From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-oa2-f0.google.com (mail-oa2-f0.google.com [74.125.231.64]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 5050B175A76 for ; Fri, 3 Jul 2026 07:16:38 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=74.125.231.64 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1783063000; cv=none; b=stqXf8aOeznEuUqiJ1yJ9RCx4fA8g30Wdbiy4Zas7B/Ebt+CPjIqjrJfrMAuFb8sZL6OWHyCh9LwopssCG2cL0eBHNPM8IkslZg1ROKHowrMnQiQRoJekicWAMrdGy6QLggOw7KNy9y43uSJnsOUbyvP5JDMN2UICqYPlxM+xnY= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1783063000; c=relaxed/simple; bh=pDWQBNRlaLxTmIvMc3NFXeZsGJx8oiS0avPZ2MjhvNQ=; h=Date:From:To:Cc:Subject:Message-ID:MIME-Version:Content-Type: Content-Disposition:In-Reply-To; b=aysPqb09GYsCBsWGRjnvhtV/7Z2sFY8bAPVDnmMB9E6JTh2rf83FHG+6grsp6sekzx3F4lE9LsBrTdBMm/5ntkuhzWf9T504FaCzY+GgJq5r6hUEKYo3g0lUHpgKFHiEBKZoyd9L521dwW0WKHSfZLswkbv18KwYwTymSPAPZNw= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=gmail.com; spf=pass smtp.mailfrom=gmail.com; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.b=muQpyQfz; arc=none smtp.client-ip=74.125.231.64 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=gmail.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=gmail.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.b="muQpyQfz" Received: by mail-oa2-f0.google.com with SMTP id 586e51a60fabf-448433e6e3eso129176fac.0 for ; Fri, 03 Jul 2026 00:16:38 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20251104; t=1783062997; x=1783667797; darn=vger.kernel.org; h=in-reply-to:content-disposition:mime-version:message-id:subject:cc :to:from:date:from:to:cc:subject:date:message-id:reply-to; bh=8/To5uu366iTzXiach4JBOKu6hRxUbkL3Sh+ENphnrM=; b=muQpyQfz8mSGPrEH/S0/huNlfsruRmRJdrwcpbUf2xFVIiR4ySTY7+mKh5ts4qn6Cc rQLNfPovFWDJ3y61uIOxdZLphR54ZyBDHKkaA+6Zg5jmtjySLSRp+C+VXwcy7cH8g3xV gDxoMc6PxrQbdcauwENd1PuRJME1+6wd+aWsve/UrUiWNZP3rAReO0EM2AO9oI4/FIBZ djBTTEfEAA85EsuFU0P2THGOCKHgFrE4LIyFt9TnZAevy3zCH6sBj/Q8+CfGk5NhPcaN UU+DlXuZ46BFMebmkmNywSo1JA8dZHHN9dpz0y2k89wqssltnQjHVps5Vzg0yGb3E9yh segg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20251104; t=1783062997; x=1783667797; h=in-reply-to:content-disposition:mime-version:message-id:subject:cc :to:from:date:x-gm-gg:x-gm-message-state:from:to:cc:subject:date :message-id:reply-to; bh=8/To5uu366iTzXiach4JBOKu6hRxUbkL3Sh+ENphnrM=; b=lr7BuXf4+qYvcBzVLwD1fCrdOQNNxa8r8huqDRBkInjfcr/fKFW9aaaOl3BLX4dPIA tscq3kWJhK2wp0OsjfgMzpFDa3fH4ivuTqirF1XaIAEPYLaAIHTa+wBTMVbeSTtRMoVr 5QUelk4Ugp29p77+ldtajiW+ERuQr8cIuAaqIwH+7GaQyn9MycW5UjCR4skqa4XX2y8y Xg9JnDlSwqV6uGmVr8LTTCckHo7Uzxc+LKuHQSebkBBaa9z2UKoFkFRmK6pG9Xu6en9z dmrQoJC7W51JkeCSJztCFM5BAfA/tLHKaCzBBaOAXLUvLhZ4xU09y/Aui3l27nAKwlvt wUTQ== X-Forwarded-Encrypted: i=1; AFNElJ/Hdfzaah23bTdTT2+aI9REFWV0KnVopmipPgdEUea8hRvgJK4wCdA1vfGLnJPT/Xt0rX7dmvvLGMwuQOw=@vger.kernel.org X-Gm-Message-State: AOJu0Yzc+vHyjN6kI2U9cSRn7LUU0KUSv0zGjKtSlm2VFvUEfkPaWgsI 9DrapIwltqW6tMFO2nU6m36eTHGgGqdQg/iWfSNwdpcTcNE2+8/yFymJ X-Gm-Gg: AfdE7clPOQzYGKxi/rbJXHM5CTsu2ewJQRY/nnGRMOj/W9wHcMsbcsqdkZO/stPFfN0 4J7nS0ClGDi02GWUFlomL6rLZhkshw7sg/Jaw1ED1ky/pNteoqTI0/YYKnpPmfp3RXl5d0MpHw3 nsykEmMIOIGSMM/HCTalaln2mZRxW0EMLMXppBXB6uCH2GPJ/YTZzrnCx6q/E1Qj7OzPlOsEo0F 3S0uZIAIuSDxtXiHXDbWDlcUKOwl2AWgNk9x2hD02Q09NsroM+b0nojbKfq5xwtWz+dcK95p4nX Hd683RByWiUbkeEeUjG+uoIsAK5JMqD2FUBMj5Qo/dId0Dkw/7eqLYL9jpoP45MRE0+rPaLzSEF QtITboGlsVVeDuFSXbMedT/Usgy+OEE02v+w1lvm0LlaoRYFxT6/jObmC/hXl7f0BmHZgaMYCWC 4RY3qQwcNnDkfr6N+kt2DD2upXLlpSfZ2i8NndXEqQ0wNLY16dB4L7Ub8f/cqNrlc= X-Received: by 2002:a05:6870:2187:b0:449:bccd:79fb with SMTP id 586e51a60fabf-44cab961fb9mr6076032fac.24.1783062997063; Fri, 03 Jul 2026 00:16:37 -0700 (PDT) Received: from ubuntu24 (dsl092-249-254.sfo4.dsl.speakeasy.net. [66.92.249.254]) by smtp.gmail.com with ESMTPSA id 586e51a60fabf-44cfb54f775sm1162626fac.10.2026.07.03.00.16.33 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 