4.1.4-stable review

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Constantine Shulyupin <[email protected]>

commit 56172d81a9bc37a69b95dd627b8d48135c9c7b31 upstream.

Excerpt from datasheet:
7.2.32 Mode Selection Register
RTD3_MD : 00=Closed , 01=Reserved , 10=Thermistor mode , 11=Voltage sense

Show temp3 only in Thermistor mode

---
 drivers/hwmon/nct7802.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -547,7 +547,7 @@ static umode_t nct7802_temp_is_visible(s
 	if (index >= 9 && index < 18 &&
 	    (reg & 0x0c) != 0x04 && (reg & 0x0c) != 0x08)	/* RD2 */
 		return 0;
-	if (index >= 18 && index < 27 && (reg & 0x30) != 0x10)	/* RD3 */
+	if (index >= 18 && index < 27 && (reg & 0x30) != 0x20)	/* RD3 */
 		return 0;
 	if (index >= 27 && index < 35)				/* local */
 		return attr->mode;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Stefan Wahren <[email protected]>

commit e8e94ed6285428ab780cd7b0df4622f71eceb39e upstream.

In order to get iio-hwmon support, the lradc must be declared as an iio provider. So fix this issue by adding the #io-channel-cells property.

---
 arch/arm/boot/dts/imx23.dtsi |    1 +
 1 file changed, 1 insertion(+)

@@ -435,6 +435,7 @@
 				interrupts = <36 37 38 39 40 41 42 43 44>;
 				status = "disabled";
 				clocks = <&clks 26>;
+				#io-channel-cells = <1>;
 			};
 
 			spdif@80054000 {

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Omar Sandoval <[email protected]>

commit 64ad6c488975d7516230cf7849190a991fd615ae upstream.

Since commit bafc9b754f75 ("vfs: More precise tests in d_invalidate"), mounted subvolumes can be deleted because d_invalidate() won't fail. However, we run into problems when we attempt to delete the default subvolume while it is mounted as the root filesystem:

	# btrfs subvol list /
	ID 257 gen 306 top level 5 path rootvol
	ID 267 gen 334 top level 5 path snap1
	# btrfs subvol get-default /
	ID 267 gen 334 top level 5 path snap1
	# btrfs inspect-internal rootid /
	267
	# mount -o subvol=/ /dev/vda1 /mnt
	# btrfs subvol del /mnt/snap1
	Delete subvolume (no-commit): '/mnt/snap1'
	ERROR: cannot delete '/mnt/snap1' - Operation not permitted
	# findmnt /
	findmnt: can't read /proc/mounts: No such file or directory
	# ls /proc
	#

Markus reported that this same scenario simply led to a kernel oops.

This happens because in btrfs_ioctl_snap_destroy(), we call d_invalidate() before we check may_destroy_subvol(), which means that we detach the submounts and drop the dentry before erroring out. Instead, we should only invalidate the dentry once the deletion has succeeded. Additionally, the shrink_dcache_sb() isn't necessary; d_invalidate() will prune the dcache for the deleted subvolume.

---
 fs/btrfs/ioctl.c |    4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

@@ -2413,8 +2413,6 @@ static noinline int btrfs_ioctl_snap_des
 		goto out_unlock_inode;
 	}
 
-	d_invalidate(dentry);
-
 	down_write(&root->fs_info->subvol_sem);
 
 	err = may_destroy_subvol(dest);

@@ -2508,7 +2506,7 @@ out_up_write:
 out_unlock_inode:
 	mutex_unlock(&inode->i_mutex);
 	if (!err) {
-		shrink_dcache_sb(root->fs_info->sb);
+		d_invalidate(dentry);
 		btrfs_invalidate_inodes(dest);
 		d_delete(dentry);
 		ASSERT(dest->send_in_progress == 0);

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Firo Yang <[email protected]>

commit 4e023612325a9034a542bfab79f78b1fe5ebb841 upstream.

Warning like this:

drivers/md/md.c: In function "update_array_info":
drivers/md/md.c:6394:26: warning: logical not is only applied to the left hand side of comparison [-Wlogical-not-parentheses]

      !mddev->persistent  != info->not_persistent||

Fix it as Neil Brown said:
mddev->persistent != !info->not_persistent ||

---
 drivers/md/md.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -6380,7 +6380,7 @@ static int update_array_info(struct mdde
 	    mddev->ctime         != info->ctime         ||
 	    mddev->level         != info->level         ||
 /*	    mddev->layout        != info->layout        || */
-	    !mddev->persistent	 != info->not_persistent||
+	    mddev->persistent	 != !info->not_persistent ||
 	    mddev->chunk_sectors != info->chunk_size >> 9 ||
 	    /* ignore bottom 8 bits of state, and allow SB_BITMAP_PRESENT to change */
 	    ((state^info->state) & 0xfffffe00)

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Filipe Manana <[email protected]>

commit c3f4a1685bb87e59c886ee68f7967eae07d4dffa upstream.

The free space entries are allocated using kmem_cache_zalloc(), through __btrfs_add_free_space(), therefore we should use kmem_cache_free() and not kfree() to avoid any confusion and any potential problem. Looking at the kfree() definition at mm/slab.c it has the following comment:

/*
* (...)
*
* Don't free memory not originally allocated by kmalloc()
* or you will run into trouble.
*/

So better be safe and use kmem_cache_free().

---
 fs/btrfs/inode-map.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -271,7 +271,7 @@ void btrfs_unpin_free_ino(struct btrfs_r
 		__btrfs_add_free_space(ctl, info->offset, count);
 free:
 		rb_erase(&info->offset_index, rbroot);
-		kfree(info);
+		kmem_cache_free(btrfs_free_space_cachep, info);
 	}
 }
 

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Filipe Manana <[email protected]>

commit ae9d8f17118551bedd797406a6768b87c2146234 upstream.

While the inode cache caching kthread is calling btrfs_unpin_free_ino(), we could have a concurrent call to btrfs_return_ino() that adds a new entry to the root's free space cache of pinned inodes. This concurrent call does not acquire the fs_info->commit_root_sem before adding a new entry if the caching state is BTRFS_CACHE_FINISHED, which is a problem because the caching kthread calls btrfs_unpin_free_ino() after setting the caching state to BTRFS_CACHE_FINISHED and therefore races with the task calling btrfs_return_ino(), which is adding a new entry, while the former (caching kthread) is navigating the cache's rbtree, removing and freeing nodes from the cache's rbtree without acquiring the spinlock that protects the rbtree.

This race resulted in memory corruption due to double free of struct btrfs_free_space objects because both tasks can end up doing freeing the same objects. Note that adding a new entry can result in merging it with other entries in the cache, in which case those entries are freed. This is particularly important as btrfs_free_space structures are also used for the block group free space caches.

This memory corruption can be detected by a debugging kernel, which reports it with the following trace:

[132408.501148] slab error in verify_redzone_free(): cache `btrfs_free_space': double free detected
[132408.505075] CPU: 15 PID: 12248 Comm: btrfs-ino-cache Tainted: G        W       4.1.0-rc5-btrfs-next-10+ #1
[132408.505075] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.8.1-0-g4adadbd-20150316_085822-nilsson.home.kraxel.org 04/01/2014
[132408.505075]  ffff880023e7d320 ffff880163d73cd8 ffffffff8145eec7 ffffffff81095dce
[132408.505075]  ffff880009735d40 ffff880163d73ce8 ffffffff81154e1e ffff880163d73d68
[132408.505075]  ffffffff81155733 ffffffffa054a95a ffff8801b6099f00 ffffffffa0505b5f
[132408.505075] Call Trace:
[132408.505075]  [<ffffffff8145eec7>] dump_stack+0x4f/0x7b
[132408.505075]  [<ffffffff81095dce>] ? console_unlock+0x356/0x3a2
[132408.505075]  [<ffffffff81154e1e>] __slab_error.isra.28+0x25/0x36
[132408.505075]  [<ffffffff81155733>] __cache_free+0xe2/0x4b6
[132408.505075]  [<ffffffffa054a95a>] ? __btrfs_add_free_space+0x2f0/0x343 [btrfs]
[132408.505075]  [<ffffffffa0505b5f>] ? btrfs_unpin_free_ino+0x8e/0x99 [btrfs]
[132408.505075]  [<ffffffff810f3b30>] ? time_hardirqs_off+0x15/0x28
[132408.505075]  [<ffffffff81084d42>] ? trace_hardirqs_off+0xd/0xf
[132408.505075]  [<ffffffff811563a1>] ? kfree+0xb6/0x14e
[132408.505075]  [<ffffffff811563d0>] kfree+0xe5/0x14e
[132408.505075]  [<ffffffffa0505b5f>] btrfs_unpin_free_ino+0x8e/0x99 [btrfs]
[132408.505075]  [<ffffffffa0505e08>] caching_kthread+0x29e/0x2d9 [btrfs]
[132408.505075]  [<ffffffffa0505b6a>] ? btrfs_unpin_free_ino+0x99/0x99 [btrfs]
[132408.505075]  [<ffffffff8106698f>] kthread+0xef/0xf7
[132408.505075]  [<ffffffff810f3b08>] ? time_hardirqs_on+0x15/0x28
[132408.505075]  [<ffffffff810668a0>] ? __kthread_parkme+0xad/0xad
[132408.505075]  [<ffffffff814653d2>] ret_from_fork+0x42/0x70
[132408.505075]  [<ffffffff810668a0>] ? __kthread_parkme+0xad/0xad
[132408.505075] ffff880023e7d320: redzone 1:0x9f911029d74e35b, redzone 2:0x9f911029d74e35b.
[132409.501654] slab: double free detected in cache 'btrfs_free_space', objp ffff880023e7d320
[132409.503355] ------------[ cut here ]------------
[132409.504241] kernel BUG at mm/slab.c:2571!

Therefore fix this by having btrfs_unpin_free_ino() acquire the lock that protects the rbtree while doing the searches and removing entries.

---
 fs/btrfs/inode-map.c |   15 +++++++++++----
 1 file changed, 11 insertions(+), 4 deletions(-)

@@ -246,6 +246,7 @@ void btrfs_unpin_free_ino(struct btrfs_r
 {
 	struct btrfs_free_space_ctl *ctl = root->free_ino_ctl;
 	struct rb_root *rbroot = &root->free_ino_pinned->free_space_offset;
+	spinlock_t *rbroot_lock = &root->free_ino_pinned->tree_lock;
 	struct btrfs_free_space *info;
 	struct rb_node *n;
 	u64 count;

@@ -254,23 +255,29 @@ void btrfs_unpin_free_ino(struct btrfs_r
 		return;
 
 	while (1) {
+		bool add_to_ctl = true;
+
+		spin_lock(rbroot_lock);
 		n = rb_first(rbroot);
-		if (!n)
+		if (!n) {
+			spin_unlock(rbroot_lock);
 			break;
+		}
 
 		info = rb_entry(n, struct btrfs_free_space, offset_index);
 		BUG_ON(info->bitmap); /* Logic error */
 
 		if (info->offset > root->ino_cache_progress)
-			goto free;
+			add_to_ctl = false;
 		else if (info->offset + info->bytes > root->ino_cache_progress)
 			count = root->ino_cache_progress - info->offset + 1;
 		else
 			count = info->bytes;
 
-		__btrfs_add_free_space(ctl, info->offset, count);
-free:
 		rb_erase(&info->offset_index, rbroot);
+		spin_unlock(rbroot_lock);
+		if (add_to_ctl)
+			__btrfs_add_free_space(ctl, info->offset, count);
 		kmem_cache_free(btrfs_free_space_cachep, info);
 	}
 }

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Filipe Manana <[email protected]>

commit e4545de5b035c7debb73d260c78377dbb69cbfb5 upstream.

If we do an append write to a file (which increases its inode's i_size) that does not have the flag BTRFS_INODE_NEEDS_FULL_SYNC set in its inode, and the previous transaction added a new hard link to the file, which sets the flag BTRFS_INODE_COPY_EVERYTHING in the file's inode, and then fsync the file, the inode's new i_size isn't logged. This has the consequence that after the fsync log is replayed, the file size remains what it was before the append write operation, which means users/applications will not be able to read the data that was successsfully fsync'ed before.

This happens because neither the inode item nor the delayed inode get their i_size updated when the append write is made - doing so would require starting a transaction in the buffered write path, something that we do not do intentionally for performance reasons.

Fix this by making sure that when the flag BTRFS_INODE_COPY_EVERYTHING is set the inode is logged with its current i_size (log the in-memory inode into the log tree).

This issue is not a recent regression and is easy to reproduce with the following test case for fstests:

seq=basename $0
seqres=$RESULT_DIR/$seq
echo "QA output created by $seq"

here=pwd
tmp=/tmp/$$
status=1 # failure is the default!

  _cleanup()
  {
          _cleanup_flakey
          rm -f $tmp.*
  }
  trap "_cleanup; exit \$status" 0 1 2 3 15

  # get standard environment, filters and checks
  . ./common/rc
  . ./common/filter
  . ./common/dmflakey

  # real QA test starts here
  _supported_fs generic
  _supported_os Linux
  _need_to_be_root
  _require_scratch
  _require_dm_flakey
  _require_metadata_journaling $SCRATCH_DEV

  _crash_and_mount()
  {
          # Simulate a crash/power loss.
          _load_flakey_table $FLAKEY_DROP_WRITES
          _unmount_flakey
          # Allow writes again and mount. This makes the fs replay its fsync log.
          _load_flakey_table $FLAKEY_ALLOW_WRITES
          _mount_flakey
  }

rm -f $seqres.full

_scratch_mkfs >> $seqres.full 2>&1
_init_flakey
_mount_flakey

  # Create the test file with some initial data and then fsync it.
  # The fsync here is only needed to trigger the issue in btrfs, as it causes the
  # the flag BTRFS_INODE_NEEDS_FULL_SYNC to be removed from the btrfs inode.
  $XFS_IO_PROG -f -c "pwrite -S 0xaa 0 32k" \
                  -c "fsync" \
                  $SCRATCH_MNT/foo | _filter_xfs_io
  sync

# Add a hard link to our file.
# On btrfs this sets the flag BTRFS_INODE_COPY_EVERYTHING on the btrfs inode,

  # which is a necessary condition to trigger the issue.
  ln $SCRATCH_MNT/foo $SCRATCH_MNT/bar

# Sync the filesystem to force a commit of the current btrfs transaction, this

  # is a necessary condition to trigger the bug on btrfs.
  sync

  # Now append more data to our file, increasing its size, and fsync the file.
  # In btrfs because the inode flag BTRFS_INODE_COPY_EVERYTHING was set and the
  # write path did not update the inode item in the btree nor the delayed inode
  # item (in memory struture) in the current transaction (created by the fsync
  # handler), the fsync did not record the inode's new i_size in the fsync
  # log/journal. This made the data unavailable after the fsync log/journal is
  # replayed.
  $XFS_IO_PROG -c "pwrite -S 0xbb 32K 32K" \
               -c "fsync" \
               $SCRATCH_MNT/foo | _filter_xfs_io

echo "File content after fsync and before crash:" od -t x1 $SCRATCH_MNT/foo

_crash_and_mount

echo "File content after crash and log replay:"
od -t x1 $SCRATCH_MNT/foo

status=0
exit

The expected file output before and after the crash/power failure expects the appended data to be available, which is:

0000000 aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa *
0100000 bb bb bb bb bb bb bb bb bb bb bb bb bb bb bb bb *
0200000

---
 fs/btrfs/tree-log.c |   14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

@@ -4161,6 +4161,7 @@ static int btrfs_log_inode(struct btrfs_
 	u64 ino = btrfs_ino(inode);
 	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
 	u64 logged_isize = 0;
+	bool need_log_inode_item = true;
 
 	path = btrfs_alloc_path();
 	if (!path)

@@ -4269,11 +4270,6 @@ static int btrfs_log_inode(struct btrfs_
 		} else {
 			if (inode_only == LOG_INODE_ALL)
 				fast_search = true;
-			ret = log_inode_item(trans, log, dst_path, inode);
-			if (ret) {
-				err = ret;
-				goto out_unlock;
-			}
 			goto log_extents;
 		}
 

@@ -4296,6 +4292,9 @@ again:
 		if (min_key.type > max_key.type)
 			break;
 
+		if (min_key.type == BTRFS_INODE_ITEM_KEY)
+			need_log_inode_item = false;
+
 		src = path->nodes[0];
 		if (ins_nr && ins_start_slot + ins_nr == path->slots[0]) {
 			ins_nr++;

@@ -4366,6 +4365,11 @@ next_slot:
 log_extents:
 	btrfs_release_path(path);
 	btrfs_release_path(dst_path);
+	if (need_log_inode_item) {
+		err = log_inode_item(trans, log, dst_path, inode);
+		if (err)
+			goto out_unlock;
+	}
 	if (fast_search) {
 		/*
 		 * Some ordered extents started by fsync might have completed

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Filipe Manana <[email protected]>

commit 497b4050e0eacd4c746dd396d14916b1e669849d upstream.

We were allocating memory with memdup_user() but we were never releasing that memory. This affected pretty much every call to the ioctl, whether it deduplicated extents or not.

This issue was reported on IRC by Julian Taylor and on the mailing list by Marcel Ritter, credit goes to them for finding the issue.

