Received-SPF: pass (google.com: domain of linux-kernel+bounces-177155-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
From: Baolin Wang <baolin.wang@linux.alibaba.com>
To: akpm@linux-foundation.org,
	hughd@google.com
Cc: willy@infradead.org,
	david@redhat.com,
	ioworker0@gmail.com,
	wangkefeng.wang@huawei.com,
	ying.huang@intel.com,
	21cnbao@gmail.com,
	ryan.roberts@arm.com,
	shy828301@gmail.com,
	ziy@nvidia.com,
	baolin.wang@linux.alibaba.com,
	linux-mm@kvack.org,
	linux-kernel@vger.kernel.org
Subject: [PATCH v2 6/7] mm: shmem: add mTHP size alignment in shmem_get_unmapped_area
Date: Mon, 13 May 2024 13:08:16 +0800
Message-Id: <664d35cfa7192adb28c133efffa81b0947f60d67.1715571279.git.baolin.wang@linux.alibaba.com>
In-Reply-To: <cover.1715571279.git.baolin.wang@linux.alibaba.com>
References: <cover.1715571279.git.baolin.wang@linux.alibaba.com>
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit

Although the top-level hugepage allocation can be turned off, anonymous shmem
can still use mTHP by configuring the sysfs interface located at
'/sys/kernel/mm/transparent_hugepage/hugepage-XXkb/shmem_enabled'. Therefore,
add alignment for mTHP size to provide a suitable alignment address in
shmem_get_unmapped_area().

Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Tested-by: Lance Yang <ioworker0@gmail.com>
---
 mm/shmem.c | 36 +++++++++++++++++++++++++++---------
 1 file changed, 27 insertions(+), 9 deletions(-)

diff --git a/mm/shmem.c b/mm/shmem.c
index b50ddf013e37..8b020ff09c72 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -2404,6 +2404,7 @@ unsigned long shmem_get_unmapped_area(struct file *file,
 	unsigned long inflated_len;
 	unsigned long inflated_addr;
 	unsigned long inflated_offset;
+	unsigned long hpage_size;
 
 	if (len > TASK_SIZE)
 		return -ENOMEM;
@@ -2422,8 +2423,6 @@ unsigned long shmem_get_unmapped_area(struct file *file,
 
 	if (shmem_huge == SHMEM_HUGE_DENY)
 		return addr;
-	if (len < HPAGE_PMD_SIZE)
-		return addr;
 	if (flags & MAP_FIXED)
 		return addr;
 	/*
@@ -2435,8 +2434,11 @@ unsigned long shmem_get_unmapped_area(struct file *file,
 	if (uaddr == addr)
 		return addr;
 
+	hpage_size = HPAGE_PMD_SIZE;
 	if (shmem_huge != SHMEM_HUGE_FORCE) {
 		struct super_block *sb;
+		unsigned long __maybe_unused hpage_orders;
+		int order = 0;
 
 		if (file) {
 			VM_BUG_ON(file->f_op != &shmem_file_operations);
@@ -2449,18 +2451,34 @@ unsigned long shmem_get_unmapped_area(struct file *file,
 			if (IS_ERR(shm_mnt))
 				return addr;
 			sb = shm_mnt->mnt_sb;
+
+			/*
+			 * Find the highest mTHP order used for anonymous shmem to
+			 * provide a suitable alignment address.
+			 */
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+			hpage_orders = READ_ONCE(huge_anon_shmem_orders_always);
+			hpage_orders |= READ_ONCE(huge_anon_shmem_orders_within_size);
+			hpage_orders |= READ_ONCE(huge_anon_shmem_orders_madvise);
+			hpage_orders |= READ_ONCE(huge_anon_shmem_orders_inherit);
+			order = highest_order(hpage_orders);
+			hpage_size = PAGE_SIZE << order;
+#endif
 		}
-		if (SHMEM_SB(sb)->huge == SHMEM_HUGE_NEVER)
+		if (SHMEM_SB(sb)->huge == SHMEM_HUGE_NEVER && !order)
 			return addr;
 	}
 
-	offset = (pgoff << PAGE_SHIFT) & (HPAGE_PMD_SIZE-1);
-	if (offset && offset + len < 2 * HPAGE_PMD_SIZE)
+	if (len < hpage_size)
+		return addr;
+
+	offset = (pgoff << PAGE_SHIFT) & (hpage_size - 1);
+	if (offset && offset + len < 2 * hpage_size)
 		return addr;
-	if ((addr & (HPAGE_PMD_SIZE-1)) == offset)
+	if ((addr & (hpage_size - 1)) == offset)
 		return addr;
 
-	inflated_len = len + HPAGE_PMD_SIZE - PAGE_SIZE;
+	inflated_len = len + hpage_size - PAGE_SIZE;
 	if (inflated_len > TASK_SIZE)
 		return addr;
 	if (inflated_len < len)
@@ -2473,10 +2491,10 @@ unsigned long shmem_get_unmapped_area(struct file *file,
 	if (inflated_addr & ~PAGE_MASK)
 		return addr;
 
-	inflated_offset = inflated_addr & (HPAGE_PMD_SIZE-1);
+	inflated_offset = inflated_addr & (hpage_size - 1);
 	inflated_addr += offset - inflated_offset;
 	if (inflated_offset > offset)
-		inflated_addr += HPAGE_PMD_SIZE;
+		inflated_addr += hpage_size;
 
 	if (inflated_addr > TASK_SIZE - len)
 		return addr;
-- 
2.39.3