>
Cc: Bjorn Andersson <andersson@...nel.org>,
	Konrad Dybcio <konradybcio@...nel.org>,
	Rob Herring <robh@...nel.org>,
	Krzysztof Kozlowski <krzk+dt@...nel.org>,
	Conor Dooley <conor+dt@...nel.org>, Johan Hovold <johan@...nel.org>,
	linux-arm-msm@...r.kernel.org, devicetree@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] arm64: dts: qcom: x1e80100: Add crypto engine

On Fri, Feb 21, 2025 at 03:07:03PM +0200, Abel Vesa wrote:
> On X Elite, there is a crypto engine IP block similar to ones found on
> SM8x50 platforms.
> 
> Describe the crypto engine and its BAM.
> 
> Signed-off-by: Abel Vesa <abel.vesa@...aro.org>
> ---
> The dt-binding schema update for the x1e80100 compatible is here
> (already picked up):
> 
> https://lore.kernel.org/all/20250213-dt-bindings-qcom-qce-x1e80100-v1-1-d17ef73a1c12@linaro.org/
> ---
> Changes in v2:
> - Added EE and channels numbers in BAM node, like Stephan suggested.
> - Added v1.7.4 compatible as well.
> - Link to v1: https://lore.kernel.org/r/20250213-x1e80100-crypto-v1-1-f93afdd4025a@linaro.org
> ---
>  arch/arm64/boot/dts/qcom/x1e80100.dtsi | 32 ++++++++++++++++++++++++++++++++
>  1 file changed, 32 insertions(+)
> 
> diff --git a/arch/arm64/boot/dts/qcom/x1e80100.dtsi b/arch/arm64/boot/dts/qcom/x1e80100.dtsi
> index 9d38436763432892ceef95daf0335d4cf446357c..71d5f5eed4511030a51fb12e453f603d294080cc 100644
> --- a/arch/arm64/boot/dts/qcom/x1e80100.dtsi
> +++ b/arch/arm64/boot/dts/qcom/x1e80100.dtsi
> @@ -3708,6 +3708,38 @@ pcie4_phy: phy@...e000 {
>  			status = "disabled";
>  		};
>  
> +		cryptobam: dma-controller@...4000 {
> +			compatible = "qcom,bam-v1.7.4", "qcom,bam-v1.7.0";
> +			reg = <0 0x01dc4000 0 0x28000>;
> +
> +			interrupts = <GIC_SPI 272 IRQ_TYPE_LEVEL_HIGH>;
> +
> +			#dma-cells = <1>;
> +
> +			iommus = <&apps_smmu 0x480 0>,
> +				 <&apps_smmu 0x481 0>;
> +
> +			qcom,ee = <0>;
> +			qcom,num-ees = <7>;
> +			num-channels = <30>;

Hm, the hardware registers report the same as SM8550 on my X1E CRD:

			qcom,num-ees = <4>;
			num-channels = <20>;

[   17.679161] bam-dma-engine 1dc4000.dma-controller: bam_init() ees 4
[   17.679163] bam-dma-engine 1dc4000.dma-controller: bam_init() num-channels 20
[   17.679164] ------------[ cut here ]------------
[   17.679165] bam-dma-engine 1dc4000.dma-controller: DT specifies num-channels = <30>, but hardware reports 20. Please fix the device tree!
[   17.679175] WARNING: CPU: 5 PID: 121 at drivers/dma/qcom/bam_dma.c:498 bam_init+0x10c/0x150 [bam_dma]

(You don't have this warning yet, it's part of an upcoming bam_dma patch
 series from me that will start validating the num-channels given in the
 device tree.)

> +			qcom,controlled-remotely;
> +		};
> +
> +		crypto: crypto@...a000 {
> +			compatible = "qcom,x1e80100-qce", "qcom,sm8150-qce", "qcom,qce";
> +			reg = <0 0x01dfa000 0 0x6000>;
> +
> +			interconnects = <&aggre2_noc MASTER_CRYPTO QCOM_ICC_TAG_ALWAYS
> +					 &mc_virt SLAVE_EBI1 QCOM_ICC_TAG_ALWAYS>;
> +			interconnect-names = "memory";
> +
> +			dmas = <&cryptobam 4>, <&cryptobam 5>;
> +			dma-names = "rx", "tx";
> +
> +			iommus = <&apps_smmu 0x480 0>,
> +				 <&apps_smmu 0x481 0>;

iommus = <&apps_smmu 0x480 0x1> seems to work fine for me, can you try
again? Not sure if this is related, but make sure to apply

https://lore.kernel.org/linux-arm-msm/20250208223112.142567-1-caleb.connolly@linaro.org/

since that can easily cause crashes for bam_dma. I added this diff for
testing and it's working fine so far:

diff --git a/arch/arm64/boot/dts/qcom/x1e80100.dtsi b/arch/arm64/boot/dts/qcom/x1e80100.dtsi
index afb8bd3b6a94..bac3d6a38055 100644
--- a/arch/arm64/boot/dts/qcom/x1e80100.dtsi
+++ b/arch/arm64/boot/dts/qcom/x1e80100.dtsi
@@ -3732,8 +3732,7 @@ cryptobam: dma-controller@...4000 {
 
 			#dma-cells = <1>;
 
-			iommus = <&apps_smmu 0x480 0>,
-				 <&apps_smmu 0x481 0>;
+			iommus = <&apps_smmu 0x480 0x1>;
 
 			qcom,ee = <0>;
 			qcom,num-ees = <4>;
@@ -3752,8 +3751,7 @@ crypto: crypto@...a000 {
 			dmas = <&cryptobam 4>, <&cryptobam 5>;
 			dma-names = "rx", "tx";
 
-			iommus = <&apps_smmu 0x480 0>,
-				 <&apps_smmu 0x481 0>;
+			iommus = <&apps_smmu 0x480 0x1>;
 		};
 
 		tcsr_mutex: hwlock@...0000 {

Thanks,
Stephan