---
 fs/btrfs/ioctl.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

@@ -2938,7 +2938,7 @@ out_unlock:
 static long btrfs_ioctl_file_extent_same(struct file *file,
 			struct btrfs_ioctl_same_args __user *argp)
 {
-	struct btrfs_ioctl_same_args *same;
+	struct btrfs_ioctl_same_args *same = NULL;
 	struct btrfs_ioctl_same_extent_info *info;
 	struct inode *src = file_inode(file);
 	u64 off;

@@ -2968,6 +2968,7 @@ static long btrfs_ioctl_file_extent_same
 
 	if (IS_ERR(same)) {
 		ret = PTR_ERR(same);
+		same = NULL;
 		goto out;
 	}
 

@@ -3038,6 +3039,7 @@ static long btrfs_ioctl_file_extent_same
 
 out:
 	mnt_drop_write_file(file);
+	kfree(same);
 	return ret;
 }
 

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Filipe Manana <[email protected]>

commit d3efe08400317888f559bbedf0e42cd31575d0ef upstream.

When we call btrfs_commit_transaction(), we splice the list "ordered" of our transaction handle into the transaction's "pending_ordered" list, but we don't re-initialize the "ordered" list of our transaction handle, this means it still points to the same elements it used to before the splice. Then we check if the current transaction's state is

btrfs_end_transaction() which simply splices again the "ordered" list of our handle into the transaction's "pending_ordered" list, leaving multiple pointers to the same ordered extents which results in list corruption when we are iterating, removing and freeing ordered extents at btrfs_wait_pending_ordered(), resulting in access to dangling pointers / use-after-free issues.
Similarly, btrfs_end_transaction() can end up in some cases calling btrfs_commit_transaction(), and both did a list splice of the transaction handle's "ordered" list into the transaction's "pending_ordered" without re-initializing the handle's "ordered" list, resulting in exactly the same problem.

This produces the following warning on a kernel with linked list debugging enabled:

[109749.265416] ------------[ cut here ]------------
[109749.266410] WARNING: CPU: 7 PID: 324 at lib/list_debug.c:59 __list_del_entry+0x5a/0x98()
[109749.267969] list_del corruption. prev->next should be ffff8800ba087e20, but was fffffff8c1f7c35d
(...)
[109749.287505] Call Trace:
[109749.288135]  [<ffffffff8145f077>] dump_stack+0x4f/0x7b
[109749.298080]  [<ffffffff81095de5>] ? console_unlock+0x356/0x3a2
[109749.331605]  [<ffffffff8104b3b0>] warn_slowpath_common+0xa1/0xbb
[109749.334849]  [<ffffffff81260642>] ? __list_del_entry+0x5a/0x98
[109749.337093]  [<ffffffff8104b410>] warn_slowpath_fmt+0x46/0x48
[109749.337847]  [<ffffffff81260642>] __list_del_entry+0x5a/0x98
[109749.338678]  [<ffffffffa053e8bf>] btrfs_wait_pending_ordered+0x46/0xdb [btrfs]
[109749.340145]  [<ffffffffa058a65f>] ? __btrfs_run_delayed_items+0x149/0x163 [btrfs]
[109749.348313]  [<ffffffffa054077d>] btrfs_commit_transaction+0x36b/0xa10 [btrfs]
[109749.349745]  [<ffffffff81087310>] ? trace_hardirqs_on+0xd/0xf
[109749.350819]  [<ffffffffa055370d>] btrfs_sync_file+0x36f/0x3fc [btrfs]
[109749.351976]  [<ffffffff8118ec98>] vfs_fsync_range+0x8f/0x9e
[109749.360341]  [<ffffffff8118ecc3>] vfs_fsync+0x1c/0x1e
[109749.368828]  [<ffffffff8118ee1d>] do_fsync+0x34/0x4e
[109749.369790]  [<ffffffff8118f045>] SyS_fsync+0x10/0x14
[109749.370925]  [<ffffffff81465197>] system_call_fastpath+0x12/0x6f
[109749.382274] ---[ end trace 48e0d07f7c03d95a ]---

On a non-debug kernel this leads to invalid memory accesses, causing a crash. Fix this by using list_splice_init() instead of list_splice() in btrfs_commit_transaction() and btrfs_end_transaction().

---
 fs/btrfs/transaction.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -758,7 +758,7 @@ static int __btrfs_end_transaction(struc
 
 	if (!list_empty(&trans->ordered)) {
 		spin_lock(&info->trans_lock);
-		list_splice(&trans->ordered, &cur_trans->pending_ordered);
+		list_splice_init(&trans->ordered, &cur_trans->pending_ordered);
 		spin_unlock(&info->trans_lock);
 	}
 

@@ -1848,7 +1848,7 @@ int btrfs_commit_transaction(struct btrf
 	}
 
 	spin_lock(&root->fs_info->trans_lock);
-	list_splice(&trans->ordered, &cur_trans->pending_ordered);
+	list_splice_init(&trans->ordered, &cur_trans->pending_ordered);
 	if (cur_trans->state >= TRANS_STATE_COMMIT_START) {
 		spin_unlock(&root->fs_info->trans_lock);
 		atomic_inc(&cur_trans->use_count);

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Filipe Manana <[email protected]>

commit ed958762644b404654a6f5d23e869f496fe127c6 upstream.

Using the clone ioctl (or extent_same ioctl, which calls the same extent cloning function as well) we end up allowing copy an inline extent from the source file into a non-zero offset of the destination file. This is something not expected and that the btrfs code is not prepared to deal with - all inline extents must be at a file offset equals to 0.

For example, the following excerpt of a test case for fstests triggers a crash/BUG_ON() on a write operation after an inline extent is cloned into a non-zero offset:

_scratch_mkfs >>$seqres.full 2>&1
_scratch_mount

  # Create our test files. File foo has the same 2K of data at offset 4K
  # as file bar has at its offset 0.
  $XFS_IO_PROG -f -s -c "pwrite -S 0xaa 0 4K" \
      -c "pwrite -S 0xbb 4k 2K" \
      -c "pwrite -S 0xcc 8K 4K" \
      $SCRATCH_MNT/foo | _filter_xfs_io

  # File bar consists of a single inline extent (2K size).
  $XFS_IO_PROG -f -s -c "pwrite -S 0xbb 0 2K" \
     $SCRATCH_MNT/bar | _filter_xfs_io

  # Now call the clone ioctl to clone the extent of file bar into file
  # foo at its offset 4K. This made file foo have an inline extent at
  # offset 4K, something which the btrfs code can not deal with in future
  # IO operations because all inline extents are supposed to start at an
  # offset of 0, resulting in all sorts of chaos.
  # So here we validate that clone ioctl returns an EOPNOTSUPP, which is
  # what it returns for other cases dealing with inlined extents.
  $CLONER_PROG -s 0 -d $((4 * 1024)) -l $((2 * 1024)) \
      $SCRATCH_MNT/bar $SCRATCH_MNT/foo

# Because of the inline extent at offset 4K, the following write made

  # the kernel crash with a BUG_ON().
  $XFS_IO_PROG -c "pwrite -S 0xdd 6K 2K" $SCRATCH_MNT/foo | _filter_xfs_io

status=0
exit

The stack trace of the BUG_ON() triggered by the last write is:

  [152154.035903] ------------[ cut here ]------------
  [152154.036424] kernel BUG at mm/page-writeback.c:2286!
  [152154.036424] invalid opcode: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC
  [152154.036424] Modules linked in: btrfs dm_flakey dm_mod crc32c_generic xor raid6_pq nfsd auth_rpcgss oid_registry nfs_acl nfs lockd grace fscache sunrpc loop fuse parport_pc acpi_cpu$
  [152154.036424] CPU: 2 PID: 17873 Comm: xfs_io Tainted: G        W       4.1.0-rc6-btrfs-next-11+ #2
  [152154.036424] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.8.1-0-g4adadbd-20150316_085822-nilsson.home.kraxel.org 04/01/2014
  [152154.036424] task: ffff880429f70990 ti: ffff880429efc000 task.ti: ffff880429efc000
  [152154.036424] RIP: 0010:[<ffffffff8111a9d5>]  [<ffffffff8111a9d5>] clear_page_dirty_for_io+0x1e/0x90
  [152154.036424] RSP: 0018:ffff880429effc68  EFLAGS: 00010246
  [152154.036424] RAX: 0200000000000806 RBX: ffffea0006a6d8f0 RCX: 0000000000000001
  [152154.036424] RDX: 0000000000000000 RSI: ffffffff81155d1b RDI: ffffea0006a6d8f0
  [152154.036424] RBP: ffff880429effc78 R08: ffff8801ce389fe0 R09: 0000000000000001
  [152154.036424] R10: 0000000000002000 R11: ffffffffffffffff R12: ffff8800200dce68
  [152154.036424] R13: 0000000000000000 R14: ffff8800200dcc88 R15: ffff8803d5736d80
  [152154.036424] FS:  00007fbf119f6700(0000) GS:ffff88043d280000(0000) knlGS:0000000000000000
  [152154.036424] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  [152154.036424] CR2: 0000000001bdc000 CR3: 00000003aa555000 CR4: 00000000000006e0
  [152154.036424] Stack:
  [152154.036424]  ffff8803d5736d80 0000000000000001 ffff880429effcd8 ffffffffa04e97c1
  [152154.036424]  ffff880429effd68 ffff880429effd60 0000000000000001 ffff8800200dc9c8
  [152154.036424]  0000000000000001 ffff8800200dcc88 0000000000000000 0000000000001000
  [152154.036424] Call Trace:
  [152154.036424]  [<ffffffffa04e97c1>] lock_and_cleanup_extent_if_need+0x147/0x18d [btrfs]
  [152154.036424]  [<ffffffffa04ea82c>] __btrfs_buffered_write+0x245/0x4c8 [btrfs]
  [152154.036424]  [<ffffffffa04ed14b>] ? btrfs_file_write_iter+0x150/0x3e0 [btrfs]
  [152154.036424]  [<ffffffffa04ed15a>] ? btrfs_file_write_iter+0x15f/0x3e0 [btrfs]
  [152154.036424]  [<ffffffffa04ed2c7>] btrfs_file_write_iter+0x2cc/0x3e0 [btrfs]
  [152154.036424]  [<ffffffff81165a4a>] __vfs_write+0x7c/0xa5
  [152154.036424]  [<ffffffff81165f89>] vfs_write+0xa0/0xe4
  [152154.036424]  [<ffffffff81166855>] SyS_pwrite64+0x64/0x82
  [152154.036424]  [<ffffffff81465197>] system_call_fastpath+0x12/0x6f
  [152154.036424] Code: 48 89 c7 e8 0f ff ff ff 5b 41 5c 5d c3 0f 1f 44 00 00 55 48 89 e5 41 54 53 48 89 fb e8 ae ef 00 00 49 89 c4 48 8b 03 a8 01 75 02 <0f> 0b 4d 85 e4 74 59 49 8b 3c 2$
  [152154.036424] RIP  [<ffffffff8111a9d5>] clear_page_dirty_for_io+0x1e/0x90
  [152154.036424]  RSP <ffff880429effc68>
  [152154.242621] ---[ end trace e3d3376b23a57041 ]---

Fix this by returning the error EOPNOTSUPP if an attempt to copy an inline extent into a non-zero offset happens, just like what is done for other scenarios that would require copying/splitting inline extents, which were introduced by the following commits:

00fdf13a2e9f ("Btrfs: fix a crash of clone with inline extents's split") 3f9e3df8da3c ("btrfs: replace error code from btrfs_drop_extents")

---
 fs/btrfs/ioctl.c |   14 ++++++++++++++
 1 file changed, 14 insertions(+)

@@ -3434,6 +3434,20 @@ process_slot:
 				u64 trim = 0;
 				u64 aligned_end = 0;
 
+				/*
+				 * Don't copy an inline extent into an offset
+				 * greater than zero. Having an inline extent
+				 * at such an offset results in chaos as btrfs
+				 * isn't prepared for such cases. Just skip
+				 * this case for the same reasons as commented
+				 * at btrfs_ioctl_clone().
+				 */
+				if (last_dest_end > 0) {
+					ret = -EOPNOTSUPP;
+					btrfs_end_transaction(trans, root);
+					goto out;
+				}
+
 				if (off > key.offset) {
 					skip = off - key.offset;
 					new_key.offset += skip;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Paul Moore <[email protected]>

commit 3324603524925c7727207027d1c15e597412d15e upstream.

At present we don't create efficient ebitmaps when importing NetLabel category bitmaps. This can present a problem when comparing ebitmaps since ebitmap_cmp() is very strict about these things and considers these wasteful ebitmaps not equal when compared to their more efficient counterparts, even if their values are the same. This isn't likely to cause problems on 64-bit systems due to a bit of luck on how NetLabel/CIPSO works and the default ebitmap size, but it can be a problem on 32-bit systems.

This patch fixes this problem by being a bit more intelligent when importing NetLabel category bitmaps by skipping over empty sections which should result in a nice, efficient ebitmap.

---
 security/selinux/ss/ebitmap.c |    6 ++++++
 1 file changed, 6 insertions(+)

@@ -153,6 +153,12 @@ int ebitmap_netlbl_import(struct ebitmap
 		if (offset == (u32)-1)
 			return 0;
 
+		/* don't waste ebitmap space if the netlabel bitmap is empty */
+		if (bitmap == 0) {
+			offset += EBITMAP_UNIT_SIZE;
+			continue;
+		}
+
 		if (e_iter == NULL ||
 		    offset >= e_iter->startbit + EBITMAP_SIZE) {
 			e_prev = e_iter;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Stephen Smalley <[email protected]>

commit 892e8cac99a71f6254f84fc662068d912e1943bf upstream.

commit 66fc13039422ba7df2d01a8ee0873e4ef965b50b ("mm: shmem_zero_setup skip security check and lockdep conflict with XFS") caused a regression for SELinux by disabling any SELinux checking of mprotect PROT_EXEC on shared anonymous mappings. However, even before that regression, the checking on such mprotect PROT_EXEC calls was inconsistent with the checking on a mmap PROT_EXEC call for a shared anonymous mapping. On a mmap, the security hook is passed a NULL file and knows it is dealing with an anonymous mapping and therefore applies an execmem check and no file checks. On a mprotect, the security hook is passed a vma with a non-NULL vm_file (as this was set from the internally-created shmem file during mmap) and therefore applies the file-based execute check and no execmem check. Since the aforementioned commit now marks the shmem zero inode with the S_PRIVATE flag, the file checks are disabled and we have no checking at all on mprotect PROT_EXEC. Add a test to the mprotect hook logic for such private inodes, and apply an execmem check in that case. This makes the mmap and mprotect checking consistent for shared anonymous mappings, as well as for /dev/zero and ashmem.

---
 security/selinux/hooks.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -3288,7 +3288,8 @@ static int file_map_prot_check(struct fi
 	int rc = 0;
 
 	if (default_noexec &&
-	    (prot & PROT_EXEC) && (!file || (!shared && (prot & PROT_WRITE)))) {
+	    (prot & PROT_EXEC) && (!file || IS_PRIVATE(file_inode(file)) ||
+				   (!shared && (prot & PROT_WRITE)))) {
 		/*
 		 * We are making executable an anonymous mapping or a
 		 * private file mapping that will also be writable.

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Miklos Szeredi <[email protected]>

commit 0ad0b3255a08020eaf50e34ef0d6df5bdf5e09ed upstream.

fc->release is called from fuse_conn_put() which was used in the error cleanup before fc->release was initialized.

[Jeremiah Mahler <[email protected]>: assign fc->release after calling fuse_conn_init(fc) instead of before.]

---
 fs/fuse/inode.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1026,6 +1026,7 @@ static int fuse_fill_super(struct super_
 		goto err_fput;
 
 	fuse_conn_init(fc);
+	fc->release = fuse_free_conn;
 
 	fc->dev = sb->s_dev;
 	fc->sb = sb;

@@ -1040,7 +1041,6 @@ static int fuse_fill_super(struct super_
 		fc->dont_mask = 1;
 	sb->s_flags |= MS_POSIXACL;
 
-	fc->release = fuse_free_conn;
 	fc->flags = d.flags;
 	fc->user_id = d.user_id;
 	fc->group_id = d.group_id;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Ilya Dryomov <[email protected]>

commit 82cd003a77173c91b9acad8033fb7931dac8d751 upstream.

struct crush_bucket_tree::num_nodes is u8, so ceph_decode_8_safe() should be used. -Wconversion catches this, but I guess it went unnoticed in all the noise it spews. The actual problem (at least for common crushmaps) isn't the u32 -> u8 truncation though - it's the advancement by 4 bytes instead of 1 in the crushmap buffer.

---
 net/ceph/osdmap.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -89,7 +89,7 @@ static int crush_decode_tree_bucket(void
 {
 	int j;
 	dout("crush_decode_tree_bucket %p to %p\n", *p, end);
-	ceph_decode_32_safe(p, end, b->num_nodes, bad);
+	ceph_decode_8_safe(p, end, b->num_nodes, bad);
 	b->node_weights = kcalloc(b->num_nodes, sizeof(u32), GFP_NOFS);
 	if (b->node_weights == NULL)
 		return -ENOMEM;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Dan Carpenter <[email protected]>

commit 7bc10388ccdd79b3d20463151a1f8e7a590a775b upstream.

There is a small memory leak on error.

---
 drivers/acpi/resource.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

@@ -660,8 +660,10 @@ static int add_region_before(u64 start,
 		return -ENOMEM;
 
 	error = request_range(start, end, space_id, flags, desc);
-	if (error)
+	if (error) {
+		kfree(reg);
 		return error;
+	}
 
 	reg->start = start;
 	reg->end = end;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: "Rafael J. Wysocki" <[email protected]>

commit 0294112ee3135fbd15eaa70015af8283642dd970 upstream.

