Linux Documentation
 help / color / mirror / Atom feed
From: Rick Edgecombe <rick.p.edgecombe@intel.com>
To: bp@alien8.de, dave.hansen@intel.com, hpa@zytor.com,
	kas@kernel.org, kvm@vger.kernel.org, linux-coco@lists.linux.dev,
	linux-doc@vger.kernel.org, linux-kernel@vger.kernel.org,
	mingo@redhat.com, nik.borisov@suse.com, pbonzini@redhat.com,
	seanjc@google.com, tglx@kernel.org, vannapurve@google.com,
	x86@kernel.org, chao.gao@intel.com, yan.y.zhao@intel.com,
	kai.huang@intel.com, tony.lindgren@linux.intel.com,
	binbin.wu@intel.com, sohil.mehta@intel.com
Cc: rick.p.edgecombe@intel.com,
	Hongyu Ning <hongyu.ning@linux.intel.com>,
	Dave Hansen <dave.hansen@linux.intel.com>
Subject: [PATCH v10 11/11] x86/virt/tdx: Optimize tdx_pamt_get/put()
Date: Wed,  2 Sep 2026 18:51:13 -0700	[thread overview]
Message-ID: <20260903015113.93343-12-rick.p.edgecombe@intel.com> (raw)
In-Reply-To: <20260903015113.93343-1-rick.p.edgecombe@intel.com>

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


      parent reply	other threads:[~2026-09-03  1:51 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-03  1:51 [PATCH v10 00/11] Dynamic PAMT Rick Edgecombe
2026-09-03  1:51 ` [PATCH v10 01/11] x86/virt/tdx: Simplify PAMT layout calculation Rick Edgecombe
2026-09-03  1:51 ` [PATCH v10 02/11] x86/virt/tdx: Allocate page bitmap for Dynamic PAMT Rick Edgecombe
2026-09-03  1:51 ` [PATCH v10 03/11] x86/virt/tdx: Add __tdx_pamt_get/put() helpers Rick Edgecombe
2026-09-03 15:28   ` Dave Hansen
2026-09-03  1:51 ` [PATCH v10 04/11] x86/virt/tdx: Allocate refcounts for Dynamic PAMT memory Rick Edgecombe
2026-09-03 15:30   ` Dave Hansen
2026-09-03 18:33     ` Edgecombe, Rick P
2026-09-03  1:51 ` [PATCH v10 05/11] x86/virt/tdx: Handle multiple callers in tdx_pamt_get/put() Rick Edgecombe
     [not found]   ` <20260903020303.349961F000E9@smtp.kernel.org>
2026-09-03 23:16     ` Edgecombe, Rick P
2026-09-03  1:51 ` [PATCH v10 06/11] KVM: TDX: Allocate PAMT memory for TD and vCPU control structures Rick Edgecombe
2026-09-03  1:51 ` [PATCH v10 07/11] x86/virt/tdx: Add APIs to support Dynamic PAMT ops from KVM's fault path Rick Edgecombe
2026-09-03  1:51 ` [PATCH v10 08/11] KVM: TDX: Get/put PAMT pages when (un)mapping private memory Rick Edgecombe
2026-09-03  1:51 ` [PATCH v10 09/11] x86/virt/tdx: Enable Dynamic PAMT Rick Edgecombe
2026-09-03 15:38   ` Dave Hansen
2026-09-03  1:51 ` [PATCH v10 10/11] Documentation/x86: Add documentation for TDX's " Rick Edgecombe
2026-09-03 15:47   ` Dave Hansen
2026-09-03 19:31     ` Edgecombe, Rick P
2026-09-03 19:36       ` Dave Hansen
2026-09-03 20:39         ` Edgecombe, Rick P
2026-09-03 20:45           ` Dave Hansen
2026-09-03  1:51 ` Rick Edgecombe [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260903015113.93343-12-rick.p.edgecombe@intel.com \
    --to=rick.p.edgecombe@intel.com \
    --cc=binbin.wu@intel.com \
    --cc=bp@alien8.de \
    --cc=chao.gao@intel.com \
    --cc=dave.hansen@intel.com \
    --cc=dave.hansen@linux.intel.com \
    --cc=hongyu.ning@linux.intel.com \
    --cc=hpa@zytor.com \
    --cc=kai.huang@intel.com \
    --cc=kas@kernel.org \
    --cc=kvm@vger.kernel.org \
    --cc=linux-coco@lists.linux.dev \
    --cc=linux-doc@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=nik.borisov@suse.com \
    --cc=pbonzini@redhat.com \
    --cc=seanjc@google.com \
    --cc=sohil.mehta@intel.com \
    --cc=tglx@kernel.org \
    --cc=tony.lindgren@linux.intel.com \
    --cc=vannapurve@google.com \
    --cc=x86@kernel.org \
    --cc=yan.y.zhao@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox