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=1794858403; b=CGSFSgI8w9dmxPjMFgypukgrhGs1E2LmJt9ZqeMZTtpo7d/QiKyKpHVn73a6qiMgiQ7d F5IvAW0mad+58hFQM0Ot3iMAgtfrG2S1S4bcXh6cxeW6AXGlwT+vrW66L8DMmI4uWH1jT MvCyokpRwDjEG8WGnH1CtmT3JebfERF/GmHFnEUgDRlzcJfhlziYJxr15LkcIAMgtuqmP K9jACkO8KElX8MR00hWupaht07oZ3OI3XbeafBA4gdRRVYFpYYMQe7DBq8cabBe3N3S7r 5Qqj0ThF1ajDVdL9LnwFzobJh60WjLbuIDpXVdHIdK+1UNyTUCJjA6unQew7YN2+4/Q== ARC-Message-Signature: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; c=relaxed/relaxed; t=1794858403; h=Received:DKIM-Signature:Message-ID:Date:MIME-Version:User-Agent: Subject:Content-Language:To:Cc:References:From:In-Reply-To: Content-Type:Content-Transfer-Encoding; bh=rb4amnB1VYxDSnvAGyLyVc2GTkn7Gl6f5gxVefQoqyA=; b=xpwlZDUnTEpp3aJ4zn93ROL+75aN2m+RTvTIG5b2Xo9Q9M6J0oH5XgBvkz9kVy11z5Uh OLPo506KcqzNDzK85fxz4o24tX1idMJ+ezi5mfwnBGEaC0QW33L3U5WIdU+b00lrNLpIK NnVXDPiK2GYfg3f1iUeZVpnLorKaWWGmAYuMu7v4MAAa+EvJRDq1M1ix4E4A07KUf3sMT mv3mGO4cYP4gcJlm80dn+xIbsNWvCRhBibdAypEBKPH7iWUsuS7R9AxEQ/BMZxTaRatIU hrDlLACp02pGa8dUY4fawgOfG6gOu85MmEztAYW4Lk8KTWHxv1OlCORIPCW9L87iEEw== 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="Whvh/yBW"; 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 88451606A31C; Mon, 16 Nov 2026 19:46:43 +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=Whvh/yBW Received: by smtp.kernel.org (Postfix) with ESMTPSA id 703D81F000C4; Mon, 16 Nov 2026 19:46:42 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1794858403; bh=rb4amnB1VYxDSnvAGyLyVc2GTkn7Gl6f5gxVefQoqyA=; h=Date:Subject:To:Cc:References:From:In-Reply-To; b=Whvh/yBWQP8P4jnBEcySSvK8dMq1EiELToOKyevOUi8LDPUsa0gdUd5im6CAlRD49 o6ZYCSMJ84ijHtui/M5Q/hhEsuKb3EXRXfBEudhNbp4d/rtuXULXn4dvidnIlCV+MN 9j0MDqdBd6STtQe7SI+r52mmHonGVoaY+pWbDIxknEzzRw7oQ3Au0v+6KX6uERmVVw 7NxxkpQc2bGdUahNmAn/XBW581869pURbIdG6tG0hncak77RW0EKZomYysa0q/R7H3 QsQZ9e/lpnq2WQ5t1InwB8P2JyXaaVz3dkGQI0u3ZB7/8AQ/pbC8Ad/jBqDJeC/+pT iTu/nugQUTBkA== Message-ID: <818e80bc-4b7c-404a-9c64-b103935eea15@kernel.org> Date: Mon, 16 Nov 2026 13:46:40 -0600 X-Mailing-List: linux-security-module@vger.kernel.org List-Id: List-Post: List-Help: List-Subscribe: List-Unsubscribe: List-Owner: Precedence: list MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Subject: Re: [PATCH 0/2] lockdown: Prevent booting with lockdown enabled but IOMMU disabled Content-Language: en-US To: Jason Gunthorpe Cc: "Joerg Roedel (AMD)" , Will Deacon , Robin Murphy , Nicolas Bouchinet , Xiu Jianfeng , Paul Moore , James Morris , "Serge E. Hallyn" , iommu@lists.linux.dev, linux-kernel@vger.kernel.org, linux-security-module@vger.kernel.org References: <20261102-lockdown-iommu-v1-0-00f3c2be4c9f@kernel.org> <20261116194430.GD2452510@ziepe.ca> From: Mario Limonciello In-Reply-To: <20261116194430.GD2452510@ziepe.ca> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit On 11/16/26 13:44, Jason Gunthorpe wrote: > On Mon, Nov 02, 2026 at 08:00:24AM -0600, Mario Limonciello (AMD) wrote: >> Lockdown is a kernel LSM that restricts access to sensitive kernel >> resources (e.g. /dev/mem, kernel modules, perf, tracing, kprobes, >> BPF access to kernel memory). However, without IOMMU, any PCI/PCIe >> device can perform DMA to arbitrary physical memory, effectively >> bypassing all lockdown protections. > > I don't see how this is enough? > > You can have an iommu but the commandline can put the whole thing in > DMA passthrough, which is the same functionality as not having an > iommu. Well that's why I was suggesting that we explicitly blocked manipulating IOMMU from the command line when lockdown is enabled. > > It is not like "any" PCI device can do this DMA. Lockdown > systematically prevents any userspace from triggering such an > uncontrolled DMA. > > This is more about preventing someone with physical control of the HW > from access all of memory, which IMHO, is not part of the threat model > lockdown should have. > > If external attacks through PCIe are part of your threat model then > lockdown is the wrong direction. Leon's driver trust series is how we > are trying to solve this issue comprehensively in a useful way. > > Jason