From mboxrd@z Thu Jan 1 00:00:00 1970 Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.remote-ip=100.103.45.18 ARC-Seal: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; cv=none; t=1794898335; b=KH86KfugD5cdKujjPzJD5GyIm8+XuiOpDMnCLvPzCRIPrgmcOqHeHDgyasHmoNV/YlrL aFUbK0IXXDPmrbgFlQizAWq7ZU9n+DKaxDPWtCU72modzy5BlNWOMpPGPFcSEP7FF3HXT dmi9itkjOd9CmYJSS7vO9kOdDSZ+X2LUaJBM6mGu+0yCoSjqMYfJTYhkyr2ro+SP/Ge8M qwdiSUg8e7fxY66p+uvCOF/GSrmIhVtvLKNrSqByWNIjxRvCdFTISWxcdwhADyF79tL6p 2kBRnOo0mGQErfrHrmi2+DQv+ZtpA7l0jYbrLutopNrBjrBZlLX5md2T6fk8MCHMLQw== ARC-Message-Signature: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; c=relaxed/relaxed; t=1794898335; h=Received:DKIM-Signature:Message-ID:From:Subject:Reply-To:To:Cc: In-Reply-To:References:Content-Type:Content-Transfer-Encoding:Date; bh=5NZ3XEgrs0JaKpU6uuTYivkYyZ9SQrxsqyQV/m+pDvY=; b=2J6iaiiUJ1zokA8UucVyczOwPyUsp/H+yr4r3mLJAFKXVDDmmjTpfq6BdirJg+i1qs1F IKDSc3Y0coqJKHfZAkK4tgYMmD25YPSbLd7GcdGQLdwMF3FQvvtidElu3GFjTsGAsBM5v 54m2MOI+6pgoMum2BPOJR+nj2o8yBAI5dc9xPFoAbhLHRCnetqgIzACmrIgYhvF4iAUIy v+8+TUbZn1tpB88MyWI+rHQp6yZDXH7HUr5mo7050uV0RY2MdUk+91fy8NfJA+HqhWtHp BHZoedjPAsi42Gq8Gj2g5MMY2vErYztGsDpQq/+gipiHTXnMUMrrlS+SPFZxKD25o0w== ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass header.d=kernel.org header.i=@kernel.org header.a=rsa-sha256 header.s=k20260515 header.b=TJpWXT1K; arc=none smtp.remote-ip=100.103.45.18 Received: from smtp.kernel.org (quasi.space.kernel.org [100.103.45.18]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (prime256v1)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 01D5F606A31C for ; Tue, 17 Nov 2026 06:52:14 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.a=rsa-sha256 header.s=k20260515 header.b=TJpWXT1K Received: by smtp.kernel.org (Postfix) with ESMTPSA id 7B6E91F000C4; Tue, 17 Nov 2026 06:52:14 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1794898334; bh=5NZ3XEgrs0JaKpU6uuTYivkYyZ9SQrxsqyQV/m+pDvY=; h=From:Subject:Reply-To:To:Cc:In-Reply-To:References:Date; b=TJpWXT1K1pcAPWZefPH28GaTDwXWIxQ3dmuNteLy+Iwrr7q0/9i9kTlCAL0NTd+Fj muZKuEABbG3XzEz86Vy8mhLHKoqKlGLTlR5U7x6GEOZG1FULC1tlBSILzQKeJfBNWI lmYP3dpyEjxk9PqWbPmiQuWOIpgZ3J0hA5OmQCqdI/ZNJEmoE+YL2+IKUf3aOjqObw +BeBEf+GZ01c31I2zyO0Lp1AqoVxzbZhiIman5Py2VH/r9sSibF07OoYfsTSKMc8ov tWOyaGZwY+jc0VQKcG1slTko2QfA8uoqXr+FwHOSqza4cXtGNypwJCzaHs3CNb3TZr jxU4FAcfUn8BA== Message-ID: From: sashiko-bot@kernel.org Subject: Re: [PATCH v3 4/5] arm64: dts: qcom: eliza-mtp: Enable iris video codec Reply-To: sashiko-reviews@lists.linux.dev To: "Wangao Wang" Cc: devicetree@vger.kernel.org, robh@kernel.org, conor+dt@kernel.org, media-ci@linuxtv.org In-Reply-To: <20261117-add_iris_for_eliza-v3-4-02a55a9272ef@oss.qualcomm.com> References: <20261117-add_iris_for_eliza-v3-0-02a55a9272ef@oss.qualcomm.com> <20261117-add_iris_for_eliza-v3-4-02a55a9272ef@oss.qualcomm.com> Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Date: Tue, 17 Nov 2026 06:52:14 +0000 X-Mailing-List: devicetree@vger.kernel.org List-Id: List-Post: List-Help: List-Subscribe: List-Unsubscribe: List-Owner: Precedence: list Thank you for your contribution! Sashiko