linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Andreas Gruenbacher <agruenba@redhat.com>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Matthew Wilcox <willy@infradead.org>,
	Dave Chinner <david@fromorbit.com>,
	linux-fsdevel@vger.kernel.org, linux-mm@kvack.org,
	linux-kernel@vger.kernel.org,
	Andreas Gruenbacher <agruenba@redhat.com>
Subject: [RFC v2 2/2] gfs2: Rework read and page fault locking
Date: Fri,  3 Jul 2020 11:53:25 +0200	[thread overview]
Message-ID: <20200703095325.1491832-3-agruenba@redhat.com> (raw)
In-Reply-To: <20200703095325.1491832-1-agruenba@redhat.com>

So far, gfs2 has taken the inode glocks inside the ->readpage and
->readahead address space operations.  Since commit d4388340ae0b ("fs:
convert mpage_readpages to mpage_readahead"), gfs2_readahead is passed
the pages to read ahead locked.  With that, the current holder of the
inode glock may be trying to lock one of those pages while
gfs2_readahead is trying to take the inode glock, resulting in a
deadlock.

Fix that by moving the lock taking to the higher-level ->read_iter file
and ->fault vm operations.  This also gets rid of an ugly lock inversion
workaround in gfs2_readpage.

The cache consistency model of filesystems like gfs2 is such that if
data is found in the page cache, the data is up to date and can be used
without taking any filesystem locks.  If a page is not cached,
filesystem locks must be taken before populating the page cache.

To avoid taking the inode glock when the data is already cached,
gfs2_file_read_iter first tries to read the data with the IOCB_NOIO flag
set.  If that fails, the inode glock is taken and the operation is
retried with the IOCB_NOIO flag cleared.

Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/aops.c | 45 +------------------------------------------
 fs/gfs2/file.c | 52 ++++++++++++++++++++++++++++++++++++++++++++++++--
 2 files changed, 51 insertions(+), 46 deletions(-)

diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c
index 72c9560f4467..68cd700a2719 100644
--- a/fs/gfs2/aops.c
+++ b/fs/gfs2/aops.c
@@ -468,21 +468,10 @@ static int stuffed_readpage(struct gfs2_inode *ip, struct page *page)
 }
 
 
-/**
- * __gfs2_readpage - readpage
- * @file: The file to read a page for
- * @page: The page to read
- *
- * This is the core of gfs2's readpage. It's used by the internal file
- * reading code as in that case we already hold the glock. Also it's
- * called by gfs2_readpage() once the required lock has been granted.
- */
-
 static int __gfs2_readpage(void *file, struct page *page)
 {
 	struct gfs2_inode *ip = GFS2_I(page->mapping->host);
 	struct gfs2_sbd *sdp = GFS2_SB(page->mapping->host);
-
 	int error;
 
 	if (i_blocksize(page->mapping->host) == PAGE_SIZE &&
@@ -505,36 +494,11 @@ static int __gfs2_readpage(void *file, struct page *page)
  * gfs2_readpage - read a page of a file
  * @file: The file to read
  * @page: The page of the file
- *
- * This deals with the locking required. We have to unlock and
- * relock the page in order to get the locking in the right
- * order.
  */
 
 static int gfs2_readpage(struct file *file, struct page *page)
 {
-	struct address_space *mapping = page->mapping;
-	struct gfs2_inode *ip = GFS2_I(mapping->host);
-	struct gfs2_holder gh;
-	int error;
-
-	unlock_page(page);
-	gfs2_holder_init(ip->i_gl, LM_ST_SHARED, 0, &gh);
-	error = gfs2_glock_nq(&gh);
-	if (unlikely(error))
-		goto out;
-	error = AOP_TRUNCATED_PAGE;
-	lock_page(page);
-	if (page->mapping == mapping && !PageUptodate(page))
-		error = __gfs2_readpage(file, page);
-	else
-		unlock_page(page);
-	gfs2_glock_dq(&gh);
-out:
-	gfs2_holder_uninit(&gh);
-	if (error && error != AOP_TRUNCATED_PAGE)
-		lock_page(page);
-	return error;
+	return __gfs2_readpage(file, page);
 }
 
 /**
@@ -598,16 +562,9 @@ static void gfs2_readahead(struct readahead_control *rac)
 {
 	struct inode *inode = rac->mapping->host;
 	struct gfs2_inode *ip = GFS2_I(inode);
-	struct gfs2_holder gh;
 
