4.9.230-rc1 review

This is the start of the stable review cycle for the 4.9.230 release. There are 24 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.

Responses should be made by Thu, 09 Jul 2020 14:57:34 +0000. Anything received after that time might be too late.

The whole patch series can be found in one patch at:

	https://www.kernel.org/pub/linux/kernel/v4.x/stable-review/patch-4.9.230-rc1.gz

or in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-4.9.y
and the diffstat can be found below.

thanks,

greg k-h

-------------
Pseudo-Shortlog of commits:

Greg Kroah-Hartman <[email protected]>

    Linux 4.9.230-rc1

Peter Jones <[email protected]>
efi: Make it possible to disable efivar_ssdt entirely

Vasily Averin <[email protected]>
netfilter: nf_conntrack_h323: lost .data_len definition for Q.931/ipv6

Hauke Mehrtens <[email protected]>

    MIPS: Add missing EHB in mtc0 -> mfc0 sequence for DSPen

Zhang Xiaoxu <[email protected]>
cifs: Fix the target file was deleted when rename failed.

Paul Aurich <[email protected]>
SMB3: Honor persistent/resilient handle flags for multiuser mounts

Paul Aurich <[email protected]>
SMB3: Honor 'seal' flag for multiuser mounts

Greg Kroah-Hartman <[email protected]>
Revert "ALSA: usb-audio: Improve frames size computation"

Chris Packham <[email protected]>

    i2c: algo-pca: Add 0x78 as SCL stuck low status for PCA9665

Hou Tao <[email protected]>
virtio-blk: free vblk-vqs in error path of virtblk_probe()

Misono Tomohiro <[email protected]>
hwmon: (acpi_power_meter) Fix potential memory leak in acpi_power_meter_add()

Chu Lin <[email protected]>
hwmon: (max6697) Make sure the OVERT mask is set correctly

Rahul Lakkireddy <[email protected]>
cxgb4: parse TC-U32 key values and masks natively

Shile Zhang <[email protected]>
sched/rt: Show the 'sched_rr_timeslice' SCHED_RR timeslice tuning knob in milliseconds

Herbert Xu <[email protected]>
crypto: af_alg - fix use-after-free in af_alg_accept() due to bh_lock_sock()

Douglas Anderson <[email protected]>
kgdb: Avoid suspicious RCU usage warning

Zqiang <[email protected]>

    usb: usbtest: fix missing kfree(dev->buf) in usbtest_disconnect

Qian Cai <[email protected]>
mm/slub: fix stack overruns with SLUB_STATS

Dongli Zhang <[email protected]>
mm/slub.c: fix corrupted freechain in deactivate_slab()

Tuomas Tynkkynen <[email protected]>
usbnet: smsc95xx: Fix use-after-free after removal

Borislav Petkov <[email protected]>
EDAC/amd64: Read back the scrub rate PCI register on F15h

Hugh Dickins <[email protected]>
mm: fix swap cache node allocation mask

Filipe Manana <[email protected]>
btrfs: fix data block group relocation failure due to concurrent scrub

Anand Jain <[email protected]>
btrfs: cow_file_range() num_bytes and disk_num_bytes are same

Filipe Manana <[email protected]>
btrfs: fix a block group ref counter leak after failure to remove block group

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

Diffstat:

 Makefile                                           |   4 +-
 arch/mips/kernel/traps.c                           |   1 +
 crypto/af_alg.c                                    |  26 ++---
 crypto/algif_aead.c                                |   9 +-
 crypto/algif_hash.c                                |   9 +-
 crypto/algif_skcipher.c                            |   9 +-
 drivers/block/virtio_blk.c                         |   1 +
 drivers/edac/amd64_edac.c                          |   2 +
 drivers/firmware/efi/Kconfig                       |  11 ++
 drivers/firmware/efi/efi.c                         |   2 +-
 drivers/hwmon/acpi_power_meter.c                   |   4 +-
 drivers/hwmon/max6697.c                            |   7 +-
 drivers/i2c/algos/i2c-algo-pca.c                   |   3 +-
 drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32.c  |  18 +--
 .../ethernet/chelsio/cxgb4/cxgb4_tc_u32_parse.h    | 122 ++++++++++++++-------
 drivers/net/usb/smsc95xx.c                         |   2 +-
 drivers/usb/misc/usbtest.c                         |   1 +
 fs/btrfs/extent-tree.c                             |  19 ++--
 fs/btrfs/inode.c                                   |  36 ++++--
 fs/cifs/connect.c                                  |   3 +
 fs/cifs/inode.c                                    |  10 +-
 include/crypto/if_alg.h                            |   4 +-
 include/linux/sched/sysctl.h                       |   1 +
 kernel/debug/debug_core.c                          |   4 +
 kernel/sched/core.c                                |   5 +-
 kernel/sched/rt.c                                  |   1 +
 kernel/sysctl.c                                    |   2 +-
 mm/slub.c                                          |  30 ++++-
 mm/swap_state.c                                    |   3 +-
 net/netfilter/nf_conntrack_h323_main.c             |   1 +
 sound/usb/card.h                                   |   4 -
 sound/usb/endpoint.c                               |  43 +-------
 sound/usb/endpoint.h                               |   1 -
 sound/usb/pcm.c                                    |   2 -
 34 files changed, 235 insertions(+), 165 deletions(-)

From: Filipe Manana <[email protected]>

[ Upstream commit 9fecd13202f520f3f25d5b1c313adb740fe19773 ]

When removing a block group, if we fail to delete the block group's item from the extent tree, we jump to the 'out' label and end up decrementing the block group's reference count once only (by 1), resulting in a counter leak because the block group at that point was already removed from the block group cache rbtree - so we have to decrement the reference count twice, once for the rbtree and once for our lookup at the start of the function.

There is a second bug where if removing the free space tree entries (the call to remove_block_group_free_space()) fails we end up jumping to the 'out_put_group' label but end up decrementing the reference count only once, when we should have done it twice, since we have already removed the block group from the block group cache rbtree. This happens because the reference count decrement for the rbtree reference happens after attempting to remove the free space tree entries, which is far away from the place where we remove the block group from the rbtree.

To make things less error prone, decrement the reference count for the rbtree immediately after removing the block group from it. This also eleminates the need for two different exit labels on error, renaming 'out_put_label' to just 'out' and removing the old 'out'.

---
 fs/btrfs/extent-tree.c | 19 +++++++++----------
 1 file changed, 9 insertions(+), 10 deletions(-)

index a83f353e44188..c0033a0d00787 100644
@@ -10645,7 +10645,7 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 	path = btrfs_alloc_path();
 	if (!path) {
 		ret = -ENOMEM;
-		goto out_put_group;
+		goto out;
 	}
 
 	/*

@@ -10684,7 +10684,7 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 		ret = btrfs_orphan_add(trans, inode);
 		if (ret) {
 			btrfs_add_delayed_iput(inode);
-			goto out_put_group;
+			goto out;
 		}
 		clear_nlink(inode);
 		/* One for the block groups ref */

@@ -10707,13 +10707,13 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 
 	ret = btrfs_search_slot(trans, tree_root, &key, path, -1, 1);
 	if (ret < 0)
-		goto out_put_group;
+		goto out;
 	if (ret > 0)
 		btrfs_release_path(path);
 	if (ret == 0) {
 		ret = btrfs_del_item(trans, tree_root, path);
 		if (ret)
-			goto out_put_group;
+			goto out;
 		btrfs_release_path(path);
 	}
 

@@ -10722,6 +10722,9 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 		 &root->fs_info->block_group_cache_tree);
 	RB_CLEAR_NODE(&block_group->cache_node);
 
+	/* Once for the block groups rbtree */
+	btrfs_put_block_group(block_group);
+
 	if (root->fs_info->first_logical_byte == block_group->key.objectid)
 		root->fs_info->first_logical_byte = (u64)-1;
 	spin_unlock(&root->fs_info->block_group_cache_lock);

@@ -10871,10 +10874,7 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 
 	ret = remove_block_group_free_space(trans, root->fs_info, block_group);
 	if (ret)
-		goto out_put_group;
-
-	/* Once for the block groups rbtree */
-	btrfs_put_block_group(block_group);
+		goto out;
 
 	ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
 	if (ret > 0)

@@ -10884,10 +10884,9 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 
 	ret = btrfs_del_item(trans, root, path);
 
-out_put_group:
+out:
 	/* Once for the lookup reference */
 	btrfs_put_block_group(block_group);
-out:
 	btrfs_free_path(path);
 	return ret;
 }

2.25.1

From: Anand Jain <[email protected]>

[ Upstream commit 3752d22fcea160cc2493e34f5e0e41cdd7fdd921 ]

This patch deletes local variable disk_num_bytes as its value is same as num_bytes in the function cow_file_range().

---
 fs/btrfs/inode.c | 17 ++++++++---------
 1 file changed, 8 insertions(+), 9 deletions(-)

index c425443c31fea..6d63050abe214 100644
@@ -947,7 +947,6 @@ static noinline int cow_file_range(struct inode *inode,
 	u64 alloc_hint = 0;
 	u64 num_bytes;
 	unsigned long ram_size;
-	u64 disk_num_bytes;
 	u64 cur_alloc_size;
 	u64 blocksize = root->sectorsize;
 	struct btrfs_key ins;

@@ -963,7 +962,6 @@ static noinline int cow_file_range(struct inode *inode,
 
 	num_bytes = ALIGN(end - start + 1, blocksize);
 	num_bytes = max(blocksize,  num_bytes);
-	disk_num_bytes = num_bytes;
 
 	/* if this is a small write inside eof, kick off defrag */
 	if (num_bytes < SZ_64K &&

@@ -992,16 +990,15 @@ static noinline int cow_file_range(struct inode *inode,
 		}
 	}
 
-	BUG_ON(disk_num_bytes >
-	       btrfs_super_total_bytes(root->fs_info->super_copy));
+	BUG_ON(num_bytes > btrfs_super_total_bytes(root->fs_info->super_copy));
 
 	alloc_hint = get_extent_allocation_hint(inode, start, num_bytes);
 	btrfs_drop_extent_cache(inode, start, start + num_bytes - 1, 0);
 
-	while (disk_num_bytes > 0) {
+	while (num_bytes > 0) {
 		unsigned long op;
 
-		cur_alloc_size = disk_num_bytes;
+		cur_alloc_size = num_bytes;
 		ret = btrfs_reserve_extent(root, cur_alloc_size, cur_alloc_size,
 					   root->sectorsize, 0, alloc_hint,
 					   &ins, 1, 1);

@@ -1058,7 +1055,7 @@ static noinline int cow_file_range(struct inode *inode,
 
 		btrfs_dec_block_group_reservations(root->fs_info, ins.objectid);
 
-		if (disk_num_bytes < cur_alloc_size)
+		if (num_bytes < cur_alloc_size)
 			break;
 
 		/* we're not doing compressed IO, don't unlock the first

@@ -1076,8 +1073,10 @@ static noinline int cow_file_range(struct inode *inode,
 					     delalloc_end, locked_page,
 					     EXTENT_LOCKED | EXTENT_DELALLOC,
 					     op);
-		disk_num_bytes -= cur_alloc_size;
-		num_bytes -= cur_alloc_size;
+		if (num_bytes < cur_alloc_size)
+			num_bytes = 0;
+		else
+			num_bytes -= cur_alloc_size;
 		alloc_hint = ins.objectid + ins.offset;
 		start += cur_alloc_size;
 	}

2.25.1

From: Filipe Manana <[email protected]>

[ Upstream commit 432cd2a10f1c10cead91fe706ff5dc52f06d642a ]

When running relocation of a data block group while scrub is running in parallel, it is possible that the relocation will fail and abort the current transaction with an -EINVAL error:

