lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20250530103941.11092-2-tao.wangtao@honor.com>
Date: Fri, 30 May 2025 18:39:38 +0800
From: wangtao <tao.wangtao@...or.com>
To: <sumit.semwal@...aro.org>, <christian.koenig@....com>,
	<kraxel@...hat.com>, <vivek.kasireddy@...el.com>, <viro@...iv.linux.org.uk>,
	<brauner@...nel.org>, <hughd@...gle.com>, <akpm@...ux-foundation.org>,
	<amir73il@...il.com>
CC: <benjamin.gaignard@...labora.com>, <Brian.Starkey@....com>,
	<jstultz@...gle.com>, <tjmercier@...gle.com>, <jack@...e.cz>,
	<baolin.wang@...ux.alibaba.com>, <linux-media@...r.kernel.org>,
	<dri-devel@...ts.freedesktop.org>, <linaro-mm-sig@...ts.linaro.org>,
	<linux-kernel@...r.kernel.org>, <linux-fsdevel@...r.kernel.org>,
	<linux-mm@...ck.org>, <bintian.wang@...or.com>, <yipengxiang@...or.com>,
	<liulu.liu@...or.com>, <feng.han@...or.com>, wangtao <tao.wangtao@...or.com>
Subject: [PATCH v3 1/4] fs: allow cross-FS copy_file_range for memory-backed files

Memory-backed files can optimize copy performance via
copy_file_range callbacks. Compared to mmap&read: reduces
GUP (get_user_pages) overhead; vs sendfile/splice: eliminates
one memory copy; supports dmabuf zero-copy implementation.

Signed-off-by: wangtao <tao.wangtao@...or.com>
---
 fs/read_write.c    | 71 +++++++++++++++++++++++++++++++++-------------
 include/linux/fs.h |  2 ++
 2 files changed, 54 insertions(+), 19 deletions(-)

diff --git a/fs/read_write.c b/fs/read_write.c
index bb0ed26a0b3a..591c6db7b785 100644
--- a/fs/read_write.c
+++ b/fs/read_write.c
@@ -1469,6 +1469,20 @@ COMPAT_SYSCALL_DEFINE4(sendfile64, int, out_fd, int, in_fd,
 }
 #endif
 
+static inline bool is_copy_memory_file_to_file(struct file *file_in,
+				struct file *file_out)
+{
+	return (file_in->f_op->fop_flags & FOP_MEMORY_FILE) &&
+		file_in->f_op->copy_file_range && file_out->f_op->write_iter;
+}
+
+static inline bool is_copy_file_to_memory_file(struct file *file_in,
+				struct file *file_out)
+{
+	return (file_out->f_op->fop_flags & FOP_MEMORY_FILE) &&
+		file_in->f_op->read_iter && file_out->f_op->copy_file_range;
+}
+
 /*
  * Performs necessary checks before doing a file copy
  *
@@ -1484,11 +1498,23 @@ static int generic_copy_file_checks(struct file *file_in, loff_t pos_in,
 	struct inode *inode_out = file_inode(file_out);
 	uint64_t count = *req_count;
 	loff_t size_in;
+	bool splice = flags & COPY_FILE_SPLICE;
+	bool has_memory_file;
 	int ret;
 
-	ret = generic_file_rw_checks(file_in, file_out);
-	if (ret)
-		return ret;
+	/* Skip generic checks, allow cross-sb copies for dma-buf/tmpfs */
+	has_memory_file = is_copy_memory_file_to_file(file_in, file_out) ||
+			  is_copy_file_to_memory_file(file_in, file_out);
+	if (!splice && has_memory_file) {
+		if (!(file_in->f_mode & FMODE_READ) ||
+		    !(file_out->f_mode & FMODE_WRITE) ||
+		    (file_out->f_flags & O_APPEND))
+			return -EBADF;
+	} else {
+		ret = generic_file_rw_checks(file_in, file_out);
+		if (ret)
+			return ret;
+	}
 
 	/*
 	 * We allow some filesystems to handle cross sb copy, but passing
@@ -1500,7 +1526,7 @@ static int generic_copy_file_checks(struct file *file_in, loff_t pos_in,
 	 * and several different sets of file_operations, but they all end up
 	 * using the same ->copy_file_range() function pointer.
 	 */
-	if (flags & COPY_FILE_SPLICE) {
+	if (splice || has_memory_file) {
 		/* cross sb splice is allowed */
 	} else if (file_out->f_op->copy_file_range) {
 		if (file_in->f_op->copy_file_range !=
@@ -1581,23 +1607,30 @@ ssize_t vfs_copy_file_range(struct file *file_in, loff_t pos_in,
 	 * same sb using clone, but for filesystems where both clone and copy
 	 * are supported (e.g. nfs,cifs), we only call the copy method.
 	 */
-	if (!splice && file_out->f_op->copy_file_range) {
-		ret = file_out->f_op->copy_file_range(file_in, pos_in,
-						      file_out, pos_out,
-						      len, flags);
-	} else if (!splice && file_in->f_op->remap_file_range && samesb) {
-		ret = file_in->f_op->remap_file_range(file_in, pos_in,
-				file_out, pos_out,
-				min_t(loff_t, MAX_RW_COUNT, len),
-				REMAP_FILE_CAN_SHORTEN);
-		/* fallback to splice */
-		if (ret <= 0)
+	if (!splice) {
+		if (is_copy_memory_file_to_file(file_in, file_out)) {
+			ret = file_in->f_op->copy_file_range(file_in, pos_in,
+					file_out, pos_out, len, flags);
+		} else if (is_copy_file_to_memory_file(file_in, file_out)) {
+			ret = file_out->f_op->copy_file_range(file_in, pos_in,
+					file_out, pos_out, len, flags);
+		} else if (file_out->f_op->copy_file_range) {
+			ret = file_out->f_op->copy_file_range(file_in, pos_in,
+							file_out, pos_out,
+							len, flags);
+		} else if (file_in->f_op->remap_file_range && samesb) {
+			ret = file_in->f_op->remap_file_range(file_in, pos_in,
+					file_out, pos_out,
+					min_t(loff_t, MAX_RW_COUNT, len),
+					REMAP_FILE_CAN_SHORTEN);
+			/* fallback to splice */
+			if (ret <= 0)
+				splice = true;
+		} else if (samesb) {
+			/* Fallback to splice for same sb copy for backward compat */
 			splice = true;
-	} else if (samesb) {
-		/* Fallback to splice for same sb copy for backward compat */
-		splice = true;
+		}
 	}
-
 	file_end_write(file_out);
 
 	if (!splice)
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 016b0fe1536e..37df1b497418 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2187,6 +2187,8 @@ struct file_operations {
 #define FOP_ASYNC_LOCK		((__force fop_flags_t)(1 << 6))
 /* File system supports uncached read/write buffered IO */
 #define FOP_DONTCACHE		((__force fop_flags_t)(1 << 7))
+/* Supports cross-FS copy_file_range for memory file */
+#define FOP_MEMORY_FILE		((__force fop_flags_t)(1 << 8))
 
 /* Wrap a directory iterator that needs exclusive inode access */
 int wrap_directory_iterator(struct file *, struct dir_context *,
-- 
2.17.1


Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