This effectively reverts the following three commits:

7bc10388ccdd ACPI / resources: free memory on error in add_region_before() 0f1b414d1907 ACPI / PNP: Avoid conflicting resource reservations b9a5e5e18fbf ACPI / init: Fix the ordering of acpi_reserve_resources()

(commit b9a5e5e18fbf introduced regressions some of which, but not all, were addressed by commit 0f1b414d1907 and commit 7bc10388ccdd was a fixup on top of the latter) and causes ACPI fixed hardware resources to be reserved at the fs_initcall_sync stage of system initialization.

The story is as follows. First, a boot regression was reported due to an apparent resource reservation ordering change after a commit that shouldn't lead to such changes. Investigation led to the conclusion that the problem happened because acpi_reserve_resources() was executed at the device_initcall() stage of system initialization which wasn't strictly ordered with respect to driver initialization (and with respect to the initialization of the pcieport driver in particular), so a random change causing the device initcalls to be run in a different order might break things.

The response to that was to attempt to run acpi_reserve_resources() as soon as we knew that ACPI would be in use (commit b9a5e5e18fbf). However, that turned out to be too early, because it caused resource reservations made by the PNP system driver to fail on at least one system and that failure was addressed by commit 0f1b414d1907.

That fix still turned out to be insufficient, though, because calling acpi_reserve_resources() before the fs_initcall stage of system initialization caused a boot regression to happen on the eCAFE EC-800-H20G/S netbook. That meant that we only could call acpi_reserve_resources() at the fs_initcall initialization stage or later, but then we might just as well call it after the PNP initalization in which case commit 0f1b414d1907 wouldn't be necessary any more.

For this reason, the changes made by commit 0f1b414d1907 are reverted (along with a memory leak fixup on top of that commit), the changes made by commit b9a5e5e18fbf that went too far are reverted too and acpi_reserve_resources() is changed into fs_initcall_sync, which will cause it to be executed after the PNP subsystem initialization (which is an fs_initcall) and before device initcalls (including the pcieport driver initialization) which should avoid the initial issue.

---
 drivers/acpi/osl.c      |   12 ++-
 drivers/acpi/resource.c |  162 ------------------------------------------------
 drivers/pnp/system.c    |   35 ++--------
 include/linux/acpi.h    |   10 --
 4 files changed, 18 insertions(+), 201 deletions(-)

@@ -175,10 +175,14 @@ static void __init acpi_request_region (
 	if (!addr || !length)
 		return;
 
-	acpi_reserve_region(addr, length, gas->space_id, 0, desc);
+	/* Resources are never freed */
+	if (gas->space_id == ACPI_ADR_SPACE_SYSTEM_IO)
+		request_region(addr, length, desc);
+	else if (gas->space_id == ACPI_ADR_SPACE_SYSTEM_MEMORY)
+		request_mem_region(addr, length, desc);
 }
 
-static void __init acpi_reserve_resources(void)
+static int __init acpi_reserve_resources(void)
 {
 	acpi_request_region(&acpi_gbl_FADT.xpm1a_event_block, acpi_gbl_FADT.pm1_event_length,
 		"ACPI PM1a_EVT_BLK");

@@ -207,7 +211,10 @@ static void __init acpi_reserve_resource
 	if (!(acpi_gbl_FADT.gpe1_block_length & 0x1))
 		acpi_request_region(&acpi_gbl_FADT.xgpe1_block,
 			       acpi_gbl_FADT.gpe1_block_length, "ACPI GPE1_BLK");
+
+	return 0;
 }
+fs_initcall_sync(acpi_reserve_resources);
 
 void acpi_os_printf(const char *fmt, ...)
 {

@@ -1838,7 +1845,6 @@ acpi_status __init acpi_os_initialize(vo
 
 acpi_status __init acpi_os_initialize1(void)
 {
-	acpi_reserve_resources();
 	kacpid_wq = alloc_workqueue("kacpid", 0, 1);
 	kacpi_notify_wq = alloc_workqueue("kacpi_notify", 0, 1);
 	kacpi_hotplug_wq = alloc_ordered_workqueue("kacpi_hotplug", 0);

@@ -7,7 +7,6 @@
  *	Bjorn Helgaas <[email protected]>
  */
 
-#include <linux/acpi.h>
 #include <linux/pnp.h>
 #include <linux/device.h>
 #include <linux/init.h>

@@ -23,41 +22,25 @@ static const struct pnp_device_id pnp_de
 	{"", 0}
 };
 
-#ifdef CONFIG_ACPI
-static bool __reserve_range(u64 start, unsigned int length, bool io, char *desc)
-{
-	u8 space_id = io ? ACPI_ADR_SPACE_SYSTEM_IO : ACPI_ADR_SPACE_SYSTEM_MEMORY;
-	return !acpi_reserve_region(start, length, space_id, IORESOURCE_BUSY, desc);
-}
-#else
-static bool __reserve_range(u64 start, unsigned int length, bool io, char *desc)
-{
-	struct resource *res;
-
-	res = io ? request_region(start, length, desc) :
-		request_mem_region(start, length, desc);
-	if (res) {
-		res->flags &= ~IORESOURCE_BUSY;
-		return true;
-	}
-	return false;
-}
-#endif
-
 static void reserve_range(struct pnp_dev *dev, struct resource *r, int port)
 {
 	char *regionid;
 	const char *pnpid = dev_name(&dev->dev);
 	resource_size_t start = r->start, end = r->end;
-	bool reserved;
+	struct resource *res;
 
 	regionid = kmalloc(16, GFP_KERNEL);
 	if (!regionid)
 		return;
 
 	snprintf(regionid, 16, "pnp %s", pnpid);
-	reserved = __reserve_range(start, end - start + 1, !!port, regionid);
-	if (!reserved)
+	if (port)
+		res = request_region(start, end - start + 1, regionid);
+	else
+		res = request_mem_region(start, end - start + 1, regionid);
+	if (res)
+		res->flags &= ~IORESOURCE_BUSY;
+	else
 		kfree(regionid);
 
 	/*

@@ -66,7 +49,7 @@ static void reserve_range(struct pnp_dev
 	 * have double reservations.
 	 */
 	dev_info(&dev->dev, "%pR %s reserved\n", r,
-		 reserved ? "has been" : "could not be");
+		 res ? "has been" : "could not be");
 }
 
 static void reserve_resources_of_dev(struct pnp_dev *dev)

@@ -332,9 +332,6 @@ int acpi_check_region(resource_size_t st
 
 int acpi_resources_are_enforced(void);
 
-int acpi_reserve_region(u64 start, unsigned int length, u8 space_id,
-			unsigned long flags, char *desc);
-
 #ifdef CONFIG_HIBERNATION
 void __init acpi_no_s4_hw_signature(void);
 #endif

@@ -530,13 +527,6 @@ static inline int acpi_check_region(reso
 	return 0;
 }
 
-static inline int acpi_reserve_region(u64 start, unsigned int length,
-				      u8 space_id, unsigned long flags,
-				      char *desc)
-{
-	return -ENXIO;
-}
-
 struct acpi_table_header;
 static inline int acpi_table_parse(char *id,
 				int (*handler)(struct acpi_table_header *))

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: "Rafael J. Wysocki" <[email protected]>

commit d3e13ff3c1aa2403d9a5f371baac088daeb8f56d upstream.

Fix a return value (which should be a negative error code) and a memory leak (the list allocated by acpi_dev_get_resources() needs to be freed on ioremap() errors too) in acpi_lpss_create_device() introduced by commit 4483d59e29fe 'ACPI / LPSS: check the result of ioremap()'.

---
 drivers/acpi/acpi_lpss.c |    7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

@@ -352,13 +352,16 @@ static int acpi_lpss_create_device(struc
 				pdata->mmio_size = resource_size(rentry->res);
 			pdata->mmio_base = ioremap(rentry->res->start,
 						   pdata->mmio_size);
-			if (!pdata->mmio_base)
-				goto err_out;
 			break;
 		}
 
 	acpi_dev_free_resource_list(&resource_list);
 
+	if (!pdata->mmio_base) {
+		ret = -ENOMEM;
+		goto err_out;
+	}
+
 	pdata->dev_desc = dev_desc;
 
 	if (dev_desc->setup)

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Lv Zheng <[email protected]>

commit c04e1fb4396d27f18296db0f914760fa7fe8223a upstream.

ACPICA commit f7b86f35416e3d1f71c3d816ff5075ddd33ed486

The following commit is reported to have broken s2ram on some platforms:
Commit: 0249ed2444d65d65fc3f3f64f398f1ad0b7e54cd
ACPICA: Add option to favor 32-bit FADT addresses.
The platform reports 2 FACS tables (which is not allowed by ACPI specification) and the new 32-bit address favor rule forces OSPMs to use the FACS table reported via FADT's X_FIRMWARE_CTRL field.

The root cause of the reported bug might be one of the followings:
1. BIOS may favor the 64-bit firmware waking vector address when the version of the FACS is greater than 0 and Linux currently only supports resuming from the real mode, so the 64-bit firmware waking vector has never been set and might be invalid to BIOS while the commit enables higher version FACS.
2. BIOS may favor the FACS reported via the "FIRMWARE_CTRL" field in the FADT while the commit doesn't set the firmware waking vector address of the FACS reported by "FIRMWARE_CTRL", it only sets the firware waking vector address of the FACS reported by "X_FIRMWARE_CTRL".

This patch excludes the cases that can trigger the bugs caused by the root cause 2.

There is no handshaking mechanism can be used by OSPM to tell BIOS which FACS is currently used. Thus the FACS reported by "FIRMWARE_CTRL" may still be used by BIOS and the 0 value of the 32-bit firmware waking vector might trigger such failure.

This patch tries to favor 32bit FACS address in another way where both the FACS reported by "FIRMWARE_CTRL" and the FACS reported by "X_FIRMWARE_CTRL" are loaded so that further commit can set firmware waking vector in the both tables to ensure we can exclude the cases that trigger the bugs caused by the root cause 2. The exclusion is split into 2 commits as this commit is also useful for dumping more ACPI tables, it won't get reverted when such exclusion is no longer necessary. Lv Zheng.

---
 drivers/acpi/acpica/aclocal.h  |    1 +
 drivers/acpi/acpica/tbfadt.c   |   21 +++++++++++++--------
 drivers/acpi/acpica/tbutils.c  |   34 +++++++++++++++++++++++-----------
 drivers/acpi/acpica/tbxfload.c |    3 ++-
 include/acpi/acpixf.h          |    9 +++++++++
 5 files changed, 48 insertions(+), 20 deletions(-)

@@ -213,6 +213,7 @@ struct acpi_table_list {
 
 #define ACPI_TABLE_INDEX_DSDT           (0)
 #define ACPI_TABLE_INDEX_FACS           (1)
+#define ACPI_TABLE_INDEX_X_FACS         (2)
 
 struct acpi_find_context {
 	char *search_for;

@@ -350,9 +350,18 @@ void acpi_tb_parse_fadt(u32 table_index)
 	/* If Hardware Reduced flag is set, there is no FACS */
 
 	if (!acpi_gbl_reduced_hardware) {
-		acpi_tb_install_fixed_table((acpi_physical_address)
-					    acpi_gbl_FADT.Xfacs, ACPI_SIG_FACS,
-					    ACPI_TABLE_INDEX_FACS);
+		if (acpi_gbl_FADT.facs) {
+			acpi_tb_install_fixed_table((acpi_physical_address)
+						    acpi_gbl_FADT.facs,
+						    ACPI_SIG_FACS,
+						    ACPI_TABLE_INDEX_FACS);
+		}
+		if (acpi_gbl_FADT.Xfacs) {
+			acpi_tb_install_fixed_table((acpi_physical_address)
+						    acpi_gbl_FADT.Xfacs,
+						    ACPI_SIG_FACS,
+						    ACPI_TABLE_INDEX_X_FACS);
+		}
 	}
 }
 

@@ -491,13 +500,9 @@ static void acpi_tb_convert_fadt(void)
 	acpi_gbl_FADT.header.length = sizeof(struct acpi_table_fadt);
 
 	/*
-	 * Expand the 32-bit FACS and DSDT addresses to 64-bit as necessary.
+	 * Expand the 32-bit DSDT addresses to 64-bit as necessary.
 	 * Later ACPICA code will always use the X 64-bit field.
 	 */
-	acpi_gbl_FADT.Xfacs = acpi_tb_select_address("FACS",
-						     acpi_gbl_FADT.facs,
-						     acpi_gbl_FADT.Xfacs);
-
 	acpi_gbl_FADT.Xdsdt = acpi_tb_select_address("DSDT",
 						     acpi_gbl_FADT.dsdt,
 						     acpi_gbl_FADT.Xdsdt);

@@ -68,7 +68,8 @@ acpi_tb_get_root_table_entry(u8 *table_e
 
 acpi_status acpi_tb_initialize_facs(void)
 {
-	acpi_status status;
+	struct acpi_table_facs *facs32;
+	struct acpi_table_facs *facs64;
 
 	/* If Hardware Reduced flag is set, there is no FACS */
 

@@ -77,11 +78,22 @@ acpi_status acpi_tb_initialize_facs(void
 		return (AE_OK);
 	}
 
-	status = acpi_get_table_by_index(ACPI_TABLE_INDEX_FACS,
-					 ACPI_CAST_INDIRECT_PTR(struct
-								acpi_table_header,
-								&acpi_gbl_FACS));
-	return (status);
+	(void)acpi_get_table_by_index(ACPI_TABLE_INDEX_FACS,
+				      ACPI_CAST_INDIRECT_PTR(struct
+							     acpi_table_header,
+							     &facs32));
+	(void)acpi_get_table_by_index(ACPI_TABLE_INDEX_X_FACS,
+				      ACPI_CAST_INDIRECT_PTR(struct
+							     acpi_table_header,
+							     &facs64));
+
+	if (acpi_gbl_use32_bit_facs_addresses) {
+		acpi_gbl_FACS = facs32 ? facs32 : facs64;
+	} else {
+		acpi_gbl_FACS = facs64 ? facs64 : facs32;
+	}
+
+	return (AE_OK);
 }
 #endif				/* !ACPI_REDUCED_HARDWARE */
 

@@ -101,7 +113,7 @@ acpi_status acpi_tb_initialize_facs(void
 u8 acpi_tb_tables_loaded(void)
 {
 
-	if (acpi_gbl_root_table_list.current_table_count >= 3) {
+	if (acpi_gbl_root_table_list.current_table_count >= 4) {
 		return (TRUE);
 	}
 

@@ -357,11 +369,11 @@ acpi_status __init acpi_tb_parse_root_ta
 	table_entry = ACPI_ADD_PTR(u8, table, sizeof(struct acpi_table_header));
 
 	/*
-	 * First two entries in the table array are reserved for the DSDT
-	 * and FACS, which are not actually present in the RSDT/XSDT - they
-	 * come from the FADT
+	 * First three entries in the table array are reserved for the DSDT
+	 * and 32bit/64bit FACS, which are not actually present in the
+	 * RSDT/XSDT - they come from the FADT
 	 */
-	acpi_gbl_root_table_list.current_table_count = 2;
+	acpi_gbl_root_table_list.current_table_count = 3;
 
 	/* Initialize the root table array from the RSDT/XSDT */
 

@@ -166,7 +166,8 @@ static acpi_status acpi_tb_load_namespac
 
 	(void)acpi_ut_acquire_mutex(ACPI_MTX_TABLES);
 	for (i = 0; i < acpi_gbl_root_table_list.current_table_count; ++i) {
-		if ((!ACPI_COMPARE_NAME
+		if (!acpi_gbl_root_table_list.tables[i].address ||
+		    (!ACPI_COMPARE_NAME
 		     (&(acpi_gbl_root_table_list.tables[i].signature),
 		      ACPI_SIG_SSDT)
 		     &&

@@ -200,6 +200,15 @@ ACPI_INIT_GLOBAL(u8, acpi_gbl_do_not_use
 ACPI_INIT_GLOBAL(u8, acpi_gbl_use32_bit_fadt_addresses, TRUE);
 
 /*
+ * Optionally use 32-bit FACS table addresses.
+ * It is reported that some platforms fail to resume from system suspending
+ * if 64-bit FACS table address is selected:
+ * https://bugzilla.kernel.org/show_bug.cgi?id=74021
+ * Default is TRUE, favor the 32-bit addresses.
+ */
+ACPI_INIT_GLOBAL(u8, acpi_gbl_use32_bit_facs_addresses, TRUE);
+
+/*
  * Optionally truncate I/O addresses to 16 bits. Provides compatibility
  * with other ACPI implementations. NOTE: During ACPICA initialization,
  * this value is set to TRUE if any Windows OSI strings have been

This particular patch is not stable. Lv has fixed it and a new patch is forthcoming, at least a patch on top of this.

If it's not in Linus's tree yet, nothing I can do about it. Please provide the git commit id for that patch when it happens.

thanks,

greg k-h

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Lv Zheng <[email protected]>

commit c04be18448355441a0c424362df65b6422e27bda upstream.

ACPICA commit 90f5332a15e9d9ba83831ca700b2b9f708274658

This patch adds a new FACS initialization flag for acpi_tb_initialize(). acpi_enable_subsystem() might be invoked several times in OS bootup process, and we don't want FACS initialization to be invoked twice. Lv Zheng.

