Skip to contentSkip to conversation

x86/tdx: Skip clearing reclaimed pages unless X86_BUG_TDX_PW_MCE is present

PATCH Superseded Adrian Hunter 2 participants 7 messages Jul 3, 2025

Subsystems (4)

v1 of 7 Superseded Latest is v7 →

lore
Share & tools
lore
#0 lore raw

cover x86/tdx: Skip clearing reclaimed pages unless X86_BUG_TDX_PW_MCE is present

Hi

Here are 2 small self-explanatory patches related to clearing TDX private pages.

Patch 1 is a minor tidy-up.

In patch 2, by skipping the clearing step, shutdown time can improve by up to 40%.

Adrian Hunter (2):
x86/tdx: Eliminate duplicate code in tdx_clear_page()
x86/tdx: Skip clearing reclaimed pages unless X86_BUG_TDX_PW_MCE is present

 arch/x86/include/asm/tdx.h  |  2 ++
 arch/x86/kvm/vmx/tdx.c      | 19 -------------------
 arch/x86/virt/vmx/tdx/tdx.c | 15 +++++++++++++++
 3 files changed, 17 insertions(+), 19 deletions(-)

Regards
Adrian

#1 lore raw

PATCH 1/2 x86/tdx: Eliminate duplicate code in tdx_clear_page()

tdx_clear_page() and reset_tdx_pages() duplicate the TDX page clearing logic. Keep the tdx_clear_page() prototype but call reset_tdx_pages() for the implementation.

---
 arch/x86/include/asm/tdx.h  |  2 ++
 arch/x86/kvm/vmx/tdx.c      | 19 -------------------
 arch/x86/virt/vmx/tdx/tdx.c |  6 ++++++
 3 files changed, 8 insertions(+), 19 deletions(-)
arch/x86/include/asm/tdx.h +2 -0
index 7ddef3a69866..e94eb0711bf1 100644
@@ -131,6 +131,8 @@ int tdx_guest_keyid_alloc(void);
 u32 tdx_get_nr_guest_keyids(void);
 void tdx_guest_keyid_free(unsigned int keyid);
 
+void tdx_clear_page(struct page *page);
+
 struct tdx_td {
 	/* TD root structure: */
 	struct page *tdr_page;
arch/x86/kvm/vmx/tdx.c +0 -19
index a08e7055d1db..098e4ee574bb 100644
@@ -276,25 +276,6 @@ static inline void tdx_disassociate_vp(struct kvm_vcpu *vcpu)
 	vcpu->cpu = -1;
 }
 
-static void tdx_clear_page(struct page *page)
-{
-	const void *zero_page = (const void *) page_to_virt(ZERO_PAGE(0));
-	void *dest = page_to_virt(page);
-	unsigned long i;
-
-	/*
-	 * The page could have been poisoned.  MOVDIR64B also clears
-	 * the poison bit so the kernel can safely use the page again.
-	 */
-	for (i = 0; i < PAGE_SIZE; i += 64)
-		movdir64b(dest + i, zero_page);
-	/*
-	 * MOVDIR64B store uses WC buffer.  Prevent following memory reads
-	 * from seeing potentially poisoned cache.
-	 */
-	__mb();
-}
-
 static void tdx_no_vcpus_enter_start(struct kvm *kvm)
 {
 	struct kvm_tdx *kvm_tdx = to_kvm_tdx(kvm);
arch/x86/virt/vmx/tdx/tdx.c +6 -0
index c7a9a087ccaf..17a7a599facd 100644
@@ -654,6 +654,12 @@ static void reset_tdx_pages(unsigned long base, unsigned long size)
 	mb();
 }
 
+void tdx_clear_page(struct page *page)
+{
+	reset_tdx_pages(page_to_phys(page), PAGE_SIZE);
+}
+EXPORT_SYMBOL_GPL(tdx_clear_page);
+
 static void tdmr_reset_pamt(struct tdmr_info *tdmr)
 {
 	tdmr_do_pamt_func(tdmr, reset_tdx_pages);
Signature · 1 line
2.48.1
#2 lore raw

Re: [PATCH 1/2] x86/tdx: Eliminate duplicate code in tdx_clear_page()

Quoted 3 lines · Adrian Hunter

tdx_clear_page() and reset_tdx_pages() duplicate the TDX page clearing logic. Keep the tdx_clear_page() prototype but call reset_tdx_pages() for the implementation.

Why keep the old prototype?

#3 lore raw

Re: [PATCH 1/2] x86/tdx: Eliminate duplicate code in tdx_clear_page()

Quoted 6 lines · Dave Hansen
Quoted 3 lines · Adrian Hunter

tdx_clear_page() and reset_tdx_pages() duplicate the TDX page clearing logic. Keep the tdx_clear_page() prototype but call reset_tdx_pages() for the implementation.

Why keep the old prototype?

What prototype would you prefer?

#4 lore raw

Re: [PATCH 1/2] x86/tdx: Eliminate duplicate code in tdx_clear_page()

Quoted 7 lines · Adrian Hunter
Quoted 5 lines · Dave Hansen
Quoted 3 lines · Adrian Hunter

tdx_clear_page() and reset_tdx_pages() duplicate the TDX page clearing logic. Keep the tdx_clear_page() prototype but call reset_tdx_pages() for the implementation.

Why keep the old prototype?

What prototype would you prefer?

My first reaction is that there should be _one_ function that does the erratum-related clearing. I'd rather have the KVM sites use the slightly cumbersome paddr/size arguments than have two functions.

Worst, case, have two functions that have similar names to make it clear that they are doing the same thing.

It is *NOT* clear that:

tdx_clear_page()
and
reset_tdx_pages()

are doing the exact same thing.

