linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Ben Hutchings <ben@decadent.org.uk>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: akpm@linux-foundation.org, "Dave Chinner" <dchinner@redhat.com>,
	"Christoph Hellwig" <hch@lst.de>, "Wen Xu" <wen.xu@gatech.edu>,
	"Carlos Maiolino" <cmaiolino@redhat.com>,
	"Darrick J. Wong" <darrick.wong@oracle.com>
Subject: [PATCH 3.16 52/63] xfs: validate cached inodes are free when allocated
Date: Sat, 22 Sep 2018 01:15:42 +0100	[thread overview]
Message-ID: <lsq.1537575342.534061407@decadent.org.uk> (raw)
In-Reply-To: <lsq.1537575341.194909669@decadent.org.uk>

3.16.58-rc1 review patch.  If anyone has any objections, please let me know.

------------------

From: Dave Chinner <dchinner@redhat.com>

commit afca6c5b2595fc44383919fba740c194b0b76aff upstream.

A recent fuzzed filesystem image cached random dcache corruption
when the reproducer was run. This often showed up as panics in
lookup_slow() on a null inode->i_ops pointer when doing pathwalks.

BUG: unable to handle kernel NULL pointer dereference at 0000000000000000
....
Call Trace:
 lookup_slow+0x44/0x60
 walk_component+0x3dd/0x9f0
 link_path_walk+0x4a7/0x830
 path_lookupat+0xc1/0x470
 filename_lookup+0x129/0x270
 user_path_at_empty+0x36/0x40
 path_listxattr+0x98/0x110
 SyS_listxattr+0x13/0x20
 do_syscall_64+0xf5/0x280
 entry_SYSCALL_64_after_hwframe+0x42/0xb7

but had many different failure modes including deadlocks trying to
lock the inode that was just allocated or KASAN reports of
use-after-free violations.

The cause of the problem was a corrupt INOBT on a v4 fs where the
root inode was marked as free in the inobt record. Hence when we
allocated an inode, it chose the root inode to allocate, found it in
the cache and re-initialised it.

We recently fixed a similar inode allocation issue caused by inobt
record corruption problem in xfs_iget_cache_miss() in commit
ee457001ed6c ("xfs: catch inode allocation state mismatch
corruption"). This change adds similar checks to the cache-hit path
to catch it, and turns the reproducer into a corruption shutdown
situation.

Reported-by: Wen Xu <wen.xu@gatech.edu>
Signed-Off-By: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
[darrick: fix typos in comment]
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
[bwh: Backported to 3.16:
 - Look up mode in XFS inode, not VFS inode
 - Use positive error codes, and EIO instead of EFSCORRUPTED]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
---
 fs/xfs/xfs_icache.c | 73 +++++++++++++++++++++++++++++----------------
 1 file changed, 48 insertions(+), 25 deletions(-)

--- a/fs/xfs/xfs_icache.c
+++ b/fs/xfs/xfs_icache.c
@@ -133,6 +133,46 @@ xfs_inode_free(
 }
 
 /*
+ * If we are allocating a new inode, then check what was returned is
+ * actually a free, empty inode. If we are not allocating an inode,
+ * then check we didn't find a free inode.
+ *
+ * Returns:
+ *	0		if the inode free state matches the lookup context
+ *	ENOENT		if the inode is free and we are not allocating
+ *	EFSCORRUPTED	if there is any state mismatch at all
+ */
+static int
+xfs_iget_check_free_state(
+	struct xfs_inode	*ip,
+	int			flags)
+{
+	if (flags & XFS_IGET_CREATE) {
+		/* should be a free inode */
+		if (ip->i_d.di_mode != 0) {
+			xfs_warn(ip->i_mount,
+"Corruption detected! Free inode 0x%llx not marked free! (mode 0x%x)",
+				ip->i_ino, ip->i_d.di_mode);
+			return EIO;
+		}
+
+		if (ip->i_d.di_nblocks != 0) {
+			xfs_warn(ip->i_mount,
+"Corruption detected! Free inode 0x%llx has blocks allocated!",
+				ip->i_ino);
+			return EIO;
+		}
+		return 0;
+	}
+
+	/* should be an allocated inode */
+	if (ip->i_d.di_mode == 0)
+		return ENOENT;
+
+	return 0;
+}
+
+/*
  * Check the validity of the inode we just found it the cache
  */
 static int
@@ -181,12 +221,12 @@ xfs_iget_cache_hit(
 	}
 