---
 drivers/acpi/acpica/utxfinit.c |   10 ++++++----
 include/acpi/actypes.h         |    1 +
 2 files changed, 7 insertions(+), 4 deletions(-)

@@ -179,10 +179,12 @@ acpi_status __init acpi_enable_subsystem
 	 * Obtain a permanent mapping for the FACS. This is required for the
 	 * Global Lock and the Firmware Waking Vector
 	 */
-	status = acpi_tb_initialize_facs();
-	if (ACPI_FAILURE(status)) {
-		ACPI_WARNING((AE_INFO, "Could not map the FACS table"));
-		return_ACPI_STATUS(status);
+	if (!(flags & ACPI_NO_FACS_INIT)) {
+		status = acpi_tb_initialize_facs();
+		if (ACPI_FAILURE(status)) {
+			ACPI_WARNING((AE_INFO, "Could not map the FACS table"));
+			return_ACPI_STATUS(status);
+		}
 	}
 #endif				/* !ACPI_REDUCED_HARDWARE */
 

@@ -572,6 +572,7 @@ typedef u64 acpi_integer;
 #define ACPI_NO_ACPI_ENABLE             0x10
 #define ACPI_NO_DEVICE_INIT             0x20
 #define ACPI_NO_OBJECT_INIT             0x40
+#define ACPI_NO_FACS_INIT               0x80
 
 /*
  * Initialization state

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Lv Zheng <[email protected]>

commit 0ea61381788a37d864f9841b0fe97d40f7058f3b upstream.

ACPICA commit 4da56eeae0749dfe8491285c1e1fad48f6efafd8

The following commit temporarily disables correct 64-bit FADT addresses favor during the period the root cause of the bug is not fixed:
Commit: 85dbd5801f62b66e2aa7826aaefcaebead44c8a6
ACPICA: Tables: Restore old behavor to favor 32-bit FADT addresses.

With enough protections, this patch re-enables 64-bit FADT addresses by default. If regressions are reported against such change, this patch should be bisected and reverted.
Note that 64-bit FACS favor and 64-bit firmware waking vector favor are excluded by this commit in order not to break OSPMs. Lv Zheng.

---
 include/acpi/acpixf.h |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -195,9 +195,9 @@ ACPI_INIT_GLOBAL(u8, acpi_gbl_do_not_use
  * address. Although ACPICA adheres to the ACPI specification which
  * requires the use of the corresponding 64-bit address if it is non-zero,
  * some machines have been found to have a corrupted non-zero 64-bit
- * address. Default is TRUE, favor the 32-bit addresses.
+ * address. Default is FALSE, do not favor the 32-bit addresses.
  */
-ACPI_INIT_GLOBAL(u8, acpi_gbl_use32_bit_fadt_addresses, TRUE);
+ACPI_INIT_GLOBAL(u8, acpi_gbl_use32_bit_fadt_addresses, FALSE);
 
 /*
  * Optionally use 32-bit FACS table addresses.

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Jiang Liu <[email protected]>

commit 1fb01ca93a1348a1469b8777326cd7632483de77 upstream.

Zoltan Boszormenyi reported this regression:
"There's a Realtek RTL8111/8168/8411 (PCI ID 10ec:8168, Subsystem ID
1565:230e) network chip on the mainboard. After the r8169 driver loaded the IRQs in the machine went berserk. Keyboard keypressed arrived with considerable latency and duplicated, so no real work was possible. The machine responded to the power button but didn't actually power down. It just stuck at the powering down message. I had to press the power button for 4 seconds to power it down.

The computer is a POS machine with a big battery inside. Because of this, either ACPI or the Realtek chip kept the bad state and after rebooting, the network chip didn't even show up in lspci. Not even the PXE ROM announced itself during boot. I had to disconnect the battery to beat some sense back to the computer.

The regression happens with 4.0.5, 4.1.0-rc8 and 4.1.0-final. 3.18.16 was good."

The regression is caused by commit 593669c2ac0f (x86/PCI/ACPI: Use common ACPI resource interfaces to simplify implementation). Since commit 593669c2ac0f, x86 PCI ACPI host bridge driver validates ACPI resources by first converting an ACPI resource to a 'struct resource' structure and then applying checks against the converted resource structure. The 'start' and 'end' fields in 'struct resource' are defined to be type of resource_size_t, which may be 32 bits or 64 bits depending on CONFIG_PHYS_ADDR_T_64BIT.

This may cause incorrect resource validation results with 32-bit kernels because 64-bit ACPI resource descriptors may get truncated when converting to 32-bit 'start' and 'end' fields in 'struct resource'. It eventually affects PCI resource allocation subsystem and makes some PCI devices and the system behave abnormally due to incorrect resource assignment.

So enhance the ACPI resource parsing interfaces to ignore ACPI resource descriptors with address/offset above 4G when running in 32-bit mode.

With the fix applied, the behavior of the machine was restored to how 3.18.16 worked, i.e. the memory range that is over 4GB is ignored again, and lspci -vvxxx shows that everything is at the same memory window as they were with 3.18.16.

---
 drivers/acpi/resource.c |   24 +++++++++++++++---------
 1 file changed, 15 insertions(+), 9 deletions(-)

@@ -193,6 +193,7 @@ static bool acpi_decode_space(struct res
 	u8 iodec = attr->granularity == 0xfff ? ACPI_DECODE_10 : ACPI_DECODE_16;
 	bool wp = addr->info.mem.write_protect;
 	u64 len = attr->address_length;
+	u64 start, end, offset = 0;
 	struct resource *res = &win->res;
 
 	/*

@@ -204,9 +205,6 @@ static bool acpi_decode_space(struct res
 		pr_debug("ACPI: Invalid address space min_addr_fix %d, max_addr_fix %d, len %llx\n",
 			 addr->min_address_fixed, addr->max_address_fixed, len);
 
-	res->start = attr->minimum;
-	res->end = attr->maximum;
-
 	/*
 	 * For bridges that translate addresses across the bridge,
 	 * translation_offset is the offset that must be added to the

@@ -214,12 +212,22 @@ static bool acpi_decode_space(struct res
 	 * primary side. Non-bridge devices must list 0 for all Address
 	 * Translation offset bits.
 	 */
-	if (addr->producer_consumer == ACPI_PRODUCER) {
-		res->start += attr->translation_offset;
-		res->end += attr->translation_offset;
-	} else if (attr->translation_offset) {
+	if (addr->producer_consumer == ACPI_PRODUCER)
+		offset = attr->translation_offset;
+	else if (attr->translation_offset)
 		pr_debug("ACPI: translation_offset(%lld) is invalid for non-bridge device.\n",
 			 attr->translation_offset);
+	start = attr->minimum + offset;
+	end = attr->maximum + offset;
+
+	win->offset = offset;
+	res->start = start;
+	res->end = end;
+	if (sizeof(resource_size_t) < sizeof(u64) &&
+	    (offset != win->offset || start != res->start || end != res->end)) {
+		pr_warn("acpi resource window ([%#llx-%#llx] ignored, not CPU addressable)\n",
+			attr->minimum, attr->maximum);
+		return false;
 	}
 
 	switch (addr->resource_type) {

@@ -236,8 +244,6 @@ static bool acpi_decode_space(struct res
 		return false;
 	}
 
-	win->offset = attr->translation_offset;
-
 	if (addr->producer_consumer == ACPI_PRODUCER)
 		res->flags |= IORESOURCE_WINDOW;
 

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Arnd Bergmann <[email protected]>

commit 357d56151976a78d90dc3dfac01777de0ef05212 upstream.

A configuration that enables earlycon but not the core console code causes a link error:

drivers/built-in.o: In function `setup_earlycon':
drivers/tty/serial/earlycon.c:70: undefined reference to `uart_parse_earlycon'

That error can be triggered by the newly added samsung earlycon support, which is missing a 'select' statement.

As suggested by Peter Hurley, solves the problem by moving the 'select SERIAL_EARLYCON' statement to the samsung console driver option, as it is done by all other console drivers.

---
 drivers/tty/serial/Kconfig |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -241,7 +241,6 @@ config SERIAL_SAMSUNG
 	tristate "Samsung SoC serial support"
 	depends on PLAT_SAMSUNG || ARCH_EXYNOS
 	select SERIAL_CORE
-	select SERIAL_EARLYCON
 	help
 	  Support for the on-chip UARTs on the Samsung S3C24XX series CPUs,
 	  providing /dev/ttySAC0, 1 and 2 (note, some machines may not

@@ -277,6 +276,7 @@ config SERIAL_SAMSUNG_CONSOLE
 	bool "Support for console on Samsung SoC serial port"
 	depends on SERIAL_SAMSUNG=y
 	select SERIAL_CORE_CONSOLE
+	select SERIAL_EARLYCON
 	help
 	  Allow selection of the S3C24XX on-board serial ports for use as
 	  an virtual console.

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Ding Wang <[email protected]>

commit 29535f7b797df35cc9b6b3bca635591cdd3dd2a8 upstream.

The current handler of MMC_BLK_CMD_ERR in mmc_blk_issue_rw_rq function may cause new coming request permanent missing when the ongoing request (previoulsy started) complete end.

The problem scenario is as follows:
(1) Request A is ongoing;
(2) Request B arrived, and finally mmc_blk_issue_rw_rq() is called;
(3) Request A encounters the MMC_BLK_CMD_ERR error;
(4) In the error handling of MMC_BLK_CMD_ERR, suppose mmc_blk_cmd_err() end request A completed and return zero. Continue the error handling, suppose mmc_blk_reset() reset device success;
(5) Continue the execution, while loop completed because variable ret

    is zero now;

(6) Finally, mmc_blk_issue_rw_rq() return without processing request B.

The process related to the missing request may wait that IO request complete forever, possibly crashing the application or hanging the system.

Fix this issue by starting new request when reset success.

---
 drivers/mmc/card/block.c |    8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

@@ -1912,9 +1912,11 @@ static int mmc_blk_issue_rw_rq(struct mm
 			break;
 		case MMC_BLK_CMD_ERR:
 			ret = mmc_blk_cmd_err(md, card, brq, req, ret);
-			if (!mmc_blk_reset(md, card->host, type))
-				break;
-			goto cmd_abort;
+			if (mmc_blk_reset(md, card->host, type))
+				goto cmd_abort;
+			if (!ret)
+				goto start_new_req;
+			break;
 		case MMC_BLK_RETRY:
 			if (retry++ < 5)
 				break;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Ulf Hansson <[email protected]>

commit 5fd26c7ecb32082745b0bd33c8e35badd1cb5a91 upstream.

Commit 3a48edc4bd68 ("mmc: sdhci: Use mmc core regulator infrastucture") changed the behavior for how to assign the ocr_avail mask for the mmc host. More precisely it started to mask the bits instead of assigning them.

Restore the behavior, but also make it clear that an OCR mask created from an external regulator overrides the other ones. The OCR mask is determined by one of the following with this priority:

1. Supported ranges of external regulator if one supplies VDD
2. Host OCR mask if set by the driver (based on DT properties)
3. The capabilities reported by the controller itself

---
 drivers/mmc/host/sdhci.c |    9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

@@ -3315,13 +3315,14 @@ int sdhci_add_host(struct sdhci_host *ho
 				   SDHCI_MAX_CURRENT_MULTIPLIER;
 	}
 
-	/* If OCR set by external regulators, use it instead */
+	/* If OCR set by host, use it instead. */
+	if (host->ocr_mask)
+		ocr_avail = host->ocr_mask;
+
+	/* If OCR set by external regulators, give it highest prio. */
 	if (mmc->ocr_avail)
 		ocr_avail = mmc->ocr_avail;
 
-	if (host->ocr_mask)
-		ocr_avail &= host->ocr_mask;
-
 	mmc->ocr_avail = ocr_avail;
 	mmc->ocr_avail_sdio = ocr_avail;
 	if (host->ocr_avail_sdio)

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Geert Uytterhoeven <[email protected]>

commit 3fc3a0be0dab352e065d1dad7d3f81953ed0d4bc upstream.

In the final iteration of commit 245bd6f6af8a62a2 ("PM / clock_ops: Add pm_clk_add_clk()"), a refcount increment was added by Grygorii Strashko. However, the accompanying IS_ERR() check operates on the wrong clock pointer, which is always zero at this point, i.e. not an error. This may lead to a NULL pointer dereference later, when __clk_get() tries to dereference an error pointer.

Check the passed clock pointer instead to fix this.

---
 drivers/base/power/clock_ops.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -93,7 +93,7 @@ static int __pm_clk_add(struct device *d
 			return -ENOMEM;
 		}
 	} else {
-		if (IS_ERR(ce->clk) || !__clk_get(clk)) {
+		if (IS_ERR(clk) || !__clk_get(clk)) {
 			kfree(ce);
 			return -ENOENT;
 		}

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Colin Ian King <[email protected]>

commit 4dc544427991e3cef38ce3ae124b7e6557063bd3 upstream.

A reorganisation of the PD allocation and deallocation in commit 9ba1377daa ("RDMA/ocrdma: Move PD resource management to driver.") introduced a double free on pd, as detected by static analysis by smatch:

drivers/infiniband/hw/ocrdma/ocrdma_verbs.c:682 ocrdma_alloc_pd()
error: double free of 'pd'^

The original call to ocrdma_mbx_dealloc_pd() (which does not kfree pd) was replaced with a call to _ocrdma_dealloc_pd() (which does kfree pd). The kfree following this call causes the double free, so just remove it to fix the problem.

---
 drivers/infiniband/hw/ocrdma/ocrdma_verbs.c |    1 -
 1 file changed, 1 deletion(-)

@@ -679,7 +679,6 @@ err:
 		ocrdma_release_ucontext_pd(uctx);
 	} else {
 		status = _ocrdma_dealloc_pd(dev, pd);
-		kfree(pd);
 	}
 exit:
 	return ERR_PTR(status);

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Arnd Bergmann <[email protected]>

commit ffb6e0c9a0572f8e5f8e9337a1b40ac2ec1493a1 upstream.

The platform_sysrq_reset_seq code was intended as a way for an embedded platform to provide its own sysrq sequence at compile time. After over two years, nobody has started using it in an upstream kernel, and the platforms that were interested in it have moved on to devicetree, which can be used to configure the sequence without requiring kernel changes. The method is also incompatible with the way that most architectures build support for multiple platforms into a single kernel.

Now the code is producing warnings when built with gcc-5.1:

drivers/tty/sysrq.c: In function 'sysrq_init':
drivers/tty/sysrq.c:959:33: warning: array subscript is above array bounds [-Warray-bounds]
key = platform_sysrq_reset_seq[i];

We could fix this, but it seems unlikely that it will ever be used, so let's just remove the code instead. We still have the option to pass the sequence either in DT, using the kernel command line, or using the /sys/module/sysrq/parameters/reset_seq file.

---
 drivers/tty/sysrq.c |   19 +------------------
 1 file changed, 1 insertion(+), 18 deletions(-)

@@ -55,9 +55,6 @@
 static int __read_mostly sysrq_enabled = CONFIG_MAGIC_SYSRQ_DEFAULT_ENABLE;
 static bool __read_mostly sysrq_always_enabled;
 
-unsigned short platform_sysrq_reset_seq[] __weak = { KEY_RESERVED };
-int sysrq_reset_downtime_ms __weak;
-
 static bool sysrq_on(void)
 {
 	return sysrq_enabled || sysrq_always_enabled;

@@ -569,6 +566,7 @@ void handle_sysrq(int key)
 EXPORT_SYMBOL(handle_sysrq);
 
 #ifdef CONFIG_INPUT
+static int sysrq_reset_downtime_ms;
 
 /* Simple translation table for the SysRq keys */
 static const unsigned char sysrq_xlate[KEY_CNT] =

@@ -949,23 +947,8 @@ static bool sysrq_handler_registered;
 
 static inline void sysrq_register_handler(void)
 {
-	unsigned short key;
 	int error;
-	int i;
-
-	/* First check if a __weak interface was instantiated. */
-	for (i = 0; i < ARRAY_SIZE(sysrq_reset_seq); i++) {
-		key = platform_sysrq_reset_seq[i];
-		if (key == KEY_RESERVED || key > KEY_MAX)
-			break;
-
-		sysrq_reset_seq[sysrq_reset_seq_len++] = key;
-	}
 
-	/*
-	 * DT configuration takes precedence over anything that would
-	 * have been defined via the __weak interface.
-	 */
 	sysrq_of_get_keyreset_config();
 
 	error = input_register_handler(&sysrq_handler);

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Naoya Horiguchi <[email protected]>

commit 641844f5616d7c6597309f560838f996466d7aac upstream.

Currently the initial value of order in dissolve_free_huge_page is 64 or 32, which leads to the following warning in static checker:

mm/hugetlb.c:1203 dissolve_free_huge_pages()
warn: potential right shift more than type allows '9,18,64'

This is a potential risk of infinite loop, because 1 << order (== 0) is used in for-loop like this:

for (pfn =3D start_pfn; pfn < end_pfn; pfn +=3D 1 << order)

      ...

