From: sashiko-bot@kernel.org
To: "Ackerley Tng via B4 Relay" <devnull+ackerleytng.google.com@kernel.org>
Cc: kvm@vger.kernel.org
Subject: Re: [PATCH v7 21/42] KVM: TDX: Make source page optional for KVM_TDX_INIT_MEM_REGION
Date: Sat, 23 May 2026 01:07:52 +0000 [thread overview]
Message-ID: <20260523010752.CA7AF1F000E9@smtp.kernel.org> (raw)
In-Reply-To: <20260522-gmem-inplace-conversion-v7-21-2f0fae496530@google.com>
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Critical] Uninitialized host kernel memory is disclosed to TDX guests when performing in-place population.
--
commit 38e76d2ce3c699d4b64b580ab16a7b4cbc2b189a
Author: Ackerley Tng <ackerleytng@google.com>
KVM: TDX: Make source page optional for KVM_TDX_INIT_MEM_REGION
Update tdx_gmem_post_populate() to handle cases where a source page is not
explicitly provided. Instead of returning -EOPNOTSUPP when src_page is NULL,
default to using the page associated with the destination PFN, allowing
in-place memory conversion.
> diff --git a/arch/x86/kvm/vmx/tdx.c b/arch/x86/kvm/vmx/tdx.c
> --- a/arch/x86/kvm/vmx/tdx.c
> +++ b/arch/x86/kvm/vmx/tdx.c
[ ... ]
> @@ -3116,8 +3116,12 @@ static int tdx_gmem_post_populate(struct kvm *kvm, gfn_t gfn, kvm_pfn_t pfn,
> if (KVM_BUG_ON(kvm_tdx->page_add_src, kvm))
> return -EIO;
>
> - if (!src_page)
> - return -EOPNOTSUPP;
> + if (!src_page) {
> + if (vm_memory_attributes)
> + return -EOPNOTSUPP;
> +
> + src_page = pfn_to_page(pfn);
> + }
[Severity: Critical]
Can this sequence lead to uninitialized host memory being exposed to the guest
if the guest_memfd folio has not yet been populated?
When KVM_TDX_INIT_MEM_REGION is called with a NULL source_addr,
__kvm_gmem_populate() attempts an in-place conversion. If the target
guest_memfd folio is unpopulated, __kvm_gmem_get_pfn() allocates a new folio
using GFP_HIGHUSER, which lacks __GFP_ZERO.
Because __kvm_gmem_populate() bypasses the normal kvm_gmem_get_pfn() zeroing
path (where clear_highpage() is called for !folio_test_uptodate()), the
allocated page contains uninitialized host memory.
Since src_page is set to the same physical page here, will the TDX module
read this uninitialized host memory using the host KeyID and encrypt it into
the guest's private memory space during the TDH.MEM.PAGE.ADD SEAMCALL?
Could a malicious unprivileged VMM exploit this regression by passing a NULL
source_addr on an unmapped guest_memfd region, causing arbitrary host kernel
memory to be disclosed to the TDX guest?
>
> kvm_tdx->page_add_src = src_page;
> ret = kvm_tdp_mmu_map_private_pfn(arg->vcpu, gfn, pfn);
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260522-gmem-inplace-conversion-v7-0-2f0fae496530@google.com?part=21
next prev parent reply other threads:[~2026-05-23 1:07 UTC|newest]
Thread overview: 58+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-05-23 0:17 [PATCH v7 00/42] guest_memfd: In-place conversion support Ackerley Tng via B4 Relay
2026-05-23 0:17 ` [PATCH v7 01/42] KVM: guest_memfd: Introduce per-gmem attributes, use to guard user mappings Ackerley Tng via B4 Relay
2026-05-23 0:17 ` [PATCH v7 02/42] KVM: Rename KVM_GENERIC_MEMORY_ATTRIBUTES to KVM_VM_MEMORY_ATTRIBUTES Ackerley Tng via B4 Relay
2026-05-23 0:17 ` [PATCH v7 03/42] KVM: Enumerate support for PRIVATE memory iff kvm_arch_has_private_mem is defined Ackerley Tng via B4 Relay
2026-05-23 0:17 ` [PATCH v7 04/42] KVM: Stub in ability to disable per-VM memory attribute tracking Ackerley Tng via B4 Relay
2026-05-23 0:17 ` [PATCH v7 05/42] KVM: guest_memfd: Wire up kvm_get_memory_attributes() to per-gmem attributes Ackerley Tng via B4 Relay
2026-05-23 0:17 ` [PATCH v7 06/42] KVM: guest_memfd: Update kvm_gmem_populate() to use gmem attributes Ackerley Tng via B4 Relay
2026-05-23 0:59 ` sashiko-bot
2026-05-23 0:17 ` [PATCH v7 07/42] KVM: guest_memfd: Only prepare folios for private pages Ackerley Tng via B4 Relay
2026-05-23 0:52 ` sashiko-bot
2026-05-27 21:22 ` Ackerley Tng
2026-05-23 0:17 ` [PATCH v7 08/42] KVM: Move kvm_supported_mem_attributes() to kvm_host.h Ackerley Tng via B4 Relay
2026-05-23 0:17 ` [PATCH v7 09/42] KVM: guest_memfd: Add base support for KVM_SET_MEMORY_ATTRIBUTES2 Ackerley Tng via B4 Relay
2026-05-23 1:01 ` sashiko-bot
2026-05-27 21:27 ` Ackerley Tng
2026-05-23 0:17 ` [PATCH v7 10/42] KVM: guest_memfd: Ensure pages are not in use before conversion Ackerley Tng via B4 Relay
2026-05-23 0:55 ` sashiko-bot
2026-05-23 0:17 ` [PATCH v7 11/42] KVM: guest_memfd: Call arch invalidate hooks on conversion Ackerley Tng via B4 Relay