   [134243.988595] BTRFS info (device sdc): found 14 extents, stage: move data extents
   [134243.999871] ------------[ cut here ]------------
   [134244.000741] BTRFS: Transaction aborted (error -22)
   [134244.001692] WARNING: CPU: 0 PID: 26954 at fs/btrfs/ctree.c:1071 __btrfs_cow_block+0x6a7/0x790 [btrfs]
   [134244.003380] Modules linked in: btrfs blake2b_generic xor raid6_pq (...)
   [134244.012577] CPU: 0 PID: 26954 Comm: btrfs Tainted: G        W         5.6.0-rc7-btrfs-next-58 #5
   [134244.014162] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.0-59-gc9ba5276e321-prebuilt.qemu.org 04/01/2014
   [134244.016184] RIP: 0010:__btrfs_cow_block+0x6a7/0x790 [btrfs]
   [134244.017151] Code: 48 c7 c7 (...)
   [134244.020549] RSP: 0018:ffffa41607863888 EFLAGS: 00010286
   [134244.021515] RAX: 0000000000000000 RBX: ffff9614bdfe09c8 RCX: 0000000000000000
   [134244.022822] RDX: 0000000000000001 RSI: ffffffffb3d63980 RDI: 0000000000000001
   [134244.024124] RBP: ffff961589e8c000 R08: 0000000000000000 R09: 0000000000000001
   [134244.025424] R10: ffffffffc0ae5955 R11: 0000000000000000 R12: ffff9614bd530d08
   [134244.026725] R13: ffff9614ced41b88 R14: ffff9614bdfe2a48 R15: 0000000000000000
   [134244.028024] FS:  00007f29b63c08c0(0000) GS:ffff9615ba600000(0000) knlGS:0000000000000000
   [134244.029491] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
   [134244.030560] CR2: 00007f4eb339b000 CR3: 0000000130d6e006 CR4: 00000000003606f0
   [134244.031997] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
   [134244.033153] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
   [134244.034484] Call Trace:
   [134244.034984]  btrfs_cow_block+0x12b/0x2b0 [btrfs]
   [134244.035859]  do_relocation+0x30b/0x790 [btrfs]
   [134244.036681]  ? do_raw_spin_unlock+0x49/0xc0
   [134244.037460]  ? _raw_spin_unlock+0x29/0x40
   [134244.038235]  relocate_tree_blocks+0x37b/0x730 [btrfs]
   [134244.039245]  relocate_block_group+0x388/0x770 [btrfs]
   [134244.040228]  btrfs_relocate_block_group+0x161/0x2e0 [btrfs]
   [134244.041323]  btrfs_relocate_chunk+0x36/0x110 [btrfs]
   [134244.041345]  btrfs_balance+0xc06/0x1860 [btrfs]
   [134244.043382]  ? btrfs_ioctl_balance+0x27c/0x310 [btrfs]
   [134244.045586]  btrfs_ioctl_balance+0x1ed/0x310 [btrfs]
   [134244.045611]  btrfs_ioctl+0x1880/0x3760 [btrfs]
   [134244.049043]  ? do_raw_spin_unlock+0x49/0xc0
   [134244.049838]  ? _raw_spin_unlock+0x29/0x40
   [134244.050587]  ? __handle_mm_fault+0x11b3/0x14b0
   [134244.051417]  ? ksys_ioctl+0x92/0xb0
   [134244.052070]  ksys_ioctl+0x92/0xb0
   [134244.052701]  ? trace_hardirqs_off_thunk+0x1a/0x1c
   [134244.053511]  __x64_sys_ioctl+0x16/0x20
   [134244.054206]  do_syscall_64+0x5c/0x280
   [134244.054891]  entry_SYSCALL_64_after_hwframe+0x49/0xbe
   [134244.055819] RIP: 0033:0x7f29b51c9dd7
   [134244.056491] Code: 00 00 00 (...)
   [134244.059767] RSP: 002b:00007ffcccc1dd08 EFLAGS: 00000202 ORIG_RAX: 0000000000000010
   [134244.061168] RAX: ffffffffffffffda RBX: 0000000000000001 RCX: 00007f29b51c9dd7
   [134244.062474] RDX: 00007ffcccc1dda0 RSI: 00000000c4009420 RDI: 0000000000000003
   [134244.063771] RBP: 0000000000000003 R08: 00005565cea4b000 R09: 0000000000000000
   [134244.065032] R10: 0000000000000541 R11: 0000000000000202 R12: 00007ffcccc2060a
   [134244.066327] R13: 00007ffcccc1dda0 R14: 0000000000000002 R15: 00007ffcccc1dec0
   [134244.067626] irq event stamp: 0
   [134244.068202] hardirqs last  enabled at (0): [<0000000000000000>] 0x0
   [134244.069351] hardirqs last disabled at (0): [<ffffffffb2abdedf>] copy_process+0x74f/0x2020
   [134244.070909] softirqs last  enabled at (0): [<ffffffffb2abdedf>] copy_process+0x74f/0x2020
   [134244.072392] softirqs last disabled at (0): [<0000000000000000>] 0x0
   [134244.073432] ---[ end trace bd7c03622e0b0a99 ]---

The -EINVAL error comes from the following chain of function calls:

__btrfs_cow_block() <-- aborts the transaction
btrfs_reloc_cow_block()
replace_file_extents()
get_new_location() <-- returns -EINVAL

When relocating a data block group, for each allocated extent of the block group, we preallocate another extent (at prealloc_file_extent_cluster()), associated with the data relocation inode, and then dirty all its pages. These preallocated extents have, and must have, the same size that extents from the data block group being relocated have.

Later before we start the relocation stage that updates pointers (bytenr field of file extent items) to point to the the new extents, we trigger writeback for the data relocation inode. The expectation is that writeback will write the pages to the previously preallocated extents, that it follows the NOCOW path. That is generally the case, however, if a scrub is running it may have turned the block group that contains those extents into RO mode, in which case writeback falls back to the COW path.

However in the COW path instead of allocating exactly one extent with the expected size, the allocator may end up allocating several smaller extents due to free space fragmentation - because we tell it at cow_file_range() that the minimum allocation size can match the filesystem's sector size. This later breaks the relocation's expectation that an extent associated to a file extent item in the data relocation inode has the same size as the respective extent pointed by a file extent item in another tree - in this case the extent to which the relocation inode poins to is smaller, causing relocation.c:get_new_location() to return -EINVAL.