So this patch fixes it by using global minimum_order calculated at boot time.

text data bss dec hex filename

   28313     469   84236  113018   1b97a mm/hugetlb.o
   28256     473   84236  112965   1b945 mm/hugetlb.o (patched)

---
 mm/hugetlb.c |   19 +++++++++++--------
 1 file changed, 11 insertions(+), 8 deletions(-)

@@ -40,6 +40,11 @@ int hugepages_treat_as_movable;
 int hugetlb_max_hstate __read_mostly;
 unsigned int default_hstate_idx;
 struct hstate hstates[HUGE_MAX_HSTATE];
+/*
+ * Minimum page order among possible hugepage sizes, set to a proper value
+ * at boot time.
+ */
+static unsigned int minimum_order __read_mostly = UINT_MAX;
 
 __initdata LIST_HEAD(huge_boot_pages);
 

@@ -1188,19 +1193,13 @@ static void dissolve_free_huge_page(stru
  */
 void dissolve_free_huge_pages(unsigned long start_pfn, unsigned long end_pfn)
 {
-	unsigned int order = 8 * sizeof(void *);
 	unsigned long pfn;
-	struct hstate *h;
 
 	if (!hugepages_supported())
 		return;
 
-	/* Set scan step to minimum hugepage size */
-	for_each_hstate(h)
-		if (order > huge_page_order(h))
-			order = huge_page_order(h);
-	VM_BUG_ON(!IS_ALIGNED(start_pfn, 1 << order));
-	for (pfn = start_pfn; pfn < end_pfn; pfn += 1 << order)
+	VM_BUG_ON(!IS_ALIGNED(start_pfn, 1 << minimum_order));
+	for (pfn = start_pfn; pfn < end_pfn; pfn += 1 << minimum_order)
 		dissolve_free_huge_page(pfn_to_page(pfn));
 }
 

@@ -1627,10 +1626,14 @@ static void __init hugetlb_init_hstates(
 	struct hstate *h;
 
 	for_each_hstate(h) {
+		if (minimum_order > huge_page_order(h))
+			minimum_order = huge_page_order(h);
+
 		/* oversize hugepages were init'ed in early boot */
 		if (!hstate_is_gigantic(h))
 			hugetlb_hstate_alloc_pages(h);
 	}
+	VM_BUG_ON(minimum_order == UINT_MAX);
 }
 
 static char * __init memfmt(char *buf, unsigned long n)

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Takashi Iwai <[email protected]>

commit fff3b16d2754a061a3549c4307a186423a0128fd upstream.

Many harddisks (mostly WD ones) have firmware problems and take too long, more than 10 seconds, to resume from suspend. And this often exceeds the default DPM watchdog timeout (12 seconds), resulting in a kernel panic out of sudden.

Since most distros just take the default as is, we should give a bit more safer value. This patch increases the default value from 12 seconds to one minute, which has been confirmed to be long enough for such problematic disks.

---
 kernel/power/Kconfig |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -187,7 +187,7 @@ config DPM_WATCHDOG
 config DPM_WATCHDOG_TIMEOUT
 	int "Watchdog timeout in seconds"
 	range 1 120
-	default 12
+	default 60
 	depends on DPM_WATCHDOG
 
 config PM_TRACE

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Jean Delvare <[email protected]>

commit 17cd5bd5391e6e7b363d66335e1bc6760ae969b9 upstream.

A 32-bit entry point to a DMI table says how many structures the table contains. The SMBIOS specification explicitly says that end-of-table markers should be ignored if they are not actually at the end of the DMI table. So only honor the end-of-table marker for tables accessed through 64-bit entry points, as they do not specify a structure count.

---
 drivers/firmware/dmi_scan.c |   13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)

@@ -89,9 +89,9 @@ static void dmi_table(u8 *buf,
 
 	/*
 	 * Stop when we have seen all the items the table claimed to have
-	 * (SMBIOS < 3.0 only) OR we reach an end-of-table marker OR we run
-	 * off the end of the table (should never happen but sometimes does
-	 * on bogus implementations.)
+	 * (SMBIOS < 3.0 only) OR we reach an end-of-table marker (SMBIOS
+	 * >= 3.0 only) OR we run off the end of the table (should never
+	 * happen but sometimes does on bogus implementations.)
 	 */
 	while ((!dmi_num || i < dmi_num) &&
 	       (data - buf + sizeof(struct dmi_header)) <= dmi_len) {

@@ -110,8 +110,13 @@ static void dmi_table(u8 *buf,
 
 		/*
 		 * 7.45 End-of-Table (Type 127) [SMBIOS reference spec v3.0.0]
+		 * For tables behind a 64-bit entry point, we have no item
+		 * count and no exact table length, so stop on end-of-table
+		 * marker. For tables behind a 32-bit entry point, we have
+		 * seen OEM structures behind the end-of-table marker on
+		 * some systems, so don't trust it.
 		 */
-		if (dm->type == DMI_ENTRY_END_OF_TABLE)
+		if (!dmi_num && dm->type == DMI_ENTRY_END_OF_TABLE)
 			break;
 
 		data += 2;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Daniel Borkmann <[email protected]>

commit b86a50c3b5414eafdbee7f34af4a201a4a7817c2 upstream.

Cleanup commit 73679e508201 ("compiler-intel.h: Remove duplicate definition") removed the double definition of __memory_barrier() intrinsics.

However, in doing so, it also removed the preceding #undef barrier by accident, meaning, the actual barrier() macro from compiler-gcc.h with inline asm is still in place as __GNUC__ is provided.

Subsequently, barrier() can never be defined as __memory_barrier() from compiler.h since it already has a definition in place and if we trust the comment in compiler-intel.h, ecc doesn't support gcc specific asm statements.

I don't have an ecc at hand (unsure if that's still used in the field?) and only found this by accident during code review, a revert of that cleanup would be simplest option.

---
 include/linux/compiler-intel.h |    2 ++
 1 file changed, 2 insertions(+)

@@ -13,10 +13,12 @@
 /* Intel ECC compiler doesn't support gcc specific asm stmts.
  * It uses intrinsics to do the equivalent things.
  */
+#undef barrier
 #undef barrier_data
 #undef RELOC_HIDE
 #undef OPTIMIZER_HIDE_VAR
 
+#define barrier() __memory_barrier()
 #define barrier_data(ptr) barrier()
 
 #define RELOC_HIDE(ptr, off)					\

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Chris Metcalf <[email protected]>

commit 2528a8b8f457d7432552d0e2b6f0f4046bb702f4 upstream.

bitmap_parselist("", &mask, nmaskbits) will erroneously set bit zero in the mask. The same bug is visible in cpumask_parselist() since it is layered on top of the bitmask code, e.g. if you boot with "isolcpus=", you will actually end up with cpu zero isolated.

The bug was introduced in commit 4b060420a596 ("bitmap, irq: add smp_affinity_list interface to /proc/irq") when bitmap_parselist() was generalized to support userspace as well as kernelspace.

---
 lib/bitmap.c |   17 +++++++++--------
 1 file changed, 9 insertions(+), 8 deletions(-)

@@ -506,12 +506,12 @@ static int __bitmap_parselist(const char
 	unsigned a, b;
 	int c, old_c, totaldigits;
 	const char __user __force *ubuf = (const char __user __force *)buf;
-	int exp_digit, in_range;
+	int at_start, in_range;
 
 	totaldigits = c = 0;
 	bitmap_zero(maskp, nmaskbits);
 	do {
-		exp_digit = 1;
+		at_start = 1;
 		in_range = 0;
 		a = b = 0;
 

@@ -540,11 +540,10 @@ static int __bitmap_parselist(const char
 				break;
 
 			if (c == '-') {
-				if (exp_digit || in_range)
+				if (at_start || in_range)
 					return -EINVAL;
 				b = 0;
 				in_range = 1;
-				exp_digit = 1;
 				continue;
 			}
 

@@ -554,16 +553,18 @@ static int __bitmap_parselist(const char
 			b = b * 10 + (c - '0');
 			if (!in_range)
 				a = b;
-			exp_digit = 0;
+			at_start = 0;
 			totaldigits++;
 		}
 		if (!(a <= b))
 			return -EINVAL;
 		if (b >= nmaskbits)
 			return -ERANGE;
-		while (a <= b) {
-			set_bit(a, maskp);
-			a++;
+		if (!at_start) {
+			while (a <= b) {
+				set_bit(a, maskp);
+				a++;
+			}
 		}
 	} while (buflen && c == ',');
 	return 0;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Vasily Averin <[email protected]>

commit d194e5d666225b04c7754471df0948f645b6ab3a upstream.

The final version of commit 637241a900cb ("kmsg: honor dmesg_restrict sysctl on /dev/kmsg") lost few hooks, as result security_syslog() are processed incorrectly:

- open of /dev/kmsg checks syslog access permissions by using check_syslog_permissions() where security_syslog() is not called if dmesg_restrict is set.

- syslog syscall and /proc/kmsg calls do_syslog() where security_syslog can be executed twice (inside check_syslog_permissions() and then directly in do_syslog())

With this patch security_syslog() is called once only in all syslog-related operations regardless of dmesg_restrict value.

---
 kernel/printk/printk.c |   11 ++++-------
 1 file changed, 4 insertions(+), 7 deletions(-)

@@ -484,11 +484,11 @@ int check_syslog_permissions(int type, b
 	 * already done the capabilities checks at open time.
 	 */
 	if (from_file && type != SYSLOG_ACTION_OPEN)
-		return 0;
+		goto ok;
 
 	if (syslog_action_restricted(type)) {
 		if (capable(CAP_SYSLOG))
-			return 0;
+			goto ok;
 		/*
 		 * For historical reasons, accept CAP_SYS_ADMIN too, with
 		 * a warning.

@@ -498,10 +498,11 @@ int check_syslog_permissions(int type, b
 				     "CAP_SYS_ADMIN but no CAP_SYSLOG "
 				     "(deprecated).\n",
 				 current->comm, task_pid_nr(current));
-			return 0;
+			goto ok;
 		}
 		return -EPERM;
 	}
+ok:
 	return security_syslog(type);
 }
 

@@ -1263,10 +1264,6 @@ int do_syslog(int type, char __user *buf
 	if (error)
 		goto out;
 
-	error = security_syslog(type);
-	if (error)
-		return error;
-
 	switch (type) {
 	case SYSLOG_ACTION_CLOSE:	/* Close log */
 		break;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Chun-Yeow Yeoh <[email protected]>

commit 8df734e865b74d9f273216482a45a38269dc767a upstream.

The csa counter has moved from sdata to beacon/presp but it is not updated accordingly for mesh and ibss. Fix this.

---
 net/mac80211/cfg.c  |    1 +
 net/mac80211/ibss.c |    1 +
 net/mac80211/mesh.c |    1 +
 3 files changed, 3 insertions(+)

@@ -3356,6 +3356,7 @@ static int ieee80211_mgmt_tx(struct wiph
 	/* Update CSA counters */
 	if (sdata->vif.csa_active &&
 	    (sdata->vif.type == NL80211_IFTYPE_AP ||
+	     sdata->vif.type == NL80211_IFTYPE_MESH_POINT ||
 	     sdata->vif.type == NL80211_IFTYPE_ADHOC) &&
 	    params->n_csa_offsets) {
 		int i;

@@ -146,6 +146,7 @@ ieee80211_ibss_build_presp(struct ieee80
 				csa_settings->chandef.chan->center_freq);
 		presp->csa_counter_offsets[0] = (pos - presp->head);
 		*pos++ = csa_settings->count;
+		presp->csa_current_counter = csa_settings->count;
 	}
 
 	/* put the remaining rates in WLAN_EID_EXT_SUPP_RATES */

@@ -680,6 +680,7 @@ ieee80211_mesh_build_beacon(struct ieee8
 		*pos++ = 0x0;
 		*pos++ = ieee80211_frequency_to_channel(
 				csa->settings.chandef.chan->center_freq);
+		bcn->csa_current_counter = csa->settings.count;
 		bcn->csa_counter_offsets[0] = hdr_len + 6;
 		*pos++ = csa->settings.count;
 		*pos++ = WLAN_EID_CHAN_SWITCH_PARAM;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Eliad Peller <[email protected]>

commit c779273b37bec14c33feeab11c4d457a24bc64e0 upstream.

commit b112889c5af8124 ("iwlwifi: mvm: add Aux ROC request/response flow") added aux ROC flow in addition to the existing ROC flow. While doing it, it moved the ROC reference release to a common work item, which is being called for both the ROC and aux ROC flows.

This resulted in invalid reference accounting, as no reference was taken in case of aux ROC, while a reference was released on completion.

Fix it by adding a reference for the aux ROC as well, and release only the relevant references on completion (according to the set bits).

While at it, convert cancel_work_sync() to flush_work(), in order to make sure the references are being cleaned properly.

---
 drivers/net/wireless/iwlwifi/mvm/debugfs.c    |    5 +++--
 drivers/net/wireless/iwlwifi/mvm/mac80211.c   |    2 +-
 drivers/net/wireless/iwlwifi/mvm/mvm.h        |    1 +
 drivers/net/wireless/iwlwifi/mvm/time-event.c |   15 +++++++++------
 4 files changed, 14 insertions(+), 9 deletions(-)

@@ -6,7 +6,7 @@
  * GPL LICENSE SUMMARY
  *
  * Copyright(c) 2012 - 2014 Intel Corporation. All rights reserved.
- * Copyright(c) 2013 - 2014 Intel Mobile Communications GmbH
+ * Copyright(c) 2013 - 2015 Intel Mobile Communications GmbH
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of version 2 of the GNU General Public License as

@@ -32,7 +32,7 @@
  * BSD LICENSE
  *
  * Copyright(c) 2012 - 2014 Intel Corporation. All rights reserved.
- * Copyright(c) 2013 - 2014 Intel Mobile Communications GmbH
+ * Copyright(c) 2013 - 2015 Intel Mobile Communications GmbH
  * All rights reserved.
  *
  * Redistribution and use in source and binary forms, with or without

@@ -1356,6 +1356,7 @@ static ssize_t iwl_dbgfs_d0i3_refs_read(
 	PRINT_MVM_REF(IWL_MVM_REF_UCODE_DOWN);
 	PRINT_MVM_REF(IWL_MVM_REF_SCAN);
 	PRINT_MVM_REF(IWL_MVM_REF_ROC);
+	PRINT_MVM_REF(IWL_MVM_REF_ROC_AUX);
 	PRINT_MVM_REF(IWL_MVM_REF_P2P_CLIENT);
 	PRINT_MVM_REF(IWL_MVM_REF_AP_IBSS);
 	PRINT_MVM_REF(IWL_MVM_REF_USER);

@@ -1404,7 +1404,7 @@ void __iwl_mvm_mac_stop(struct iwl_mvm *
 	 * The work item could be running or queued if the
 	 * ROC time event stops just as we get here.
 	 */
-	cancel_work_sync(&mvm->roc_done_wk);
+	flush_work(&mvm->roc_done_wk);
 
 	iwl_trans_stop_device(mvm->trans);
 

@@ -275,6 +275,7 @@ enum iwl_mvm_ref_type {
 	IWL_MVM_REF_UCODE_DOWN,
 	IWL_MVM_REF_SCAN,
 	IWL_MVM_REF_ROC,
+	IWL_MVM_REF_ROC_AUX,
 	IWL_MVM_REF_P2P_CLIENT,
 	IWL_MVM_REF_AP_IBSS,
 	IWL_MVM_REF_USER,

@@ -6,7 +6,7 @@
  * GPL LICENSE SUMMARY
  *
  * Copyright(c) 2012 - 2014 Intel Corporation. All rights reserved.
- * Copyright(c) 2013 - 2014 Intel Mobile Communications GmbH
+ * Copyright(c) 2013 - 2015 Intel Mobile Communications GmbH
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of version 2 of the GNU General Public License as

@@ -32,7 +32,7 @@
  * BSD LICENSE
  *
  * Copyright(c) 2012 - 2014 Intel Corporation. All rights reserved.
- * Copyright(c) 2013 - 2014 Intel Mobile Communications GmbH
+ * Copyright(c) 2013 - 2015 Intel Mobile Communications GmbH
  * All rights reserved.
  *
  * Redistribution and use in source and binary forms, with or without

@@ -108,12 +108,14 @@ void iwl_mvm_roc_done_wk(struct work_str
 	 * in the case that the time event actually completed in the firmware
 	 * (which is handled in iwl_mvm_te_handle_notif).
 	 */
-	if (test_and_clear_bit(IWL_MVM_STATUS_ROC_RUNNING, &mvm->status))
+	if (test_and_clear_bit(IWL_MVM_STATUS_ROC_RUNNING, &mvm->status)) {
 		queues |= BIT(IWL_MVM_OFFCHANNEL_QUEUE);
-	if (test_and_clear_bit(IWL_MVM_STATUS_ROC_AUX_RUNNING, &mvm->status))
+		iwl_mvm_unref(mvm, IWL_MVM_REF_ROC);
+	}
+	if (test_and_clear_bit(IWL_MVM_STATUS_ROC_AUX_RUNNING, &mvm->status)) {
 		queues |= BIT(mvm->aux_queue);
-
-	iwl_mvm_unref(mvm, IWL_MVM_REF_ROC);
+		iwl_mvm_unref(mvm, IWL_MVM_REF_ROC_AUX);
+	}
 
 	synchronize_net();
 

@@ -393,6 +395,7 @@ static int iwl_mvm_aux_roc_te_handle_not
 	} else if (le32_to_cpu(notif->action) == TE_V2_NOTIF_HOST_EVENT_START) {
 		set_bit(IWL_MVM_STATUS_ROC_AUX_RUNNING, &mvm->status);
 		te_data->running = true;
+		iwl_mvm_ref(mvm, IWL_MVM_REF_ROC_AUX);
 		ieee80211_ready_on_channel(mvm->hw); /* Start TE */
 	} else {
 		IWL_DEBUG_TE(mvm,

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Michal Kazior <[email protected]>

commit 6cbfb1bb66e4e85da5db78e8ff429a85bd84ce64 upstream.