03 Jul 2026 00:16:36 -0700 (PDT) Date: Fri, 3 Jul 2026 15:16:29 +0800 From: Xing Loong To: Krzysztof Kozlowski Cc: Jens Wiklander , Krzysztof Kozlowski , Conor Dooley , Rob Herring , Sumit Garg , op-tee@lists.trustedfirmware.org, devicetree@vger.kernel.org, linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH 2/3] dt-bindings: firmware: add mbedtee,rpc binding Message-ID: Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <70068ae8-ede1-41d2-899b-98ed11e19953@kernel.org> On 01/07/2026 16:05, Krzysztof Kozlowski wrote: > On 01/07/2026 15:25, Xing Loong wrote: > > Add YAML devicetree binding for the MbedTEE Trusted Execution > > Drop YAML, there is no such thing as YAML binding. Will drop. > > Environment driver. > > We don't take bindings for drivers but for hardware or firmware. Please > describe these instead. Will rewrite the description to describe what MbedTEE is as firmware, not the driver or the binding. > > The binding covers two platform configurations: > > - ARM/AArch64 (TrustZone, SMC): two reserved-memory regions > > (rpc-t2r-ring and rpc-t2r-shm) plus a GIC SPI edge interrupt > > for TEE-to-REE notifications. > > - RISC-V (IMSIC): three reserved-memory regions, adding > > rpc-r2t-ring for REE-to-TEE command submissions; no interrupts > > property (T2R notifications use IMSIC MSI allocated at runtime). > > > > Signed-off-by: Xing Loong > > --- > > .../bindings/firmware/mbedtee,rpc.yaml | 221 ++++++++++++++++++ > > 1 file changed, 221 insertions(+) > > create mode 100644 Documentation/devicetree/bindings/firmware/mbedtee,rpc.yaml > > > > diff --git a/Documentation/devicetree/bindings/firmware/mbedtee,rpc.yaml b/Documentation/devicetree/bindings/firmware/mbedtee,rpc.yaml > > new file mode 100644 > > index 0000000..08ae255 > > --- /dev/null > > +++ b/Documentation/devicetree/bindings/firmware/mbedtee,rpc.yaml > > @@ -0,0 +1,221 @@ > > +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) > > +%YAML 1.2 > > +--- > > +$id: http://devicetree.org/schemas/firmware/mbedtee,rpc.yaml# > > +$schema: http://devicetree.org/meta-schemas/core.yaml# > > + > > +title: MbedTEE Trusted Execution Environment > > + > > +maintainers: > > + - Xing Loong > > + > > +description: | > > + MbedTEE is a Trusted Execution Environment for embedded systems. > > + This binding describes the shared-memory regions used for RPC > > Describe firmware, not the binding. It's redundant to say what the > binding is about, just say what is the hardware. Will rewrite, removing "driver", "YAML" and "This binding describes" wording throughout. > > + communication between the Linux REE driver and MbedTEE OS. > > + > > + The REE and TEE CPUs sharing the RPC memory must be in a > > + hardware-coherent domain (same CPU cluster, coherent caches). > > + > > + Two or three reserved-memory regions are required: > > + > > + rpc-t2r-ring ring buffer for TEE-to-REE notifications (all platforms) > > + rpc-t2r-shm shared memory for TEE-to-REE RPC payloads (all platforms) > > + rpc-r2t-ring ring buffer for REE-to-TEE command submissions (RISC-V only) > > + > > + On ARM/AArch64 the transport uses SMC calls; TEE-to-REE > > + notifications use a GIC SPI edge interrupt. > > + > > + On RISC-V the TEE notifies the REE via IMSIC MSI; the REE submits > > + commands via shared-memory rpc-r2t-ring that the TEE polls. No > > + REE-to-TEE interrupt is used. No SBI ecall is involved. > > + > > +properties: > > + $nodename: > > + const: mbedtee > > Drop, why would it be relevant? Will drop. > > + > > + compatible: > > + const: mbedtee,rpc > > Feels way too generic. First, Google results on mbedtee are basically > non-existing, so what sort of company is that? > > Second, rpc is just not specific enough. Please carefully read writing > bindings doc. Will rename to "mbedtee,tee". mbedtee is an open-source TEE project (https://github.com/mbedtee), not a company. Patch 1/3 adds the