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=1793518139; b=ZVX0KqF9/8rySpjRvapEpmmeN+ePYqYLVS89KoHvXgNda+EwUUyjY4tsgeCw6NP2wxCt pWvqgrmvoo3yYS9yqJsjptN1b25/bPPWcrAULR/QzioU9QwMufPlfXkrVQvIaWKMTL4Mc k3CGnDEn0Gb/QNN3UE0QwTFFFLnN2U6pJIfpj1Co4dJJLqGHYpiAvNgv4dGPZOFN6Y+ut rbLR1xkpMnq/VFkfxsI9D8FVOKsHahwY0NtXebIshtMNdmcinKEsGE4ZJmBxtXRtAEf9T cRXukCTt3XYMlCPzmufSNht1l8nmMZ93G0S/ol19igjY/LGoMh9pUoE6+bBX1VwEOrQ== ARC-Message-Signature: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; c=relaxed/relaxed; t=1793518139; h=Received:DKIM-Signature:X-Mailer:From:To:Cc:Subject:In-Reply-To: References:Date:Message-ID:MIME-Version:Content-Type; bh=ZEkTE9TgeFOprp8X/PqQu5dF595QKTp33C0W7LGmaFo=; b=fpN8Grzt2ZfCFC4oybytFQn7IyTR4wiFYtIKaRKvHhMA6QTjnGNE6aN61o1Epa/03Vvk lMq50r424h8TyQCd8hTiji0CiZ/kq1yH8ukMM5sdThDkQk/ILgIs9dguQyB33seKC8WMq AmBNAbDf3EGAj4Tlo27JPt0sYz90J+TqSkScCnsldRKZQx8uRyN5ejOHQbq8EREOQ6RrL Aa3Xe0EknjT8T2sH18ZVJXUSyo2hEUEGYyVmT5tPKoUKlyxQkD54Xg4691LhtyX74QjbQ i4UA3GhBFAFKqX6tp0j31ChqRulblFaJ1f136HPkKojwPWyuzQ2imAcAIgrei2tfprg== 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=K8xgiLEf; 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 D8EA6606A324; Sun, 1 Nov 2026 07:28:58 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key, unprotected) header.d=kernel.org header.i=@kernel.org header.a=rsa-sha256 header.s=k20260515 header.b=K8xgiLEf Received: by smtp.kernel.org (Postfix) with ESMTPSA id 734D41F000C1; Sun, 1 Nov 2026 07:28:51 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1793518138; bh=ZEkTE9TgeFOprp8X/PqQu5dF595QKTp33C0W7LGmaFo=; h=From:To:Cc:Subject:In-Reply-To:References:Date; b=K8xgiLEfZj/gyEaaz2uZMZEtINC0VjibSRUzhyA0UtxtVj7GBaePTRNRR5fWz/W9l xbPqnCcpLaOo0aRXU3RMM0zMN4VGVdy0m0nQmDoPizWSzBvPjHm9AkJpHifvSWqqVA 3G7EEBcuXL7pKNXHrvz/ow7TSqx7rpRsWv4eFNlxaGzKnkewO7xsNcu8elCrWHzfjo eLhX7+EkulhuuyBzXYopkKEo5/fGiKchuInGtrXvWKoy/dUVy+Lb++wylvey4+ztZE 9G8ROFw68lG4BH3Zg1puVCJ4P6+9BpO4NxGFID+CzmVs+ykCBaiKornyJlcnjea3/t x5PLjVjP9f2vg== X-Mailer: emacs 31.1 (via feedmail 11-beta-1 I) From: Aneesh Kumar K.V To: Ackerley Tng , linux-coco@lists.linux.dev Cc: Alex Williamson , David Woodhouse , David Hildenbrand , Jason Gunthorpe , "Joerg Roedel (AMD)" , Kevin Tian , Paolo Bonzini , Robin Murphy , Sean Christopherson , Will Deacon , Alexey Kardashevskiy , Xu Yilun , Catalin Marinas , Suzuki K Poulose , Steven Price , Fred Griffoul , iommu@lists.linux.dev, kvm@vger.kernel.org, patches@lists.linux.dev Subject: Re: [RFC PATCH v2 02/13] KVM: guest_memfd: Support private/shared conversion of device memory In-Reply-To: References: <20261017073149.1303738-1-aneesh.kumar@kernel.org> <20261017073149.1303738-3-aneesh.kumar@kernel.org> Date: Sun, 01 Nov 2026 12:58:48 +0530 Message-ID: X-Mailing-List: iommu@lists.linux.dev List-Id: List-Post: List-Help: List-Subscribe: List-Unsubscribe: List-Owner: Precedence: list MIME-Version: 1.0 Content-Type: text/plain Ackerley Tng writes: > Ackerley Tng writes: > >> "Aneesh Kumar K.V (Arm)" writes: >> >>> >>> [...snip...] >>> >>> +static int kvm_gmem_device_convert(struct inode *inode, pgoff_t start, >>> + pgoff_t nr_pages, const struct guest_memfd_device_request *req, >>> + const struct kvm_memory_slot *slot) >>> +{ >>> + struct gmem_inode *gi = GMEM_I(inode); >>> + u64 attrs = req ? KVM_MEMORY_ATTRIBUTE_PRIVATE : 0; >>> + bool prepared = false; >>> + int ret; >>> + >>> + MA_STATE(mas, &gi->attributes, start, start + nr_pages - 1); >>> + >>> + filemap_invalidate_lock(inode->i_mapping); >>> + if (__kvm_gmem_range_has_attributes(inode, start, nr_pages, attrs)) { >>> + ret = 0; >>> + goto out; >>> + } >>> + >>> + if (req && !