It was possible for mac80211 to be coerced into an unexpected flow causing sdata union to become corrupted. Station pointer was put into sdata->u.vlan.sta memory location while it was really master AP's sdata->u.ap.next_beacon. This led to station entry being later freed as next_beacon before __sta_info_flush() in ieee80211_stop_ap() and a subsequent invalid pointer dereference crash.

The problem was that ieee80211_ptr->use_4addr wasn't cleared on interface type changes.

This could be reproduced with the following steps:

 # host A and host B have just booted; no
 # wpa_s/hostapd running; all vifs are down
 host A> iw wlan0 set type station
 host A> iw wlan0 set 4addr on
 host A> printf 'interface=wlan0\nssid=4addrcrash\nchannel=1\nwds_sta=1' > /tmp/hconf
 host A> hostapd -B /tmp/conf
 host B> iw wlan0 set 4addr on
 host B> ifconfig wlan0 up
 host B> iw wlan0 connect -w hostAssid
 host A> pkill hostapd
 # host A crashed:

 [  127.928192] BUG: unable to handle kernel NULL pointer dereference at 00000000000006c8
 [  127.929014] IP: [<ffffffff816f4f32>] __sta_info_flush+0xac/0x158
 ...
 [  127.934578]  [<ffffffff8170789e>] ieee80211_stop_ap+0x139/0x26c
 [  127.934578]  [<ffffffff8100498f>] ? dump_trace+0x279/0x28a
 [  127.934578]  [<ffffffff816dc661>] __cfg80211_stop_ap+0x84/0x191
 [  127.934578]  [<ffffffff816dc7ad>] cfg80211_stop_ap+0x3f/0x58
 [  127.934578]  [<ffffffff816c5ad6>] nl80211_stop_ap+0x1b/0x1d
 [  127.934578]  [<ffffffff815e53f8>] genl_family_rcv_msg+0x259/0x2b5

Note: This isn't a revert of f8cdddb8d61d ("cfg80211: check iface combinations only when iface is running") as far as functionality is considered because b6a550156bc ("cfg80211/mac80211:
move more combination checks to mac80211") moved the logic somewhere else already.

---
 net/wireless/util.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -944,7 +944,7 @@ int cfg80211_change_iface(struct cfg8021
 	     ntype == NL80211_IFTYPE_P2P_CLIENT))
 		return -EBUSY;
 
-	if (ntype != otype && netif_running(dev)) {
+	if (ntype != otype) {
 		dev->ieee80211_ptr->use_4addr = false;
 		dev->ieee80211_ptr->mesh_id_up_len = 0;
 		wdev_lock(dev->ieee80211_ptr);

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Michal Kazior <[email protected]>

commit ab499db80fcf07c18e4053f91a619500f663e90e upstream.

There was a possible race between ieee80211_reconfig() and ieee80211_delayed_tailroom_dec(). This could result in inability to transmit data if driver crashed during roaming or rekeying and subsequent skbs with insufficient tailroom appeared.

This race was probably never seen in the wild because a device driver would have to crash AND recover within 0.5s which is very unlikely.

I was able to prove this race exists after changing the delay to 10s locally and crashing ath10k via debugfs immediately after GTK rekeying. In case of ath10k the counter went below
0. This was harmless but other drivers which actually require tailroom (e.g. for WEP ICV or MMIC) could end up with the counter at 0 instead of >0 and introduce insufficient skb tailroom failures because mac80211 would not resize skbs appropriately anymore.

---
 net/mac80211/main.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -249,6 +249,7 @@ static void ieee80211_restart_work(struc
 {
 	struct ieee80211_local *local =
 		container_of(work, struct ieee80211_local, restart_work);
+	struct ieee80211_sub_if_data *sdata;
 
 	/* wait for scan work complete */
 	flush_workqueue(local->workqueue);

@@ -257,6 +258,8 @@ static void ieee80211_restart_work(struc
 	     "%s called with hardware scan in progress\n", __func__);
 
 	rtnl_lock();
+	list_for_each_entry(sdata, &local->interfaces, list)
+		flush_delayed_work(&sdata->dec_tailroom_needed_wk);
 	ieee80211_scan_cancel(local);
 	ieee80211_reconfig(local);
 	rtnl_unlock();

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Alexander Duyck <[email protected]>

commit 8084b86dcfbc4b4822868c1dbdb429b5c08154e2 upstream.

When the VLAN_HLEN was added to the calculation for the maximum frame size there seems to have been a number of issues added to the driver.

The first issue is that in some cases the maximum frame size for a device never really reached the actual maximum frame size as the VLAN header length was not included the calculation for that value. As a result some parts only supported a maximum frame size of either 1496 in the case of parts that didn't support jumbo frames, and 8996 in the case of the parts that do.

The second issue is the fact that there were several checks that weren't updated so as a result setting an MTU of 1500 was treated as enabling jumbo frames as the calculated value was 1522 instead of 1518. I have addressed those by replacing ETH_FRAME_LEN with VLAN_ETH_FRAME_LEN where appropriate.

The final issue was the fact that lowering the MTU below 1500 would cause the driver to allocate 2K buffers for the rings. This is an old issue that was fixed several years ago in igb/ixgbe and I am addressing now by just replacing == with a <= so that we always just round up to 1522 for anything that isn't a jumbo frame.

---
 drivers/net/ethernet/intel/e1000e/82571.c   |    2 +-
 drivers/net/ethernet/intel/e1000e/ich8lan.c |   10 +++++-----
 drivers/net/ethernet/intel/e1000e/netdev.c  |   18 ++++++++----------
 3 files changed, 14 insertions(+), 16 deletions(-)

@@ -2010,7 +2010,7 @@ const struct e1000_info e1000_82573_info
 	.flags2			= FLAG2_DISABLE_ASPM_L1
 				  | FLAG2_DISABLE_ASPM_L0S,
 	.pba			= 20,
-	.max_hw_frame_size	= ETH_FRAME_LEN + ETH_FCS_LEN,
+	.max_hw_frame_size	= VLAN_ETH_FRAME_LEN + ETH_FCS_LEN,
 	.get_variants		= e1000_get_variants_82571,
 	.mac_ops		= &e82571_mac_ops,
 	.phy_ops		= &e82_phy_ops_m88,

@@ -1563,7 +1563,7 @@ static s32 e1000_get_variants_ich8lan(st
 	    ((adapter->hw.mac.type >= e1000_pch2lan) &&
 	     (!(er32(CTRL_EXT) & E1000_CTRL_EXT_LSECCK)))) {
 		adapter->flags &= ~FLAG_HAS_JUMBO_FRAMES;
-		adapter->max_hw_frame_size = ETH_FRAME_LEN + ETH_FCS_LEN;
+		adapter->max_hw_frame_size = VLAN_ETH_FRAME_LEN + ETH_FCS_LEN;
 
 		hw->mac.ops.blink_led = NULL;
 	}

@@ -5681,7 +5681,7 @@ const struct e1000_info e1000_ich8_info
 				  | FLAG_HAS_FLASH
 				  | FLAG_APME_IN_WUC,
 	.pba			= 8,
-	.max_hw_frame_size	= ETH_FRAME_LEN + ETH_FCS_LEN,
+	.max_hw_frame_size	= VLAN_ETH_FRAME_LEN + ETH_FCS_LEN,
 	.get_variants		= e1000_get_variants_ich8lan,
 	.mac_ops		= &ich8_mac_ops,
 	.phy_ops		= &ich8_phy_ops,

@@ -5754,7 +5754,7 @@ const struct e1000_info e1000_pch2_info
 	.flags2			= FLAG2_HAS_PHY_STATS
 				  | FLAG2_HAS_EEE,
 	.pba			= 26,
-	.max_hw_frame_size	= 9018,
+	.max_hw_frame_size	= 9022,
 	.get_variants		= e1000_get_variants_ich8lan,
 	.mac_ops		= &ich8_mac_ops,
 	.phy_ops		= &ich8_phy_ops,

@@ -5774,7 +5774,7 @@ const struct e1000_info e1000_pch_lpt_in
 	.flags2			= FLAG2_HAS_PHY_STATS
 				  | FLAG2_HAS_EEE,
 	.pba			= 26,
-	.max_hw_frame_size	= 9018,
+	.max_hw_frame_size	= 9022,
 	.get_variants		= e1000_get_variants_ich8lan,
 	.mac_ops		= &ich8_mac_ops,
 	.phy_ops		= &ich8_phy_ops,

@@ -5794,7 +5794,7 @@ const struct e1000_info e1000_pch_spt_in
 	.flags2			= FLAG2_HAS_PHY_STATS
 				  | FLAG2_HAS_EEE,
 	.pba			= 26,
-	.max_hw_frame_size	= 9018,
+	.max_hw_frame_size	= 9022,
 	.get_variants		= e1000_get_variants_ich8lan,
 	.mac_ops		= &ich8_mac_ops,
 	.phy_ops		= &ich8_phy_ops,

@@ -3807,7 +3807,7 @@ void e1000e_reset(struct e1000_adapter *
 	/* reset Packet Buffer Allocation to default */
 	ew32(PBA, pba);
 
-	if (adapter->max_frame_size > ETH_FRAME_LEN + ETH_FCS_LEN) {
+	if (adapter->max_frame_size > (VLAN_ETH_FRAME_LEN + ETH_FCS_LEN)) {
 		/* To maintain wire speed transmits, the Tx FIFO should be
 		 * large enough to accommodate two full transmit packets,
 		 * rounded up to the next 1KB and expressed in KB.  Likewise,

@@ -4196,9 +4196,9 @@ static int e1000_sw_init(struct e1000_ad
 {
 	struct net_device *netdev = adapter->netdev;
 
-	adapter->rx_buffer_len = ETH_FRAME_LEN + VLAN_HLEN + ETH_FCS_LEN;
+	adapter->rx_buffer_len = VLAN_ETH_FRAME_LEN + ETH_FCS_LEN;
 	adapter->rx_ps_bsize0 = 128;
-	adapter->max_frame_size = netdev->mtu + ETH_HLEN + ETH_FCS_LEN;
+	adapter->max_frame_size = netdev->mtu + VLAN_ETH_HLEN + ETH_FCS_LEN;
 	adapter->min_frame_size = ETH_ZLEN + ETH_FCS_LEN;
 	adapter->tx_ring_count = E1000_DEFAULT_TXD;
 	adapter->rx_ring_count = E1000_DEFAULT_RXD;

@@ -5781,17 +5781,17 @@ struct rtnl_link_stats64 *e1000e_get_sta
 static int e1000_change_mtu(struct net_device *netdev, int new_mtu)
 {
 	struct e1000_adapter *adapter = netdev_priv(netdev);
-	int max_frame = new_mtu + VLAN_HLEN + ETH_HLEN + ETH_FCS_LEN;
+	int max_frame = new_mtu + VLAN_ETH_HLEN + ETH_FCS_LEN;
 
 	/* Jumbo frame support */
-	if ((max_frame > ETH_FRAME_LEN + ETH_FCS_LEN) &&
+	if ((max_frame > (VLAN_ETH_FRAME_LEN + ETH_FCS_LEN)) &&
 	    !(adapter->flags & FLAG_HAS_JUMBO_FRAMES)) {
 		e_err("Jumbo Frames not supported.\n");
 		return -EINVAL;
 	}
 
 	/* Supported frame sizes */
-	if ((new_mtu < ETH_ZLEN + ETH_FCS_LEN + VLAN_HLEN) ||
+	if ((new_mtu < (VLAN_ETH_ZLEN + ETH_FCS_LEN)) ||
 	    (max_frame > adapter->max_hw_frame_size)) {
 		e_err("Unsupported MTU setting\n");
 		return -EINVAL;

@@ -5831,10 +5831,8 @@ static int e1000_change_mtu(struct net_d
 		adapter->rx_buffer_len = 4096;
 
 	/* adjust allocation if LPE protects us, and we aren't using SBP */
-	if ((max_frame == ETH_FRAME_LEN + ETH_FCS_LEN) ||
-	    (max_frame == ETH_FRAME_LEN + VLAN_HLEN + ETH_FCS_LEN))
-		adapter->rx_buffer_len = ETH_FRAME_LEN + VLAN_HLEN
-		    + ETH_FCS_LEN;
+	if (max_frame <= (VLAN_ETH_FRAME_LEN + ETH_FCS_LEN))
+		adapter->rx_buffer_len = VLAN_ETH_FRAME_LEN + ETH_FCS_LEN;
 
 	if (netif_running(netdev))
 		e1000e_up(adapter);

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Damian Eppel <[email protected]>

commit 56a94f13919c0db5958611b388e1581b4852f3c9 upstream.

Whilst testing cpu hotplug events on kernel configured with DEBUG_PREEMPT and DEBUG_ATOMIC_SLEEP we get following BUG message, caused by calling request_irq() and free_irq() in the context of hotplug notification (which is in this case atomic context).

[   40.785859] CPU1: Software reset
[   40.786660] BUG: sleeping function called from invalid context at mm/slub.c:1241
[   40.786668] in_atomic(): 1, irqs_disabled(): 128, pid: 0, name: swapper/1
[   40.786678] Preemption disabled at:[<  (null)>]   (null)
[   40.786681]
[   40.786692] CPU: 1 PID: 0 Comm: swapper/1 Not tainted 3.19.0-rc4-00024-g7dca860 #36
[   40.786698] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   40.786728] [<c0014a00>] (unwind_backtrace) from [<c0011980>] (show_stack+0x10/0x14)
[   40.786747] [<c0011980>] (show_stack) from [<c0449ba0>] (dump_stack+0x70/0xbc)
[   40.786767] [<c0449ba0>] (dump_stack) from [<c00c6124>] (kmem_cache_alloc+0xd8/0x170)
[   40.786785] [<c00c6124>] (kmem_cache_alloc) from [<c005d6f8>] (request_threaded_irq+0x64/0x128)
[   40.786804] [<c005d6f8>] (request_threaded_irq) from [<c0350b8c>] (exynos4_local_timer_setup+0xc0/0x13c)
[   40.786820] [<c0350b8c>] (exynos4_local_timer_setup) from [<c0350ca8>] (exynos4_mct_cpu_notify+0x30/0xa8)
[   40.786838] [<c0350ca8>] (exynos4_mct_cpu_notify) from [<c003b330>] (notifier_call_chain+0x44/0x84)
[   40.786857] [<c003b330>] (notifier_call_chain) from [<c0022fd4>] (__cpu_notify+0x28/0x44)
[   40.786873] [<c0022fd4>] (__cpu_notify) from [<c0013714>] (secondary_start_kernel+0xec/0x150)
[   40.786886] [<c0013714>] (secondary_start_kernel) from [<40008764>] (0x40008764)

Interrupts cannot be requested/freed in the CPU_STARTING/CPU_DYING notifications which run on the hotplugged cpu with interrupts and preemption disabled.

To avoid the issue, request the interrupts for all possible cpus in the boot code. The interrupts are marked NO_AUTOENABLE to avoid a racy request_irq/disable_irq() sequence. The flag prevents the request_irq() code from enabling the interrupt immediately.

The interrupt is then enabled in the CPU_STARTING notifier of the hotplugged cpu and again disabled with disable_irq_nosync() in the CPU_DYING notifier.

[ tglx: Massaged changelog to match the patch ]

---
 drivers/clocksource/exynos_mct.c |   43 +++++++++++++++++++++++++++------------
 1 file changed, 30 insertions(+), 13 deletions(-)

@@ -466,15 +466,12 @@ static int exynos4_local_timer_setup(str
 	exynos4_mct_write(TICK_BASE_CNT, mevt->base + MCT_L_TCNTB_OFFSET);
 
 	if (mct_int_type == MCT_INT_SPI) {
-		evt->irq = mct_irqs[MCT_L0_IRQ + cpu];
-		if (request_irq(evt->irq, exynos4_mct_tick_isr,
-				IRQF_TIMER | IRQF_NOBALANCING,
-				evt->name, mevt)) {
-			pr_err("exynos-mct: cannot register IRQ %d\n",
-				evt->irq);
+
+		if (evt->irq == -1)
 			return -EIO;
-		}
-		irq_force_affinity(mct_irqs[MCT_L0_IRQ + cpu], cpumask_of(cpu));
+
+		irq_force_affinity(evt->irq, cpumask_of(cpu));
+		enable_irq(evt->irq);
 	} else {
 		enable_percpu_irq(mct_irqs[MCT_L0_IRQ], 0);
 	}

@@ -487,10 +484,12 @@ static int exynos4_local_timer_setup(str
 static void exynos4_local_timer_stop(struct clock_event_device *evt)
 {
 	evt->set_mode(CLOCK_EVT_MODE_UNUSED, evt);
-	if (mct_int_type == MCT_INT_SPI)
-		free_irq(evt->irq, this_cpu_ptr(&percpu_mct_tick));
-	else
+	if (mct_int_type == MCT_INT_SPI) {
+		if (evt->irq != -1)
+			disable_irq_nosync(evt->irq);
+	} else {
 		disable_percpu_irq(mct_irqs[MCT_L0_IRQ]);
+	}
 }
 
 static int exynos4_mct_cpu_notify(struct notifier_block *self,

@@ -522,7 +521,7 @@ static struct notifier_block exynos4_mct
 
 static void __init exynos4_timer_resources(struct device_node *np, void __iomem *base)
 {
-	int err;
+	int err, cpu;
 	struct mct_clock_event_device *mevt = this_cpu_ptr(&percpu_mct_tick);
 	struct clk *mct_clk, *tick_clk;
 

@@ -549,7 +548,25 @@ static void __init exynos4_timer_resourc
 		WARN(err, "MCT: can't request IRQ %d (%d)\n",
 		     mct_irqs[MCT_L0_IRQ], err);
 	} else {
-		irq_set_affinity(mct_irqs[MCT_L0_IRQ], cpumask_of(0));
+		for_each_possible_cpu(cpu) {
+			int mct_irq = mct_irqs[MCT_L0_IRQ + cpu];
+			struct mct_clock_event_device *pcpu_mevt =
+				per_cpu_ptr(&percpu_mct_tick, cpu);
+
+			pcpu_mevt->evt.irq = -1;
+
+			irq_set_status_flags(mct_irq, IRQ_NOAUTOEN);
+			if (request_irq(mct_irq,
+					exynos4_mct_tick_isr,
+					IRQF_TIMER | IRQF_NOBALANCING,
+					pcpu_mevt->name, pcpu_mevt)) {
+				pr_err("exynos-mct: cannot register IRQ (cpu%d)\n",
+									cpu);
+
+				continue;
+			}
+			pcpu_mevt->evt.irq = mct_irq;
+		}
 	}
 
 	err = register_cpu_notifier(&exynos4_mct_cpu_nb);

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Dmitry Tunin <[email protected]>

commit 4fa9dabcffc8e16601307d3d56b58c68d9716ba4 upstream.

