Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Qi Zheng <zhengqi.arch@bytedance.com>
To:     akpm@linux-foundation.org, david@redhat.com,
        kirill.shutemov@linux.intel.com, mika.penttila@nextfour.com,
        jgg@nvidia.com, tglx@linutronix.de, willy@infradead.org
Cc:     linux-kernel@vger.kernel.org, linux-mm@kvack.org,
        muchun.song@linux.dev, Qi Zheng <zhengqi.arch@bytedance.com>
Subject: [RFC PATCH 1/7] mm: use ptep_clear() in non-present cases
Date:   Thu, 25 Aug 2022 18:10:31 +0800
Message-Id: <20220825101037.96517-2-zhengqi.arch@bytedance.com>
In-Reply-To: <20220825101037.96517-1-zhengqi.arch@bytedance.com>
References: <20220825101037.96517-1-zhengqi.arch@bytedance.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

After commit 08d5b29eac7d ("mm: ptep_clear() page table helper"),
the ptep_clear() can be used to track the clearing of PTE entries,
but it skips some places since the page table check does not care
about non-present PTE entries.

Subsequent patches need to use ptep_clear() to track all clearing
PTE entries, so this patch makes ptep_clear() used for all cases
including clearing non-present PTE entries.

Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
---
 include/linux/pgtable.h | 2 +-
 mm/memory.c             | 2 +-
 mm/mprotect.c           | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/include/linux/pgtable.h b/include/linux/pgtable.h
index 3cdc16cfd867..9745684b0cdb 100644
--- a/include/linux/pgtable.h
+++ b/include/linux/pgtable.h
@@ -428,7 +428,7 @@ static inline void pte_clear_not_present_full(struct mm_struct *mm,
 					      pte_t *ptep,
 					      int full)
 {
-	pte_clear(mm, address, ptep);
+	ptep_clear(mm, address, ptep);
 }
 #endif
 
diff --git a/mm/memory.c b/mm/memory.c
index 1c6027adc542..207e0ee657e9 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -3655,7 +3655,7 @@ static vm_fault_t pte_marker_clear(struct vm_fault *vmf)
 	 * none pte.  Otherwise it means the pte could have changed, so retry.
 	 */
 	if (is_pte_marker(*vmf->pte))
-		pte_clear(vmf->vma->vm_mm, vmf->address, vmf->pte);
+		ptep_clear(vmf->vma->vm_mm, vmf->address, vmf->pte);
 	pte_unmap_unlock(vmf->pte, vmf->ptl);
 	return 0;
 }
diff --git a/mm/mprotect.c b/mm/mprotect.c
index ba5592655ee3..1a01bd22a4ed 100644
--- a/mm/mprotect.c
+++ b/mm/mprotect.c
@@ -201,7 +201,7 @@ static unsigned long change_pte_range(struct mmu_gather *tlb,
 				 * fault will trigger without uffd trapping.
 				 */
 				if (uffd_wp_resolve) {
-					pte_clear(vma->vm_mm, addr, pte);
+					ptep_clear(vma->vm_mm, addr, pte);
 					pages++;
 				}
 				continue;
-- 
2.20.1