From: sashiko-bot@kernel.org
To: "Paolo Bonzini" <pbonzini@redhat.com>
Cc: kvm@vger.kernel.org
Subject: Re: [PATCH 26/31] KVM: x86/mmu: Issue memory fault exit if walk failed due to memory attribute
Date: Fri, 18 Sep 2026 14:46:59 +0000 [thread overview]
Message-ID: <20260918144700.4ADE51F00898@smtp.kernel.org> (raw)
In-Reply-To: <20260918135030.171564-27-pbonzini@redhat.com>
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] kvm: x86/mmu: lost memory attribute fault on GVA to GPA translation during emulation
--
commit a5b7ea3b514c528d86717bda8577133bab95a9d8
Author: Nicolas Saenz Julienne <nsaenz@amazon.com>
KVM: x86/mmu: Issue memory fault exit if walk failed due to memory attribute
This commit modifies the guest page table walk to issue a user-space
memory fault exit if it fails due to memory attributes, instead of
injecting a page fault. It introduces the memory_attributes_fault
flag to the guest_walker structure.
> diff --git a/arch/x86/kvm/mmu/paging_tmpl.h b/arch/x86/kvm/mmu/paging_tmpl.h
> index b031123fcb369..cd055bfb1eb6b 100644
> --- a/arch/x86/kvm/mmu/paging_tmpl.h
> +++ b/arch/x86/kvm/mmu/paging_tmpl.h
[ ... ]
> @@ -820,6 +831,12 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, struct kvm_page_fault *fault
> * The page is not mapped by the guest. Let the guest handle it.
> */
> if (!r) {
> + if (walker.memory_attributes_fault) {
> + kvm_prepare_memory_fault_exit(vcpu, walker.gfn << PAGE_SHIFT,
> + PAGE_SIZE, false, false, false);
> + return -EFAULT;
> + }
> +
> if (!fault->prefetch)
> __kvm_inject_emulated_page_fault(vcpu, &walker.fault, true);
[Severity: High]
Does FNAME(gva_to_gpa)() also need to handle the new memory_attributes_fault
flag?
FNAME(gva_to_gpa)() uses a local walker to translate guest virtual addresses
by calling FNAME(walk_addr_generic)():
static gpa_t FNAME(gva_to_gpa)(struct kvm_vcpu *vcpu, struct kvm_pagewalk *w,
gpa_t addr, u64 access,
struct x86_exception *exception)
{
struct guest_walker walker;
gpa_t gpa = INVALID_GPA;
int r;
[ ... ]
r = FNAME(walk_addr_generic)(&walker, vcpu, w, addr, access);
if (r) {
gpa = gfn_to_gpa(walker.gfn);
gpa |= addr & ~PAGE_MASK;
} else if (exception)
*exception = walker.fault;
return gpa;
}
If FNAME(walk_addr_generic)() detects a memory attribute fault when reading
guest page tables, it returns 0 with walker.memory_attributes_fault set.
However, FNAME(gva_to_gpa)() appears to ignore the new flag and only copies
walker.fault to the output exception.
Can this result in the instruction emulator improperly injecting a spurious
page fault into the guest instead of exiting to userspace with
KVM_EXIT_MEMORY_FAULT when guest page tables reside in memory restricted by
KVM memory attributes?
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260918135030.171564-2-pbonzini@redhat.com?part=26
next prev parent reply other threads:[~2026-09-18 14:47 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-09-18 13:49 [PATCH v3 00/28] KVM: x86: Introduce memory protection attributes Paolo Bonzini
2026-09-18 13:50 ` [PATCH 01/31] KVM: x86/hyperv: do not overwrite hc->ingpa for slow SIGNAL_EVENT hypercall Paolo Bonzini
2026-09-18 13:50 ` [PATCH 02/31] KVM: selftests: Take into account mixed memory fault flags Paolo Bonzini
2026-09-18 13:50 ` [PATCH 03/31] KVM: Define and communicate KVM_EXIT_MEMORY_FAULT RWX flags to userspace Paolo Bonzini
2026-09-18 13:50 ` [PATCH 04/31] KVM: selftests: Test address translation for Hyper-V direct L2 hypercalls Paolo Bonzini
2026-09-18 13:50 ` [PATCH 05/31] KVM: apply nGPA->GPA translation to KVM_HC_CLOCK_PAIRING Paolo Bonzini
2026-09-18 13:50 ` [PATCH 06/31] KVM: x86: Introduce memory fault on invalid hypercalls reads/writes Paolo Bonzini
2026-09-18 14:11 ` sashiko-bot
2026-09-21 16:56 ` Vitaly Kuznetsov
2026-09-18 13:50 ` [PATCH 07/31] KVM: selftests: test hypercall memory fault exits Paolo Bonzini
2026-09-18 13:50 ` [PATCH 08/31] KVM: x86/mmu: intersect writability from __kvm_faultin_pfn with fault->map_writable Paolo Bonzini
2026-09-18 13:50 ` [PATCH 09/31] KVM: x86/mmu: Extend map_writable to a full ACC_* mask Paolo Bonzini
2026-09-18 13:50 ` [PATCH 10/31] KVM: x86/mmu: Init memslot hugepage information for non-private_mem VMs too Paolo Bonzini
2026-09-18 13:50 ` [PATCH 11/31] KVM: pass kvm == NULL case to kvm_arch_has_private_mem Paolo Bonzini
2026-09-18 13:50 ` [PATCH 12/31] KVM: adjust for presence of more than one attribute Paolo Bonzini
2026-09-18 13:50 ` [PATCH 13/31] KVM: Introduce NR/NW/NX memory attributes Paolo Bonzini
2026-09-18 13:50 ` [PATCH 14/31] KVM: Include memory protections in result of gfn->hva conversion Paolo Bonzini
2026-09-18 13:50 ` [PATCH 15/31] KVM: Introduce kvm_fetch_guest_page() and use it for x86 Paolo Bonzini
2026-09-18 13:50 ` [PATCH 16/31] KVM: Take memory protections into account for memory read/write/fetch Paolo Bonzini
2026-09-18 13:50 ` [PATCH 17/31] KVM: Take memory protections into account for __kvm_vcpu_map Paolo Bonzini
2026-09-18 13:50 ` [PATCH 18/31] KVM: Encapsulate memattrs array into anonymous struct Paolo Bonzini
2026-09-18 13:50 ` [PATCH 19/31] KVM: loongarch: do full validity check on the gfn-to-hva cache Paolo Bonzini
2026-09-18 13:50 ` [PATCH 20/31] KVM: Introduce kvm_check_gen()/kvm_memslots_check_gen() Paolo Bonzini
2026-09-18 14:26 ` sashiko-bot
2026-09-18 13:50 ` [PATCH 21/31] KVM: Introduce a generation number for memory attributes Paolo Bonzini
2026-09-18 14:42 ` sashiko-bot
2026-09-18 13:50 ` [PATCH 22/31] KVM: Take memory protections into account for accesses with cached gfn->hva Paolo Bonzini
2026-09-18 14:42 ` sashiko-bot
2026-09-18 13:50 ` [PATCH 23/31] KVM: pfncache: Fail to refresh if it contains memory protections Paolo Bonzini
2026-09-18 13:50 ` [PATCH 24/31] KVM: x86/mmu: Do not prefetch sptes on gfns backed by memory attributes Paolo Bonzini
2026-09-18 13:50 ` [PATCH 25/31] KVM: x86/mmu: Take memory protection attributes into account during faults Paolo Bonzini
2026-09-18 14:57 ` sashiko-bot
2026-09-18 13:50 ` [PATCH 26/31] KVM: x86/mmu: Issue memory fault exit if walk failed due to memory attribute Paolo Bonzini
2026-09-18 14:46 ` sashiko-bot [this message]
2026-09-18 13:50 ` [PATCH 27/31] KVM: let kvm_arch_post_set_memory_attributes drop mmu_lock Paolo Bonzini
2026-09-18 13:50 ` [PATCH 28/31] KVM: x86/mmu: Obsolete all roots if memattr contains gPTEs Paolo Bonzini
2026-09-18 14:57 ` sashiko-bot
2026-09-18 13:50 ` [PATCH 29/31] KVM: x86: selftests: Introduce memory protection attributes test Paolo Bonzini
2026-09-18 13:50 ` [PATCH 30/31] KVM: x86: selftests: Introduce memory attributes PTE test Paolo Bonzini
2026-09-18 13:50 ` [PATCH 31/31] KVM: x86: selftests: Introduce memory attributes side-channel tests Paolo Bonzini
2026-09-18 14:56 ` sashiko-bot
2026-09-18 15:20 ` [PATCH v3 00/28] KVM: x86: Introduce memory protection attributes Paolo Bonzini
2026-09-21 16:56 ` Vitaly Kuznetsov
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=20260918144700.4ADE51F00898@smtp.kernel.org \
--to=sashiko-bot@kernel.org \
--cc=kvm@vger.kernel.org \
--cc=pbonzini@redhat.com \
--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