[PATCH v10 00/11] Dynamic PAMT

[PATCH v10 00/11] Dynamic PAMT

[Rick Edgecombe]

Hi,

This is another revision of the Dynamic PAMT TDX series. The biggest 
change from the last version resulted from a request from Dave to more 
clearly comment and name things *DPAMT* in the DPAMT specific parts of the 
code, rather than PAMT. Discretion was used, and I left some of the PAMT 
names (the API names previously agreed with KVM were not updated, and the 
names associated with the TDX ABI itself were left matching the spec, 
etc). Comments were also strengthened around the earlier get/put patches 
to clarify their temporary limitations. There were no functional changes. 
To help if anyone wants to scrutinize the diff, a rebased branch of v9 is 
pushed here [0].

Our CI AI system has gained some extra checks, which I thought were better
to apply. So that was done too. The changes were all trivial.

Lastly, based on offlist discussion with Dave and Kiryl, I author flipped 
the arch/x86 patches to me and updated them to have a single Intel 
signoff. I left the KVM patches as they were, based on my best expectation 
of Sean's preference.


We have a ton of RBs at this point (8 different people, thanks!), which I 
left in place due to the trivial depth of the changes. All KVM patches 
have Sean's ack. The following escaped Dave's RB last time:
[PATCH 03/11] x86/virt/tdx: Add __tdx_pamt_get/put() helpers
[PATCH 04/11] x86/virt/tdx: Allocate refcounts for Dynamic PAMT memory
[PATCH 09/11] x86/virt/tdx: Enable Dynamic PAMT
[PATCH 10/11] Documentation/x86: Add documentation for TDX's Dynamic PAMT

This should be in pretty good shape, Dave please strongly consider 
applying.

Background
==========
Dynamic PAMT is a TDX feature that allows saving memory by allocating some 
of its page tracking metadata dynamically, instead of statically at boot. 
These static allocations take roughly 0.4% of system memory. The savings 
are variable depending on system and TDX usage, but could be up to 100x. 
For more Dynamic PAMT background, please refer to [1]. For more analysis 
of the savings in different scenarios, see the v6 coverleter[2].

It occurred to me that since the Dynamic PAMT effort began, RAM has become
much more expensive. Consequently, this feature is even more valuable now.
It would be good to enable it for TDX users.

Base
====
This is based on v7.3-rc1. A branch of this series can be found here: [3].

Testing
=======
Since there were not any changes which would be expected to have 
functional impact, I just did some regression testing with our standard 
automated testing for this one. I left Tested-by tags per Hongyu's
preference.


[0] https://github.com/intel-staging/tdx/tree/dpamt_v9_on_v7.3-rc1
[1] https://lore.kernel.org/lkml/20250918232224.2202592-1-rick.p.edgecombe@intel.com/
[2] https://lore.kernel.org/lkml/20260526023515.288829-1-rick.p.edgecombe@intel.com/
[3] https://github.com/intel-staging/tdx/tree/dpamt_v10


Kiryl Shutsemau (2):
  KVM: TDX: Allocate PAMT memory for TD and vCPU control structures
  KVM: TDX: Get/put PAMT pages when (un)mapping private memory

Rick Edgecombe (9):
  x86/virt/tdx: Simplify PAMT layout calculation
  x86/virt/tdx: Allocate page bitmap for Dynamic PAMT
  x86/virt/tdx: Add __tdx_pamt_get/put() helpers
  x86/virt/tdx: Allocate refcounts for Dynamic PAMT memory
  x86/virt/tdx: Handle multiple callers in tdx_pamt_get/put()
  x86/virt/tdx: Add APIs to support Dynamic PAMT ops from KVM's fault
    path
  x86/virt/tdx: Enable Dynamic PAMT
  Documentation/x86: Add documentation for TDX's Dynamic PAMT
  x86/virt/tdx: Optimize tdx_pamt_get/put()

 .../admin-guide/kernel-parameters.txt         |  10 +
 Documentation/arch/x86/tdx.rst                |  27 +
 arch/x86/include/asm/kvm-x86-ops.h            |   1 +
 arch/x86/include/asm/kvm_host.h               |   1 +
 arch/x86/include/asm/tdx.h                    |  23 +
 arch/x86/include/asm/tdx_global_metadata.h    |   9 +-
 arch/x86/kvm/mmu/mmu.c                        |   4 +
 arch/x86/kvm/vmx/tdx.c                        |  99 ++--
 arch/x86/kvm/vmx/tdx.h                        |   2 +
 arch/x86/virt/vmx/tdx/tdx.c                   | 460 +++++++++++++++---
 arch/x86/virt/vmx/tdx/tdx.h                   |   2 +
 arch/x86/virt/vmx/tdx/tdx_global_metadata.c   |  23 +-
 12 files changed, 566 insertions(+), 95 deletions(-)

-- 
2.55.0

[PATCH v10 01/11] x86/virt/tdx: Simplify PAMT layout calculation

[Rick Edgecombe]

For each memory region that the TDX module might use (called TDMR), three
separate traditional PAMT allocations are needed. There is one for each
supported page size (1GB, 2MB, 4KB). These store information on each page
in the TDMR. In Linux, they are allocated out of one physically contiguous
block, in order to more efficiently use some internal TDX module
bookkeeping resources. So some simple math is needed to break the single
large allocation into three smaller allocations for each page size.

There are some commonalities in the math needed to calculate the base and
size for each smaller allocation, and so an effort was made to share logic
across the three. Unfortunately doing this turned out unnaturally tortured,
with a loop iterating over the three page sizes, only to call into a
function with case statements for each page size. In the future Dynamic
PAMT will add more logic that is special to the 4KB page size, making the
benefit of the math sharing even more questionable.

Three is not a very high number, so get rid of the loop and just duplicate
the small calculation three times. In doing so, setup for future Dynamic
PAMT changes.

Since the loop that iterates over it is gone, further simplify the code by
dropping the array of intermediate size and base storage. Just store the
values to their final locations.

AI was used under supervision to collect/apply feedback, review code and
workshop logs.

Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Reviewed-by: Binbin Wu <binbin.wu@linux.intel.com>
Reviewed-by: Kiryl Shutsemau (Meta) <kas@kernel.org>
Reviewed-by: Chao Gao <chao.gao@intel.com>
Reviewed-by: Yan Zhao <yan.y.zhao@intel.com>
Reviewed-by: Tony Lindgren <tony.lindgren@linux.intel.com>
Reviewed-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Vishal Annapurve <vannapurve@google.com>
Acked-by: Sohil Mehta <sohil.mehta@intel.com>
---
v10:
 - Fix whitespace change in tdmrs_set_up_pamt_all() (AI nit checker)

v8:
 - Fix PAMT capitalization in comment (Sohil)
 - Delete stale info in log, that Chao mentioned last time (Binbin)

v7:
 - Remove accidental whitespace changes (Kiryl)
 - Drop stale sentence in log (Chao)
 - Better patch subject (Yan)
 - Drop Assisted-by tag and cover AI use in log (Dave)
---
 arch/x86/virt/vmx/tdx/tdx.c | 90 ++++++++++++-------------------------
 1 file changed, 28 insertions(+), 62 deletions(-)

diff --git a/arch/x86/virt/vmx/tdx/tdx.c b/arch/x86/virt/vmx/tdx/tdx.c
index 1b9ff749dd8e0..51b191bf12258 100644
--- a/arch/x86/virt/vmx/tdx/tdx.c
+++ b/arch/x86/virt/vmx/tdx/tdx.c
@@ -514,31 +514,21 @@ static __init int fill_out_tdmrs(struct list_head *tmb_list,
  * Calculate PAMT size given a TDMR and a page size.  The returned
  * PAMT size is always aligned up to 4K page boundary.
  */
-static __init unsigned long tdmr_get_pamt_sz(struct tdmr_info *tdmr, int pgsz,
-					     u16 pamt_entry_size)
+static __init unsigned long tdmr_get_pamt_sz(struct tdmr_info *tdmr, int pgsz)
 {
 	unsigned long pamt_sz, nr_pamt_entries;
+	const int tdx_pg_size_shift[TDX_PS_NR] = { PAGE_SHIFT, PMD_SHIFT, PUD_SHIFT };
+	const u16 pamt_entry_size[TDX_PS_NR] = {
+		tdx_sysinfo.tdmr.pamt_4k_entry_size,
+		tdx_sysinfo.tdmr.pamt_2m_entry_size,
+		tdx_sysinfo.tdmr.pamt_1g_entry_size,
+	};
 
-	switch (pgsz) {
-	case TDX_PS_4K:
-		nr_pamt_entries = tdmr->size >> PAGE_SHIFT;
-		break;
-	case TDX_PS_2M:
-		nr_pamt_entries = tdmr->size >> PMD_SHIFT;
-		break;
-	case TDX_PS_1G:
-		nr_pamt_entries = tdmr->size >> PUD_SHIFT;
-		break;
-	default:
-		WARN_ON_ONCE(1);
-		return 0;
-	}
+	nr_pamt_entries = tdmr->size >> tdx_pg_size_shift[pgsz];
+	pamt_sz = nr_pamt_entries * pamt_entry_size[pgsz];
 
-	pamt_sz = nr_pamt_entries * pamt_entry_size;
 	/* TDX requires PAMT size must be 4K aligned */
-	pamt_sz = ALIGN(pamt_sz, PAGE_SIZE);
-
-	return pamt_sz;
+	return PAGE_ALIGN(pamt_sz);
 }
 
 /*
@@ -576,15 +566,11 @@ static __init int tdmr_get_nid(struct tdmr_info *tdmr, struct list_head *tmb_lis
  * within @tdmr, and set up PAMTs for @tdmr.
  */
 static __init int tdmr_set_up_pamt(struct tdmr_info *tdmr,
-				   struct list_head *tmb_list,
-				   u16 pamt_entry_size[])
+				   struct list_head *tmb_list)
 {
-	unsigned long pamt_base[TDX_PS_NR];
-	unsigned long pamt_size[TDX_PS_NR];
-	unsigned long tdmr_pamt_base;
 	unsigned long tdmr_pamt_size;
 	struct page *pamt;
-	int pgsz, nid;
+	int nid;
 
 	nid = tdmr_get_nid(tdmr, tmb_list);
 
@@ -592,12 +578,10 @@ static __init int tdmr_set_up_pamt(struct tdmr_info *tdmr,
 	 * Calculate the PAMT size for each TDX supported page size
 	 * and the total PAMT size.
 	 */
-	tdmr_pamt_size = 0;
-	for (pgsz = TDX_PS_4K; pgsz < TDX_PS_NR; pgsz++) {
-		pamt_size[pgsz] = tdmr_get_pamt_sz(tdmr, pgsz,
-					pamt_entry_size[pgsz]);
-		tdmr_pamt_size += pamt_size[pgsz];
-	}
+	tdmr->pamt_4k_size = tdmr_get_pamt_sz(tdmr, TDX_PS_4K);
+	tdmr->pamt_2m_size = tdmr_get_pamt_sz(tdmr, TDX_PS_2M);
+	tdmr->pamt_1g_size = tdmr_get_pamt_sz(tdmr, TDX_PS_1G);
+	tdmr_pamt_size = tdmr->pamt_4k_size + tdmr->pamt_2m_size + tdmr->pamt_1g_size;
 
 	/*
 	 * Allocate one chunk of physically contiguous memory for all
@@ -606,25 +590,17 @@ static __init int tdmr_set_up_pamt(struct tdmr_info *tdmr,
 	 */
 	pamt = alloc_contig_pages(tdmr_pamt_size >> PAGE_SHIFT, GFP_KERNEL,
 			nid, &node_online_map);
+
+	/*
+	 * tdmr->pamt_4k_base is still zero so the error
+	 * path of the caller will skip freeing the PAMT.
+	 */
 	if (!pamt)
 		return -ENOMEM;
 
-	/*
-	 * Break the contiguous allocation back up into the
-	 * individual PAMTs for each page size.
-	 */
-	tdmr_pamt_base = page_to_pfn(pamt) << PAGE_SHIFT;
-	for (pgsz = TDX_PS_4K; pgsz < TDX_PS_NR; pgsz++) {
-		pamt_base[pgsz] = tdmr_pamt_base;
-		tdmr_pamt_base += pamt_size[pgsz];
-	}
-
-	tdmr->pamt_4k_base = pamt_base[TDX_PS_4K];
-	tdmr->pamt_4k_size = pamt_size[TDX_PS_4K];
-	tdmr->pamt_2m_base = pamt_base[TDX_PS_2M];
-	tdmr->pamt_2m_size = pamt_size[TDX_PS_2M];
-	tdmr->pamt_1g_base = pamt_base[TDX_PS_1G];
-	tdmr->pamt_1g_size = pamt_size[TDX_PS_1G];
+	tdmr->pamt_4k_base = page_to_phys(pamt);
+	tdmr->pamt_2m_base = tdmr->pamt_4k_base + tdmr->pamt_4k_size;
+	tdmr->pamt_1g_base = tdmr->pamt_2m_base + tdmr->pamt_2m_size;
 
 	return 0;
 }
@@ -655,10 +631,7 @@ static __init void tdmr_do_pamt_func(struct tdmr_info *tdmr,
 	tdmr_get_pamt(tdmr, &pamt_base, &pamt_size);
 
 	/* Do nothing if PAMT hasn't been allocated for this TDMR */
-	if (!pamt_size)
-		return;
-
-	if (WARN_ON_ONCE(!pamt_base))
+	if (!pamt_base)
 		return;
 
 	pamt_func(pamt_base, pamt_size);
@@ -684,14 +657,12 @@ static __init void tdmrs_free_pamt_all(struct tdmr_info_list *tdmr_list)
 
 /* Allocate and set up PAMTs for all TDMRs */
 static __init int tdmrs_set_up_pamt_all(struct tdmr_info_list *tdmr_list,
-					struct list_head *tmb_list,
-					u16 pamt_entry_size[])
+					struct list_head *tmb_list)
 {
 	int i, ret = 0;
 
 	for (i = 0; i < tdmr_list->nr_consumed_tdmrs; i++) {
-		ret = tdmr_set_up_pamt(tdmr_entry(tdmr_list, i), tmb_list,
-				pamt_entry_size);
+		ret = tdmr_set_up_pamt(tdmr_entry(tdmr_list, i), tmb_list);
 		if (ret)
 			goto err;
 	}
@@ -968,18 +939,13 @@ static __init int construct_tdmrs(struct list_head *tmb_list,
 				  struct tdmr_info_list *tdmr_list,
 				  struct tdx_sys_info_tdmr *sysinfo_tdmr)
 {
-	u16 pamt_entry_size[TDX_PS_NR] = {
-		sysinfo_tdmr->pamt_4k_entry_size,
-		sysinfo_tdmr->pamt_2m_entry_size,
-		sysinfo_tdmr->pamt_1g_entry_size,
-	};
 	int ret;
 
 	ret = fill_out_tdmrs(tmb_list, tdmr_list);
 	if (ret)
 		return ret;
 
-	ret = tdmrs_set_up_pamt_all(tdmr_list, tmb_list, pamt_entry_size);
+	ret = tdmrs_set_up_pamt_all(tdmr_list, tmb_list);
 	if (ret)
 		return ret;
 
-- 
2.55.0

[PATCH v10 02/11] x86/virt/tdx: Allocate page bitmap for Dynamic PAMT

[Rick Edgecombe]

The TDX Physical Address Metadata Table (PAMT) holds data about the
physical memory used by TDX, and must be allocated by the kernel during
TDX module initialization.

The exact size of the required PAMT memory is determined by the TDX module
and may vary between TDX module versions. Currently it is approximately
0.4% of the system memory. This is a significant commitment, especially if
it is not known upfront whether the machine will run any TDX guests.

Each memory region that the TDX module might use needs three separate PAMT
allocations. One for each supported page size (1GB, 2MB, 4KB). The
TDX module supports a new feature designed to reduce PAMT overhead called
Dynamic PAMT (DPAMT). Under DPAMT the 4KB level is allocated dynamically
during runtime, while the 1GB and 2MB levels remain allocated on TDX
module initialization.

However, in the details, DPAMT still needs some smaller per 4KB page
scoped data (currently it is 1 bit per page). The TDX module exposes the
number of bits as a separate piece of metadata than the 4KB static
allocation for normal PAMT. Although the size is enumerated differently,
it is handed to the TDX module in the same way the 4KB page size PAMT
allocation is for normal PAMT.

Begin to implement DPAMT in the kernel by reading the bits-per-page needed
for DPAMT. Calculate the size needed for the bitmap, and use it instead of
the 4KB size determined for normal PAMT, in the case of DPAMT.

The existing metadata reading code was generated by a script, but the
current plan is to stop generating this code, as the script has continued
to need adjustments. So add manually written code and adjust the comment
about it being autogenerated to be more generic. Start to adopt a more
normal kernel code style without the ternary statements and if
conditionals assignments that the auto generated code has.

AI was used under supervision to collect/apply feedback, review code and
workshop logs.

Based on a patch originally by Kiryl Shutsemau.

Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Reviewed-by: Binbin Wu <binbin.wu@linux.intel.com>
Reviewed-by: Tony Lindgren <tony.lindgren@linux.intel.com>
Reviewed-by: Yan Zhao <yan.y.zhao@intel.com>
Reviewed-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Vishal Annapurve <vannapurve@google.com>
Acked-by: Sohil Mehta <sohil.mehta@intel.com>
---
v10:
 - Remove other "autogenerated" references. (AI nit checker)
 - Change "Dynamic PAMT" to "DPAMT" in comments, and in the second
   reference in the logs. (Dave)

v9:
 - Switch to byte size hint in metadata field id (Sashiko)

v7:
 - Re-order pamt size calculations for greater readability (Kiryl)
 - Move comment to its own line in tdx_supports_dynamic_pamt() (Yan,
   Sohil)
 - Log tweak (Sohil)
 - Make comment in metadata reading more appropriate (Sohil)
 - Drop Assisted-by tag and cover AI use in log (Dave)
 - Move tdx_supports_dynamic_pamt() to not static inline to reduce
   churn in later changes

v6:
 - Improve comment (Binbin)
 - Log tweaks
 - Mark tdmr_get_pamt_bitmap_sz() __init in response to upstream
   changes
 - Switch to more normal kernel code style, even though it differs from
   the existing auto generated code.
---
 arch/x86/include/asm/tdx.h                  |  2 ++
 arch/x86/include/asm/tdx_global_metadata.h  |  9 ++++---
 arch/x86/virt/vmx/tdx/tdx.c                 | 29 +++++++++++++++++++--
 arch/x86/virt/vmx/tdx/tdx_global_metadata.c | 23 +++++++++++++++-
 4 files changed, 57 insertions(+), 6 deletions(-)

diff --git a/arch/x86/include/asm/tdx.h b/arch/x86/include/asm/tdx.h
index 89e97d5761d89..d414064436221 100644
--- a/arch/x86/include/asm/tdx.h
+++ b/arch/x86/include/asm/tdx.h
@@ -118,6 +118,8 @@ static inline bool tdx_supports_runtime_update(const struct tdx_sys_info *sysinf
 	return sysinfo->features.tdx_features0 & TDX_FEATURES0_TD_PRESERVING;
 }
 
+bool tdx_supports_dynamic_pamt(const struct tdx_sys_info *sysinfo);
+
 int tdx_guest_keyid_alloc(void);
 u32 tdx_get_nr_guest_keyids(void);
 void tdx_guest_keyid_free(unsigned int keyid);
diff --git a/arch/x86/include/asm/tdx_global_metadata.h b/arch/x86/include/asm/tdx_global_metadata.h
index 41150d546589c..8a3cc1a2a41e2 100644
--- a/arch/x86/include/asm/tdx_global_metadata.h
+++ b/arch/x86/include/asm/tdx_global_metadata.h
@@ -1,7 +1,7 @@
 /* SPDX-License-Identifier: GPL-2.0 */
-/* Automatically generated TDX global metadata structures. */
-#ifndef _X86_VIRT_TDX_AUTO_GENERATED_TDX_GLOBAL_METADATA_H
-#define _X86_VIRT_TDX_AUTO_GENERATED_TDX_GLOBAL_METADATA_H
+/* TDX global metadata structures. */
+#ifndef _X86_VIRT_TDX_TDX_GLOBAL_METADATA_H
+#define _X86_VIRT_TDX_TDX_GLOBAL_METADATA_H
 
 #include <linux/types.h>
 
@@ -21,6 +21,9 @@ struct tdx_sys_info_tdmr {
 	u16 pamt_4k_entry_size;
 	u16 pamt_2m_entry_size;
 	u16 pamt_1g_entry_size;
+
+	/* Optional metadata, if DPAMT is supported */
+	u8  pamt_page_bitmap_entry_bits;
 };
 
 struct tdx_sys_info_td_ctrl {
diff --git a/arch/x86/virt/vmx/tdx/tdx.c b/arch/x86/virt/vmx/tdx/tdx.c
index 51b191bf12258..54074ddcce858 100644
--- a/arch/x86/virt/vmx/tdx/tdx.c
+++ b/arch/x86/virt/vmx/tdx/tdx.c
@@ -510,6 +510,18 @@ static __init int fill_out_tdmrs(struct list_head *tmb_list,
 	return 0;
 }
 
+static __init unsigned long tdmr_get_pamt_bitmap_sz(struct tdmr_info *tdmr)
+{
+	unsigned long pamt_sz, nr_pamt_entries;
+	int bits_per_entry;
+
+	bits_per_entry = tdx_sysinfo.tdmr.pamt_page_bitmap_entry_bits;
+	nr_pamt_entries = tdmr->size >> PAGE_SHIFT;
+	pamt_sz = DIV_ROUND_UP(nr_pamt_entries * bits_per_entry, BITS_PER_BYTE);
+
+	return PAGE_ALIGN(pamt_sz);
+}
+
 /*
  * Calculate PAMT size given a TDMR and a page size.  The returned
  * PAMT size is always aligned up to 4K page boundary.
@@ -578,9 +590,16 @@ static __init int tdmr_set_up_pamt(struct tdmr_info *tdmr,
 	 * Calculate the PAMT size for each TDX supported page size
 	 * and the total PAMT size.
 	 */
-	tdmr->pamt_4k_size = tdmr_get_pamt_sz(tdmr, TDX_PS_4K);
-	tdmr->pamt_2m_size = tdmr_get_pamt_sz(tdmr, TDX_PS_2M);
 	tdmr->pamt_1g_size = tdmr_get_pamt_sz(tdmr, TDX_PS_1G);
+	tdmr->pamt_2m_size = tdmr_get_pamt_sz(tdmr, TDX_PS_2M);
+
+	if (tdx_supports_dynamic_pamt(&tdx_sysinfo)) {
+		/* With DPAMT, PAMT_4K is replaced with a bitmap */
+		tdmr->pamt_4k_size = tdmr_get_pamt_bitmap_sz(tdmr);
+	} else {
+		tdmr->pamt_4k_size = tdmr_get_pamt_sz(tdmr, TDX_PS_4K);
+	}
+
 	tdmr_pamt_size = tdmr->pamt_4k_size + tdmr->pamt_2m_size + tdmr->pamt_1g_size;
 
 	/*
@@ -1971,6 +1990,12 @@ u64 tdh_phymem_page_wbinvd_hkid(u64 hkid, kvm_pfn_t pfn)
 }
 EXPORT_SYMBOL_FOR_KVM(tdh_phymem_page_wbinvd_hkid);
 
+bool tdx_supports_dynamic_pamt(const struct tdx_sys_info *sysinfo)
+{
+	/* To be enabled when kernel is ready. */
+	return false;
+}
+
 void tdx_sys_disable(void)
 {
 	struct tdx_module_args args = {};
diff --git a/arch/x86/virt/vmx/tdx/tdx_global_metadata.c b/arch/x86/virt/vmx/tdx/tdx_global_metadata.c
index e49c300f23d43..98ebf17aab1ce 100644
--- a/arch/x86/virt/vmx/tdx/tdx_global_metadata.c
+++ b/arch/x86/virt/vmx/tdx/tdx_global_metadata.c
@@ -1,6 +1,6 @@
 // SPDX-License-Identifier: GPL-2.0
 /*
- * Automatically generated functions to read TDX global metadata.
+ * Functions to read TDX global metadata.
  *
  * This file doesn't compile on its own as it lacks of inclusion
  * of SEAMCALL wrapper primitive which reads global metadata.
@@ -33,6 +33,18 @@ static __init int get_tdx_sys_info_features(struct tdx_sys_info_features *sysinf
 	return ret;
 }
 
+static __init int get_tdx_sys_info_tdmr_dpamt(struct tdx_sys_info_tdmr *sysinfo_tdmr)
+{
+	int ret;
+	u64 val;
+
+	ret = read_sys_metadata_field(0x9100000000000013, &val);
+	if (!ret)
+		sysinfo_tdmr->pamt_page_bitmap_entry_bits = val;
+
+	return ret;
+}
+
 static __init int get_tdx_sys_info_tdmr(struct tdx_sys_info_tdmr *sysinfo_tdmr)
 {
 	int ret = 0;
@@ -129,5 +141,14 @@ static __init int get_tdx_sys_info(struct tdx_sys_info *sysinfo)
 	ret = ret ?: get_tdx_sys_info_td_ctrl(&sysinfo->td_ctrl);
 	ret = ret ?: get_tdx_sys_info_td_conf(&sysinfo->td_conf);
 
+	/*
+	 * The kernel supports using TDX without DPAMT, so
+	 * avoid reporting failure if it's not supported. Don't
+	 * try to support buggy TDX modules that advertise
+	 * DPAMT but don't expose the metadata.
+	 */
+	if (!ret && tdx_supports_dynamic_pamt(sysinfo))
+		ret = get_tdx_sys_info_tdmr_dpamt(&sysinfo->tdmr);
+
 	return ret;
 }
-- 
2.55.0

[PATCH v10 03/11] x86/virt/tdx: Add __tdx_pamt_get/put() helpers

[Rick Edgecombe]

Add helpers to use when allocating or preparing pages that are handed to
the TDX module for use as control/S-EPT pages, and thus need Dynamic PAMT
(DPAMT) adjustments.

The TDX module tracks some state for each page of physical memory that it
might use. It calls this state the PAMT. It includes separate state for
each page size a physical page could be utilized at within the TDX module
(1GB, 2MB, 4KB). In DPAMT, only the 4KB page size state is
allocated dynamically.

KVM will need to hand pages to the TDX module that it will use at 4KB
granularity. So these pages will need DPAMT backing added before they are
used by the TDX module, and removed afterwards.

Create __tdx_pamt_get/put() to handle installing DPAMT 4KB backing for
pages that are already allocated (such as KVM's use of S-EPT page tables
or guest private memory). Have them take a pfn instead of a struct page,
as future changes will want to use these helpers for guest pages which are
tracked by PFN.

Also add __tdx_alloc_control_page() and __tdx_free_control_page() to handle
both page allocation and DPAMT installation. Make them behave like
normal alloc/free functions where allocation can fail in the case of no
memory, but free (with any necessary DPAMT release) always
succeeds. Do this so they can eventually support the existing TDX flows
that require teardowns to succeed.

Don't CLFLUSH the DPAMT pages handed to the TDX module, as is done
for some other SEAMCALLs, as the TDX docs specify that this is only
needed on "TD private memory or TD control structure page".

Since these allocations will be easily user triggerable, account the
memory.

Only one pair of DPAMT pages is required for each 2MB-aligned physical
region, so multiple callers could trip over each other if they try to
manage the shared backing for two separate 4KB pages contained in one.
To build the logic up iteratively, don't do anything to handle pages
from the same 2MB region yet. Functionality to handle this will be added
before DPAMT can be enabled.

AI was used under supervision to collect/apply feedback, split patches,
review code and workshop logs.

Based on a patch originally by Kiryl Shutsemau.

Co-developed-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Reviewed-by: Chao Gao <chao.gao@intel.com>
Reviewed-by: Binbin Wu <binbin.wu@linux.intel.com>
Reviewed-by: Tony Lindgren <tony.lindgren@linux.intel.com>
Reviewed-by: Vishal Annapurve <vannapurve@google.com>
Acked-by: Sohil Mehta <sohil.mehta@intel.com>
---
v10:
 - Clarify in comments about the limitations of the helpers added. (Dave)
 - Add comments in pamt_2mb_arg() about what is going on (Dave)
 - Prepend helpers with __ (Dave)
 - Align tdx_module_args args (Dave)
 - Standardize "TDX Module" (Dave)
 - Remove exports for newly private __ functions.
 - Move TDX_DPAMT_ENTRY_PAGE_CNT out of header to sole file that uses it.
   (AI nit checker)
 - Standardize for loop iterators declaration. (Nit checker)
 - Change "Dynamic PAMT" to "DPAMT" in comments, and in the second
   reference in the logs. (Dave)

v9:
 - Address about how the log doesn't fully cover the limitations at this
   stage. (Sashiko)

v7:
 - Comment improvements (Chao)
 - Drop unneeded addition of mm.h header include (Binbin)
 - Log clarity, code style nits (Sohil)
 - Drop Assisted-by tag and cover AI use in log (Dave)
---
 arch/x86/virt/vmx/tdx/tdx.c | 197 ++++++++++++++++++++++++++++++++++++
 arch/x86/virt/vmx/tdx/tdx.h |   2 +
 2 files changed, 199 insertions(+)

diff --git a/arch/x86/virt/vmx/tdx/tdx.c b/arch/x86/virt/vmx/tdx/tdx.c
index 54074ddcce858..9caaa5de28817 100644
--- a/arch/x86/virt/vmx/tdx/tdx.c
+++ b/arch/x86/virt/vmx/tdx/tdx.c
@@ -46,6 +46,9 @@
 #include "seamcall_internal.h"
 #include "tdx.h"
 
+/* Number of DPAMT pages to be provided to TDX module per 2MB region of PA */
+#define TDX_DPAMT_ENTRY_PAGE_CNT 2
+
 struct tdx_module_state {
 	bool initialized;
 	bool sysinit_done;
@@ -1996,6 +1999,200 @@ bool tdx_supports_dynamic_pamt(const struct tdx_sys_info *sysinfo)
 	return false;
 }
 
+static int alloc_pamt_array(struct page **pamt_pages)
+{
+	int i, j;
+
+	for (i = 0; i < TDX_DPAMT_ENTRY_PAGE_CNT; i++) {
+		pamt_pages[i] = alloc_page(GFP_KERNEL_ACCOUNT);
+		if (!pamt_pages[i])
+			goto err;
+	}
+
+	return 0;
+
+err:
+	for (j = 0; j < i; j++)
+		__free_page(pamt_pages[j]);
+
+	return -ENOMEM;
+}
+
+static void free_pamt_array(struct page **pamt_pages)
+{
+	int i;
+
+	for (i = 0; i < TDX_DPAMT_ENTRY_PAGE_CNT; i++) {
+		/*
+		 * Reset pages unconditionally to cover cases
+		 * where they were passed to the TDX module.
+		 */
+		tdx_quirk_reset_paddr(page_to_phys(pamt_pages[i]), PAGE_SIZE);
+
+		__free_page(pamt_pages[i]);
+	}
+}
+
+/* Helper for building DPAMT seamcall() arguments. */
+static u64 pamt_2mb_arg(kvm_pfn_t pfn)
+{
+	/* Find the 2MB-wide DPAMT region for 'pfn': */
+	unsigned long hpa_2mb = ALIGN_DOWN(pfn << PAGE_SHIFT, PMD_SIZE);
+
+	/*
+	 * TDX ABI requires specifying the page level the installed DPAMT
+	 * backing will cover, even though today only 2MB is supported.
+	 */
+	return hpa_2mb | TDX_PS_2M;
+}
+
+/* Add PAMT backing for the 2MB region surrounding the given pfn. */
+static u64 tdh_phymem_pamt_add(kvm_pfn_t pfn, struct page **pamt_pages)
+{
+	struct tdx_module_args args = {
+		.rcx = pamt_2mb_arg(pfn),
+		.rdx = page_to_phys(pamt_pages[0]),
+		.r8  = page_to_phys(pamt_pages[1]),
+	};
+
+	return seamcall(TDH_PHYMEM_PAMT_ADD, &args);
+}
+
+/* Remove PAMT backing for the 2MB region surrounding the given pfn. */
+static u64 tdh_phymem_pamt_remove(kvm_pfn_t pfn, struct page **pamt_pages)
+{
+	struct tdx_module_args args = {
+		.rcx = pamt_2mb_arg(pfn),
+	};
+	u64 ret;
+
+	ret = seamcall_ret(TDH_PHYMEM_PAMT_REMOVE, &args);
+	if (ret)
+		return ret;
+
+	/* Copy PAMT pages out of the struct per the TDX ABI */
+	pamt_pages[0] = phys_to_page(args.rdx);
+	pamt_pages[1] = phys_to_page(args.r8);
+
+	return 0;
+}
+
+/*
+ * Allocate DPAMT memory for the 2MB aligned region  surrounding
+ * the given page.
+ *
+ * Only call this when the pfn is known not to already have Dynamic
+ * PAMT pages in the TDX module for it.
+ *
+ * Effectively it is not (yet) like a get, and more like a manual
+ * manipulation of the DPAMT backing for the 2MB aligned range
+ * covered by the pfn.
+ */
+static int __tdx_pamt_get(kvm_pfn_t pfn)
+{
+	struct page *pamt_pages[TDX_DPAMT_ENTRY_PAGE_CNT];
+	u64 tdx_status;
+	int ret;
+
+	if (!tdx_supports_dynamic_pamt(&tdx_sysinfo))
+		return 0;
+
+	ret = alloc_pamt_array(pamt_pages);
+	if (ret)
+		return ret;
+
+	tdx_status = tdh_phymem_pamt_add(pfn, pamt_pages);
+	if (tdx_status != TDX_SUCCESS) {
+		ret = -EIO;
+		goto out_free;
+	}
+
+	return 0;
+
+out_free:
+	free_pamt_array(pamt_pages);
+
+	return ret;
+}
+
+/*
+ * Free DPAMT memory for the 2MB aligned region surrounding the
+ * given page. Only call this when the pfn is known to already
+ * have DPAMT pages in the TDX module for it, and no other pfns
+ * in the aligned 2MB physical region still need it.
+ *
+ * Don't make multiple calls concurrently of __tdx_pamt_get/put(),
+ * as there is no protections from races.
+ *
+ * Effectively it is not (yet) like a refcounted put, and more like a
+ * manual manipulation of the DPAMT backing for the 2MB aligned
+ * range covered by the pfn.
+ */
+static void __tdx_pamt_put(kvm_pfn_t pfn)
+{
+	struct page *pamt_pages[TDX_DPAMT_ENTRY_PAGE_CNT] = {};
+	u64 tdx_status;
+
+	if (!tdx_supports_dynamic_pamt(&tdx_sysinfo))
+		return;
+
+	tdx_status = tdh_phymem_pamt_remove(pfn, pamt_pages);
+
+	/*
+	 * Don't free pamt_pages as it could hold garbage when
+	 * tdh_phymem_pamt_remove() fails.  Don't panic/BUG_ON(), as
+	 * there is no risk of data corruption, but do yell loudly as
+	 * failure indicates a kernel bug, memory is being leaked, and
+	 * the dangling PAMT entry may cause future operations to fail.
+	 */
+	if (WARN_ON_ONCE(tdx_status != TDX_SUCCESS))
+		return;
+
+	free_pamt_array(pamt_pages);
+}
+
+/*
+ * Return a page that can be gifted to the TDX module for use as a "control"
+ * page, i.e. pages that are used for control structures for a given TDX
+ * guest, and thus obtain TDX protections, including PAMT tracking.
+ *
+ * This function is currently only safe to call once. And not safe to call
+ * if __tdx_pamt_get() is called before or after.
+ */
+static struct page * __maybe_unused __tdx_alloc_control_page(void)
+{
+	struct page *page;
+
+	page = alloc_page(GFP_KERNEL_ACCOUNT);
+	if (!page)
+		return NULL;
+
+	if (__tdx_pamt_get(page_to_pfn(page))) {
+		__free_page(page);
+		return NULL;
+	}
+
+	return page;
+}
+
+/*
+ * Free a page that was gifted to the TDX module for use as a control
+ * page. After this, the page is no longer protected by TDX.
+ *
+ * Like __tdx_pamt_put(), this is currently only safe to call this when
+ * a page is already known to have DPAMT pages in the TDX module for
+ * it, and no other pages in the aligned 2MB physical region will
+ * still need the backing.
+ */
+static void __maybe_unused __tdx_free_control_page(struct page *page)
+{
+	if (!page)
+		return;
+
+	__tdx_pamt_put(page_to_pfn(page));
+	__free_page(page);
+}
+
 void tdx_sys_disable(void)
 {
 	struct tdx_module_args args = {};
diff --git a/arch/x86/virt/vmx/tdx/tdx.h b/arch/x86/virt/vmx/tdx/tdx.h
index bdfd0e1e337ac..a886c54decaad 100644
--- a/arch/x86/virt/vmx/tdx/tdx.h
+++ b/arch/x86/virt/vmx/tdx/tdx.h
@@ -48,6 +48,8 @@
 #define TDH_SYS_CONFIG			45
 #define TDH_SYS_SHUTDOWN		52
 #define TDH_SYS_UPDATE			53
+#define TDH_PHYMEM_PAMT_ADD		58
+#define TDH_PHYMEM_PAMT_REMOVE		59
 #define TDH_SYS_DISABLE			69
 
 /*
-- 
2.55.0

[PATCH v10 04/11] x86/virt/tdx: Allocate refcounts for Dynamic PAMT memory

[Rick Edgecombe]

The PAMT memory holds metadata for all possible TDX protected memory. Each
physical address range is covered by PAMT entries at three levels (1GB,
2MB, 4KB). With Dynamic PAMT (DPAMT), the 4KB level of PAMT is allocated
on demand. The kernel supplies the TDX module with page pairs to store the
4KB level entries, which cover 2MB of host physical memory. The kernel must
provide this page pair before using pages from the range for TDX. If this
is not done, SEAMCALLs that give the pages to be protected by the TDX
module will fail.

Allocate reference counters for every 2MB range to track TDX memory usage.
This can be used to handle concurrent get/put callers, in order to
accurately determine when the dynamic 4KB level of DPAMT needs to be
allocated and when it can be freed.

This allocation will currently consume 2MB for every 1TB of address
space from 0 to max_pfn. The allocation size will depend on how the RAM is
physically laid out. In a worst case scenario where the entire 52 bit
address space is covered this would be 8GB. Then the DPAMT refcount
allocations could hypothetically cause the savings from DPAMT to go
negative on exotic platforms with sparse, small amounts of memory.

Future changes could reduce this refcount overhead to be only allocating
refcounts for physical ranges that contain memory that TDX can use.
However, this is left for future work.

AI was used under supervision to collect/apply feedback, review code and
workshop logs.

Based on a patch originally by Kiryl Shutsemau.

Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Reviewed-by: Binbin Wu <binbin.wu@linux.intel.com>
Reviewed-by: Chao Gao <chao.gao@intel.com>
Reviewed-by: Yan Zhao <yan.y.zhao@intel.com>
Reviewed-by: Tony Lindgren <tony.lindgren@linux.intel.com>
Reviewed-by: Vishal Annapurve <vannapurve@google.com>
Acked-by: Sohil Mehta <sohil.mehta@intel.com>
---
v10:
 - Change "Dynamic PAMT" to "DPAMT" in comments, and in the second
   reference in the logs. (Dave)
 - Rename some other pamt gunk to dpamt (Dave)

v7:
 - Annotate functions __init. (Chao)
 - Log tweaks. (Yan)
 - Standardize on memory units in the text. (Sohil)
 - Delete unneeded comment. (Sohil)
 - Use vzalloc(). (Sohil)
 - Drop Assisted-by tag and cover AI use in log. (Dave)

v6:
 - Remove confusing reference to allocating PAMT memory in
   pamt_refcounts comment. (Yan)
 - Rename "metadata" function names that really deal with refcounts, as
   metadata already has a different meaning in TDX.
 - Move tdx_find_pamt_refcount() to this patch to aid in reviewability.
---
 arch/x86/virt/vmx/tdx/tdx.c | 57 +++++++++++++++++++++++++++++++++++--
 1 file changed, 54 insertions(+), 3 deletions(-)

diff --git a/arch/x86/virt/vmx/tdx/tdx.c b/arch/x86/virt/vmx/tdx/tdx.c
index 9caaa5de28817..305289bd673be 100644
--- a/arch/x86/virt/vmx/tdx/tdx.c
+++ b/arch/x86/virt/vmx/tdx/tdx.c
@@ -30,6 +30,7 @@
 #include <linux/suspend.h>
 #include <linux/syscore_ops.h>
 #include <linux/idr.h>
+#include <linux/vmalloc.h>
 #include <asm/page.h>
 #include <asm/special_insns.h>
 #include <asm/msr-index.h>
@@ -66,6 +67,14 @@ static DEFINE_PER_CPU(bool, tdx_lp_initialized);
 
 static struct tdmr_info_list tdx_tdmr_list;
 
+/*
+ * On a machine with DPAMT, the kernel maintains a reference counter
+ * for every 2MB range. The counter indicates how many users there are for
+ * the DPAMT at the 2MB range. The kernel allocates DPAMT refcounts at
+ * initialization.
+ */
+static atomic_t *dpamt_refcounts;
+
 /* All TDX-usable memory regions.  Protected by mem_hotplug_lock. */
 static LIST_HEAD(tdx_memlist);
 
@@ -255,6 +264,42 @@ static struct syscore tdx_syscore = {
 	.ops = &tdx_syscore_ops,
 };
 
+/*
+ * Allocate DPAMT reference counters for all physical memory.
+ *
+ * It consumes 2MB for every 1TB of physical memory.
+ */
+static __init int init_dpamt_refcounts(void)
+{
+	size_t size = DIV_ROUND_UP(max_pfn, PTRS_PER_PTE) * sizeof(*dpamt_refcounts);
+
+	if (!tdx_supports_dynamic_pamt(&tdx_sysinfo))
+		return 0;
+
+	dpamt_refcounts = vzalloc(size);
+	if (!dpamt_refcounts)
+		return -ENOMEM;
+
+	return 0;
+}
+
+static __init void free_dpamt_refcounts(void)
+{
+	if (!tdx_supports_dynamic_pamt(&tdx_sysinfo))
+		return;
+
+	vfree(dpamt_refcounts);
+	dpamt_refcounts = NULL;
+}
+
+static __maybe_unused atomic_t *tdx_find_dpamt_refcount(unsigned long pfn)
+{
+	/* Find which PMD a PFN is in. */
+	unsigned long index = pfn >> (PMD_SHIFT - PAGE_SHIFT);
+
+	return &dpamt_refcounts[index];
+}
+
 /*
  * Add a memory region as a TDX memory block.  The caller must make sure
  * all memory regions are added in address ascending order and don't
@@ -1155,10 +1200,14 @@ static __init int init_tdx_module(void)
 	 */
 	get_online_mems();
 
-	ret = build_tdx_memlist(&tdx_memlist);
+	ret = init_dpamt_refcounts();
 	if (ret)
 		goto out_put_tdxmem;
 
+	ret = build_tdx_memlist(&tdx_memlist);
+	if (ret)
+		goto err_free_dpamt_refcounts;
+
 	/* Allocate enough space for constructing TDMRs */
 	ret = alloc_tdmr_list(&tdx_tdmr_list, &tdx_sysinfo.tdmr);
 	if (ret)
@@ -1208,6 +1257,8 @@ static __init int init_tdx_module(void)
 	free_tdmr_list(&tdx_tdmr_list);
 err_free_tdxmem:
 	free_tdx_memlist(&tdx_memlist);
+err_free_dpamt_refcounts:
+	free_dpamt_refcounts();
 	goto out_put_tdxmem;
 }
 
@@ -2159,7 +2210,7 @@ static void __tdx_pamt_put(kvm_pfn_t pfn)
  * This function is currently only safe to call once. And not safe to call
  * if __tdx_pamt_get() is called before or after.
  */
-static struct page * __maybe_unused __tdx_alloc_control_page(void)
+static __maybe_unused struct page *__tdx_alloc_control_page(void)
 {
 	struct page *page;
 
@@ -2184,7 +2235,7 @@ static struct page * __maybe_unused __tdx_alloc_control_page(void)
  * it, and no other pages in the aligned 2MB physical region will
  * still need the backing.
  */
-static void __maybe_unused __tdx_free_control_page(struct page *page)
+static __maybe_unused void __tdx_free_control_page(struct page *page)
 {
 	if (!page)
 		return;
-- 
2.55.0

[PATCH v10 05/11] x86/virt/tdx: Handle multiple callers in tdx_pamt_get/put()

[Rick Edgecombe]

__tdx_pamt_get()/__tdx_pamt_put() unconditionally add or remove Dynamic
PAMT (DPAMT) backing for the 2MB region covering the passed page. However,
multiple callers can add or remove 4KB pages that fall within the same 2MB
region and in that scenario only a single PAMT entry is required.

Make the helpers handle only adding/removing DPAMT backing when required,
by refcounting each 2MB range. Gate the actual DPAMT add and remove on
refcount transitions (0->1 and 1->0). Serialize the refcount check and
SEAMCALL with a global spinlock so the read-decide-act sequence is atomic.
This also avoids TDX module BUSY errors, as the DPAMT add and remove
SEAMCALLs take internal TDX module locks for the 2MB ranges of the
specified PFN and the PAMT page pair PFNs. So simultaneous attempts on
the same 2MB ranges of the PFNs would otherwise encounter an error, which
would not be handleable in the put case.

The lock is global and heavyweight. Use simple conditional logic to keep
correctness obvious. This will be optimized in a later change.

The dpamt_refcounts[] are atomic_t's. They do not strictly need to be
because all access is protected by pamt_lock. The overhead of an atomic_t
in this situation is minuscule compared to the global lock. Leave the
atomic_t in place to enable future optimization with minimal churn.

Since the DPAMT helpers are broadly functional now, drop the "__"
to rename them tdx_pamt_get/put() and tdx_alloc/free_control_page().
Export them for use in KVM in subsequent changes.

AI was used under supervision to collect/apply feedback, split patches,
review code and workshop logs.

Based on a patch originally by Kiryl Shutsemau.

Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Reviewed-by: Binbin Wu <binbin.wu@linux.intel.com>
Reviewed-by: Chao Gao <chao.gao@intel.com>
Reviewed-by: Yan Zhao <yan.y.zhao@intel.com>
Reviewed-by: Tony Lindgren <tony.lindgren@linux.intel.com>
Reviewed-by: Nikolay Borisov <nik.borisov@suse.com>
Reviewed-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Vishal Annapurve <vannapurve@google.com>
Acked-by: Sohil Mehta <sohil.mehta@intel.com>
---
v10:
 - Rename the helper functions in this patch after changes to patch
   "x86/virt/tdx: Add tdx_alloc/free_control_page() helpers".
 - Remove the comments around helper limitations, leaving to be as they
   were in v9, except for the "TDX module" wording fix.
 - Export tdx_alloc/free_control_page() in this patch as a result of
   changes in "x86/virt/tdx: Add tdx_alloc/free_control_page()
   helpers" to make them private functions initially. Since they now become
   more broadly functional here, do the export in this patch.
 - For consistency, also move the tdx_pamt_get/put() export here too.
 - Change "Dynamic PAMT" to "DPAMT" in comments, and in the second
   reference in the logs. (Dave)
 - Rename "pamt" gunk to "dpamt". (Dave)

v8:
 - Fix PAMT capitalization in comment (Sohil)

v7:
 - Convert scoped_guard() blocks to use normal spin_un/lock() for the
   sake of making next patches diff cleaner
 - Drop __maybe_unused from tdx_find_pamt_refcount() (Binbin)
 - Switch to atomic_inc_not_zero() (Dave)
 - Justify use of atomic_t in log (Sohil)
 - Log/comments (Yan)
 - Drop Assisted-by tag and cover AI use in log (Dave)
---
 arch/x86/include/asm/tdx.h  |   6 +++
 arch/x86/virt/vmx/tdx/tdx.c | 102 ++++++++++++++++++++----------------
 2 files changed, 63 insertions(+), 45 deletions(-)

diff --git a/arch/x86/include/asm/tdx.h b/arch/x86/include/asm/tdx.h
index d414064436221..f7442ad20e46d 100644
--- a/arch/x86/include/asm/tdx.h
+++ b/arch/x86/include/asm/tdx.h
@@ -120,12 +120,18 @@ static inline bool tdx_supports_runtime_update(const struct tdx_sys_info *sysinf
 
 bool tdx_supports_dynamic_pamt(const struct tdx_sys_info *sysinfo);
 
+int tdx_pamt_get(kvm_pfn_t pfn);
+void tdx_pamt_put(kvm_pfn_t pfn);
+
 int tdx_guest_keyid_alloc(void);
 u32 tdx_get_nr_guest_keyids(void);
 void tdx_guest_keyid_free(unsigned int keyid);
 
 void tdx_quirk_reset_paddr(unsigned long base, unsigned long size);
 
+struct page *tdx_alloc_control_page(void);
+void tdx_free_control_page(struct page *page);
+
 struct tdx_td {
 	/* TD root structure: */
 	struct page *tdr_page;
diff --git a/arch/x86/virt/vmx/tdx/tdx.c b/arch/x86/virt/vmx/tdx/tdx.c
index 305289bd673be..c347600a0aabb 100644
--- a/arch/x86/virt/vmx/tdx/tdx.c
+++ b/arch/x86/virt/vmx/tdx/tdx.c
@@ -292,7 +292,7 @@ static __init void free_dpamt_refcounts(void)
 	dpamt_refcounts = NULL;
 }
 
-static __maybe_unused atomic_t *tdx_find_dpamt_refcount(unsigned long pfn)
+static atomic_t *tdx_find_dpamt_refcount(unsigned long pfn)
 {
 	/* Find which PMD a PFN is in. */
 	unsigned long index = pfn >> (PMD_SHIFT - PAGE_SHIFT);
@@ -2097,7 +2097,7 @@ static u64 pamt_2mb_arg(kvm_pfn_t pfn)
 	return hpa_2mb | TDX_PS_2M;
 }
 
-/* Add PAMT backing for the 2MB region surrounding the given pfn. */
+/* Add DPAMT backing for the 2MB region surrounding the given pfn. */
 static u64 tdh_phymem_pamt_add(kvm_pfn_t pfn, struct page **pamt_pages)
 {
 	struct tdx_module_args args = {
@@ -2109,7 +2109,7 @@ static u64 tdh_phymem_pamt_add(kvm_pfn_t pfn, struct page **pamt_pages)
 	return seamcall(TDH_PHYMEM_PAMT_ADD, &args);
 }
 
-/* Remove PAMT backing for the 2MB region surrounding the given pfn. */
+/* Remove DPAMT backing for the 2MB region surrounding the given pfn. */
 static u64 tdh_phymem_pamt_remove(kvm_pfn_t pfn, struct page **pamt_pages)
 {
 	struct tdx_module_args args = {
@@ -2128,20 +2128,14 @@ static u64 tdh_phymem_pamt_remove(kvm_pfn_t pfn, struct page **pamt_pages)
 	return 0;
 }
 
-/*
- * Allocate DPAMT memory for the 2MB aligned region  surrounding
- * the given page.
- *
- * Only call this when the pfn is known not to already have Dynamic
- * PAMT pages in the TDX module for it.
- *
- * Effectively it is not (yet) like a get, and more like a manual
- * manipulation of the DPAMT backing for the 2MB aligned range
- * covered by the pfn.
- */
-static int __tdx_pamt_get(kvm_pfn_t pfn)
+/* Serializes adding/removing DPAMT memory */
+static DEFINE_SPINLOCK(dpamt_lock);
+
+/* Bump DPAMT refcount for the given pfn and allocate DPAMT backing if needed. */
+int tdx_pamt_get(kvm_pfn_t pfn)
 {
 	struct page *pamt_pages[TDX_DPAMT_ENTRY_PAGE_CNT];
+	atomic_t *dpamt_refcount;
 	u64 tdx_status;
 	int ret;
 
@@ -2152,41 +2146,59 @@ static int __tdx_pamt_get(kvm_pfn_t pfn)
 	if (ret)
 		return ret;
 
+	dpamt_refcount = tdx_find_dpamt_refcount(pfn);
+
+	spin_lock(&dpamt_lock);
+
+	/*
+	 * If the DPAMT entry is already added (i.e. refcount >= 1),
+	 * then just increment the refcount.
+	 */
+	if (atomic_inc_not_zero(dpamt_refcount))
+		goto out_free;
+
+	/* Try to add the PAMT page and take the refcount 0->1. */
 	tdx_status = tdh_phymem_pamt_add(pfn, pamt_pages);
-	if (tdx_status != TDX_SUCCESS) {
+	if (WARN_ON_ONCE(tdx_status != TDX_SUCCESS)) {
 		ret = -EIO;
 		goto out_free;
 	}
 
+	atomic_set(dpamt_refcount, 1);
+	spin_unlock(&dpamt_lock);
 	return 0;
 
 out_free:
+	spin_unlock(&dpamt_lock);
 	free_pamt_array(pamt_pages);
 
 	return ret;
 }
+EXPORT_SYMBOL_FOR_KVM(tdx_pamt_get);
 
-/*
- * Free DPAMT memory for the 2MB aligned region surrounding the
- * given page. Only call this when the pfn is known to already
- * have DPAMT pages in the TDX module for it, and no other pfns
- * in the aligned 2MB physical region still need it.
- *
- * Don't make multiple calls concurrently of __tdx_pamt_get/put(),
- * as there is no protections from races.
- *
- * Effectively it is not (yet) like a refcounted put, and more like a
- * manual manipulation of the DPAMT backing for the 2MB aligned
- * range covered by the pfn.
- */
-static void __tdx_pamt_put(kvm_pfn_t pfn)
+/* Drop DPAMT refcount for the given pfn and free DPAMT backing if needed. */
+void tdx_pamt_put(kvm_pfn_t pfn)
 {
 	struct page *pamt_pages[TDX_DPAMT_ENTRY_PAGE_CNT] = {};
+	atomic_t *dpamt_refcount;
 	u64 tdx_status;
 
 	if (!tdx_supports_dynamic_pamt(&tdx_sysinfo))
 		return;
 
+	dpamt_refcount = tdx_find_dpamt_refcount(pfn);
+
+	spin_lock(&dpamt_lock);
+	/*
+	 * If there is more than 1 reference on the DPAMT entry, don't
+	 * remove it yet. Just decrement the refcount.
+	 */
+	if (atomic_read(dpamt_refcount) > 1) {
+		atomic_dec(dpamt_refcount);
+		goto out_unlock;
+	}
+
+	/* Try to remove the pamt page and take the refcount 1->0. */
 	tdx_status = tdh_phymem_pamt_remove(pfn, pamt_pages);
 
 	/*
@@ -2194,23 +2206,26 @@ static void __tdx_pamt_put(kvm_pfn_t pfn)
 	 * tdh_phymem_pamt_remove() fails.  Don't panic/BUG_ON(), as
 	 * there is no risk of data corruption, but do yell loudly as
 	 * failure indicates a kernel bug, memory is being leaked, and
-	 * the dangling PAMT entry may cause future operations to fail.
+	 * the dangling DPAMT entry may cause future operations to fail.
 	 */
 	if (WARN_ON_ONCE(tdx_status != TDX_SUCCESS))
-		return;
+		goto out_unlock;
 
+	atomic_set(dpamt_refcount, 0);
+	spin_unlock(&dpamt_lock);
 	free_pamt_array(pamt_pages);
+	return;
+out_unlock:
+	spin_unlock(&dpamt_lock);
 }
+EXPORT_SYMBOL_FOR_KVM(tdx_pamt_put);
 
 /*
  * Return a page that can be gifted to the TDX module for use as a "control"
  * page, i.e. pages that are used for control structures for a given TDX
- * guest, and thus obtain TDX protections, including PAMT tracking.
- *
- * This function is currently only safe to call once. And not safe to call
- * if __tdx_pamt_get() is called before or after.
+ * guest, and thus obtain TDX protections, including DPAMT tracking.
  */
-static __maybe_unused struct page *__tdx_alloc_control_page(void)
+struct page *tdx_alloc_control_page(void)
 {
 	struct page *page;
 
@@ -2218,31 +2233,28 @@ static __maybe_unused struct page *__tdx_alloc_control_page(void)
 	if (!page)
 		return NULL;
 
-	if (__tdx_pamt_get(page_to_pfn(page))) {
+	if (tdx_pamt_get(page_to_pfn(page))) {
 		__free_page(page);
 		return NULL;
 	}
 
 	return page;
 }
+EXPORT_SYMBOL_FOR_KVM(tdx_alloc_control_page);
 
 /*
  * Free a page that was gifted to the TDX module for use as a control
  * page. After this, the page is no longer protected by TDX.
- *
- * Like __tdx_pamt_put(), this is currently only safe to call this when
- * a page is already known to have DPAMT pages in the TDX module for
- * it, and no other pages in the aligned 2MB physical region will
- * still need the backing.
  */
-static __maybe_unused void __tdx_free_control_page(struct page *page)
+void tdx_free_control_page(struct page *page)
 {
 	if (!page)
 		return;
 
-	__tdx_pamt_put(page_to_pfn(page));
+	tdx_pamt_put(page_to_pfn(page));
 	__free_page(page);
 }
+EXPORT_SYMBOL_FOR_KVM(tdx_free_control_page);
 
 void tdx_sys_disable(void)
 {
-- 
2.55.0

[PATCH v10 06/11] KVM: TDX: Allocate PAMT memory for TD and vCPU control structures

[Rick Edgecombe]

From: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>

Use control page helpers for allocating and freeing TD control structures,
such that these operations can work for Dynamic PAMT.

The TDX module tracks some state for each page of physical memory that it
might use. It calls this state the PAMT. It includes separate state for
each page size a physical page could be utilized at within the TDX module
(1GB, 2MB, 4KB). In Dynamic PAMT, only the 4KB page size state is
allocated dynamically. So the kernel must ensure PAMT backing is installed
for any 4KB page being gifted to the TDX module, and must tear down the
backing when all associated gifted pages are reclaimed.

TD scoped control pages (TDR, TDCS) and vCPU scoped control pages (TDVPR,
TDCX) are all handed to the TDX module at 4KB page size and are therefore
subject to this requirement. Replace the raw alloc_page()/__free_page()
calls for these pages with tdx_alloc/free_control_page().

Switching between special Dynamic PAMT operations or normal page
alloc/free operations is handled internally in
tdx_alloc/free_control_page(). So don't check for Dynamic PAMT around these
calls. Just call them unconditionally. Similarly, drop the NULL checks
before freeing, as tdx_free_control_page() handles NULL internally.

No functional change intended when DPAMT is not in use.

Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
[sean: handle alloc+free+reclaim in one patch]
Signed-off-by: Sean Christopherson <seanjc@google.com>
[rick: enhance log, reviewing, rebase, with help from AI tooling]
Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Reviewed-by: Binbin Wu <binbin.wu@linux.intel.com>
Reviewed-by: Chao Gao <chao.gao@intel.com>
Reviewed-by: Yan Zhao <yan.y.zhao@intel.com>
Reviewed-by: Tony Lindgren <tony.lindgren@linux.intel.com>
Reviewed-by: Nikolay Borisov <nik.borisov@suse.com>
Reviewed-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Vishal Annapurve <vannapurve@google.com>
Acked-by: Sohil Mehta <sohil.mehta@intel.com>
Acked-by: Sean Christopherson <seanjc@google.com>
---
v7:
 - Fixup tags (Sean)
 - Missing word in log (Binbin)
 - Log smoothness (Yan)
---
 arch/x86/kvm/vmx/tdx.c | 35 ++++++++++++++---------------------
 1 file changed, 14 insertions(+), 21 deletions(-)

diff --git a/arch/x86/kvm/vmx/tdx.c b/arch/x86/kvm/vmx/tdx.c
index b272c20586a74..3592596b5649a 100644
--- a/arch/x86/kvm/vmx/tdx.c
+++ b/arch/x86/kvm/vmx/tdx.c
@@ -362,7 +362,7 @@ static void tdx_reclaim_control_page(struct page *ctrl_page)
 	if (tdx_reclaim_page(ctrl_page))
 		return;
 
-	__free_page(ctrl_page);
+	tdx_free_control_page(ctrl_page);
 }
 
 struct tdx_flush_vp_arg {
@@ -589,7 +589,7 @@ static void tdx_reclaim_td_control_pages(struct kvm *kvm)
 
 	tdx_quirk_reset_paddr(page_to_phys(kvm_tdx->td.tdr_page), PAGE_SIZE);
 
-	__free_page(kvm_tdx->td.tdr_page);
+	tdx_free_control_page(kvm_tdx->td.tdr_page);
 	kvm_tdx->td.tdr_page = NULL;
 }
 
@@ -2456,7 +2456,7 @@ static int __tdx_td_init(struct kvm *kvm, struct td_params *td_params,
 
 	ret = -ENOMEM;
 
-	tdr_page = alloc_page(GFP_KERNEL_ACCOUNT);
+	tdr_page = tdx_alloc_control_page();
 	if (!tdr_page)
 		goto free_hkid;
 
@@ -2469,7 +2469,7 @@ static int __tdx_td_init(struct kvm *kvm, struct td_params *td_params,
 		goto free_tdr;
 
 	for (i = 0; i < kvm_tdx->td.tdcs_nr_pages; i++) {
-		tdcs_pages[i] = alloc_page(GFP_KERNEL_ACCOUNT);
+		tdcs_pages[i] = tdx_alloc_control_page();
 		if (!tdcs_pages[i])
 			goto free_tdcs;
 	}
@@ -2587,10 +2587,8 @@ static int __tdx_td_init(struct kvm *kvm, struct td_params *td_params,
 teardown:
 	/* Only free pages not yet added, so start at 'i' */
 	for (; i < kvm_tdx->td.tdcs_nr_pages; i++) {
-		if (tdcs_pages[i]) {
-			__free_page(tdcs_pages[i]);
-			tdcs_pages[i] = NULL;
-		}
+		tdx_free_control_page(tdcs_pages[i]);
+		tdcs_pages[i] = NULL;
 	}
 	if (!kvm_tdx->td.tdcs_pages)
 		kfree(tdcs_pages);
@@ -2605,16 +2603,13 @@ static int __tdx_td_init(struct kvm *kvm, struct td_params *td_params,
 	free_cpumask_var(packages);
 
 free_tdcs:
-	for (i = 0; i < kvm_tdx->td.tdcs_nr_pages; i++) {
-		if (tdcs_pages[i])
-			__free_page(tdcs_pages[i]);
-	}
+	for (i = 0; i < kvm_tdx->td.tdcs_nr_pages; i++)
+		tdx_free_control_page(tdcs_pages[i]);
 	kfree(tdcs_pages);
 	kvm_tdx->td.tdcs_pages = NULL;
 
 free_tdr:
-	if (tdr_page)
-		__free_page(tdr_page);
+	tdx_free_control_page(tdr_page);
 	kvm_tdx->td.tdr_page = NULL;
 
 free_hkid:
@@ -2948,7 +2943,7 @@ static int tdx_td_vcpu_init(struct kvm_vcpu *vcpu, u64 vcpu_rcx)
 	int ret, i;
 	u64 err;
 
-	page = alloc_page(GFP_KERNEL_ACCOUNT);
+	page = tdx_alloc_control_page();
 	if (!page)
 		return -ENOMEM;
 	tdx->vp.tdvpr_page = page;
@@ -2968,7 +2963,7 @@ static int tdx_td_vcpu_init(struct kvm_vcpu *vcpu, u64 vcpu_rcx)
 	}
 
 	for (i = 0; i < kvm_tdx->td.tdcx_nr_pages; i++) {
-		page = alloc_page(GFP_KERNEL_ACCOUNT);
+		page = tdx_alloc_control_page();
 		if (!page) {
 			ret = -ENOMEM;
 			goto free_tdcx;
@@ -2990,7 +2985,7 @@ static int tdx_td_vcpu_init(struct kvm_vcpu *vcpu, u64 vcpu_rcx)
 			 * method, but the rest are freed here.
 			 */
 			for (; i < kvm_tdx->td.tdcx_nr_pages; i++) {
-				__free_page(tdx->vp.tdcx_pages[i]);
+				tdx_free_control_page(tdx->vp.tdcx_pages[i]);
 				tdx->vp.tdcx_pages[i] = NULL;
 			}
 			return -EIO;
@@ -3018,16 +3013,14 @@ static int tdx_td_vcpu_init(struct kvm_vcpu *vcpu, u64 vcpu_rcx)
 
 free_tdcx:
 	for (i = 0; i < kvm_tdx->td.tdcx_nr_pages; i++) {
-		if (tdx->vp.tdcx_pages[i])
-			__free_page(tdx->vp.tdcx_pages[i]);
+		tdx_free_control_page(tdx->vp.tdcx_pages[i]);
 		tdx->vp.tdcx_pages[i] = NULL;
 	}
 	kfree(tdx->vp.tdcx_pages);
 	tdx->vp.tdcx_pages = NULL;
 
 free_tdvpr:
-	if (tdx->vp.tdvpr_page)
-		__free_page(tdx->vp.tdvpr_page);
+	tdx_free_control_page(tdx->vp.tdvpr_page);
 	tdx->vp.tdvpr_page = NULL;
 	tdx->vp.tdvpr_pa = 0;
 
-- 
2.55.0

[PATCH v10 07/11] x86/virt/tdx: Add APIs to support Dynamic PAMT ops from KVM's fault path

[Rick Edgecombe]

When handling an EPT violation, KVM holds a spinlock while manipulating
the EPT. Before entering the spinlock it doesn't know how many EPT page
tables will need to be installed or whether a huge page will be used. For
this reason it allocates a worst case number of page tables that it might
need as part of servicing the EPT violation.

Under Dynamic PAMT (DPAMT) these pre-allocated pages will potentially need
to have DPAMT backing pages installed for them. KVM already has helpers to
manage topping up page caches before taking the MMU lock, but they cannot
be passed from KVM to arch/x86 code.

The problem of how and when to install the DPAMT backing pages for the
pages given to the TDX module during the fault path has had a lot of
design attempts.
 - Extracting KVM's MMU caches requires too much inlined code added to
   headers.
 - A few varieties of installing DPAMT backing when allocating the
   S-EPT page tables. (see links)
 - Using mempool_t to transfer the pages between KVM and arch/x86 doesn't
   work because the component is designed more around maintaining a pool
   of pages, rather than topping up a continually drained cache.

So don't do these as they all had various problems. Instead just create a
small simple data structure to use for handing a pre-allocated list of
pages between KVM and arch/x86 code. Model this on KVM's existing MMU
memory caches.

Add a tdx_pamt_cache arg to tdx_pamt_get() so it can draw pages from a
cache when needed. Not all DPAMT page installations will happen under
spinlock, for example TD and vCPU scoped control pages. So have
tdx_pamt_get() maintain the existing behavior of allocating from the page
allocator when NULL is passed for the struct tdx_pamt_cache arg. This
prevents excess allocations for cases where it can be avoided.

Export the new helpers for KVM.

AI was used under supervision to review code and workshop logs.

Co-developed-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Reviewed-by: Kiryl Shutsemau (Meta) <kas@kernel.org>
Reviewed-by: Binbin Wu <binbin.wu@linux.intel.com>
Reviewed-by: Chao Gao <chao.gao@intel.com>
Reviewed-by: Yan Zhao <yan.y.zhao@intel.com>
Reviewed-by: Tony Lindgren <tony.lindgren@linux.intel.com>
Reviewed-by: Nikolay Borisov <nik.borisov@suse.com>
Reviewed-by: Dave Hansen <dave.hansen@linux.intel.com>
Acked-by: Sohil Mehta <sohil.mehta@intel.com>
Link: https://lore.kernel.org/kvm/aXENNKjAKTM9UJNH@google.com/
Link: https://lore.kernel.org/kvm/20260129011517.3545883-20-seanjc@google.com/
Link: https://lore.kernel.org/kvm/aYW5CbUvZrLogsWF@yzhao56-desk.sh.intel.com/
---
v10:
 - Change "Dynamic PAMT" to "DPAMT" in comments, and in the second
   reference in the logs. (Dave)
 - x86/tdx -> x86/virt/tdx in title to match the others (AI nit checker)

v7:
 - Log/comment tweaks (Yan)
 - Drop Assisted-by tag and cover AI use in log (Dave)

v6:
 - Filled out log from Sean's series
---
 arch/x86/include/asm/tdx.h  | 16 +++++++++-
 arch/x86/virt/vmx/tdx/tdx.c | 61 ++++++++++++++++++++++++++++++++++---
 2 files changed, 71 insertions(+), 6 deletions(-)

diff --git a/arch/x86/include/asm/tdx.h b/arch/x86/include/asm/tdx.h
index f7442ad20e46d..8c7839d61296d 100644
--- a/arch/x86/include/asm/tdx.h
+++ b/arch/x86/include/asm/tdx.h
@@ -120,7 +120,21 @@ static inline bool tdx_supports_runtime_update(const struct tdx_sys_info *sysinf
 
 bool tdx_supports_dynamic_pamt(const struct tdx_sys_info *sysinfo);
 
-int tdx_pamt_get(kvm_pfn_t pfn);
+/* Simple structure for pre-allocating DPAMT pages outside of spinlocks. */
+struct tdx_pamt_cache {
+	struct list_head page_list;
+	int cnt;
+};
+
+static inline void tdx_init_pamt_cache(struct tdx_pamt_cache *cache)
+{
+	INIT_LIST_HEAD(&cache->page_list);
+	cache->cnt = 0;
+}
+
+void tdx_free_pamt_cache(struct tdx_pamt_cache *cache);
+int tdx_topup_pamt_cache(struct tdx_pamt_cache *cache, unsigned long npages);
+int tdx_pamt_get(kvm_pfn_t pfn, struct tdx_pamt_cache *cache);
 void tdx_pamt_put(kvm_pfn_t pfn);
 
 int tdx_guest_keyid_alloc(void);
diff --git a/arch/x86/virt/vmx/tdx/tdx.c b/arch/x86/virt/vmx/tdx/tdx.c
index c347600a0aabb..39865a2da5822 100644
--- a/arch/x86/virt/vmx/tdx/tdx.c
+++ b/arch/x86/virt/vmx/tdx/tdx.c
@@ -2050,12 +2050,33 @@ bool tdx_supports_dynamic_pamt(const struct tdx_sys_info *sysinfo)
 	return false;
 }
 
-static int alloc_pamt_array(struct page **pamt_pages)
+static struct page *tdx_alloc_page_pamt_cache(struct tdx_pamt_cache *cache)
+{
+	struct page *page;
+
+	page = list_first_entry_or_null(&cache->page_list, struct page, lru);
+	if (page) {
+		list_del(&page->lru);
+		cache->cnt--;
+	}
+
+	return page;
+}
+
+static struct page *alloc_dpamt_page(struct tdx_pamt_cache *cache)
+{
+	if (cache)
+		return tdx_alloc_page_pamt_cache(cache);
+
+	return alloc_page(GFP_KERNEL_ACCOUNT);
+}
+
+static int alloc_pamt_array(struct page **pamt_pages, struct tdx_pamt_cache *cache)
 {
 	int i, j;
 
 	for (i = 0; i < TDX_DPAMT_ENTRY_PAGE_CNT; i++) {
-		pamt_pages[i] = alloc_page(GFP_KERNEL_ACCOUNT);
+		pamt_pages[i] = alloc_dpamt_page(cache);
 		if (!pamt_pages[i])
 			goto err;
 	}
@@ -2132,7 +2153,7 @@ static u64 tdh_phymem_pamt_remove(kvm_pfn_t pfn, struct page **pamt_pages)
 static DEFINE_SPINLOCK(dpamt_lock);
 
 /* Bump DPAMT refcount for the given pfn and allocate DPAMT backing if needed. */
-int tdx_pamt_get(kvm_pfn_t pfn)
+int tdx_pamt_get(kvm_pfn_t pfn, struct tdx_pamt_cache *cache)
 {
 	struct page *pamt_pages[TDX_DPAMT_ENTRY_PAGE_CNT];
 	atomic_t *dpamt_refcount;
@@ -2142,7 +2163,7 @@ int tdx_pamt_get(kvm_pfn_t pfn)
 	if (!tdx_supports_dynamic_pamt(&tdx_sysinfo))
 		return 0;
 
-	ret = alloc_pamt_array(pamt_pages);
+	ret = alloc_pamt_array(pamt_pages, cache);
 	if (ret)
 		return ret;
 
@@ -2220,6 +2241,36 @@ void tdx_pamt_put(kvm_pfn_t pfn)
 }
 EXPORT_SYMBOL_FOR_KVM(tdx_pamt_put);
 
+void tdx_free_pamt_cache(struct tdx_pamt_cache *cache)
+{
+	struct page *page;
+
+	while ((page = tdx_alloc_page_pamt_cache(cache)))
+		__free_page(page);
+}
+EXPORT_SYMBOL_FOR_KVM(tdx_free_pamt_cache);
+
+int tdx_topup_pamt_cache(struct tdx_pamt_cache *cache, unsigned long npages)
+{
+	if (WARN_ON_ONCE(!tdx_supports_dynamic_pamt(&tdx_sysinfo)))
+		return 0;
+
+	npages *= TDX_DPAMT_ENTRY_PAGE_CNT;
+
+	while (cache->cnt < npages) {
+		struct page *page = alloc_page(GFP_KERNEL_ACCOUNT);
+
+		if (!page)
+			return -ENOMEM;
+
+		list_add(&page->lru, &cache->page_list);
+		cache->cnt++;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_FOR_KVM(tdx_topup_pamt_cache);
+
 /*
  * Return a page that can be gifted to the TDX module for use as a "control"
  * page, i.e. pages that are used for control structures for a given TDX
@@ -2233,7 +2284,7 @@ struct page *tdx_alloc_control_page(void)
 	if (!page)
 		return NULL;
 
-	if (tdx_pamt_get(page_to_pfn(page))) {
+	if (tdx_pamt_get(page_to_pfn(page), NULL)) {
 		__free_page(page);
 		return NULL;
 	}
-- 
2.55.0

[PATCH v10 08/11] KVM: TDX: Get/put PAMT pages when (un)mapping private memory

[Rick Edgecombe]

From: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>

Add Dynamic PAMT support to KVM's S-EPT MMU by "getting" a PAMT page when
adding guest memory (PAGE.ADD or PAGE.AUG), and "putting" the page when
removing guest memory (PAGE.REMOVE).

To access the per-vCPU PAMT caches without plumbing @vcpu throughout the
TDP MMU, begrudgingly use kvm_get_running_vcpu() to get the vCPU, and bug
the VM if KVM attempts to set an S-EPT leaf without an active vCPU.  KVM
only supports creating _new_ mappings in page (pre)fault paths, all of
which require an active vCPU.

The PAMT memory holds metadata for TDX protected memory. With Dynamic
PAMT, PAMT_4K is allocated on demand. The kernel supplies the TDX module
with a few pages that cover 2MB of host physical memory.

Releases are balanced via tdx_pamt_put(): every control-page free goes
through tdx_free_control_page(), and guest data pages are put directly on
the successful tdh_mem_page_remove() path and in the
tdx_mem_page_add/aug() error path.

Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Co-developed-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
[rick: enhance log, reviewing, rebase, with help from AI tooling]
Co-developed-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Reviewed-by: Binbin Wu <binbin.wu@linux.intel.com>
Reviewed-by: Tony Lindgren <tony.lindgren@linux.intel.com>
Reviewed-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Vishal Annapurve <vannapurve@google.com>
Acked-by: Sohil Mehta <sohil.mehta@intel.com>
Acked-by: Sean Christopherson <seanjc@google.com>
---
v8:
 - Remove extra new line around *topup_external_cache() (Sean)
 - Make +1/-1 logic clearer by just coding the easily optimizable math.
   (Nikolay, Sean)
 - Do KVM_BUG_ON() on tdx_pamt_get() return. (Sean)

v7:
 - Don't do to_tdx() before NULL check for readability (Binbin, Sean)
 - Fixup tags (Sean)
 - Export tdx_supports_dynamic_pamt() since it's used in KVM here and no
   longer an static inline.
---
 arch/x86/include/asm/kvm-x86-ops.h |  1 +
 arch/x86/include/asm/kvm_host.h    |  1 +
 arch/x86/kvm/mmu/mmu.c             |  4 ++
 arch/x86/kvm/vmx/tdx.c             | 64 ++++++++++++++++++++++++++----
 arch/x86/kvm/vmx/tdx.h             |  2 +
 arch/x86/virt/vmx/tdx/tdx.c        |  1 +
 6 files changed, 65 insertions(+), 8 deletions(-)

diff --git a/arch/x86/include/asm/kvm-x86-ops.h b/arch/x86/include/asm/kvm-x86-ops.h
index e213c9ae3e301..5c358c40eae8a 100644
--- a/arch/x86/include/asm/kvm-x86-ops.h
+++ b/arch/x86/include/asm/kvm-x86-ops.h
@@ -99,6 +99,7 @@ KVM_X86_OP_OPTIONAL_RET0(tdp_has_smep)
 KVM_X86_OP(load_mmu_pgd)
 KVM_X86_OP_OPTIONAL_RET0(set_external_spte)
 KVM_X86_OP_OPTIONAL(free_external_spt)
+KVM_X86_OP_OPTIONAL_RET0(topup_external_cache)
 KVM_X86_OP(has_wbinvd_exit)
 KVM_X86_OP(get_l2_tsc_offset)
 KVM_X86_OP(get_l2_tsc_multiplier)
diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index 683bb8bf43a94..57d37491c7c09 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -1644,6 +1644,7 @@ struct kvm_x86_ops {
 	/* Update external page tables for page table about to be freed. */
 	void (*free_external_spt)(struct kvm *kvm, struct kvm_mmu_page *sp);
 
+	int (*topup_external_cache)(struct kvm_vcpu *vcpu, int min_nr_spts);
 
 	bool (*has_wbinvd_exit)(void);
 
diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index 064ecc33b9267..2816861ca279c 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -617,6 +617,10 @@ static int mmu_topup_memory_caches(struct kvm_vcpu *vcpu, bool maybe_indirect)
 					       PT64_ROOT_MAX_LEVEL);
 		if (r)
 			return r;
+
+		r = kvm_x86_call(topup_external_cache)(vcpu, PT64_ROOT_MAX_LEVEL);
+		if (r)
+			return r;
 	}
 	r = kvm_mmu_topup_memory_cache(&vcpu->arch.mmu_shadow_page_cache,
 				       PT64_ROOT_MAX_LEVEL);
diff --git a/arch/x86/kvm/vmx/tdx.c b/arch/x86/kvm/vmx/tdx.c
index 3592596b5649a..7c79f4d4bf99a 100644
--- a/arch/x86/kvm/vmx/tdx.c
+++ b/arch/x86/kvm/vmx/tdx.c
@@ -681,6 +681,8 @@ int tdx_vcpu_create(struct kvm_vcpu *vcpu)
 	if (!irqchip_split(vcpu->kvm))
 		return -EINVAL;
 
+	tdx_init_pamt_cache(&tdx->pamt_cache);
+
 	fpstate_set_confidential(&vcpu->arch.guest_fpu);
 	vcpu->arch.apic->guest_apic_protected = true;
 	INIT_LIST_HEAD(&tdx->vt.pi_wakeup_list);
@@ -866,6 +868,8 @@ void tdx_vcpu_free(struct kvm_vcpu *vcpu)
 	struct vcpu_tdx *tdx = to_tdx(vcpu);
 	int i;
 
+	tdx_free_pamt_cache(&tdx->pamt_cache);
+
 	if (vcpu->cpu != -1) {
 		KVM_BUG_ON(tdx->state == VCPU_TD_STATE_INITIALIZED, vcpu->kvm);
 		tdx_flush_vp_on_cpu(vcpu);
@@ -1618,6 +1622,17 @@ void tdx_load_mmu_pgd(struct kvm_vcpu *vcpu, hpa_t root_hpa, int pgd_level)
 	td_vmcs_write64(to_tdx(vcpu), SHARED_EPT_POINTER, root_hpa);
 }
 
+static int tdx_topup_external_pamt_cache(struct kvm_vcpu *vcpu, int min_nr_spts)
+{
+	/*
+	 * Minus one page to exclude the root SPT, but plus one page for a
+	 * possible 4KB private mapping.
+	 */
+	min_nr_spts += -1 + 1;
+
+	return tdx_topup_pamt_cache(&to_tdx(vcpu)->pamt_cache, min_nr_spts);
+}
+
 static int tdx_mem_page_add(struct kvm *kvm, gfn_t gfn, enum pg_level level,
 			    kvm_pfn_t pfn)
 {
@@ -1676,16 +1691,28 @@ static struct page *tdx_spte_to_sept_pt(struct kvm *kvm, gfn_t gfn,
 static int tdx_sept_map_nonleaf_spte(struct kvm *kvm, gfn_t gfn,
 				     enum pg_level level, u64 new_spte)
 {
+	struct kvm_vcpu *vcpu = kvm_get_running_vcpu();
 	gpa_t gpa = gfn_to_gpa(gfn);
 	u64 err, entry, level_state;
 	struct page *sept_pt;
+	int ret;
+
+	if (KVM_BUG_ON(!vcpu, kvm))
+		return -EIO;
 
 	sept_pt = tdx_spte_to_sept_pt(kvm, gfn, new_spte, level);
 	if (!sept_pt)
 		return -EIO;
 
+	ret = tdx_pamt_get(page_to_pfn(sept_pt), &to_tdx(vcpu)->pamt_cache);
+	if (KVM_BUG_ON(ret, kvm))
+		return ret;
+
 	err = tdh_mem_sept_add(&to_kvm_tdx(kvm)->td, gpa, level, sept_pt,
 			       &entry, &level_state);
+	if (err)
+		tdx_pamt_put(page_to_pfn(sept_pt));
+
 	if (unlikely(tdx_operand_busy(err)))
 		return -EBUSY;
 
@@ -1698,8 +1725,13 @@ static int tdx_sept_map_nonleaf_spte(struct kvm *kvm, gfn_t gfn,
 static int tdx_sept_map_leaf_spte(struct kvm *kvm, gfn_t gfn, enum pg_level level,
 				  u64 new_spte)
 {
+	struct kvm_vcpu *vcpu = kvm_get_running_vcpu();
 	struct kvm_tdx *kvm_tdx = to_kvm_tdx(kvm);
 	kvm_pfn_t pfn = spte_to_pfn(new_spte);
+	int ret;
+
+	if (KVM_BUG_ON(!vcpu, kvm))
+		return -EIO;
 
 	/* TODO: handle large pages. */
 	if (KVM_BUG_ON(level != PG_LEVEL_4K, kvm))
@@ -1707,6 +1739,10 @@ static int tdx_sept_map_leaf_spte(struct kvm *kvm, gfn_t gfn, enum pg_level leve
 
 	WARN_ON_ONCE((new_spte & VMX_EPT_RWX_MASK) != VMX_EPT_RWX_MASK);
 
+	ret = tdx_pamt_get(pfn, &to_tdx(vcpu)->pamt_cache);
+	if (KVM_BUG_ON(ret, kvm))
+		return ret;
+
 	/*
 	 * Ensure pre_fault_allowed is read by kvm_arch_vcpu_pre_fault_memory()
 	 * before kvm_tdx->state.  Userspace must not be allowed to pre-fault
@@ -1719,10 +1755,15 @@ static int tdx_sept_map_leaf_spte(struct kvm *kvm, gfn_t gfn, enum pg_level leve
 	 * If the TD isn't finalized/runnable, then userspace is initializing
 	 * the VM image via KVM_TDX_INIT_MEM_REGION; ADD the page to the TD.
 	 */
-	if (unlikely(kvm_tdx->state != TD_STATE_RUNNABLE))
-		return tdx_mem_page_add(kvm, gfn, level, pfn);
+	if (likely(kvm_tdx->state == TD_STATE_RUNNABLE))
+		ret = tdx_mem_page_aug(kvm, gfn, level, pfn);
+	else
+		ret = tdx_mem_page_add(kvm, gfn, level, pfn);
 
-	return tdx_mem_page_aug(kvm, gfn, level, pfn);
+	if (ret)
+		tdx_pamt_put(pfn);
+
+	return ret;
 }
 
 /*
@@ -1819,6 +1860,7 @@ static int tdx_sept_remove_leaf_spte(struct kvm *kvm, gfn_t gfn,
 		return -EIO;
 
 	tdx_quirk_reset_paddr(PFN_PHYS(pfn), PAGE_SIZE);
+	tdx_pamt_put(pfn);
 	return 0;
 }
 
@@ -1862,6 +1904,8 @@ static int tdx_sept_set_private_spte(struct kvm *kvm, gfn_t gfn, u64 old_spte,
  */
 static void tdx_sept_free_private_spt(struct kvm *kvm, struct kvm_mmu_page *sp)
 {
+	struct page *sept_pt = virt_to_page(sp->external_spt);
+
 	/*
 	 * KVM doesn't (yet) zap page table pages in mirror page table while
 	 * TD is active, though guest pages mapped in mirror page table could be
@@ -1875,15 +1919,15 @@ static void tdx_sept_free_private_spt(struct kvm *kvm, struct kvm_mmu_page *sp)
 	 * the page to prevent the kernel from accessing the encrypted page.
 	 */
 	if (KVM_BUG_ON(is_hkid_assigned(to_kvm_tdx(kvm)), kvm) ||
-	    tdx_reclaim_page(virt_to_page(sp->external_spt)))
+	    tdx_reclaim_page(sept_pt))
 		goto out;
 
 	/*
-	 * Immediately free the S-EPT page because RCU-time free is unnecessary
-	 * after TDH.PHYMEM.PAGE.RECLAIM ensures there are no outstanding
-	 * readers.
+	 * Immediately free the S-EPT page as the TDX subsystem doesn't support
+	 * freeing pages from RCU callbacks, and more importantly because
+	 * TDH.PHYMEM.PAGE.RECLAIM ensures there are no outstanding readers.
 	 */
-	free_page((unsigned long)sp->external_spt);
+	tdx_free_control_page(sept_pt);
 out:
 	sp->external_spt = NULL;
 }
@@ -3480,6 +3524,10 @@ int __init tdx_hardware_setup(void)
 
 	vt_x86_ops.set_external_spte = tdx_sept_set_private_spte;
 	vt_x86_ops.free_external_spt = tdx_sept_free_private_spt;
+
+	if (tdx_supports_dynamic_pamt(tdx_sysinfo))
+		vt_x86_ops.topup_external_cache = tdx_topup_external_pamt_cache;
+
 	vt_x86_ops.protected_apic_has_interrupt = tdx_protected_apic_has_interrupt;
 	return 0;
 
diff --git a/arch/x86/kvm/vmx/tdx.h b/arch/x86/kvm/vmx/tdx.h
index ac8323a68b163..fd368e3ee060b 100644
--- a/arch/x86/kvm/vmx/tdx.h
+++ b/arch/x86/kvm/vmx/tdx.h
@@ -72,6 +72,8 @@ struct vcpu_tdx {
 
 	u64 map_gpa_next;
 	u64 map_gpa_end;
+
+	struct tdx_pamt_cache pamt_cache;
 };
 
 void tdh_vp_rd_failed(struct vcpu_tdx *tdx, char *uclass, u32 field, u64 err);
diff --git a/arch/x86/virt/vmx/tdx/tdx.c b/arch/x86/virt/vmx/tdx/tdx.c
index 39865a2da5822..ff00ee6d5705a 100644
--- a/arch/x86/virt/vmx/tdx/tdx.c
+++ b/arch/x86/virt/vmx/tdx/tdx.c
@@ -2049,6 +2049,7 @@ bool tdx_supports_dynamic_pamt(const struct tdx_sys_info *sysinfo)
 	/* To be enabled when kernel is ready. */
 	return false;
 }
+EXPORT_SYMBOL_FOR_KVM(tdx_supports_dynamic_pamt);
 
 static struct page *tdx_alloc_page_pamt_cache(struct tdx_pamt_cache *cache)
 {
-- 
2.55.0

[PATCH v10 09/11] x86/virt/tdx: Enable Dynamic PAMT

[Rick Edgecombe]

The Physical Address Metadata Table (PAMT) holds TDX metadata for
physical memory and must be allocated by the kernel during TDX module
initialization. Dynamic PAMT (DPAMT) is a TDX module feature that can
reduce this memory use by allocating part of the PAMT dynamically.

The TDX module exposes whether DPAMT is supported via a bit in the
'features0' metadata. Unfortunately, the TDX module exposes the feature as
supported even when it does not support using it with the number of keyids
currently configured in the BIOS. Since no TDX modules exist today with
that issue fixed, make the feature default off to prevent users from
upgrading their kernel and encountering TDX erroring out when trying to
enable DPAMT.

For the decision of whether to make it a boot time option and/or compile
time option, consider that DPAMT's memory savings are significant enough
to make it a good default configuration. That is most TDX users should
want it unless they have strange keyid configurations.

The feature increases the kernel size by 2KB (when TDX is configured in the
build).

All pieces are in place to enable DPAMT if it is supported and the user
passes a kernel parameter.

AI was used to review code. It made a style suggestion.

Based on a patch originally by Kiryl Shutsemau.

Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Reviewed-by: Tony Lindgren <tony.lindgren@linux.intel.com>
Reviewed-by: Vishal Annapurve <vannapurve@google.com>
Acked-by: Sohil Mehta <sohil.mehta@intel.com>
---
v10:
 - Change "Dynamic PAMT" to "DPAMT" at the second reference in the
   logs. (Dave)
 - Adjust tdx_supports_dynamic_pamt() to check multiple conditions in a
   more tip style. (AI nit checker)
v8:
 - Order tdx_dpamt in kernel-parameters.txt (Sohil)
 - Make tdx_enable_dpamt static (Sashiko)

v7:
 - Add kernel parameter following some twists and turns, deriving
   originally from a comment by (Chao)
---
 .../admin-guide/kernel-parameters.txt         |  7 ++++++
 arch/x86/include/asm/tdx.h                    |  1 +
 arch/x86/virt/vmx/tdx/tdx.c                   | 23 +++++++++++++++++--
 3 files changed, 29 insertions(+), 2 deletions(-)

diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
index 68647ff4bdd24..f32aa58f744a1 100644
--- a/Documentation/admin-guide/kernel-parameters.txt
+++ b/Documentation/admin-guide/kernel-parameters.txt
@@ -7582,6 +7582,13 @@ Kernel parameters
 
 	tdfx=		[HW,DRM]
 
+	tdx_dpamt=
+			[X86] Controls whether TDX will use Dynamic PAMT
+			to save memory, when supported.
+
+			Valid parameters: "on", "off"
+			Default: "off"
+
 	test_suspend=	[SUSPEND]
 			Format: { "mem" | "standby" | "freeze" }[,N]
 			Specify "mem" (for Suspend-to-RAM) or "standby" (for
diff --git a/arch/x86/include/asm/tdx.h b/arch/x86/include/asm/tdx.h
index 8c7839d61296d..e186dfe5bf885 100644
--- a/arch/x86/include/asm/tdx.h
+++ b/arch/x86/include/asm/tdx.h
@@ -36,6 +36,7 @@
 /* Bit definitions of TDX_FEATURES0 metadata field */
 #define TDX_FEATURES0_TD_PRESERVING	BIT_ULL(1)
 #define TDX_FEATURES0_NO_RBP_MOD	BIT_ULL(18)
+#define TDX_FEATURES0_DYNAMIC_PAMT	BIT_ULL(36)
 
 #ifndef __ASSEMBLER__
 
diff --git a/arch/x86/virt/vmx/tdx/tdx.c b/arch/x86/virt/vmx/tdx/tdx.c
index ff00ee6d5705a..3daa8c63f9c51 100644
--- a/arch/x86/virt/vmx/tdx/tdx.c
+++ b/arch/x86/virt/vmx/tdx/tdx.c
@@ -50,6 +50,8 @@
 /* Number of DPAMT pages to be provided to TDX module per 2MB region of PA */
 #define TDX_DPAMT_ENTRY_PAGE_CNT 2
 
+static bool tdx_enable_dpamt __ro_after_init;
+
 struct tdx_module_state {
 	bool initialized;
 	bool sysinit_done;
@@ -1031,6 +1033,8 @@ static __init int construct_tdmrs(struct list_head *tmb_list,
 	return ret;
 }
 
+#define TDX_SYS_CONFIG_DYNAMIC_PAMT	BIT(16)
+
 static __init int config_tdx_module(struct tdmr_info_list *tdmr_list,
 				    u64 global_keyid)
 {
@@ -1059,6 +1063,12 @@ static __init int config_tdx_module(struct tdmr_info_list *tdmr_list,
 	args.rcx = __pa(tdmr_pa_array);
 	args.rdx = tdmr_list->nr_consumed_tdmrs;
 	args.r8 = global_keyid;
+
+	if (tdx_supports_dynamic_pamt(&tdx_sysinfo)) {
+		pr_info("Enable Dynamic PAMT\n");
+		args.r8 |= TDX_SYS_CONFIG_DYNAMIC_PAMT;
+	}
+
 	ret = seamcall_prerr(TDH_SYS_CONFIG, &args);
 
 	/* Free the array as it is not required anymore. */
@@ -2046,8 +2056,10 @@ EXPORT_SYMBOL_FOR_KVM(tdh_phymem_page_wbinvd_hkid);
 
 bool tdx_supports_dynamic_pamt(const struct tdx_sys_info *sysinfo)
 {
-	/* To be enabled when kernel is ready. */
-	return false;
+	if (!tdx_enable_dpamt)
+		return false;
+
+	return sysinfo->features.tdx_features0 & TDX_FEATURES0_DYNAMIC_PAMT;
 }
 EXPORT_SYMBOL_FOR_KVM(tdx_supports_dynamic_pamt);
 
@@ -2308,6 +2320,13 @@ void tdx_free_control_page(struct page *page)
 }
 EXPORT_SYMBOL_FOR_KVM(tdx_free_control_page);
 
+static int __init tdx_dpamt_setup(char *str)
+{
+	return kstrtobool(str, &tdx_enable_dpamt) == 0;
+}
+
+__setup("tdx_dpamt=", tdx_dpamt_setup);
+
 void tdx_sys_disable(void)
 {
 	struct tdx_module_args args = {};
-- 
2.55.0

[PATCH v10 10/11] Documentation/x86: Add documentation for TDX's Dynamic PAMT

[Rick Edgecombe]

Expand TDX documentation to include information on the Dynamic PAMT
feature.

The new section explains PAMT support in the TDX module and how Dynamic
PAMT affects the kernel memory use.

Some of the docs verbiage was provided by Dave Hansen (see link). AI was
used under supervision to review the docs.

Based on a patch originally by Kiryl Shutsemau.

Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Reviewed-by: Binbin Wu <binbin.wu@linux.intel.com>
Reviewed-by: Tony Lindgren <tony.lindgren@linux.intel.com>
Acked-by: Sohil Mehta <sohil.mehta@intel.com>
Link: https://lore.kernel.org/all/8e40862e-891a-4cad-8bb8-06ad25ad6061@intel.com/
---
v10:
 - Updated verbiage from Dave, but unify 4k to 4KB to be consistent with
   the rest of the series.

v7:
 - Spell out PAMT acronym (Binbin)
 - Drop Assisted-by tag and cover AI use in log (Dave)
 - Add info about kernel parameter

v6:
 - Add missing word (Binbin)
 - Use "::" instead of ":"
 - Make format of dmesg example accurate
---
 .../admin-guide/kernel-parameters.txt         |  3 +++
 Documentation/arch/x86/tdx.rst                | 27 +++++++++++++++++++
 2 files changed, 30 insertions(+)

diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
index f32aa58f744a1..e7558b8382acf 100644
--- a/Documentation/admin-guide/kernel-parameters.txt
+++ b/Documentation/admin-guide/kernel-parameters.txt
@@ -7589,6 +7589,9 @@ Kernel parameters
 			Valid parameters: "on", "off"
 			Default: "off"
 
+			For details see:
+			Documentation/arch/x86/tdx.rst
+
 	test_suspend=	[SUSPEND]
 			Format: { "mem" | "standby" | "freeze" }[,N]
 			Specify "mem" (for Suspend-to-RAM) or "standby" (for
diff --git a/Documentation/arch/x86/tdx.rst b/Documentation/arch/x86/tdx.rst
index 3303499ad4c6f..de09967ce8d8c 100644
--- a/Documentation/arch/x86/tdx.rst
+++ b/Documentation/arch/x86/tdx.rst
@@ -200,6 +200,33 @@ reflects the TCB of the currently running TDX module and therefore
 changes after an update. By contrast, TEE_TCB_SVN reflects the TCB at TD
 launch time and is not affected.
 
+Dynamic PAMT
+------------
+
+The Physical Address Metadata Table (PAMT) is metadata in which the TDX
+module keeps data about each physical page (think struct page). Space
+for it is allocated by the VMM, consumes up to about 0.4% of system
+memory and needs to be supplied to the TDX module when the TDX module is
+first loaded.
+
+Dynamic PAMT is an add-on feature that allows a VMM to dynamically
+allocate the part of the PAMT which tracks 4KB pages. This reduces the
+amount of memory that TDX consumes while TDs are not in use.
+
+When Dynamic PAMT is in use, dmesg shows it like::
+
+  [..] virt/tdx: Enable Dynamic PAMT
+  [..] virt/tdx: 10092 KB allocated for PAMT
+  [..] virt/tdx: TDX-Module initialized
+
+Dynamic PAMT is only enabled when supported and the ``tdx_dpamt=`` kernel
+parameter is set to "on". The feature is off by default because TDX module
+internal details prevent Dynamic PAMT from working on all keyid partitioning
+configurations. When the TDX module is fixed to include these constraints in
+its enumeration of Dynamic PAMT support, kernel support can be changed to
+default on. For more information, consult the Intel TDX documentation about
+Dynamic PAMT.
+
 TDX Interaction to Other Kernel Components
 ------------------------------------------
 
-- 
2.55.0

[PATCH v10 11/11] x86/virt/tdx: Optimize tdx_pamt_get/put()

[Rick Edgecombe]

The Dynamic PAMT (DPAMT) get/put helpers use a global spinlock to
serialize all refcount updates and SEAMCALL invocations. This gives correct
behavior for concurrent callers, but leads to contention. It is especially
bad from the KVM side, which is designed to allow faulting in EPT under a
shared lock. With the global spinlock, not only is the lock an exclusive
one, but it is for all TDs instead of just a single one.

But taking the global lock each time is actually unnecessary. Only the 0->1
and 1->0 refcount transitions actually need the lock (to pair with
SEAMCALLs that actually add and remove with the DPAMT pages). The common
case of incrementing or decrementing a non-zero refcount can be done
locklessly.

So create a fast and slow path. Check the refcount outside the lock and
only take it for the slow path (0->1 and 1->0 transitions).

On the put side make the refcount adjustment and lock taking atomic so if
a 'get' happens between them, it doesn't cause the DPAMT to be freed
incorrectly. On the get side there is no technique for doing the refcount
adjustment and lock atomically, so check the refcount again inside the
lock.

AI was used under supervision to collect/apply feedback, review code and
workshop logs. It assisted in identifying/evaluating the stale
conditionals for the races resolved from the atomic_dec_and_lock() change.
Separate from atomic_dec_and_lock() fallout, it suggested to change
atomic_inc() to atomic_set(pamt_refcount, 1) in the put error path for the
sake of being more precise, which Kiryl had also suggested in the past.
The model also suggested updated comments following the
atomic_dec_and_lock() change based on some directed prompting. The
comments were subsequently edited or further prompted for fine tuning.

Based on a patch originally by Kiryl Shutsemau.

Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Reviewed-by: Chao Gao <chao.gao@intel.com>
Reviewed-by: Tony Lindgren <tony.lindgren@linux.intel.com>
Reviewed-by: Nikolay Borisov <nik.borisov@suse.com>
Reviewed-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Vishal Annapurve <vannapurve@google.com>
Acked-by: Sohil Mehta <sohil.mehta@intel.com>
---
v10:
 - Change "Dynamic PAMT" to "DPAMT" in comments. (Dave)

v7:
 - Drop Assisted-by tag and cover AI use in log. (Dave)
 - Move to end of the series.
 - Use atomic_inc_not_zero() in this patch inside the spin_lock(), as
   suggested on the non-optimized patch by (Dave).

v6:
 - Fix "tdx_pamt_add()" typo to "tdx_pamt_get()" in lost-race comment
 - Fix error path bug: set ret = -EIO and use WARN_ON_ONCE() instead of
   pr_err() for unexpected PAMT.ADD failures (Sean)
 - Use "set the refcount 0->1" wording to match atomic_set() usage
 - Wrap comments to 80 columns
 - Switch to atomic_dec_and_lock() and remove handling of races that are
   no longer needed as a result. Adjust comments as appropriate. (Dave)
 - Adjustments from dropping error helper patches
---
 arch/x86/virt/vmx/tdx/tdx.c | 44 ++++++++++++++++++++++++-------------
 1 file changed, 29 insertions(+), 15 deletions(-)

diff --git a/arch/x86/virt/vmx/tdx/tdx.c b/arch/x86/virt/vmx/tdx/tdx.c
index 3daa8c63f9c51..2acab6e5f5f87 100644
--- a/arch/x86/virt/vmx/tdx/tdx.c
+++ b/arch/x86/virt/vmx/tdx/tdx.c
@@ -2176,28 +2176,41 @@ int tdx_pamt_get(kvm_pfn_t pfn, struct tdx_pamt_cache *cache)
 	if (!tdx_supports_dynamic_pamt(&tdx_sysinfo))
 		return 0;
 
-	ret = alloc_pamt_array(pamt_pages, cache);
-	if (ret)
-		return ret;
-
 	dpamt_refcount = tdx_find_dpamt_refcount(pfn);
 
-	spin_lock(&dpamt_lock);
-
 	/*
 	 * If the DPAMT entry is already added (i.e. refcount >= 1),
 	 * then just increment the refcount.
 	 */
+	if (atomic_inc_not_zero(dpamt_refcount))
+		return 0;
+
+	ret = alloc_pamt_array(pamt_pages, cache);
+	if (ret)
+		return ret;
+
+	spin_lock(&dpamt_lock);
+
+	/*
+	 * Unlike tdx_pamt_put() which uses atomic_dec_and_lock() to
+	 * atomically handle the 1->0 transition, the get side has no
+	 * equivalent combined primitive for 0->1. Recheck under the
+	 * lock since another get may have already done the 0->1
+	 * transition after both saw atomic_inc_not_zero() fail.
+	 */
 	if (atomic_inc_not_zero(dpamt_refcount))
 		goto out_free;
 
-	/* Try to add the PAMT page and take the refcount 0->1. */
 	tdx_status = tdh_phymem_pamt_add(pfn, pamt_pages);
 	if (WARN_ON_ONCE(tdx_status != TDX_SUCCESS)) {
 		ret = -EIO;
 		goto out_free;
 	}
 
+	/*
+	 * The refcount is zero, and this locked path is the
+	 * only way to increase it from 0->1.
+	 */
 	atomic_set(dpamt_refcount, 1);
 	spin_unlock(&dpamt_lock);
 	return 0;
@@ -2222,17 +2235,13 @@ void tdx_pamt_put(kvm_pfn_t pfn)
 
 	dpamt_refcount = tdx_find_dpamt_refcount(pfn);
 
-	spin_lock(&dpamt_lock);
 	/*
 	 * If there is more than 1 reference on the DPAMT entry, don't
 	 * remove it yet. Just decrement the refcount.
 	 */
-	if (atomic_read(dpamt_refcount) > 1) {
-		atomic_dec(dpamt_refcount);
-		goto out_unlock;
-	}
+	if (!atomic_dec_and_lock(dpamt_refcount, &dpamt_lock))
+		return;
 
-	/* Try to remove the pamt page and take the refcount 1->0. */
 	tdx_status = tdh_phymem_pamt_remove(pfn, pamt_pages);
 
 	/*
@@ -2242,10 +2251,15 @@ void tdx_pamt_put(kvm_pfn_t pfn)
 	 * failure indicates a kernel bug, memory is being leaked, and
 	 * the dangling DPAMT entry may cause future operations to fail.
 	 */
-	if (WARN_ON_ONCE(tdx_status != TDX_SUCCESS))
+	if (WARN_ON_ONCE(tdx_status != TDX_SUCCESS)) {
+		/*
+		 * atomic_dec_and_lock() already decremented it to 0,
+		 * but the DPAMT entry still exists since REMOVE failed.
+		 */
+		atomic_set(dpamt_refcount, 1);
 		goto out_unlock;
+	}
 
-	atomic_set(dpamt_refcount, 0);
 	spin_unlock(&dpamt_lock);
 	free_pamt_array(pamt_pages);
 	return;
-- 
2.55.0

Re: [PATCH v10 03/11] x86/virt/tdx: Add __tdx_pamt_get/put() helpers

[Dave Hansen]

On 9/2/26 18:51, Rick Edgecombe wrote:
> Create __tdx_pamt_get/put() to handle installing DPAMT 4KB backing for
> pages that are already allocated (such as KVM's use of S-EPT page tables
> or guest private memory). Have them take a pfn instead of a struct page,
> as future changes will want to use these helpers for guest pages which are
> tracked by PFN.

Looks fine with the v10 changes:

Reviewed-by: Dave Hansen <dave.hansen@linux.intel.com>

Re: [PATCH v10 04/11] x86/virt/tdx: Allocate refcounts for Dynamic PAMT memory

[Dave Hansen]

On 9/2/26 18:51, Rick Edgecombe wrote:
> @@ -2159,7 +2210,7 @@ static void __tdx_pamt_put(kvm_pfn_t pfn)
>   * This function is currently only safe to call once. And not safe to call
>   * if __tdx_pamt_get() is called before or after.
>   */
> -static struct page * __maybe_unused __tdx_alloc_control_page(void)
> +static __maybe_unused struct page *__tdx_alloc_control_page(void)
>  {
>  	struct page *page;
>  
> @@ -2184,7 +2235,7 @@ static struct page * __maybe_unused __tdx_alloc_control_page(void)
>   * it, and no other pages in the aligned 2MB physical region will
>   * still need the backing.
>   */
> -static void __maybe_unused __tdx_free_control_page(struct page *page)
> +static __maybe_unused void __tdx_free_control_page(struct page *page)
>  {

What's with all the __maybe_unused order flipping?

Re: [PATCH v10 09/11] x86/virt/tdx: Enable Dynamic PAMT

[Dave Hansen]

This patch is doing too much.

Let's just replace it with a patch that enables DPAMT when it's
supported by the module. I can apply that.

Then, we can talk about how a disable command-line knob would work.

Re: [PATCH v10 10/11] Documentation/x86: Add documentation for TDX's Dynamic PAMT

[Dave Hansen]

On 9/2/26 18:51, Rick Edgecombe wrote:
> +Dynamic PAMT is only enabled when supported and the ``tdx_dpamt=`` kernel
> +parameter is set to "on". The feature is off by default because TDX module
> +internal details prevent Dynamic PAMT from working on all keyid partitioning
> +configurations. When the TDX module is fixed to include these constraints in
> +its enumeration of Dynamic PAMT support, kernel support can be changed to
> +default on. For more information, consult the Intel TDX documentation about
> +Dynamic PAMT.

What's the end user visible effect of hitting one of these troublesome
"keyid partitioning configurations"?

Re: [PATCH v10 04/11] x86/virt/tdx: Allocate refcounts for Dynamic PAMT memory

[Edgecombe, Rick P]

On Thu, 2026-09-03 at 08:30 -0700, Dave Hansen wrote:
> On 9/2/26 18:51, Rick Edgecombe wrote:
> > @@ -2159,7 +2210,7 @@ static void __tdx_pamt_put(kvm_pfn_t pfn)
> >    * This function is currently only safe to call once. And not safe to call
> >    * if __tdx_pamt_get() is called before or after.
> >    */
> > -static struct page * __maybe_unused __tdx_alloc_control_page(void)
> > +static __maybe_unused struct page *__tdx_alloc_control_page(void)
> >   {
> >   	struct page *page;
> >   
> > @@ -2184,7 +2235,7 @@ static struct page * __maybe_unused __tdx_alloc_control_page(void)
> >    * it, and no other pages in the aligned 2MB physical region will
> >    * still need the backing.
> >    */
> > -static void __maybe_unused __tdx_free_control_page(struct page *page)
> > +static __maybe_unused void __tdx_free_control_page(struct page *page)
> >   {
> 
> What's with all the __maybe_unused order flipping?

Oops, it was applied to the wrong patch. Should be the previous one.

While applying some other changes in the middle of the series, I got build
warning on the order of the attribute and the rest of the signature being wrong.
Somehow the CI bisectability test missed it. Then I applied the change to the
wrong patch.

Re: [PATCH v10 10/11] Documentation/x86: Add documentation for TDX's Dynamic PAMT

[Edgecombe, Rick P]

On Thu, 2026-09-03 at 08:47 -0700, Dave Hansen wrote:
> On 9/2/26 18:51, Rick Edgecombe wrote:
> > +Dynamic PAMT is only enabled when supported and the ``tdx_dpamt=`` kernel
> > +parameter is set to "on". The feature is off by default because TDX module
> > +internal details prevent Dynamic PAMT from working on all keyid partitioning
> > +configurations. When the TDX module is fixed to include these constraints in
> > +its enumeration of Dynamic PAMT support, kernel support can be changed to
> > +default on. For more information, consult the Intel TDX documentation about
> > +Dynamic PAMT.
> 
> What's the end user visible effect of hitting one of these troublesome
> "keyid partitioning configurations"?

DPAMT gets turned on via the TDH_SYS_CONFIG SEAMCALL in config_tdx_module().
Inside the TDX module, this SEAMCALL does some checks about the number of
configured keyids, and will fail the call if it doesn't like the configured
keyids. For a server with 52 bit physical address width, it should require at
least 16 hkids (4 bits). If a lower number of bits is set, TDH_SYS_CONFIG will
fail and it will cascade into a TDX init failure.

The TDX module exposes if DPAMT is supported via a features0 bit. If DPAMT is
not supported due to the hkids limits, the module still reports it as supported.
The module folks agreed to change the behavior of this bit to say if DPAMT is
actually supported (i.e. do the keyid bit checks itself).

So we have:
1. If you are inclined to think support bits should actually report whether
something is supported, then this kinda is a TDX module bug that we are working
around by making DPAMT an opt-in.
2. The bug will only be hit if anyone has configured a low number of hkids.

We could optimistically think the TDX module with the fix will be distributed
before anyone with this configuration hits it? And if that is wrong, add the
keyid checking or opt-in as a fix. Then maybe we save some code long term.

Re: [PATCH v10 10/11] Documentation/x86: Add documentation for TDX's Dynamic PAMT

[Dave Hansen]

On 9/3/26 12:31, Edgecombe, Rick P wrote:
> So we have:
> 1. If you are inclined to think support bits should actually report whether
> something is supported, then this kinda is a TDX module bug that we are working
> around by making DPAMT an opt-in.
> 2. The bug will only be hit if anyone has configured a low number of hkids.
> 
> We could optimistically think the TDX module with the fix will be distributed
> before anyone with this configuration hits it? And if that is wrong, add the
> keyid checking or opt-in as a fix. Then maybe we save some code long term.

How about we just make the kernel tolerate DPAMT enabling failures. If a
user makes this weird hkid config, then they'll get temporarily higher
memory consumption until they fix their config or get a new TDX module.

That seems way better than trying to work around this with opt-ins or
permanent documentation for a temporary problem.

Re: [PATCH v10 10/11] Documentation/x86: Add documentation for TDX's Dynamic PAMT

[Edgecombe, Rick P]

On Thu, 2026-09-03 at 12:36 -0700, Dave Hansen wrote:
> On 9/3/26 12:31, Edgecombe, Rick P wrote:
> > So we have:
> > 1. If you are inclined to think support bits should actually report whether
> > something is supported, then this kinda is a TDX module bug that we are working
> > around by making DPAMT an opt-in.
> > 2. The bug will only be hit if anyone has configured a low number of hkids.
> > 
> > We could optimistically think the TDX module with the fix will be distributed
> > before anyone with this configuration hits it? And if that is wrong, add the
> > keyid checking or opt-in as a fix. Then maybe we save some code long term.
> 
> How about we just make the kernel tolerate DPAMT enabling failures. If a
> user makes this weird hkid config, then they'll get temporarily higher
> memory consumption until they fix their config or get a new TDX module.
> 
> That seems way better than trying to work around this with opt-ins or
> permanent documentation for a temporary problem.

Tolerating failure would be a bit ugly. It is not a dedicated operation to
enable Dynamic PAMT. It happens along with a bunch of other stuff in
TDH_SYS_CONFIG. So you would need to try once, then make an guess that maybe it
failed because of Dynamic PAMT, and try again without Dynamic PAMT.

This ugly hack might be a bit better to get the same behavior? Basically do the
check TDX module wanted VMMs to do, but in a more convenient way for
the kernel:https://lore.kernel.org/lkml/20260526023515.288829-11-rick.p.edgecombe@intel.com/

Maybe add a:
/*
 * Work around bug in TDX module where Dynamic PAMT support is reported 
 * incorrectly. Remove someday when the modules are fixed.
 */

Re: [PATCH v10 10/11] Documentation/x86: Add documentation for TDX's Dynamic PAMT

[Dave Hansen]

On 9/3/26 13:39, Edgecombe, Rick P wrote:
> This ugly hack might be a bit better to get the same behavior? Basically do the
> check TDX module wanted VMMs to do, but in a more convenient way for
> the kernel:https://lore.kernel.org/lkml/20260526023515.288829-11-
> rick.p.edgecombe@intel.com/

OK, so no good options for kernel fixes. That leaves us with:

 1. The TDX module gets fixed
 2. Users learn not to set incompatible configs

Neither one of which is the kernel's problem, honestly. There's only so
much we can do.

Re: [PATCH v10 05/11] x86/virt/tdx: Handle multiple callers in tdx_pamt_get/put()

[Edgecombe, Rick P]

On Thu, 2026-09-03 at 02:03 +0000, sashiko-bot@kernel.org wrote:
> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> - [High] tdx_find_dpamt_refcount() uses an unchecked PFN to index the dpamt_refcounts array, leading to out-of-bounds memory access.
> --
> 
> commit 413476a1747fac53ffdaf87030c42cb8b25bc668
> Author: Rick Edgecombe <rick.p.edgecombe@intel.com>
> 
> x86/virt/tdx: Handle multiple callers in tdx_pamt_get/put()
> 
> This patch refcounts the dynamic PAMT additions for 2MB regions so multiple
> callers can safely share them. It also drops the "__" prefix from the
> helpers and exports them for use by KVM.
> 
> > diff --git a/arch/x86/virt/vmx/tdx/tdx.c b/arch/x86/virt/vmx/tdx/tdx.c
> > index e8bdeacd989d6..a126e395f5f6a 100644
> > --- a/arch/x86/virt/vmx/tdx/tdx.c
> > +++ b/arch/x86/virt/vmx/tdx/tdx.c
> > @@ -292,7 +292,7 @@ static __init void free_dpamt_refcounts(void)
> >   	dpamt_refcounts = NULL;
> >   }
> >   
> > -static __maybe_unused atomic_t *tdx_find_dpamt_refcount(unsigned long pfn)
> > +static atomic_t *tdx_find_dpamt_refcount(unsigned long pfn)
> >   {
> >   	/* Find which PMD a PFN is in. */
> >   	unsigned long index = pfn >> (PMD_SHIFT - PAGE_SHIFT);
> > 
> >   	return &dpamt_refcounts[index];
> 
> [Severity: High]
> Does this index calculation need bounds checking?
> 
> If memory is hotplugged after boot, increasing the global max_pfn, and KVM
> allocates private guest memory that lands in this hotplugged region (e.g.,
> via guest_memfd), the returned PFN could be greater than the boot-time
> max_pfn used to size the dpamt_refcounts array. This would lead to an
> out-of-bounds pointer being returned here.

TDX's support for memory hotplug is limited. It will reject onlining memory that
was not present when setting up TDX initially to prevent it from being used for
TDX.

> 
> [ ... ]
> 
> > @@ -2150,41 +2144,59 @@ static int __tdx_pamt_get(kvm_pfn_t pfn)
> [ ... ]
> >   out_free:
> > +	spin_unlock(&dpamt_lock);
> >   	free_pamt_array(pamt_pages);
> >   
> >   	return ret;
> >   }
> > +EXPORT_SYMBOL_FOR_KVM(tdx_pamt_get);
> 
> [Severity: High]
> Is it safe to export this function without adding bounds checking to
> tdx_find_dpamt_refcount()?
> 
> Exporting this helper allows external callers like KVM to pass in arbitrary
> or hotplugged PFNs. If a hotplugged PFN is passed, the resulting
> out-of-bounds pointer from tdx_find_dpamt_refcount() will be modified by
> atomic_inc_not_zero() and atomic_set() within this function, potentially
> causing memory corruption or a host kernel panic.

This is not a real functional issue unless something passes an invalid pfn in. 

In the past there was an optimization to only populate the refcounts where there
was actually memory (i.e. leave gaps in the allocation). But now with the
current code, we could more reasonably warn by just checking max_pfn.
Although... I guess it still wouldn't warn about invalid PFNs, just refcount
bounds.

I think I'll leave it alone at this point unless somebody shouts. I don't know
about bounds checking, especially for the put side where there is no return
code. But it could be cheap thing to warn about.