linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: hch <hch@lst.de>
To: Christian Brauner <brauner@kernel.org>
Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com>,
	"viro@zeniv.linux.org.uk" <viro@zeniv.linux.org.uk>,
	"axboe@kernel.dk" <axboe@kernel.dk>,
	"djwong@kernel.org" <djwong@kernel.org>,
	"ebiggers@google.com" <ebiggers@google.com>,
	"linux-fsdevel@vger.kernel.org" <linux-fsdevel@vger.kernel.org>,
	"linux-block@vger.kernel.org" <linux-block@vger.kernel.org>,
	Xiao Ni <xni@redhat.com>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	"Rafael J. Wysocki" <rafael@kernel.org>,
	Danilo Krummrich <dakr@kernel.org>
Subject: Re: [PATCH] fs: move the bdex_statx call to vfs_getattr_nosec
Date: Tue, 22 Apr 2025 16:15:05 +0200	[thread overview]
Message-ID: <20250422141505.GA25426@lst.de> (raw)
In-Reply-To: <20250422081736.GA674@lst.de>

Turns out this doesn't work.  We used to have the request_mask, but it
got removed in 25fbcd62d2e1 ("bdev: use bdev_io_min() for statx block
size") so that stat can expose the block device min I/O size in
st_blkdev, and as the blksize doesn't have it's own request_mask flag
is hard to special case.

So maybe the better question is why devtmpfs even calls into
vfs_getattr?  As far as I can tell handle_remove is only ever called on
the actual devtmpfs file system, so we don't need to go through the
VFS to query i_mode.  i.e. the patch should also fix the issue.  The
modify_change is probably not needed either, but for now I'm aiming
for the minimal fix.

diff --git a/drivers/base/devtmpfs.c b/drivers/base/devtmpfs.c
index 6dd1a8860f1c..53fb0829eb7b 100644
--- a/drivers/base/devtmpfs.c
+++ b/drivers/base/devtmpfs.c
@@ -296,7 +296,7 @@ static int delete_path(const char *nodepath)
 	return err;
 }
 
-static int dev_mynode(struct device *dev, struct inode *inode, struct kstat *stat)
+static int dev_mynode(struct device *dev, struct inode *inode)
 {
 	/* did we create it */
 	if (inode->i_private != &thread)
@@ -304,13 +304,13 @@ static int dev_mynode(struct device *dev, struct inode *inode, struct kstat *sta
 
 	/* does the dev_t match */
 	if (is_blockdev(dev)) {
-		if (!S_ISBLK(stat->mode))
+		if (!S_ISBLK(inode->i_mode))
 			return 0;
 	} else {
-		if (!S_ISCHR(stat->mode))
+		if (!S_ISCHR(inode->i_mode))
 			return 0;
 	}
-	if (stat->rdev != dev->devt)
+	if (inode->i_rdev != dev->devt)
 		return 0;
 
 	/* ours */
@@ -321,8 +321,7 @@ static int handle_remove(const char *nodename, struct device *dev)
 {
 	struct path parent;
 	struct dentry *dentry;
-	struct kstat stat;
-	struct path p;
+	struct inode *inode;
 	int deleted = 0;
 	int err;
 
@@ -330,11 +329,8 @@ static int handle_remove(const char *nodename, struct device *dev)
 	if (IS_ERR(dentry))
 		return PTR_ERR(dentry);
 
-	p.mnt = parent.mnt;
-	p.dentry = dentry;
-	err = vfs_getattr(&p, &stat, STATX_TYPE | STATX_MODE,
-			  AT_STATX_SYNC_AS_STAT);
-	if (!err && dev_mynode(dev, d_inode(dentry), &stat)) {
+	inode = d_inode(dentry);
+	if (dev_mynode(dev, inode)) {
 		struct iattr newattrs;
 		/*
 		 * before unlinking this node, reset permissions
@@ -342,7 +338,7 @@ static int handle_remove(const char *nodename, struct device *dev)
 		 */
 		newattrs.ia_uid = GLOBAL_ROOT_UID;
 		newattrs.ia_gid = GLOBAL_ROOT_GID;
-		newattrs.ia_mode = stat.mode & ~0777;
+		newattrs.ia_mode = inode->i_mode & ~0777;
 		newattrs.ia_valid =
 			ATTR_UID|ATTR_GID|ATTR_MODE;
 		inode_lock(d_inode(dentry));

  parent reply	other threads:[~2025-04-22 14:15 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-04-17  6:40 [PATCH] fs: move the bdex_statx call to vfs_getattr_nosec Christoph Hellwig
2025-04-17  8:01 ` Christian Brauner
2025-04-17  8:14 ` Christian Brauner
2025-04-17 16:20 ` Darrick J. Wong
2025-04-22  5:03 ` Shinichiro Kawasaki
2025-04-22  5:51   ` hch
2025-04-22  7:25     ` Christian Brauner
2025-04-22  7:27       ` hch
2025-04-22  8:15     ` Christian Brauner
2025-04-22  8:17       ` hch
2025-04-22  9:29         ` Christian Brauner
2025-04-22 14:15         ` hch [this message]
2025-04-22 15:35           ` Christian Brauner
2025-04-23  1:31           ` Shinichiro Kawasaki
2025-04-23  4:30           ` Jain, Ayush
2025-04-22 10:31     ` Shinichiro Kawasaki
2025-04-22 14:18     ` Heiko Carstens
     [not found] ` <d6dc234d922d8beda65f2a1eed1e2de6a50c978f.camel@linux.ibm.com>
2025-04-25 13:39   ` Christoph Hellwig

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=20250422141505.GA25426@lst.de \
    --to=hch@lst.de \
    --cc=axboe@kernel.dk \
    --cc=brauner@kernel.org \
    --cc=dakr@kernel.org \
    --cc=djwong@kernel.org \
    --cc=ebiggers@google.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=linux-block@vger.kernel.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=rafael@kernel.org \
    --cc=shinichiro.kawasaki@wdc.com \
    --cc=viro@zeniv.linux.org.uk \
    --cc=xni@redhat.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;
as well as URLs for NNTP newsgroup(s).