2026-05-23 0:17 ` [PATCH v7 12/42] KVM: guest_memfd: Return early if range already has requested attributes Ackerley Tng via B4 Relay
2026-05-23 0:17 ` [PATCH v7 13/42] KVM: guest_memfd: Advertise KVM_SET_MEMORY_ATTRIBUTES2 ioctl Ackerley Tng via B4 Relay
2026-05-23 0:17 ` [PATCH v7 14/42] KVM: guest_memfd: Handle lru_add fbatch refcounts during conversion safety check Ackerley Tng via B4 Relay
2026-05-23 0:17 ` [PATCH v7 15/42] KVM: guest_memfd: Use actual size for invalidation in kvm_gmem_release() Ackerley Tng via B4 Relay
2026-05-23 0:17 ` [PATCH v7 16/42] KVM: guest_memfd: Determine invalidation filter from memory attributes Ackerley Tng via B4 Relay
2026-05-23 1:06 ` sashiko-bot
[not found] ` <CAEvNRgH21BoKT1mOQzgmKHKpDi4xbwtbMuenGv5U1ZUSENrJmg@mail.gmail.com>
2026-05-29 20:44 ` Sean Christopherson
2026-05-23 0:17 ` [PATCH v7 17/42] KVM: Move KVM_VM_MEMORY_ATTRIBUTES config definition to x86 Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 18/42] KVM: Let userspace disable per-VM mem attributes, enable per-gmem attributes Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 19/42] KVM: guest_memfd: Enable INIT_SHARED on guest_memfd for x86 Coco VMs Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 20/42] KVM: SEV: Make 'uaddr' parameter optional for KVM_SEV_SNP_LAUNCH_UPDATE Ackerley Tng via B4 Relay
2026-05-23 0:55 ` sashiko-bot
2026-05-27 23:31 ` Ackerley Tng
2026-05-23 0:18 ` [PATCH v7 21/42] KVM: TDX: Make source page optional for KVM_TDX_INIT_MEM_REGION Ackerley Tng via B4 Relay
2026-05-23 1:07 ` sashiko-bot [this message]
2026-05-23 0:18 ` [PATCH v7 22/42] KVM: selftests: Create gmem fd before "regular" fd when adding memslot Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 23/42] KVM: selftests: Rename guest_memfd{,_offset} to gmem_{fd,offset} Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 24/42] KVM: selftests: Add support for mmap() on guest_memfd in core library Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 25/42] KVM: selftests: Add selftests global for guest memory attributes capability Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 26/42] KVM: selftests: Add helpers for calling ioctls on guest_memfd Ackerley Tng via B4 Relay
2026-05-23 0:42 ` sashiko-bot
2026-05-23 0:18 ` [PATCH v7 27/42] KVM: selftests: Test basic single-page conversion flow Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 28/42] KVM: selftests: Test conversion flow when INIT_SHARED Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 29/42] KVM: selftests: Test conversion precision in guest_memfd Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 30/42] KVM: selftests: Test conversion before allocation Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 31/42] KVM: selftests: Convert with allocated folios in different layouts Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 32/42] KVM: selftests: Test that truncation does not change shared/private status Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 33/42] KVM: selftests: Test that shared/private status is consistent across processes Ackerley Tng via B4 Relay
2026-05-23 1:11 ` sashiko-bot
2026-05-23 0:18 ` [PATCH v7 34/42] KVM: selftests: Test conversion with elevated page refcount Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 35/42] KVM: selftests: Reset shared memory after hole-punching Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 36/42] KVM: selftests: Provide function to look up guest_memfd details from gpa Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 37/42] KVM: selftests: Provide common function to set memory attributes Ackerley Tng via B4 Relay
2026-05-23 1:35 ` sashiko-bot
2026-05-23 0:18 ` [PATCH v7 38/42] KVM: selftests: Check fd/flags provided to mmap() when setting up memslot Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 39/42] KVM: selftests: Make TEST_EXPECT_SIGBUS thread-safe Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 40/42] KVM: selftests: Update private_mem_conversions_test to mmap() guest_memfd Ackerley Tng via B4 Relay
2026-05-23 0:18 ` [PATCH v7 41/42] KVM: selftests: Add script to exercise private_mem_conversions_test Ackerley Tng via B4 Relay
2026-05-23 1:15 ` sashiko-bot
2026-05-23 0:18 ` [PATCH v7 42/42] KVM: selftests: Update private memory exits test to work with per-gmem attributes Ackerley Tng via B4 Relay
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260523010752.CA7AF1F000E9@smtp.kernel.org \
--to=sashiko-bot@kernel.org \
--cc=devnull+ackerleytng.google.com@kernel.org \
--cc=kvm@vger.kernel.org \
--cc=sashiko-reviews@lists.linux.dev \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox