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: Andreas Gruenbacher <agruenba@redhat.com>,
	cluster-devel@redhat.com, linux-kernel@vger.kernel.org,
	Alexander Viro <viro@zeniv.linux.org.uk>, Jan Kara <jack@suse.cz>,
	Matthew Wilcox <willy@infradead.org>
Subject: [RFC 4/9] gfs2: Fix mmap + page fault deadlocks (part 1)
Date: Mon, 31 May 2021 19:01:18 +0200	[thread overview]
Message-ID: <20210531170123.243771-5-agruenba@redhat.com> (raw)
In-Reply-To: <20210531170123.243771-1-agruenba@redhat.com>

When the buffer passed to a read or write system call is memory mapped
to the same file, a page fault can occur in filemap_fault.  In that
case, the task will already be holding the inode glock, and trying to
take the same lock again will result in a BUG in add_to_queue().

Fix that by recognizing the self-recursion case.  Either skip the lock
taking (when the glock is held in a compatible way), or fail the
operation.

Likewise, a request to un-share a copy-on-write page can *probably*
happen in similar situations, so treat the locking in gfs2_page_mkwrite
in the same way.

A future patch will handle these case more gracefully by retrying
operations instead of failing them, along with addressing more complex
deadlock scenarios.

Reported-by: Jan Kara <jack@suse.cz>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/file.c | 40 ++++++++++++++++++++++++++++++----------
 1 file changed, 30 insertions(+), 10 deletions(-)

diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c
index 6d77743f11a4..7d88abb4629b 100644
--- a/fs/gfs2/file.c
+++ b/fs/gfs2/file.c
@@ -423,6 +423,7 @@ static vm_fault_t gfs2_page_mkwrite(struct vm_fault *vmf)
 	struct page *page = vmf->page;
 	struct inode *inode = file_inode(vmf->vma->vm_file);
 	struct gfs2_inode *ip = GFS2_I(inode);
+	struct gfs2_holder *outer_gh = gfs2_glock_is_locked_by_me(ip->i_gl);
 	struct gfs2_sbd *sdp = GFS2_SB(inode);
 	struct gfs2_alloc_parms ap = { .aflags = 0, };
 	u64 offset = page_offset(page);
@@ -436,10 +437,18 @@ static vm_fault_t gfs2_page_mkwrite(struct vm_fault *vmf)
 	sb_start_pagefault(inode->i_sb);
 
 	gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh);
-	err = gfs2_glock_nq(&gh);
-	if (err) {
-		ret = block_page_mkwrite_return(err);
-		goto out_uninit;
+	if (likely(!outer_gh)) {
+		err = gfs2_glock_nq(&gh);
+		if (err) {
+			ret = block_page_mkwrite_return(err);
+			goto out_uninit;
+		}
+	} else {
+		if (!gfs2_holder_is_compatible(outer_gh, LM_ST_EXCLUSIVE)) {
+			/* We could try to upgrade outer_gh here. */
+			ret = VM_FAULT_SIGBUS;
+			goto out_uninit;
+		}
 	}
 
 	/* Check page index against inode size */
@@ -540,7 +549,8 @@ static vm_fault_t gfs2_page_mkwrite(struct vm_fault *vmf)
 out_quota_unlock:
 	gfs2_quota_unlock(ip);
 out_unlock:
-	gfs2_glock_dq(&gh);
+	if (likely(!outer_gh))
+		gfs2_glock_dq(&gh);
 out_uninit:
 	gfs2_holder_uninit(&gh);
 	if (ret == VM_FAULT_LOCKED) {
@@ -555,6 +565,7 @@ 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 *outer_gh = gfs2_glock_is_locked_by_me(ip->i_gl);
 	struct gfs2_holder gh;
 	vm_fault_t ret;
 	u16 state;
@@ -562,13 +573,22 @@ static vm_fault_t gfs2_fault(struct vm_fault *vmf)
 
 	state = (vmf->flags & FAULT_FLAG_WRITE) ? LM_ST_EXCLUSIVE : LM_ST_SHARED;
 	gfs2_holder_init(ip->i_gl, state, 0, &gh);
-	err = gfs2_glock_nq(&gh);
-	if (err) {
-		ret = block_page_mkwrite_return(err);
-		goto out_uninit;
+	if (likely(!outer_gh)) {
+		err = gfs2_glock_nq(&gh);
+		if (err) {
+			ret = block_page_mkwrite_return(err);
+			goto out_uninit;
+		}
+	} else {
+		if (!gfs2_holder_is_compatible(outer_gh, state)) {
+			/* We could try to upgrade outer_gh here. */
+			ret = VM_FAULT_SIGBUS;
+			goto out_uninit;
+		}
 	}
 	ret = filemap_fault(vmf);
-	gfs2_glock_dq(&gh);
+	if (likely(!outer_gh))
+		gfs2_glock_dq(&gh);
 out_uninit:
 	gfs2_holder_uninit(&gh);
 	return ret;
-- 
2.26.3


  parent reply	other threads:[~2021-05-31 17:09 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-31 17:01 [RFC 0/9] gfs2: handle page faults during read and write Andreas Gruenbacher
2021-05-31 17:01 ` [RFC 1/9] gfs2: Clean up the error handling in gfs2_page_mkwrite Andreas Gruenbacher
2021-05-31 17:01 ` [RFC 2/9] gfs2: Add wrapper for iomap_file_buffered_write Andreas Gruenbacher
2021-05-31 17:01 ` [RFC 3/9] gfs2: Add gfs2_holder_is_compatible helper Andreas Gruenbacher
2021-05-31 17:01 ` Andreas Gruenbacher [this message]
2021-06-01  6:00   ` [RFC 4/9] gfs2: Fix mmap + page fault deadlocks (part 1) Linus Torvalds
2021-06-02 11:16     ` Andreas Gruenbacher
2021-06-11 16:25       ` Al Viro
2021-06-12 21:05         ` Al Viro
2021-06-12 21:35           ` Al Viro
2021-06-13  8:44             ` [Cluster-devel] " Steven Whitehouse
2021-05-31 17:01 ` [RFC 5/9] iov_iter: Add iov_iter_fault_in_writeable() Andreas Gruenbacher
2021-05-31 17:12   ` Al Viro
2021-06-12 21:12     ` Al Viro
2021-06-12 21:33       ` Linus Torvalds
2021-06-12 21:47         ` Al Viro
2021-06-12 23:17           ` Linus Torvalds
2021-06-12 23:38             ` Al Viro
2021-05-31 17:01 ` [RFC 6/9] gfs2: Add wrappers for accessing journal_info Andreas Gruenbacher
2021-05-31 17:01 ` [RFC 7/9] gfs2: Encode glock holding and retry flags in journal_info Andreas Gruenbacher
2021-05-31 17:01 ` [RFC 8/9] gfs2: Add LM_FLAG_OUTER glock holder flag Andreas Gruenbacher
2021-05-31 17:01 ` [RFC 9/9] gfs2: Fix mmap + page fault deadlocks (part 2) Andreas Gruenbacher
2021-06-01  5:47   ` Linus Torvalds
     [not found] ` <CAHk-=wgX=fZ+y=SxBsod8CvZmZ0-X7vZ6dV6EgLPkpBXbt=nQQ@mail.gmail.com>
2021-05-31 20:35   ` [RFC 0/9] gfs2: handle page faults during read and write Andreas Gruenbacher

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=20210531170123.243771-5-agruenba@redhat.com \
    --to=agruenba@redhat.com \
    --cc=cluster-devel@redhat.com \
    --cc=jack@suse.cz \
    --cc=linux-kernel@vger.kernel.org \
    --cc=torvalds@linux-foundation.org \
    --cc=viro@zeniv.linux.org.uk \
    --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).