AI review found 2 potential issues to consider: - [High] arm64: dts: qcom: eliza-mtp: iris_probe() fails due to child-node iommus - [High] media: iris: NULL deref in iris_set_slice_count() from incomplete eliza_data -- commit 0429ecb16555afb36c6bba89f407a14fa28f8a19 Author: Wangao Wang arm64: dts: qcom: eliza-mtp: Enable iris video codec This commit enables the iris video codec device tree node on the Eliza MTP board. > diff --git a/arch/arm64/boot/dts/qcom/eliza-mtp.dts b/arch/arm64/boot/dts= /qcom/eliza-mtp.dts > index ee05192adcfb9..e5a698602dfc7 100644 > --- a/arch/arm64/boot/dts/qcom/eliza-mtp.dts > +++ b/arch/arm64/boot/dts/qcom/eliza-mtp.dts > @@ -430,6 +430,10 @@ &gpu_zap_shader { > firmware-name =3D "qcom/eliza/gen71700_zap.mbn"; > }; > =20 > +&iris { > + status =3D "okay"; > +}; [Severity: High] Does enabling &iris cause iris_probe() to fail with -ENOMEM? In commit "arm64: dts: qcom: eliza: Add iris video codec node", the video-codec@aa00000 node places its iommus phandles inside non-pixel and pixel child nodes rather than on the video-codec node itself: arch/arm64/boot/dts/qcom/eliza.dtsi { ... iris: video-codec@aa00000 { ... non-pixel { iommus =3D <&apps_smmu 0x1940 0x0>; iommu-ranges =3D <0x25800000 0xba800000>; }; pixel { iommus =3D <&apps_smmu 0x1947 0x0>; }; }; } The upstream iris driver does not parse those child nodes and instead operates directly on the top-level platform device. Because the parent node has no iommus property, pdev->dev is not attached to apps_smmu and use_dma_iommu(dev) is false, so iris_probe() -> iris_reserve_iova_region() -> dma_iova_try_alloc() fails immediately. This layout also fails dtbs_check against Documentation/devicetree/bindings/media/qcom,sm8550-iris= .yaml, which requires iommus on the video-codec node and sets unevaluatedProperties: false. [Severity: High] Can enabling &iris lead to a NULL pointer dereference in iris_set_slice_count() when encoding with multiple slices? Enabling this node binds the device to eliza_data (added in commit "media: iris: Add support for Eliza" in drivers/media/platform/qcom/iris/iris_platform_vpu3x.c), which omits .slice_caps, .wd_intr_mask, and .icc_ib_multiplier unlike the other VPU3x platform descriptors. When a user configures multi-slice mode (V4L2_MPEG_VIDEO_MULTI_SLICE_MODE_MAX_MB or V4L2_MPEG_VIDEO_MULTI_SLICE_MODE_MAX_BYTES), iris_set_slice_count() dereferences inst->core->iris_platform_data->slice_caps unconditionally: drivers/media/platform/qcom/iris/iris_ctrls.c:iris_set_slice_count() { ... slice_caps =3D inst->core->iris_platform_data->slice_caps; ... if (fps > slice_caps->max_slice_frame_rate || (rc_type !=3D HFI_RC_OFF && rc_type !=3D HFI_RC_CBR_CFR && ... } Leaving .wd_intr_mask and .icc_ib_multiplier zero-initialized in eliza_data also causes iris_vpu_clear_interrupt() and iris_vpu_watchdog() to ignore hardware watchdog interrupts, and causes iris_set_icc_bw() to vote 0 peak bandwidth on the video-mem interconnect path. --=20 Sashiko AI review =C2=B7 https://sashiko.dev/#/patchset/20261117-add_iris_f= or_eliza-v3-0-02a55a9272ef@oss.qualcomm.com?part=3D4