__kvm_gmem_range_has_attributes(inode, start, nr_pages, 0)) { >>> + ret = -EEXIST; >>> + goto out; >>> + } >>> + >>> + ret = kvm_gmem_mas_preallocate(&mas, attrs, start, nr_pages); >>> + if (ret) >>> + goto out; >>> + >>> + kvm_gmem_invalidate_start(inode, start, start + nr_pages, KVM_FILTER_SHARED); >>> + if (req && READ_ONCE(gi->device.kvm->vm_dead)) >>> + ret = -EIO; >>> + else >>> + ret = gi->provider_ops->prepare_conversion(gi->provider, req); >>> + if (!ret) { >>> + prepared = true; >>> + if (req) >>> + ret = kvm_gmem_device_make_private(inode, slot, req, nr_pages); >> >> Is there a way to avoid making private from within the conversion flow? >> >> kvm_gmem_device_map() below already does it as part of vcpu_run. >> >> If we do the above, we'd be aligned with regular conversions, where >> to-shared conversions eagerly update arch state (make_shared) but >> to-private notifications to the firmware are done lazily. >> > > I was a bit confused above. IIUC to handle > RMI_EXIT_VDEV_VALIDATE_MAPPING, the conversion is not done from outside > the kernel, instead it is done in the vcpu_run completion hook. > All of that has changed in the latest series. kvm_complete_vdev_map_validate() now does the following. For the dependent changes in other subsystems, you may want to fetch the Git branch mentioned in the cover letter. static void kvm_complete_vdev_map_validate(struct kvm_vcpu *vcpu) { struct kvm *kvm = vcpu->kvm; struct realm_rec *rec = &vcpu->arch.rec; struct realm *realm = &kvm->arch.realm; struct kvm_memory_slot *slot; unsigned long start = rec->run->exit.dev_mem_base; unsigned long top = rec->run->exit.dev_mem_top; unsigned long rd = virt_to_phys(realm->rd); unsigned long rec_phys = virt_to_phys(rec->rec_page); int idx, ret; if (vcpu->run->cca_exit.response != KVM_ARM64_TIO_RESPONSE_APPROVE || top <= start || !PAGE_ALIGNED(start | top | rec->run->exit.dev_mem_pa) || !kvm_realm_is_private_address(realm, start) || !kvm_realm_is_private_address(realm, top - 1)) goto reject; idx = srcu_read_lock(&kvm->srcu); slot = gfn_to_memslot(kvm, gpa_to_gfn(start)); if (!slot || (slot->flags & KVM_MEMSLOT_INVALID) || !(slot->flags & KVM_MEMSLOT_GMEM_DEVICE) || gpa_to_gfn(top - 1) >= slot->base_gfn + slot->npages || slot->arch.device.vdev_id != rec->run->exit.vdev_id_1) goto out_reject; /* Userspace installs the mapping through guest_memfd before approval. */ if (!kvm_gmem_range_has_attributes(slot, gpa_to_gfn(start), gpa_to_gfn(top), KVM_MEMORY_ATTRIBUTE_PRIVATE)) goto out_reject; while (start < top) { unsigned long next, rmi_ret; ret = rmi_rtt_dev_validate(rd, rec_phys, start, top, &next, &rmi_ret); if (ret || rmi_ret || next <= start || next > top) goto out_reject; start = next; } srcu_read_unlock(&kvm->srcu, idx); return; out_reject: srcu_read_unlock(&kvm->srcu, idx); reject: rec->run->enter.flags |= REC_ENTER_FLAG_DEV_MEM_RESPONSE; } -aneesh