 	/*
-	 * If lookup is racing with unlink return an error immediately.
+	 * Check the inode free state is valid. This also detects lookup
+	 * racing with unlinks.
 	 */
-	if (ip->i_d.di_mode == 0 && !(flags & XFS_IGET_CREATE)) {
-		error = ENOENT;
+	error = xfs_iget_check_free_state(ip, flags);
+	if (error)
 		goto out_error;
-	}
 
 	/*
 	 * If IRECLAIMABLE is set, we've torn down the VFS inode already.
@@ -295,29 +335,12 @@ xfs_iget_cache_miss(
 
 
 	/*
-	 * If we are allocating a new inode, then check what was returned is
-	 * actually a free, empty inode. If we are not allocating an inode,
-	 * the check we didn't find a free inode.
+	 * Check the inode free state is valid. This also detects lookup
+	 * racing with unlinks.
 	 */
-	if (flags & XFS_IGET_CREATE) {
-		if (ip->i_d.di_mode != 0) {
-			xfs_warn(mp,
-"Corruption detected! Free inode 0x%llx not marked free on disk",
-				ino);
-			error = EIO;
-			goto out_destroy;
-		}
-		if (ip->i_d.di_nblocks != 0) {
-			xfs_warn(mp,
-"Corruption detected! Free inode 0x%llx has blocks allocated!",
-				ino);
-			error = EIO;
-			goto out_destroy;
-		}
-	} else if (ip->i_d.di_mode == 0) {
-		error = ENOENT;
+	error = xfs_iget_check_free_state(ip, flags);
+	if (error)
 		goto out_destroy;
-	}
 
 	/*
 	 * Preload the radix tree so we can insert safely under the


  parent reply	other threads:[~2018-09-22  0:24 UTC|newest]

Thread overview: 71+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-09-22  0:15 [PATCH 3.16 00/63] 3.16.58-rc1 review Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 51/63] xfs: catch inode allocation state mismatch corruption Ben Hutchings
2018-09-22  5:25   ` Dave Chinner
2018-09-22 20:57     ` Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 07/63] usbip: usbip_host: refine probe and disconnect debug msgs to be useful Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 41/63] USB: yurex: fix out-of-bounds uaccess in read handler Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 54/63] seccomp: create internal mode-setting function Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 45/63] x86/paravirt: Fix spectre-v2 mitigations for paravirt guests Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 36/63] jbd2: don't mark block as modified if the handle is out of credits Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 18/63] sr: pass down correctly sized SCSI sense buffer Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 06/63] usbip: usbip_host: fix to hold parent lock for device_attach() calls Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 29/63] ext4: make sure bitmaps and the inode table don't overlap with bg descriptors Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 63/63] mm: get rid of vmacache_flush_all() entirely Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 26/63] ext4: verify the depth of extent tree in ext4_find_extent() Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 38/63] Fix up non-directory creation in SGID directories Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 56/63] seccomp: split mode setting routines Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 13/63] futex: Remove unnecessary warning from get_futex_key Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 21/63] Bluetooth: hidp: buffer overflow in hidp_process_report Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 32/63] ext4: always verify the magic number in xattr blocks Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 23/63] xfs: set format back to extents if xfs_bmap_extents_to_btree Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 08/63] usbip: usbip_host: delete device from busid_table after rebind Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 60/63] x86/cpu/AMD: Fix erratum 1076 (CPB bit) Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 57/63] seccomp: add "seccomp" syscall Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 20/63] scsi: sg: allocate with __GFP_ZERO in sg_build_indirect() Ben Hutchings
2018-09-22  0:19   ` syzbot
2018-09-22  0:15 ` [PATCH 3.16 04/63] net: Set sk_prot_creator when cloning sockets to the right proto Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 33/63] ext4: never move the system.data xattr out of the inode body Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 55/63] seccomp: extract check/assign mode helpers Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 16/63] KVM: x86: pass kvm_vcpu to kvm_read_guest_virt and kvm_write_guest_virt_system Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 35/63] ext4: add more inode number paranoia checks Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 42/63] ALSA: rawmidi: Change resized buffers atomically Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 27/63] ext4: always check block group bounds in ext4_init_block_bitmap() Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 25/63] ext4: fix check to prevent initializing reserved inodes Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 14/63] KVM: x86: Emulator ignores LDTR/TR extended base on LLDT/LTR Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 15/63] KVM: x86: introduce linear_{read,write}_system Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 61/63] x86/cpu/intel: Add Knights Mill to Intel family Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 40/63] infiniband: fix a possible use-after-free bug Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 09/63] usbip: usbip_host: run rebind from exit when module is removed Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 11/63] usbip: usbip_host: fix bad unlock balance during stub_probe() Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 12/63] futex: Remove requirement for lock_page() in get_futex_key() Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 30/63] ext4: fix false negatives *and* false positives in ext4_check_descriptors() Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 37/63] ext4: avoid running out of journal credits when appending to an inline file Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 47/63] uas: replace WARN_ON_ONCE() with lockdep_assert_held() Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 44/63] x86/speculation: Protect against userspace-userspace spectreRSB Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 49/63] btrfs: relocation: Only remove reloc rb_trees if reloc control has been initialized Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 24/63] ext4: only look at the bg_flags field if it is valid Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 53/63] xfs: don't call xfs_da_shrink_inode with NULL bp Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 17/63] kvm: x86: use correct privilege level for sgdt/sidt/fxsave/fxrstor access Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 43/63] x86/speculation: Clean up various Spectre related details Ben Hutchings
2018-09-22  0:15 ` Ben Hutchings [this message]
2018-09-22  5:26   ` [PATCH 3.16 52/63] xfs: validate cached inodes are free when allocated Dave Chinner
2018-09-22 20:57     ` Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 48/63] video: uvesafb: Fix integer overflow in allocation Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 19/63] jfs: Fix inconsistency between memory allocation and ea_buf->max_size Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 22/63] scsi: libsas: defer ata device eh commands to libata Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 05/63] usbip: fix error handling in stub_probe() Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 10/63] usbip: usbip_host: fix NULL-ptr deref and use-after-free errors Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 46/63] cdrom: Fix info leak/OOB read in cdrom_ioctl_drive_status Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 28/63] ext4: don't allow r/w mounts if metadata blocks overlap the superblock Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 03/63] Revert "vti4: Don't override MTU passed on link creation via IFLA_MTU" Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 34/63] ext4: clear i_data in ext4_inode_info when removing inline data Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 01/63] x86/fpu: Fix the 'nofxsr' boot parameter to also clear X86_FEATURE_FXSR_OPT Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 02/63] x86/fpu: Default eagerfpu if FPU and FXSR are enabled Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 58/63] x86/process: Optimize TIF checks in __switch_to_xtra() Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 31/63] ext4: add corruption check in ext4_xattr_set_entry() Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 39/63] x86/entry/64: Remove %ebx handling from error_entry/exit Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 59/63] x86/process: Correct and optimize TIF_BLOCKSTEP switch Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 50/63] hfsplus: fix NULL dereference in hfsplus_lookup() Ben Hutchings
2018-09-22  0:15 ` [PATCH 3.16 62/63] KVM: x86: introduce num_emulated_msrs Ben Hutchings
2018-09-22 12:28 ` [PATCH 3.16 00/63] 3.16.58-rc1 review Guenter Roeck
2018-09-22 21:03   ` Ben Hutchings

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=lsq.1537575342.534061407@decadent.org.uk \
    --to=ben@decadent.org.uk \
    --cc=akpm@linux-foundation.org \
    --cc=cmaiolino@redhat.com \
    --cc=darrick.wong@oracle.com \
    --cc=dchinner@redhat.com \
    --cc=hch@lst.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=stable@vger.kernel.org \
    --cc=wen.xu@gatech.edu \
    /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).