I think I'd probably rename reset_tdx_pages() to tdx_reset_paddr() and then use it in KVM like this:

	tdx_quirk_reset_paddr(page_to_phys(page), PAGE_SIZE);

The alternative would be to retain a function that keeps the 'struct page' as an argument. Something like:

tdx_quirk_reset_paddr(unsigned long base, unsigned long size)
and
tdx_quirk_reset_page(struct page *page)

But the way proposed in this patch is very much a half measure.

#5 lore raw

PATCH 2/2 x86/tdx: Skip clearing reclaimed pages unless X86_BUG_TDX_PW_MCE is present

In reply to Adrian Hunter

Avoid clearing reclaimed TDX private pages unless the platform is affected by the X86_BUG_TDX_PW_MCE erratum. This significantly reduces VM shutdown time on unaffected systems.

Background

KVM currently clears reclaimed TDX private pages using MOVDIR64B, which:

- Clears the TD Owner bit (which identifies TDX private memory) and integrity metadata without triggering integrity violations.
- Clears poison from cache lines without consuming it, avoiding MCEs on access (refer TDX Module Base spec. 16.5. Handling Machine Check Events during Guest TD Operation).

The TDX module also uses MOVDIR64B to initialize private pages before use. If cache flushing is needed, it sets TDX_FEATURES.CLFLUSH_BEFORE_ALLOC. However, KVM currently flushes unconditionally, refer commit 94c477a751c7b ("x86/virt/tdx: Add SEAMCALL wrappers to add TD private pages")

In contrast, when private pages are reclaimed, the TDX Module handles flushing via the TDH.PHYMEM.CACHE.WB SEAMCALL.

Problem

Clearing all private pages during VM shutdown is costly. For guests with a large amount of memory it can take minutes.

Solution

TDX Module Base Architecture spec. documents that private pages reclaimed from a TD should be initialized using MOVDIR64B, in order to avoid integrity violation or TD bit mismatch detection when later being read using a shared HKID, refer April 2025 spec. "Page Initialization" in section "8.6.2. Platforms not Using ACT: Required Cache Flush and Initialization by the Host VMM"

That is an overstatement and will be clarified in coming versions of the spec. In fact, as outlined in "Table 16.2: Non-ACT Platforms Checks on Memory" and "Table 16.3: Non-ACT Platforms Checks on Memory Reads in Li Mode" in the same spec, there is no issue accessing such reclaimed pages using a shared key that does not have integrity enabled. Linux always uses KeyID 0 which never has integrity enabled. KeyID 0 is also the TME KeyID which disallows integrity, refer "TME Policy/Encryption Algorithm" bit description in "Intel Architecture Memory Encryption Technologies" spec version 1.6 April 2025. So there is no need to clear pages to avoid integrity violations.

There remains a risk of poison consumption. However, in the context of TDX, it is expected that there would be a machine check associated with the original poisoning. On some platforms that results in a panic. However platforms may support "SEAM_NR" Machine Check capability, in which case Linux machine check handler marks the page as poisoned, which prevents it from being allocated anymore, refer commit 7911f145de5fe ("x86/mce:
Implement recovery for errors in TDX/SEAM non-root mode")

Improvement

By skipping the clearing step on unaffected platforms, shutdown time can improve by up to 40%.

On platforms with the X86_BUG_TDX_PW_MCE erratum (SPR and EMR), continue clearing because these platforms may trigger poison on partial writes to previously-private pages, even with KeyID 0, refer commit 1e536e1068970 ("x86/cpu: Detect TDX partial write machine check erratum")

---
 arch/x86/virt/vmx/tdx/tdx.c | 9 +++++++++
 1 file changed, 9 insertions(+)
arch/x86/virt/vmx/tdx/tdx.c +9 -0
index 17a7a599facd..a030679c9239 100644
@@ -642,6 +642,15 @@ static void reset_tdx_pages(unsigned long base, unsigned long size)
 	const void *zero_page = (const void *)page_address(ZERO_PAGE(0));
 	unsigned long phys, end;
 
+	/*
+	 * Linux uses only KeyID 0 which can read or write pages
+	 * that were formerly TDX private pages, without poisoning
+	 * memory. However on platforms with the partial-write errata,
+	 * poisoning still happens for partial writes.
+	 */
+	if (!boot_cpu_has_bug(X86_BUG_TDX_PW_MCE))
+		return;
+
 	end = base + size;
 	for (phys = base; phys < end; phys += 64)
 		movdir64b(__va(phys), zero_page);
Signature · 1 line
2.48.1
#6 lore raw

arch/x86/virt/vmx/tdx/tdx.c static void reset_tdx_pages(unsigned long base, unsigned long size) · line 643

Quoted 3 lines · Adrian Hunter
arch/x86/virt/vmx/tdx/tdx.c +0 -0
@@ -642,6 +642,15 @@ static void reset_tdx_pages(unsigned long base, unsigned long size)
 	const void *zero_page = (const void *)page_address(ZERO_PAGE(0));
 	unsigned long phys, end;

If the only possible use for reset_tdx_pages() is for the erratum, then it needs to be in the function name.

arch/x86/virt/vmx/tdx/tdx.c static void reset_tdx_pages(unsigned long base, unsigned long size) · line 652

Quoted 8 lines
arch/x86/virt/vmx/tdx/tdx.c +8 -0
+	/*
+	 * Linux uses only KeyID 0 which can read or write pages
+	 * that were formerly TDX private pages, without poisoning
+	 * memory. However on platforms with the partial-write errata,
+	 * poisoning still happens for partial writes.
+	 */
+	if (!boot_cpu_has_bug(X86_BUG_TDX_PW_MCE))
+		return;

Would this be more clear?

/*
* Typically, any write to the page will convert it from TDX
* private back to normal kernel memory. Systems with the
* erratum need to do the conversion explicitly. */