For example, if we are relocating a data block group X that has a logical address of X and the block group has an extent allocated at the logical address X + 128KiB with a size of 64KiB:

1) At prealloc_file_extent_cluster() we allocate an extent for the data relocation inode with a size of 64KiB and associate it to the file offset 128KiB (X + 128KiB - X) of the data relocation inode. This preallocated extent was allocated at block group Z;

2) A scrub running in parallel turns block group Z into RO mode and starts scrubing its extents;

3) Relocation triggers writeback for the data relocation inode;

4) When running delalloc (btrfs_run_delalloc_range()), we try first the NOCOW path because the data relocation inode has BTRFS_INODE_PREALLOC set in its flags. However, because block group Z is in RO mode, the NOCOW path (run_delalloc_nocow()) falls back into the COW path, by calling cow_file_range();

5) At cow_file_range(), in the first iteration of the while loop we call btrfs_reserve_extent() to allocate a 64KiB extent and pass it a minimum allocation size of 4KiB (fs_info->sectorsize). Due to free space fragmentation, btrfs_reserve_extent() ends up allocating two extents of 32KiB each, each one on a different iteration of that while loop;

6) Writeback of the data relocation inode completes;

7) Relocation proceeds and ends up at relocation.c:replace_file_extents(), with a leaf which has a file extent item that points to the data extent from block group X, that has a logical address (bytenr) of X + 128KiB and a size of 64KiB. Then it calls get_new_location(), which does a lookup in the data relocation tree for a file extent item starting at offset 128KiB (X + 128KiB - X) and belonging to the data relocation inode. It finds a corresponding file extent item, however that item points to an extent that has a size of 32KiB, which doesn't match the expected size of 64KiB, resuling in -EINVAL being returned from this function and propagated up to __btrfs_cow_block(), which aborts the current transaction.

To fix this make sure that at cow_file_range() when we call the allocator we pass it a minimum allocation size corresponding the desired extent size if the inode belongs to the data relocation tree, otherwise pass it the filesystem's sector size as the minimum allocation size.

---
 fs/btrfs/inode.c | 19 ++++++++++++++++++-
 1 file changed, 18 insertions(+), 1 deletion(-)

index 6d63050abe214..dfc0b3adf57af 100644
@@ -947,6 +947,7 @@ static noinline int cow_file_range(struct inode *inode,
 	u64 alloc_hint = 0;
 	u64 num_bytes;
 	unsigned long ram_size;
+	u64 min_alloc_size;
 	u64 cur_alloc_size;
 	u64 blocksize = root->sectorsize;
 	struct btrfs_key ins;

@@ -995,12 +996,28 @@ static noinline int cow_file_range(struct inode *inode,
 	alloc_hint = get_extent_allocation_hint(inode, start, num_bytes);
 	btrfs_drop_extent_cache(inode, start, start + num_bytes - 1, 0);
 
+	/*
+	 * Relocation relies on the relocated extents to have exactly the same
+	 * size as the original extents. Normally writeback for relocation data
+	 * extents follows a NOCOW path because relocation preallocates the
+	 * extents. However, due to an operation such as scrub turning a block
+	 * group to RO mode, it may fallback to COW mode, so we must make sure
+	 * an extent allocated during COW has exactly the requested size and can
+	 * not be split into smaller extents, otherwise relocation breaks and
+	 * fails during the stage where it updates the bytenr of file extent
+	 * items.
+	 */
+	if (root->root_key.objectid == BTRFS_DATA_RELOC_TREE_OBJECTID)
+		min_alloc_size = num_bytes;
+	else
+		min_alloc_size = root->sectorsize;
+
 	while (num_bytes > 0) {
 		unsigned long op;
 
 		cur_alloc_size = num_bytes;
 		ret = btrfs_reserve_extent(root, cur_alloc_size, cur_alloc_size,
-					   root->sectorsize, 0, alloc_hint,
+					   min_alloc_size, 0, alloc_hint,
 					   &ins, 1, 1);
 		if (ret < 0)
 			goto out_unlock;

2.25.1

From: Hugh Dickins <[email protected]>

[ Upstream commit 243bce09c91b0145aeaedd5afba799d81841c030 ]

Chris Murphy reports that a slightly overcommitted load, testing swap and zram along with i915, splats and keeps on splatting, when it had better fail less noisily:

gnome-shell: page allocation failure: order:0, mode:0x400d0(__GFP_IO|__GFP_FS|__GFP_COMP|__GFP_RECLAIMABLE), nodemask=(null),cpuset=/,mems_allowed=0

  CPU: 2 PID: 1155 Comm: gnome-shell Not tainted 5.7.0-1.fc33.x86_64 #1
  Call Trace:
    dump_stack+0x64/0x88
    warn_alloc.cold+0x75/0xd9
    __alloc_pages_slowpath.constprop.0+0xcfa/0xd30
    __alloc_pages_nodemask+0x2df/0x320
    alloc_slab_page+0x195/0x310
    allocate_slab+0x3c5/0x440
    ___slab_alloc+0x40c/0x5f0
    __slab_alloc+0x1c/0x30
    kmem_cache_alloc+0x20e/0x220
    xas_nomem+0x28/0x70
    add_to_swap_cache+0x321/0x400
    __read_swap_cache_async+0x105/0x240
    swap_cluster_readahead+0x22c/0x2e0
    shmem_swapin+0x8e/0xc0
    shmem_swapin_page+0x196/0x740
    shmem_getpage_gfp+0x3a2/0xa60
    shmem_read_mapping_page_gfp+0x32/0x60
    shmem_get_pages+0x155/0x5e0 [i915]
    __i915_gem_object_get_pages+0x68/0xa0 [i915]
    i915_vma_pin+0x3fe/0x6c0 [i915]
    eb_add_vma+0x10b/0x2c0 [i915]
    i915_gem_do_execbuffer+0x704/0x3430 [i915]
    i915_gem_execbuffer2_ioctl+0x1ea/0x3e0 [i915]
    drm_ioctl_kernel+0x86/0xd0 [drm]
    drm_ioctl+0x206/0x390 [drm]
    ksys_ioctl+0x82/0xc0
    __x64_sys_ioctl+0x16/0x20
    do_syscall_64+0x5b/0xf0
    entry_SYSCALL_64_after_hwframe+0x44/0xa9

Reported on 5.7, but it goes back really to 3.1: when shmem_read_mapping_page_gfp() was implemented for use by i915, and allowed for __GFP_NORETRY and __GFP_NOWARN flags in most places, but missed swapin's "& GFP_KERNEL" mask for page tree node allocation in __read_swap_cache_async() - that was to mask off HIGHUSER_MOVABLE bits from what page cache uses, but GFP_RECLAIM_MASK is now what's needed.