Lenovo G30-50 does not have a hardware wireless switch and wireless is always blocked.

---
 drivers/platform/x86/ideapad-laptop.c |    7 +++++++
 1 file changed, 7 insertions(+)

@@ -837,6 +837,13 @@ static const struct dmi_system_id no_hw_
 		},
 	},
 	{
+		.ident = "Lenovo G50-30",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
+			DMI_MATCH(DMI_PRODUCT_VERSION, "Lenovo G50-30"),
+		},
+	},
+	{
 		.ident = "Lenovo Yoga 2 11 / 13 / Pro",
 		.matches = {
 			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Arnd Bergmann <[email protected]>

commit 4b200b4604bec3388426159f1656109d19fadf6e upstream.

This fixes a several year old regression that I found while trying to get the Yoga 3 11 to work. The ideapad_rfk_set function is meant to send a command to the embedded controller through ACPI, but as of c1f73658ed, it sends the index of the rfkill device instead of the command, and ignores the opcode field.

This changes it back to the original behavior, which indeed flips the rfkill state as seen in the debugfs interface.

---
 drivers/platform/x86/ideapad-laptop.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -464,8 +464,9 @@ static const struct ideapad_rfk_data ide
 static int ideapad_rfk_set(void *data, bool blocked)
 {
 	struct ideapad_rfk_priv *priv = data;
+	int opcode = ideapad_rfk_data[priv->dev].opcode;
 
-	return write_ec_cmd(priv->priv->adev->handle, priv->dev, !blocked);
+	return write_ec_cmd(priv->priv->adev->handle, opcode, !blocked);
 }
 
 static struct rfkill_ops ideapad_rfk_ops = {

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Jingoo Han <[email protected]>

commit 294240ffe784e951dc2ef070da04fa31ef6db3a0 upstream.

When kzalloc() is called under spin_lock(), GFP_ATOMIC should be used to avoid sleeping allocation.
The call tree is:
of_pci_range_to_resource()

    --> pci_register_io_range() <-- takes spin_lock(&io_range_lock);
       --> kzalloc()

---
 drivers/of/address.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -712,7 +712,7 @@ int __weak pci_register_io_range(phys_ad
 	}
 
 	/* add the range to the list */
-	range = kzalloc(sizeof(*range), GFP_KERNEL);
+	range = kzalloc(sizeof(*range), GFP_ATOMIC);
 	if (!range) {
 		err = -ENOMEM;
 		goto end_register;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Miklos Szeredi <[email protected]>

commit cdb672795876d7bc1870aed9a2d7cb59f43d1d96 upstream.

If jffs2 can deadlock on overlayfs readdir because it takes the same lock on ->iterate() as in ->lookup().

Fix by moving whiteout checking outside iterate_dir(). Optimized by collecting potential whiteouts (DT_CHR) in a temporary list and if non-empty iterating throug these and checking for a 0/0 chardev.

---
 fs/overlayfs/readdir.c |   77 +++++++++++++++++++++++++++++++------------------
 1 file changed, 49 insertions(+), 28 deletions(-)

@@ -23,6 +23,7 @@ struct ovl_cache_entry {
 	u64 ino;
 	struct list_head l_node;
 	struct rb_node node;
+	struct ovl_cache_entry *next_maybe_whiteout;
 	bool is_whiteout;
 	char name[];
 };

@@ -39,7 +40,7 @@ struct ovl_readdir_data {
 	struct rb_root root;
 	struct list_head *list;
 	struct list_head middle;
-	struct dentry *dir;
+	struct ovl_cache_entry *first_maybe_whiteout;
 	int count;
 	int err;
 };

@@ -79,7 +80,7 @@ static struct ovl_cache_entry *ovl_cache
 	return NULL;
 }
 
-static struct ovl_cache_entry *ovl_cache_entry_new(struct dentry *dir,
+static struct ovl_cache_entry *ovl_cache_entry_new(struct ovl_readdir_data *rdd,
 						   const char *name, int len,
 						   u64 ino, unsigned int d_type)
 {

@@ -98,29 +99,8 @@ static struct ovl_cache_entry *ovl_cache
 	p->is_whiteout = false;
 
 	if (d_type == DT_CHR) {
-		struct dentry *dentry;
-		const struct cred *old_cred;
-		struct cred *override_cred;
-
-		override_cred = prepare_creds();
-		if (!override_cred) {
-			kfree(p);
-			return NULL;
-		}
-
-		/*
-		 * CAP_DAC_OVERRIDE for lookup
-		 */
-		cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE);
-		old_cred = override_creds(override_cred);
-
-		dentry = lookup_one_len(name, dir, len);
-		if (!IS_ERR(dentry)) {
-			p->is_whiteout = ovl_is_whiteout(dentry);
-			dput(dentry);
-		}
-		revert_creds(old_cred);
-		put_cred(override_cred);
+		p->next_maybe_whiteout = rdd->first_maybe_whiteout;
+		rdd->first_maybe_whiteout = p;
 	}
 	return p;
 }

@@ -148,7 +128,7 @@ static int ovl_cache_entry_add_rb(struct
 			return 0;
 	}
 
-	p = ovl_cache_entry_new(rdd->dir, name, len, ino, d_type);
+	p = ovl_cache_entry_new(rdd, name, len, ino, d_type);
 	if (p == NULL)
 		return -ENOMEM;
 

@@ -169,7 +149,7 @@ static int ovl_fill_lower(struct ovl_rea
 	if (p) {
 		list_move_tail(&p->l_node, &rdd->middle);
 	} else {
-		p = ovl_cache_entry_new(rdd->dir, name, namelen, ino, d_type);
+		p = ovl_cache_entry_new(rdd, name, namelen, ino, d_type);
 		if (p == NULL)
 			rdd->err = -ENOMEM;
 		else

@@ -219,6 +199,43 @@ static int ovl_fill_merge(struct dir_con
 		return ovl_fill_lower(rdd, name, namelen, offset, ino, d_type);
 }
 
+static int ovl_check_whiteouts(struct dentry *dir, struct ovl_readdir_data *rdd)
+{
+	int err;
+	struct ovl_cache_entry *p;
+	struct dentry *dentry;
+	const struct cred *old_cred;
+	struct cred *override_cred;
+
+	override_cred = prepare_creds();
+	if (!override_cred)
+		return -ENOMEM;
+
+	/*
+	 * CAP_DAC_OVERRIDE for lookup
+	 */
+	cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE);
+	old_cred = override_creds(override_cred);
+
+	err = mutex_lock_killable(&dir->d_inode->i_mutex);
+	if (!err) {
+		while (rdd->first_maybe_whiteout) {
+			p = rdd->first_maybe_whiteout;
+			rdd->first_maybe_whiteout = p->next_maybe_whiteout;
+			dentry = lookup_one_len(p->name, dir, p->len);
+			if (!IS_ERR(dentry)) {
+				p->is_whiteout = ovl_is_whiteout(dentry);
+				dput(dentry);
+			}
+		}
+		mutex_unlock(&dir->d_inode->i_mutex);
+	}
+	revert_creds(old_cred);
+	put_cred(override_cred);
+
+	return err;
+}
+
 static inline int ovl_dir_read(struct path *realpath,
 			       struct ovl_readdir_data *rdd)
 {

@@ -229,7 +246,7 @@ static inline int ovl_dir_read(struct pa
 	if (IS_ERR(realfile))
 		return PTR_ERR(realfile);
 
-	rdd->dir = realpath->dentry;
+	rdd->first_maybe_whiteout = NULL;
 	rdd->ctx.pos = 0;
 	do {
 		rdd->count = 0;

@@ -238,6 +255,10 @@ static inline int ovl_dir_read(struct pa
 		if (err >= 0)
 			err = rdd->err;
 	} while (!err && rdd->count);
+
+	if (!err && rdd->first_maybe_whiteout)
+		err = ovl_check_whiteouts(realpath->dentry, rdd);
+
 	fput(realfile);
 
 	return err;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Konstantin Khlebnikov <[email protected]>

commit c8fff7bc5bba6bd59cad40441c189c4efe7190f6 upstream.

Node 0 might be offline as well as any other numa node, in this case kernel cannot handle memory allocation and crashes.

---
 drivers/of/base.c  |    2 +-
 include/linux/of.h |    5 ++++-
 2 files changed, 5 insertions(+), 2 deletions(-)

@@ -89,7 +89,7 @@ EXPORT_SYMBOL(of_n_size_cells);
 #ifdef CONFIG_NUMA
 int __weak of_node_to_nid(struct device_node *np)
 {
-	return numa_node_id();
+	return NUMA_NO_NODE;
 }
 #endif
 

@@ -673,7 +673,10 @@ static inline void of_property_clear_fla
 #if defined(CONFIG_OF) && defined(CONFIG_NUMA)
 extern int of_node_to_nid(struct device_node *np);
 #else
-static inline int of_node_to_nid(struct device_node *device) { return 0; }
+static inline int of_node_to_nid(struct device_node *device)
+{
+	return NUMA_NO_NODE;
+}
 #endif
 
 static inline struct device_node *of_find_matching_node(

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Stefan Wahren <[email protected]>

commit 7cb81136d2efe0f5ed9d965857f4756a15e6c338 upstream.

key/value pairs in a JSON object must be separated by a comma. After adding the properties "accuracy" and "phase" the JSON output of /sys/kernel/debug/clk/clk_dump is invalid.

So add the missing commas to fix it.

---
 drivers/clk/clk.c |    5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

@@ -230,11 +230,12 @@ static void clk_dump_one(struct seq_file
 	if (!c)
 		return;
 
+	/* This should be JSON format, i.e. elements separated with a comma */
 	seq_printf(s, "\"%s\": { ", c->name);
 	seq_printf(s, "\"enable_count\": %d,", c->enable_count);
 	seq_printf(s, "\"prepare_count\": %d,", c->prepare_count);
-	seq_printf(s, "\"rate\": %lu", clk_core_get_rate(c));
-	seq_printf(s, "\"accuracy\": %lu", clk_core_get_accuracy(c));
+	seq_printf(s, "\"rate\": %lu,", clk_core_get_rate(c));
+	seq_printf(s, "\"accuracy\": %lu,", clk_core_get_accuracy(c));
 	seq_printf(s, "\"phase\": %d", clk_core_get_phase(c));
 }
 

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Krzysztof Kozlowski <[email protected]>

commit e0cdcda508f110b7ec190dc7c5eb2869ba73a535 upstream.

of_clk_get_from_provider() returns ERR_PTR on failure. The dra7-atl-clock driver was not checking its return value and immediately used it in __clk_get_hw(). __clk_get_hw() dereferences supplied clock, if it is not NULL, so in that case it would dereference an ERR_PTR.

---
 drivers/clk/ti/clk-dra7-atl.c |    5 +++++
 1 file changed, 5 insertions(+)

@@ -252,6 +252,11 @@ static int of_dra7_atl_clk_probe(struct
 		}
 
 		clk = of_clk_get_from_provider(&clkspec);
+		if (IS_ERR(clk)) {
+			pr_err("%s: failed to get atl clock %d from provider\n",
+			       __func__, i);
+			return PTR_ERR(clk);
+		}
 
 		cdesc = to_atl_desc(__clk_get_hw(clk));
 		cdesc->cinfo = cinfo;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Hai Li <[email protected]>

commit 6d451367bfa16fc103604bacd258f534c65d1540 upstream.

Since the parent rate has been recalculated, pixel RCG clock should rely on it to find the correct M/N values during set_rate, instead of calling __clk_round_rate() to its parent again.

---
 drivers/clk/qcom/clk-rcg2.c |    9 +++------
 1 file changed, 3 insertions(+), 6 deletions(-)

@@ -530,19 +530,16 @@ static int clk_pixel_set_rate(struct clk
 	struct clk_rcg2 *rcg = to_clk_rcg2(hw);
 	struct freq_tbl f = *rcg->freq_tbl;
 	const struct frac_entry *frac = frac_table_pixel;
-	unsigned long request, src_rate;
+	unsigned long request;
 	int delta = 100000;
 	u32 mask = BIT(rcg->hid_width) - 1;
 	u32 hid_div;
-	int index = qcom_find_src_index(hw, rcg->parent_map, f.src);
-	struct clk *parent = clk_get_parent_by_index(hw->clk, index);
 
 	for (; frac->num; frac++) {
 		request = (rate * frac->den) / frac->num;
 
-		src_rate = __clk_round_rate(parent, request);
-		if ((src_rate < (request - delta)) ||
-			(src_rate > (request + delta)))
+		if ((parent_rate < (request - delta)) ||
+			(parent_rate > (request + delta)))
 			continue;
 
 		regmap_read(rcg->clkr.regmap, rcg->cmd_rcgr + CFG_REG,

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Trond Myklebust <[email protected]>

commit 1ca018d28d96d07788474abf66a5f3e9594841f5 upstream.

pnfs_do_write() expects the call to pnfs_write_through_mds() to free the pgio header and to release the layout segment before exiting. The problem is that nfs_pgio_data_destroy() doesn't actually do this; it only frees the memory allocated by nfs_generic_pgio().

Ditto for pnfs_do_read()...

Fix in both cases is to add a call to hdr->release(hdr).

---
 fs/nfs/pnfs.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -1865,6 +1865,7 @@ pnfs_write_through_mds(struct nfs_pageio
 		mirror->pg_recoalesce = 1;
 	}
 	nfs_pgio_data_destroy(hdr);
+	hdr->release(hdr);
 }
 
 static enum pnfs_try_status

@@ -1979,6 +1980,7 @@ pnfs_read_through_mds(struct nfs_pageio_
 		mirror->pg_recoalesce = 1;
 	}
 	nfs_pgio_data_destroy(hdr);
+	hdr->release(hdr);
 }
 
 /*

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Trond Myklebust <[email protected]>

commit d620876990f02788d5a663075df007ffb91bdfad upstream.

hdr->good_bytes needs to be set to the length of the request, not zero.

---
 fs/nfs/flexfilelayout/flexfilelayout.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -631,7 +631,7 @@ static void ff_layout_reset_write(struct
 			nfs_direct_set_resched_writes(hdr->dreq);
 			/* fake unstable write to let common nfs resend pages */
 			hdr->verf.committed = NFS_UNSTABLE;
-			hdr->good_bytes = 0;
+			hdr->good_bytes = hdr->args.count;
 		}
 		return;
 	}

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Chuck Lever <[email protected]>

commit d683cc49daf7c5afca8cd9654aaa1bf63cdf2ad9 upstream.

When encoding the NFSACL SETACL operation, reserve just the estimated size of the ACL rather than a fixed maximum. This eliminates needless zero padding on the wire that the server ignores.

---
 fs/nfs/nfs3xdr.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1342,7 +1342,7 @@ static void nfs3_xdr_enc_setacl3args(str
 	if (args->npages != 0)
 		xdr_write_pages(xdr, args->pages, 0, args->len);
 	else
-		xdr_reserve_space(xdr, NFS_ACL_INLINE_BUFSIZE);
+		xdr_reserve_space(xdr, args->len);
 
 	error = nfsacl_encode(xdr->buf, base, args->inode,
 			    (args->mask & NFS_ACL) ?

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Olga Kornievskaia <[email protected]>

commit e8d975e73e5fa05f983fbf2723120edcf68e0b38 upstream.

Problem: When an operation like WRITE receives a BAD_STATEID, even though recovery code clears the RECLAIM_NOGRACE recovery flag before recovering the open state, because of clearing delegation state for the associated inode, nfs_inode_find_state_and_recover() gets called and it makes the same state with RECLAIM_NOGRACE flag again. As a results, when we restart looking over the open states, we end up in the infinite loop instead of breaking out in the next test of state flags.

Solution: unset the RECLAIM_NOGRACE set because of
calling of nfs_inode_find_state_and_recover() after returning from calling recover_open() function.