-	gfs2_holder_init(ip->i_gl, LM_ST_SHARED, 0, &gh);
-	if (gfs2_glock_nq(&gh))
-		goto out_uninit;
 	if (!gfs2_is_stuffed(ip))
 		mpage_readahead(rac, gfs2_block_map);
-	gfs2_glock_dq(&gh);
-out_uninit:
-	gfs2_holder_uninit(&gh);
 }
 
 /**
diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c
index fe305e4bfd37..bebde537ac8c 100644
--- a/fs/gfs2/file.c
+++ b/fs/gfs2/file.c
@@ -558,8 +558,29 @@ static vm_fault_t gfs2_page_mkwrite(struct vm_fault *vmf)
 	return block_page_mkwrite_return(ret);
 }
 
+static vm_fault_t gfs2_fault(struct vm_fault *vmf)
+{
+	struct inode *inode = file_inode(vmf->vma->vm_file);
+	struct gfs2_inode *ip = GFS2_I(inode);
+	struct gfs2_holder gh;
+	vm_fault_t ret;
+	int err;
+
+	gfs2_holder_init(ip->i_gl, LM_ST_SHARED, 0, &gh);
+	err = gfs2_glock_nq(&gh);
+	if (err) {
+		ret = block_page_mkwrite_return(err);
+		goto out_uninit;
+	}
+	ret = filemap_fault(vmf);
+	gfs2_glock_dq(&gh);
+out_uninit:
+	gfs2_holder_uninit(&gh);
+	return ret;
+}
+
 static const struct vm_operations_struct gfs2_vm_ops = {
-	.fault = filemap_fault,
+	.fault = gfs2_fault,
 	.map_pages = filemap_map_pages,
 	.page_mkwrite = gfs2_page_mkwrite,
 };
@@ -824,6 +845,9 @@ static ssize_t gfs2_file_direct_write(struct kiocb *iocb, struct iov_iter *from)
 
 static ssize_t gfs2_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
 {
+	struct gfs2_inode *ip;
+	struct gfs2_holder gh;
+	size_t written = 0;
 	ssize_t ret;
 
 	if (iocb->ki_flags & IOCB_DIRECT) {
@@ -832,7 +856,31 @@ static ssize_t gfs2_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
 			return ret;
 		iocb->ki_flags &= ~IOCB_DIRECT;
 	}
-	return generic_file_read_iter(iocb, to);
+	iocb->ki_flags |= IOCB_NOIO;
+	ret = generic_file_read_iter(iocb, to);
+	iocb->ki_flags &= ~IOCB_NOIO;
+	if (ret >= 0) {
+		if (!iov_iter_count(to))
+			return ret;
+		written = ret;
+	} else {
+		if (ret != -EAGAIN)
+			return ret;
+		if (iocb->ki_flags & IOCB_NOWAIT)
+			return ret;
+	}
+	ip = GFS2_I(iocb->ki_filp->f_mapping->host);
+	gfs2_holder_init(ip->i_gl, LM_ST_SHARED, 0, &gh);
+	ret = gfs2_glock_nq(&gh);
+	if (ret)
+		goto out_uninit;
+	ret = generic_file_read_iter(iocb, to);
+	if (ret > 0)
+		written += ret;
+	gfs2_glock_dq(&gh);
+out_uninit:
+	gfs2_holder_uninit(&gh);
+	return written ? written : ret;
 }
 
 /**
-- 
2.26.2


  parent reply	other threads:[~2020-07-03  9:53 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-07-03  9:53 [RFC v2 0/2] Fix gfs2 readahead deadlocks Andreas Gruenbacher
2020-07-03  9:53 ` [RFC v2 1/2] fs: Add IOCB_NOIO flag for generic_file_read_iter Andreas Gruenbacher
2020-07-03 11:41   ` Matthew Wilcox
2020-07-05 15:08     ` Andreas Gruenbacher
2020-07-03  9:53 ` Andreas Gruenbacher [this message]
2020-07-03 11:38   ` [RFC v2 2/2] gfs2: Rework read and page fault locking Matthew Wilcox
2020-07-03 11:44     ` Matthew Wilcox
2020-07-03 19:24 ` [RFC v2 0/2] Fix gfs2 readahead deadlocks Linus Torvalds
2020-07-03 19:26   ` Linus Torvalds

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=20200703095325.1491832-3-agruenba@redhat.com \
    --to=agruenba@redhat.com \
    --cc=david@fromorbit.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=torvalds@linux-foundation.org \
    --cc=willy@infradead.org \
    /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).