---
 mm/swap_state.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

index 35d7e0ee1c77c..f5cb6b23cedaf 100644
@@ -19,6 +19,7 @@
 #include <linux/migrate.h>
 
 #include <asm/pgtable.h>
+#include "internal.h"
 
 /*
  * swapper_space is a fiction, retained to simplify the path through

@@ -326,7 +327,7 @@ struct page *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
 		/*
 		 * call radix_tree_preload() while we can wait.
 		 */
-		err = radix_tree_maybe_preload(gfp_mask & GFP_KERNEL);
+		err = radix_tree_maybe_preload(gfp_mask & GFP_RECLAIM_MASK);
 		if (err)
 			break;
 

2.25.1

From: Borislav Petkov <[email protected]>

[ Upstream commit ee470bb25d0dcdf126f586ec0ae6dca66cb340a4 ]

Commit:

da92110dfdfa ("EDAC, amd64_edac: Extend scrub rate support to F15hM60h")

added support for F15h, model 0x60 CPUs but in doing so, missed to read back SCRCTRL PCI config register on F15h CPUs which are *not* model 0x60. Add that read so that doing

  $ cat /sys/devices/system/edac/mc/mc0/sdram_scrub_rate

can show the previously set DRAM scrub rate.

---
 drivers/edac/amd64_edac.c | 2 ++
 1 file changed, 2 insertions(+)

index 1c5f23224b3cb..020dd07d1c23a 100644
@@ -243,6 +243,8 @@ static int get_scrub_rate(struct mem_ctl_info *mci)
 
 		if (pvt->model == 0x60)
 			amd64_read_pci_cfg(pvt->F2, F15H_M60H_SCRCTRL, &scrubval);
+		else
+			amd64_read_pci_cfg(pvt->F3, SCRCTRL, &scrubval);
 	} else
 		amd64_read_pci_cfg(pvt->F3, SCRCTRL, &scrubval);
 

2.25.1

From: Tuomas Tynkkynen <[email protected]>

[ Upstream commit b835a71ef64a61383c414d6bf2896d2c0161deca ]

Syzbot reports an use-after-free in workqueue context:

BUG: KASAN: use-after-free in mutex_unlock+0x19/0x40 kernel/locking/mutex.c:737
 mutex_unlock+0x19/0x40 kernel/locking/mutex.c:737
 __smsc95xx_mdio_read drivers/net/usb/smsc95xx.c:217 [inline]
 smsc95xx_mdio_read+0x583/0x870 drivers/net/usb/smsc95xx.c:278
 check_carrier+0xd1/0x2e0 drivers/net/usb/smsc95xx.c:644
 process_one_work+0x777/0xf90 kernel/workqueue.c:2274
 worker_thread+0xa8f/0x1430 kernel/workqueue.c:2420
 kthread+0x2df/0x300 kernel/kthread.c:255

It looks like that smsc95xx_unbind() is freeing the structures that are still in use by the concurrently running workqueue callback. Thus switch to using cancel_delayed_work_sync() to ensure the work callback really is no longer active.

---
 drivers/net/usb/smsc95xx.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index e719ecd69d01b..6852ebafd4d3b 100644
@@ -1327,7 +1327,7 @@ static void smsc95xx_unbind(struct usbnet *dev, struct usb_interface *intf)
 	struct smsc95xx_priv *pdata = (struct smsc95xx_priv *)(dev->data[0]);
 
 	if (pdata) {
-		cancel_delayed_work(&pdata->carrier_check);
+		cancel_delayed_work_sync(&pdata->carrier_check);
 		netif_dbg(dev, ifdown, dev->net, "free pdata\n");
 		kfree(pdata);
 		pdata = NULL;

2.25.1

From: Dongli Zhang <[email protected]>

[ Upstream commit 52f23478081ae0dcdb95d1650ea1e7d52d586829 ]

The slub_debug is able to fix the corrupted slab freelist/page. However, alloc_debug_processing() only checks the validity of current and next freepointer during allocation path. As a result, once some objects have their freepointers corrupted, deactivate_slab() may lead to page fault.

Below is from a test kernel module when 'slub_debug=PUF,kmalloc-128 slub_nomerge'. The test kernel corrupts the freepointer of one free object on purpose. Unfortunately, deactivate_slab() does not detect it when iterating the freechain.

  BUG: unable to handle page fault for address: 00000000123456f8
  #PF: supervisor read access in kernel mode
  #PF: error_code(0x0000) - not-present page
  PGD 0 P4D 0
  Oops: 0000 [#1] SMP PTI
  ... ...
  RIP: 0010:deactivate_slab.isra.92+0xed/0x490
  ... ...
  Call Trace:
   ___slab_alloc+0x536/0x570
   __slab_alloc+0x17/0x30
   __kmalloc+0x1d9/0x200
   ext4_htree_store_dirent+0x30/0xf0
   htree_dirblock_to_tree+0xcb/0x1c0
   ext4_htree_fill_tree+0x1bc/0x2d0
   ext4_readdir+0x54f/0x920
   iterate_dir+0x88/0x190
   __x64_sys_getdents+0xa6/0x140
   do_syscall_64+0x49/0x170
   entry_SYSCALL_64_after_hwframe+0x44/0xa9

Therefore, this patch adds extra consistency check in deactivate_slab(). Once an object's freepointer is corrupted, all following objects starting at this object are isolated.