corresponding "mbedtee" entry to vendor-prefixes.yaml. > > + > > + interrupts: > > + description: > > + GIC interrupt used by the TEE to notify the REE of pending RPC > > + responses (ARM/AArch64 only). Not present on RISC-V platforms which > > + use IMSIC platform MSI interrupts allocated dynamically at runtime. > > Please read writing bindings doc. Will drop the description. Will add maxItems: 1 instead. > > + > > + msi-parent: > > + maxItems: 1 > > + description: > > + IMSIC MSI controller used by the Linux driver to allocate the > > Again drivers... Will drop the description. Will keep only maxItems: 1. > > + TEE-to-REE notification interrupt on RISC-V platforms. Not present on > > + ARM/AArch64 platforms, which use the interrupts property. > > + > > + memory-region: > > + minItems: 2 > > + maxItems: 3 > > + description: > > + References to reserved-memory regions for REE<->TEE communication. > > + Entries must match memory-region-names order. > > Obvious. Please do not come with your own style of bindings. Will drop the description. Understood, will follow the standard style used by existing bindings. > > + > > + memory-region-names: > > + minItems: 2 > > + maxItems: 3 > > Why is this flexible? Will constrain it via the allOf: if/then branches (ARM: fixed at 2, RISC-V: fixed at 3). > > + items: > > + enum: > > + - rpc-t2r-ring > > rpc is redundant, drop Will drop the "rpc-" prefix. Will use t2r-ring, t2r-shm, r2t-ring. > > + - rpc-t2r-shm > > + - rpc-r2t-ring > > + > > +required: > > + - compatible > > + > > +allOf: > > + - if: > > + required: > > + - interrupts > > + then: > > + required: > > + - interrupts > > + - memory-region > > + - memory-region-names > > + properties: > > + msi-parent: false > > + memory-region: > > + minItems: 2 > > + maxItems: 2 > > + memory-region-names: > > + items: > > + - const: rpc-t2r-ring > > + - const: rpc-t2r-shm > > + else: > > + required: > > + - msi-parent > > + - memory-region > > + - memory-region-names > > So memory-region is always required? Yes. Will move memory-region and memory-region-names to the top-level required: list to make this explicit. The interrupts and msi-parent properties will also be made mutually exclusive per platform (msi-parent: false on ARM, interrupts: false on RISC-V). > > + properties: > > + memory-region: > > + minItems: 3 > > + maxItems: 3 > > + memory-region-names: > > + items: > > + - const: rpc-t2r-ring > > + - const: rpc-t2r-shm > > + - const: rpc-r2t-ring > > Your top level schema said that. You only need minItems. Will drop the redundant constraints from the else branch. > > + > > +additionalProperties: false > > + > > +examples: > > + - | > > + /* ARM TrustZone (SMC) */ > > + #include > > + / { > > + #address-cells = <2>; > > + #size-cells = <2>; > > + > > + gic: interrupt-controller@2f000000 { > > + compatible = "arm,gic-v3"; > > + reg = <0 0x2f000000 0 0x10000>, > > + <0 0x2f100000 0 0x200000>; > > + interrupt-controller; > > + #interrupt-cells = <3>; > > + }; > > + > > + reserved-memory { > > + #address-cells = <2>; > > + #size-cells = <2>; > > + ranges; > > + > > + mbedtee_t2r_ring: rpc-t2r-ring@85f10000 { > > + reg = <0 0x85f10000 0 0x20000>; > > + no-map; > > + }; > > + > > + mbedtee_t2r_shm: rpc-t2r-shm@85f30000 { > > + reg = <0 0x85f30000 0 0x40000>; > > + no-map; > > + }; > > + }; > > None of the above is relevant, drop. Will drop all platform infrastructure nodes. Will follow the minimal example pattern seen in linaro,optee-tz.yaml. > > + > > + firmware { > > + mbedtee { > > + compatible = "mbedtee,rpc"; > > + interrupt-parent = <&gic>; > > + interrupts = ; > > + memory-region = <&mbedtee_t2r_ring>, <&mbedtee_t2r_shm>; > > + memory-region-names = "rpc-t2r-ring", "rpc-t2r-shm"; > > + }; > > + }; > > + }; > > + > > Best regards, > Krzysztof Thanks for the thorough review. Will send v3 after incorporating all of the above. Best regards, Xing Loong