From: Dmitry Monakhov <dmonakhov@openvz.org>
To: linux-kernel@vger.kernel.org
Cc: linux-fsdevel@vger.kernel.org, viro@zeniv.linux.org.uk,
Dmitry Monakhov <dmonakhov@openvz.org>
Subject: [PATCH 15/16] pipe: use is_xxx_kiocb instead of filp->fl_flags
Date: Sat, 4 Apr 2015 23:13:24 +0400 [thread overview]
Message-ID: <1428174805-853-16-git-send-email-dmonakhov@openvz.org> (raw)
In-Reply-To: <1428174805-853-1-git-send-email-dmonakhov@openvz.org>
Also fix other long standing issues caused by fcntl(,F_SETFL,):
- One can disable O_DIRECT for pipe[1] (paketized IO), but can not enable it again.
- Currently we do not set O_APPEND on pipe[1] (IMHO it is wrong, but let it be)
so it is reasonable to completely prohibit change O_APPEND flag on both
end's of pipe. Add ->check_flags method in order to diallow O_APPEND toggling.
Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
---
fs/fcntl.c | 6 ++++--
fs/pipe.c | 20 +++++++++++++++-----
2 files changed, 19 insertions(+), 7 deletions(-)
diff --git a/fs/fcntl.c b/fs/fcntl.c
index ee85cd4..0bdc9c7 100644
--- a/fs/fcntl.c
+++ b/fs/fcntl.c
@@ -51,9 +51,11 @@ static int setfl(int fd, struct file * filp, unsigned long arg)
if (arg & O_NDELAY)
arg |= O_NONBLOCK;
+ /* allowed only for inodes with ->direct_io method or write pipe */
if (arg & O_DIRECT) {
- if (!filp->f_mapping || !filp->f_mapping->a_ops ||
- !filp->f_mapping->a_ops->direct_IO)
+ if ((!filp->f_mapping || !filp->f_mapping->a_ops ||
+ !filp->f_mapping->a_ops->direct_IO) &&
+ !(get_pipe_info(filp) && (filp->f_flags | O_WRONLY)))
return -EINVAL;
}
diff --git a/fs/pipe.c b/fs/pipe.c
index 2d084f2..95b5fe4 100644
--- a/fs/pipe.c
+++ b/fs/pipe.c
@@ -301,7 +301,7 @@ pipe_read(struct kiocb *iocb, struct iov_iter *to)
*/
if (ret)
break;
- if (filp->f_flags & O_NONBLOCK) {
+ if (is_nonblock_kiocb(iocb)) {
ret = -EAGAIN;
break;
}
@@ -329,9 +329,9 @@ pipe_read(struct kiocb *iocb, struct iov_iter *to)
return ret;
}
-static inline int is_packetized(struct file *file)
+static inline int is_packetized(struct kiocb *kiocb)
{
- return (file->f_flags & O_DIRECT) != 0;
+ return is_direct_kiocb(kiocb);
}
static ssize_t
@@ -427,7 +427,7 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from)
buf->offset = 0;
buf->len = copied;
buf->flags = 0;
- if (is_packetized(filp)) {
+ if (is_packetized(iocb)) {
buf->ops = &packet_pipe_buf_ops;
buf->flags = PIPE_BUF_FLAG_PACKET;
}
@@ -439,7 +439,7 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from)
}
if (bufs < pipe->buffers)
continue;
- if (filp->f_flags & O_NONBLOCK) {
+ if (is_nonblock_kiocb(iocb)) {
if (!ret)
ret = -EAGAIN;
break;
@@ -943,6 +943,15 @@ err:
return ret;
}
+/* XXX: Currently it is not possible distinguish read side from write one */
+static int pipe_check_flags(int flags)
+{
+ if (flags & O_APPEND)
+ return -EINVAL;
+
+ return 0;
+}
+
const struct file_operations pipefifo_fops = {
.open = fifo_open,
.llseek = no_llseek,
@@ -954,6 +963,7 @@ const struct file_operations pipefifo_fops = {
.unlocked_ioctl = pipe_ioctl,
.release = pipe_release,
.fasync = pipe_fasync,
+ .check_flags = pipe_check_flags,
};
/*
--
1.7.1
next prev parent reply other threads:[~2015-04-04 19:13 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-04-04 19:13 [PATCH 00/16] fs: fixup racy check file->f_flags for xxx_write_iter Dmitry Monakhov
2015-04-04 19:13 ` [PATCH 01/16] fs: save file->f_flags to kiocb->ki_flags Dmitry Monakhov
2015-04-04 19:13 ` [PATCH 02/16] vfs: check kiocb->ki_flags instead filp->fl_flags Dmitry Monakhov
2015-04-04 21:36 ` Al Viro
2015-04-05 11:03 ` Dmitry Monakhov
2015-04-05 18:11 ` Al Viro
2015-04-05 21:54 ` Al Viro
2015-04-04 19:13 ` [PATCH 03/16] ext4: use is_xxx_kiocb instead of filp->fl_flags Dmitry Monakhov
2015-04-04 19:13 ` [PATCH 04/16] 9p: " Dmitry Monakhov
2015-04-04 19:13 ` [PATCH 05/16] btrfs: " Dmitry Monakhov
2015-04-04 19:13 ` [PATCH 06/16] ceph: " Dmitry Monakhov
2015-04-04 19:13 ` [PATCH 07/16] cifs: " Dmitry Monakhov
2015-04-04 19:13 ` [PATCH 08/16] gfs2: " Dmitry Monakhov
2015-04-07 13:11 ` [Cluster-devel] " Steven Whitehouse
2015-04-04 19:13 ` [PATCH 09/16] nfs: " Dmitry Monakhov
2015-04-04 19:13 ` [PATCH 10/16] ntfs: " Dmitry Monakhov
2015-04-04 19:13 ` [PATCH 11/16] ocfs2: " Dmitry Monakhov
2015-04-04 19:13 ` [PATCH 12/16] udf: " Dmitry Monakhov
2015-04-04 19:13 ` [PATCH 13/16] xfs: " Dmitry Monakhov
2015-04-04 19:13 ` [PATCH 14/16] fuse: " Dmitry Monakhov
2015-04-04 19:13 ` Dmitry Monakhov [this message]
2015-04-04 19:13 ` [PATCH 16/16] splice: fix race beween splice_write vs fcntl(,F_SETFL,) Dmitry Monakhov
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=1428174805-853-16-git-send-email-dmonakhov@openvz.org \
--to=dmonakhov@openvz.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=viro@zeniv.linux.org.uk \
/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;
as well as URLs for NNTP newsgroup(s).