---
 mm/slub.c | 27 +++++++++++++++++++++++++++
 1 file changed, 27 insertions(+)

index 3d45713187a47..c20c87099a99d 100644
@@ -624,6 +624,20 @@ static void slab_fix(struct kmem_cache *s, char *fmt, ...)
 	va_end(args);
 }
 
+static bool freelist_corrupted(struct kmem_cache *s, struct page *page,
+			       void *freelist, void *nextfree)
+{
+	if ((s->flags & SLAB_CONSISTENCY_CHECKS) &&
+	    !check_valid_pointer(s, page, nextfree)) {
+		object_err(s, page, freelist, "Freechain corrupt");
+		freelist = NULL;
+		slab_fix(s, "Isolate corrupted freechain");
+		return true;
+	}
+
+	return false;
+}
+
 static void print_trailer(struct kmem_cache *s, struct page *page, u8 *p)
 {
 	unsigned int off;	/* Offset of last byte */

@@ -1305,6 +1319,11 @@ static inline void inc_slabs_node(struct kmem_cache *s, int node,
 static inline void dec_slabs_node(struct kmem_cache *s, int node,
 							int objects) {}
 
+static bool freelist_corrupted(struct kmem_cache *s, struct page *page,
+			       void *freelist, void *nextfree)
+{
+	return false;
+}
 #endif /* CONFIG_SLUB_DEBUG */
 
 /*

@@ -2016,6 +2035,14 @@ static void deactivate_slab(struct kmem_cache *s, struct page *page,
 		void *prior;
 		unsigned long counters;
 
+		/*
+		 * If 'nextfree' is invalid, it is possible that the object at
+		 * 'freelist' is already corrupted.  So isolate all objects
+		 * starting at 'freelist'.
+		 */
+		if (freelist_corrupted(s, page, freelist, nextfree))
+			break;
+
 		do {
 			prior = page->freelist;
 			counters = page->counters;

2.25.1

From: Qian Cai <[email protected]>

[ Upstream commit a68ee0573991e90af2f1785db309206408bad3e5 ]

There is no need to copy SLUB_STATS items from root memcg cache to new memcg cache copies. Doing so could result in stack overruns because the store function only accepts 0 to clear the stat and returns an error for everything else while the show method would print out the whole stat.

Then, the mismatch of the lengths returns from show and store methods happens in memcg_propagate_slab_attrs():

	else if (root_cache->max_attr_size < ARRAY_SIZE(mbuf))
		buf = mbuf;

max_attr_size is only 2 from slab_attr_store(), then, it uses mbuf[64] in show_stat() later where a bounch of sprintf() would overrun the stack variable. Fix it by always allocating a page of buffer to be used in show_stat() if SLUB_STATS=y which should only be used for debug purpose.

  # echo 1 > /sys/kernel/slab/fs_cache/shrink
  BUG: KASAN: stack-out-of-bounds in number+0x421/0x6e0
  Write of size 1 at addr ffffc900256cfde0 by task kworker/76:0/53251

  Hardware name: HPE ProLiant DL385 Gen10/ProLiant DL385 Gen10, BIOS A40 07/10/2019
  Workqueue: memcg_kmem_cache memcg_kmem_cache_create_func
  Call Trace:
    number+0x421/0x6e0
    vsnprintf+0x451/0x8e0
    sprintf+0x9e/0xd0
    show_stat+0x124/0x1d0
    alloc_slowpath_show+0x13/0x20
    __kmem_cache_create+0x47a/0x6b0

addr ffffc900256cfde0 is located in stack of task kworker/76:0/53251 at offset 0 in frame:
process_one_work+0x0/0xb90

this frame has 1 object:
[32, 72) 'lockdep_map'

  Memory state around the buggy address:
   ffffc900256cfc80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
   ffffc900256cfd00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
  >ffffc900256cfd80: 00 00 00 00 00 00 00 00 00 00 00 00 f1 f1 f1 f1
                                                         ^
   ffffc900256cfe00: 00 00 00 00 00 f2 f2 f2 00 00 00 00 00 00 00 00
   ffffc900256cfe80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
  ==================================================================
  Kernel panic - not syncing: stack-protector: Kernel stack is corrupted in: __kmem_cache_create+0x6ac/0x6b0
  Workqueue: memcg_kmem_cache memcg_kmem_cache_create_func
  Call Trace:
    __kmem_cache_create+0x6ac/0x6b0

---
 mm/slub.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

index c20c87099a99d..454c1d550ad22 100644
@@ -5526,7 +5526,8 @@ static void memcg_propagate_slab_attrs(struct kmem_cache *s)
 		 */
 		if (buffer)
 			buf = buffer;
-		else if (root_cache->max_attr_size < ARRAY_SIZE(mbuf))
+		else if (root_cache->max_attr_size < ARRAY_SIZE(mbuf) &&
+			 !IS_ENABLED(CONFIG_SLUB_STATS))
 			buf = mbuf;
 		else {
 			buffer = (char *) get_zeroed_page(GFP_KERNEL);

2.25.1

From: Zqiang <[email protected]>

[ Upstream commit 28ebeb8db77035e058a510ce9bd17c2b9a009dba ]

BUG: memory leak
unreferenced object 0xffff888055046e00 (size 256):
  comm "kworker/2:9", pid 2570, jiffies 4294942129 (age 1095.500s)
  hex dump (first 32 bytes):
    00 70 04 55 80 88 ff ff 18 bb 5a 81 ff ff ff ff  .p.U......Z.....
    f5 96 78 81 ff ff ff ff 37 de 8e 81 ff ff ff ff  ..x.....7.......
  backtrace:
    [<00000000d121dccf>] kmemleak_alloc_recursive
include/linux/kmemleak.h:43 [inline]
    [<00000000d121dccf>] slab_post_alloc_hook mm/slab.h:586 [inline]
    [<00000000d121dccf>] slab_alloc_node mm/slub.c:2786 [inline]
    [<00000000d121dccf>] slab_alloc mm/slub.c:2794 [inline]
    [<00000000d121dccf>] kmem_cache_alloc_trace+0x15e/0x2d0 mm/slub.c:2811
    [<000000005c3c3381>] kmalloc include/linux/slab.h:555 [inline]
    [<000000005c3c3381>] usbtest_probe+0x286/0x19d0
drivers/usb/misc/usbtest.c:2790
    [<000000001cec6910>] usb_probe_interface+0x2bd/0x870
drivers/usb/core/driver.c:361
    [<000000007806c118>] really_probe+0x48d/0x8f0 drivers/base/dd.c:551
    [<00000000a3308c3e>] driver_probe_device+0xfc/0x2a0 drivers/base/dd.c:724
    [<000000003ef66004>] __device_attach_driver+0x1b6/0x240
drivers/base/dd.c:831
    [<00000000eee53e97>] bus_for_each_drv+0x14e/0x1e0 drivers/base/bus.c:431
    [<00000000bb0648d0>] __device_attach+0x1f9/0x350 drivers/base/dd.c:897
    [<00000000838b324a>] device_initial_probe+0x1a/0x20 drivers/base/dd.c:944
    [<0000000030d501c1>] bus_probe_device+0x1e1/0x280 drivers/base/bus.c:491
    [<000000005bd7adef>] device_add+0x131d/0x1c40 drivers/base/core.c:2504
    [<00000000a0937814>] usb_set_configuration+0xe84/0x1ab0
drivers/usb/core/message.c:2030
    [<00000000e3934741>] generic_probe+0x6a/0xe0 drivers/usb/core/generic.c:210
    [<0000000098ade0f1>] usb_probe_device+0x90/0xd0
drivers/usb/core/driver.c:266
    [<000000007806c118>] really_probe+0x48d/0x8f0 drivers/base/dd.c:551
    [<00000000a3308c3e>] driver_probe_device+0xfc/0x2a0 drivers/base/dd.c:724

---
 drivers/usb/misc/usbtest.c | 1 +
 1 file changed, 1 insertion(+)

index e31f72b3a22c5..54b8e8645e0bb 100644
@@ -2771,6 +2771,7 @@ static void usbtest_disconnect(struct usb_interface *intf)
 
 	usb_set_intfdata(intf, NULL);
 	dev_dbg(&intf->dev, "disconnect\n");
+	kfree(dev->buf);
 	kfree(dev);
 }
 

2.25.1

From: Douglas Anderson <[email protected]>

[ Upstream commit 440ab9e10e2e6e5fd677473ee6f9e3af0f6904d6 ]

At times when I'm using kgdb I see a splat on my console about suspicious RCU usage. I managed to come up with a case that could reproduce this that looked like this:

WARNING: suspicious RCU usage
5.7.0-rc4+ #609 Not tainted
-----------------------------
kernel/pid.c:395 find_task_by_pid_ns() needs rcu_read_lock() protection!

other info that might help us debug this:

rcu_scheduler_active = 2, debug_locks = 1
3 locks held by swapper/0/1:
#0: ffffff81b6b8e988 (&dev->mutex){....}-{3:3}, at: __device_attach+0x40/0x13c #1: ffffffd01109e9e8 (dbg_master_lock){....}-{2:2}, at: kgdb_cpu_enter+0x20c/0x7ac #2: ffffffd01109ea90 (dbg_slave_lock){....}-{2:2}, at: kgdb_cpu_enter+0x3ec/0x7ac

  stack backtrace:
  CPU: 7 PID: 1 Comm: swapper/0 Not tainted 5.7.0-rc4+ #609
  Hardware name: Google Cheza (rev3+) (DT)
  Call trace:
   dump_backtrace+0x0/0x1b8
   show_stack+0x1c/0x24
   dump_stack+0xd4/0x134
   lockdep_rcu_suspicious+0xf0/0x100
   find_task_by_pid_ns+0x5c/0x80
   getthread+0x8c/0xb0
   gdb_serial_stub+0x9d4/0xd04
   kgdb_cpu_enter+0x284/0x7ac
   kgdb_handle_exception+0x174/0x20c
   kgdb_brk_fn+0x24/0x30
   call_break_hook+0x6c/0x7c
   brk_handler+0x20/0x5c
   do_debug_exception+0x1c8/0x22c
   el1_sync_handler+0x3c/0xe4
   el1_sync+0x7c/0x100
   rpmh_rsc_probe+0x38/0x420
   platform_drv_probe+0x94/0xb4
   really_probe+0x134/0x300
   driver_probe_device+0x68/0x100
   __device_attach_driver+0x90/0xa8
   bus_for_each_drv+0x84/0xcc
   __device_attach+0xb4/0x13c
   device_initial_probe+0x18/0x20
   bus_probe_device+0x38/0x98
   device_add+0x38c/0x420

If I understand properly we should just be able to blanket kgdb under one big RCU read lock and the problem should go away. We'll add it to the beast-of-a-function known as kgdb_cpu_enter().

With this I no longer get any splats and things seem to work fine.

---
 kernel/debug/debug_core.c | 4 ++++
 1 file changed, 4 insertions(+)

index 9c939c6bf21cb..321ccdbb73649 100644
@@ -488,6 +488,7 @@ static int kgdb_cpu_enter(struct kgdb_state *ks, struct pt_regs *regs,
 		arch_kgdb_ops.disable_hw_break(regs);
 
 acquirelock:
+	rcu_read_lock();
 	/*
 	 * Interrupts will be restored by the 'trap return' code, except when
 	 * single stepping.

@@ -542,6 +543,7 @@ return_normal:
 			atomic_dec(&slaves_in_kgdb);
 			dbg_touch_watchdogs();
 			local_irq_restore(flags);
+			rcu_read_unlock();
 			return 0;
 		}
 		cpu_relax();

@@ -560,6 +562,7 @@ return_normal:
 		raw_spin_unlock(&dbg_master_lock);
 		dbg_touch_watchdogs();
 		local_irq_restore(flags);
+		rcu_read_unlock();
 
 		goto acquirelock;
 	}

@@ -677,6 +680,7 @@ kgdb_restore:
 	raw_spin_unlock(&dbg_master_lock);
 	dbg_touch_watchdogs();
 	local_irq_restore(flags);
+	rcu_read_unlock();
 
 	return kgdb_info[cpu].ret_state;
 }

2.25.1

From: Herbert Xu <[email protected]>

commit 34c86f4c4a7be3b3e35aa48bd18299d4c756064d upstream.

The locking in af_alg_release_parent is broken as the BH socket lock can only be taken if there is a code-path to handle the case where the lock is owned by process-context. Instead of adding such handling, we can fix this by changing the ref counts to atomic_t.

This patch also modifies the main refcnt to include both normal and nokey sockets. This way we don't have to fudge the nokey ref count when a socket changes from nokey to normal.