---
 fs/nfs/nfs4state.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -1482,6 +1482,8 @@ restart:
 					spin_unlock(&state->state_lock);
 				}
 				nfs4_put_open_state(state);
+				clear_bit(NFS4CLNT_RECLAIM_NOGRACE,
+					&state->flags);
 				spin_lock(&sp->so_lock);
 				goto restart;
 			}

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Jeff Layton <[email protected]>

commit 764ad8ba8cd4c6f836fca9378f8c5121aece0842 upstream.

The current buffer is much too small if you have a relatively long hostname. Bring it up to the size of the one that SETCLIENTID has.

---
 include/linux/nfs_xdr.h |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1142,7 +1142,7 @@ struct nfs41_state_protection {
 	struct nfs4_op_map allow;
 };
 
-#define NFS4_EXCHANGE_ID_LEN	(48)
+#define NFS4_EXCHANGE_ID_LEN	(127)
 struct nfs41_exchange_id_args {
 	struct nfs_client		*client;
 	nfs4_verifier			*verifier;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Trond Myklebust <[email protected]>

commit c70701131f7a8edea91fc49d11796d342cff7c62 upstream.

If a write attempt fails, and the write is queued up for resending to the server, as opposed to being dropped, then we need to set the appropriate flag so that nfs_file_fsync() does the right thing.

---
 fs/nfs/pnfs.c  |    1 +
 fs/nfs/write.c |    1 +
 2 files changed, 2 insertions(+)

@@ -1821,6 +1821,7 @@ int pnfs_write_done_resend_to_mds(struct
 	/* Resend all requests through the MDS */
 	nfs_pageio_init_write(&pgio, hdr->inode, FLUSH_STABLE, true,
 			      hdr->completion_ops);
+	set_bit(NFS_CONTEXT_RESEND_WRITES, &hdr->args.context->flags);
 	return nfs_pageio_resend(&pgio, hdr);
 }
 EXPORT_SYMBOL_GPL(pnfs_write_done_resend_to_mds);

@@ -1289,6 +1289,7 @@ static void nfs_initiate_write(struct nf
 static void nfs_redirty_request(struct nfs_page *req)
 {
 	nfs_mark_request_dirty(req);
+	set_bit(NFS_CONTEXT_RESEND_WRITES, &req->wb_context->flags);
 	nfs_unlock_request(req);
 	nfs_end_page_writeback(req);
 	nfs_release_request(req);

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Jeff Layton <[email protected]>

commit a24221dca1868101c9b4b5adde4a6a5b1a3a64a7 upstream.

If we have two tasks racing to update a mirror's credentials, then they can end up leaking one (or more) sets of credentials. The first task will set mirror->cred and then the second task will just overwrite it.

Use a cmpxchg to ensure that the creds are only set once. If we get to the point where we would set mirror->cred and find that they're already set, then we just release the creds that were just found.

---
 fs/nfs/flexfilelayout/flexfilelayoutdev.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -324,7 +324,8 @@ static int ff_layout_update_mirror_cred(
 				__func__, PTR_ERR(cred));
 			return PTR_ERR(cred);
 		} else {
-			mirror->cred = cred;
+			if (cmpxchg(&mirror->cred, NULL, cred))
+				put_rpccred(cred);
 		}
 	}
 	return 0;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Jeff Layton <[email protected]>

commit 0c8315dd56577445dd1afe6b9cfa06b7efdf2f82 upstream.

A ds can be associated with more than one mirror, but we currently skip setting a mirror's credentials if we find that it's already set up with a connected client.

The upshot is that we can end up sending DS writes with MDS credentials instead of properly setting them up. Fix nfs4_ff_layout_prepare_ds to always verify that the mirror's credentials are set up, even when we have a DS that's already connected.

---
 fs/nfs/flexfilelayout/flexfilelayoutdev.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -387,7 +387,7 @@ nfs4_ff_layout_prepare_ds(struct pnfs_la
 	/* matching smp_wmb() in _nfs4_pnfs_v3/4_ds_connect */
 	smp_rmb();
 	if (ds->ds_clp)
-		goto out;
+		goto out_update_creds;
 
 	flavor = nfs4_ff_layout_choose_authflavor(mirror);
 

@@ -431,7 +431,7 @@ nfs4_ff_layout_prepare_ds(struct pnfs_la
 			}
 		}
 	}
-
+out_update_creds:
 	if (ff_layout_update_mirror_cred(mirror, ds))
 		ds = NULL;
 out:

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Trond Myklebust <[email protected]>

commit 88de6af24f2b48b06c514d3c3d0a8f22fafe30bd upstream.

req->rq_private_buf isn't initialised when xprt_setup_backchannel calls xprt_free_allocation.

---
 net/sunrpc/backchannel_rqst.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -60,7 +60,7 @@ static void xprt_free_allocation(struct
 
 	dprintk("RPC:        free allocations for req= %p\n", req);
 	WARN_ON_ONCE(test_bit(RPC_BC_PA_IN_USE, &req->rq_bc_pa_state));
-	xbufp = &req->rq_private_buf;
+	xbufp = &req->rq_rcv_buf;
 	free_page((unsigned long)xbufp->head[0].iov_base);
 	xbufp = &req->rq_snd_buf;
 	free_page((unsigned long)xbufp->head[0].iov_base);

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Al Viro <[email protected]>

commit a84b69cb6e0a41e86bc593904faa6def3b957343 upstream.

If we'd already sent a request and decide to abort it, we *must* issue TFLUSH properly and not just blindly reuse the tag, or we'll get seriously screwed when response eventually arrives and we confuse it for response to later request that had reused the same tag.

---
 net/9p/client.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -843,7 +843,8 @@ static struct p9_req_t *p9_client_zc_rpc
 	if (err < 0) {
 		if (err == -EIO)
 			c->status = Disconnected;
-		goto reterr;
+		if (err != -ERESTARTSYS)
+			goto reterr;
 	}
 	if (req->status == REQ_STATUS_ERROR) {
 		p9_debug(P9_DEBUG_ERROR, "req_status error %d\n", req->t_err);

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Al Viro <[email protected]>

commit 0a73d0a204a4a04a1e110539c5a524ae51f91d6d upstream.

---
 fs/9p/vfs_inode.c      |    3 +--
 fs/9p/vfs_inode_dotl.c |    3 +--
 2 files changed, 2 insertions(+), 4 deletions(-)

@@ -540,8 +540,7 @@ static struct inode *v9fs_qid_iget(struc
 	unlock_new_inode(inode);
 	return inode;
 error:
-	unlock_new_inode(inode);
-	iput(inode);
+	iget_failed(inode);
 	return ERR_PTR(retval);
 
 }

@@ -149,8 +149,7 @@ static struct inode *v9fs_qid_iget_dotl(
 	unlock_new_inode(inode);
 	return inode;
 error:
-	unlock_new_inode(inode);
-	iput(inode);
+	iget_failed(inode);
 	return ERR_PTR(retval);
 
 }

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Ilya Dryomov <[email protected]>

commit 5a60e87603c4c533492c515b7f62578189b03c9c upstream.

rbd_obj_request_create() is called on the main I/O path, so we need to use GFP_NOIO to make sure allocation doesn't blow back on us. Not all callers need this, but I'm still hardcoding the flag inside rather than making it a parameter because a) this is going to stable, and b) those callers shouldn't really use rbd_obj_request_create() and will be fixed in the future.

More memory allocation fixes will follow.

---
 drivers/block/rbd.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -2001,11 +2001,11 @@ static struct rbd_obj_request *rbd_obj_r
 	rbd_assert(obj_request_type_valid(type));
 
 	size = strlen(object_name) + 1;
-	name = kmalloc(size, GFP_KERNEL);
+	name = kmalloc(size, GFP_NOIO);
 	if (!name)
 		return NULL;
 
-	obj_request = kmem_cache_zalloc(rbd_obj_request_cache, GFP_KERNEL);
+	obj_request = kmem_cache_zalloc(rbd_obj_request_cache, GFP_NOIO);
 	if (!obj_request) {
 		kfree(name);
 		return NULL;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Chris Wilson <[email protected]>

commit 8b572a4200828b4e75cc22ed2f494b58d5372d65 upstream.

In needs_ilk_vtd_wa(), we pass in the GPU device but compared it against the ids for the mobile GPU and the mobile host bridge. That latter is impossible and so likely was just a typo for the desktop GPU device id (which is also buggy).

Fixes commit da88a5f7f7d434e2cde1b3e19d952e6d84533662

Author: Chris Wilson <[email protected]>
Date:   Wed Feb 13 09:31:53 2013 +0000

drm/i915: Disable WC PTE updates to w/a buggy IOMMU on ILK

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91127
References: https://bugzilla.freedesktop.org/show_bug.cgi?id=60391

---
 drivers/char/agp/intel-gtt.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -581,7 +581,7 @@ static inline int needs_ilk_vtd_wa(void)
 	/* Query intel_iommu to see if we need the workaround. Presumably that
 	 * was loaded first.
 	 */
-	if ((gpu_devid == PCI_DEVICE_ID_INTEL_IRONLAKE_M_HB ||
+	if ((gpu_devid == PCI_DEVICE_ID_INTEL_IRONLAKE_D_IG ||
 	     gpu_devid == PCI_DEVICE_ID_INTEL_IRONLAKE_M_IG) &&
 	     intel_iommu_gfx_mapped)
 		return 1;

4.1-stable review patch. If anyone has any objections, please let me know.

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

From: Russell King <[email protected]>

commit 5d6bed2a9c8bc161bff4cc7cede00f2e0e27a7e7 upstream.

v3.18 changed handle_IRQ() to call __handle_domain_irq(), which now rejects attempts to deliver IRQ0. Since IRQ 0 is used as the timer interrupt (just like the PIT on x86), this causes boot to fail as the bogomips calibration never completes.

Fix this by shuffling all interrupts up by one.

---
 arch/arm/mach-dove/include/mach/irqs.h |  118 ++++++++++++++++-----------------
 arch/arm/mach-dove/irq.c               |    8 +-
 2 files changed, 63 insertions(+), 63 deletions(-)

@@ -14,73 +14,73 @@
 /*
  * Dove Low Interrupt Controller
  */
-#define IRQ_DOVE_BRIDGE		0
-#define IRQ_DOVE_H2C		1
-#define IRQ_DOVE_C2H		2
-#define IRQ_DOVE_NAND		3
-#define IRQ_DOVE_PDMA		4
-#define IRQ_DOVE_SPI1		5
-#define IRQ_DOVE_SPI0		6
-#define IRQ_DOVE_UART_0		7
-#define IRQ_DOVE_UART_1		8
-#define IRQ_DOVE_UART_2		9
-#define IRQ_DOVE_UART_3		10
-#define IRQ_DOVE_I2C		11
-#define IRQ_DOVE_GPIO_0_7	12
-#define IRQ_DOVE_GPIO_8_15	13
-#define IRQ_DOVE_GPIO_16_23	14
-#define IRQ_DOVE_PCIE0_ERR	15
-#define IRQ_DOVE_PCIE0		16
-#define IRQ_DOVE_PCIE1_ERR	17
-#define IRQ_DOVE_PCIE1		18
-#define IRQ_DOVE_I2S0		19
-#define IRQ_DOVE_I2S0_ERR	20
-#define IRQ_DOVE_I2S1		21
-#define IRQ_DOVE_I2S1_ERR	22
-#define IRQ_DOVE_USB_ERR	23
-#define IRQ_DOVE_USB0		24
-#define IRQ_DOVE_USB1		25
-#define IRQ_DOVE_GE00_RX	26
-#define IRQ_DOVE_GE00_TX	27
-#define IRQ_DOVE_GE00_MISC	28
-#define IRQ_DOVE_GE00_SUM	29
-#define IRQ_DOVE_GE00_ERR	30
-#define IRQ_DOVE_CRYPTO		31
+#define IRQ_DOVE_BRIDGE		(1 + 0)
+#define IRQ_DOVE_H2C		(1 + 1)
+#define IRQ_DOVE_C2H		(1 + 2)
+#define IRQ_DOVE_NAND		(1 + 3)
+#define IRQ_DOVE_PDMA		(1 + 4)
+#define IRQ_DOVE_SPI1		(1 + 5)
+#define IRQ_DOVE_SPI0		(1 + 6)
+#define IRQ_DOVE_UART_0		(1 + 7)
+#define IRQ_DOVE_UART_1		(1 + 8)
+#define IRQ_DOVE_UART_2		(1 + 9)
+#define IRQ_DOVE_UART_3		(1 + 10)
+#define IRQ_DOVE_I2C		(1 + 11)
+#define IRQ_DOVE_GPIO_0_7	(1 + 12)
+#define IRQ_DOVE_GPIO_8_15	(1 + 13)
+#define IRQ_DOVE_GPIO_16_23	(1 + 14)
+#define IRQ_DOVE_PCIE0_ERR	(1 + 15)
+#define IRQ_DOVE_PCIE0		(1 + 16)
+#define IRQ_DOVE_PCIE1_ERR	(1 + 17)
+#define IRQ_DOVE_PCIE1		(1 + 18)
+#define IRQ_DOVE_I2S0		(1 + 19)
+#define IRQ_DOVE_I2S0_ERR	(1 + 20)
+#define IRQ_DOVE_I2S1		(1 + 21)
+#define IRQ_DOVE_I2S1_ERR	(1 + 22)
+#define IRQ_DOVE_USB_ERR	(1 + 23)
+#define IRQ_DOVE_USB0		(1 + 24)
+#define IRQ_DOVE_USB1		(1 + 25)
+#define IRQ_DOVE_GE00_RX	(1 + 26)
+#define IRQ_DOVE_GE00_TX	(1 + 27)
+#define IRQ_DOVE_GE00_MISC	(1 + 28)
+#define IRQ_DOVE_GE00_SUM	(1 + 29)
+#define IRQ_DOVE_GE00_ERR	(1 + 30)
+#define IRQ_DOVE_CRYPTO		(1 + 31)
 
 /*
  * Dove High Interrupt Controller
  */
-#define IRQ_DOVE_AC97		32
-#define IRQ_DOVE_PMU		33
-#define IRQ_DOVE_CAM		34
-#define IRQ_DOVE_SDIO0		35
-#define IRQ_DOVE_SDIO1		36
-#define IRQ_DOVE_SDIO0_WAKEUP	37
-#define IRQ_DOVE_SDIO1_WAKEUP	38
-#define IRQ_DOVE_XOR_00		39
-#define IRQ_DOVE_XOR_01		40
-#define IRQ_DOVE_XOR0_ERR	41
-#define IRQ_DOVE_XOR_10		42
-#define IRQ_DOVE_XOR_11		43
-#define IRQ_DOVE_XOR1_ERR	44
-#define IRQ_DOVE_LCD_DCON	45
-#define IRQ_DOVE_LCD1		46
-#define IRQ_DOVE_LCD0		47
-#define IRQ_DOVE_GPU		48
-#define IRQ_DOVE_PERFORM_MNTR	49
-#define IRQ_DOVE_VPRO_DMA1	51
-#define IRQ_DOVE_SSP_TIMER	54
-#define IRQ_DOVE_SSP		55
-#define IRQ_DOVE_MC_L2_ERR	56
-#define IRQ_DOVE_CRYPTO_ERR	59
-#define IRQ_DOVE_GPIO_24_31	60
-#define IRQ_DOVE_HIGH_GPIO	61
-#define IRQ_DOVE_SATA		62
+#define IRQ_DOVE_AC97		(1 + 32)
+#define IRQ_DOVE_PMU		(1 + 33)
+#define IRQ_DOVE_CAM		(1 + 34)
+#define IRQ_DOVE_SDIO0		(1 + 35)
+#define IRQ_DOVE_SDIO1		(1 + 36)
+#define IRQ_DOVE_SDIO0_WAKEUP	(1 + 37)
+#define IRQ_DOVE_SDIO1_WAKEUP	(1 + 38)
+#define IRQ_DOVE_XOR_00		(1 + 39)
+#define IRQ_DOVE_XOR_01		(1 + 40)
+#define IRQ_DOVE_XOR0_ERR	(1 + 41)
+#define IRQ_DOVE_XOR_10		(1 + 42)
+#define IRQ_DOVE_XOR_11		(1 + 43)
+#define IRQ_DOVE_XOR1_ERR	(1 + 44)
+#define IRQ_DOVE_LCD_DCON	(1 + 45)
+#define IRQ_DOVE_LCD1		(1 + 46)
+#define IRQ_DOVE_LCD0		(1 + 47)
+#define IRQ_DOVE_GPU		(1 + 48)
+#define IRQ_DOVE_PERFORM_MNTR	(1 + 49)
+#define IRQ_DOVE_VPRO_DMA1	(1 + 51)
+#define IRQ_DOVE_SSP_TIMER	(1 + 54)
+#define IRQ_DOVE_SSP		(1 + 55)
+#define IRQ_DOVE_MC_L2_ERR	(1 + 56)
+#define IRQ_DOVE_CRYPTO_ERR	(1 + 59)
+#define IRQ_DOVE_GPIO_24_31	(1 + 60)
+#define IRQ_DOVE_HIGH_GPIO	(1 + 61)
+#define IRQ_DOVE_SATA		(1 + 62)
 
 /*
  * DOVE General Purpose Pins
  */
-#define IRQ_DOVE_GPIO_START	64
+#define IRQ_DOVE_GPIO_START	65
 #define NR_GPIO_IRQS		64
 
 /*