Credits go to Mauricio Faria de Oliveira who diagnosed this bug and sent a patch for it:

https://lore.kernel.org/linux-crypto/[email protected]/

---
 crypto/af_alg.c         |   26 +++++++++++---------------
 crypto/algif_aead.c     |    9 +++------
 crypto/algif_hash.c     |    9 +++------
 crypto/algif_skcipher.c |    9 +++------
 include/crypto/if_alg.h |    4 ++--
 5 files changed, 22 insertions(+), 35 deletions(-)

@@ -130,21 +130,15 @@ EXPORT_SYMBOL_GPL(af_alg_release);
 void af_alg_release_parent(struct sock *sk)
 {
 	struct alg_sock *ask = alg_sk(sk);
-	unsigned int nokey = ask->nokey_refcnt;
-	bool last = nokey && !ask->refcnt;
+	unsigned int nokey = atomic_read(&ask->nokey_refcnt);
 
 	sk = ask->parent;
 	ask = alg_sk(sk);
 
-	local_bh_disable();
-	bh_lock_sock(sk);
-	ask->nokey_refcnt -= nokey;
-	if (!last)
-		last = !--ask->refcnt;
-	bh_unlock_sock(sk);
-	local_bh_enable();
+	if (nokey)
+		atomic_dec(&ask->nokey_refcnt);
 
-	if (last)
+	if (atomic_dec_and_test(&ask->refcnt))
 		sock_put(sk);
 }
 EXPORT_SYMBOL_GPL(af_alg_release_parent);

@@ -189,7 +183,7 @@ static int alg_bind(struct socket *sock,
 
 	err = -EBUSY;
 	lock_sock(sk);
-	if (ask->refcnt | ask->nokey_refcnt)
+	if (atomic_read(&ask->refcnt))
 		goto unlock;
 
 	swap(ask->type, type);

@@ -238,7 +232,7 @@ static int alg_setsockopt(struct socket
 	int err = -EBUSY;
 
 	lock_sock(sk);
-	if (ask->refcnt)
+	if (atomic_read(&ask->refcnt) != atomic_read(&ask->nokey_refcnt))
 		goto unlock;
 
 	type = ask->type;

@@ -305,12 +299,14 @@ int af_alg_accept(struct sock *sk, struc
 
 	sk2->sk_family = PF_ALG;
 
-	if (nokey || !ask->refcnt++)
+	if (atomic_inc_return_relaxed(&ask->refcnt) == 1)
 		sock_hold(sk);
-	ask->nokey_refcnt += nokey;
+	if (nokey) {
+		atomic_inc(&ask->nokey_refcnt);
+		atomic_set(&alg_sk(sk2)->nokey_refcnt, 1);
+	}
 	alg_sk(sk2)->parent = sk;
 	alg_sk(sk2)->type = type;
-	alg_sk(sk2)->nokey_refcnt = nokey;
 
 	newsock->ops = type->ops;
 	newsock->state = SS_CONNECTED;

@@ -747,7 +747,7 @@ static int aead_check_key(struct socket
 	struct alg_sock *ask = alg_sk(sk);
 
 	lock_sock(sk);
-	if (ask->refcnt)
+	if (!atomic_read(&ask->nokey_refcnt))
 		goto unlock_child;
 
 	psk = ask->parent;

@@ -759,11 +759,8 @@ static int aead_check_key(struct socket
 	if (!tfm->has_key)
 		goto unlock;
 
-	if (!pask->refcnt++)
-		sock_hold(psk);
-
-	ask->refcnt = 1;
-	sock_put(psk);
+	atomic_dec(&pask->nokey_refcnt);
+	atomic_set(&ask->nokey_refcnt, 0);
 
 	err = 0;
 

@@ -308,7 +308,7 @@ static int hash_check_key(struct socket
 	struct alg_sock *ask = alg_sk(sk);
 
 	lock_sock(sk);
-	if (ask->refcnt)
+	if (!atomic_read(&ask->nokey_refcnt))
 		goto unlock_child;
 
 	psk = ask->parent;

@@ -320,11 +320,8 @@ static int hash_check_key(struct socket
 	if (crypto_ahash_get_flags(tfm) & CRYPTO_TFM_NEED_KEY)
 		goto unlock;
 
-	if (!pask->refcnt++)
-		sock_hold(psk);
-
-	ask->refcnt = 1;
-	sock_put(psk);
+	atomic_dec(&pask->nokey_refcnt);
+	atomic_set(&ask->nokey_refcnt, 0);
 
 	err = 0;
 

@@ -774,7 +774,7 @@ static int skcipher_check_key(struct soc
 	struct alg_sock *ask = alg_sk(sk);
 
 	lock_sock(sk);
-	if (ask->refcnt)
+	if (!atomic_read(&ask->nokey_refcnt))
 		goto unlock_child;
 
 	psk = ask->parent;

@@ -786,11 +786,8 @@ static int skcipher_check_key(struct soc
 	if (!tfm->has_key)
 		goto unlock;
 
-	if (!pask->refcnt++)
-		sock_hold(psk);
-
-	ask->refcnt = 1;
-	sock_put(psk);
+	atomic_dec(&pask->nokey_refcnt);
+	atomic_set(&ask->nokey_refcnt, 0);
 
 	err = 0;
 

@@ -30,8 +30,8 @@ struct alg_sock {
 
 	struct sock *parent;
 
-	unsigned int refcnt;
-	unsigned int nokey_refcnt;
+	atomic_t refcnt;
+	atomic_t nokey_refcnt;
 
 	const struct af_alg_type *type;
 	void *private;

From: Shile Zhang <[email protected]>

[ Upstream commit 975e155ed8732cb81f55c021c441ae662dd040b5 ]

We added the 'sched_rr_timeslice_ms' SCHED_RR tuning knob in this commit:

ce0dbbbb30ae ("sched/rt: Add a tuning knob to allow changing SCHED_RR timeslice")

... which name suggests to users that it's in milliseconds, while in reality it's being set in milliseconds but the result is shown in jiffies.

This is obviously confusing when HZ is not 1000, it makes it appear like the value set failed, such as HZ=100:

root# echo 100 > /proc/sys/kernel/sched_rr_timeslice_ms root# cat /proc/sys/kernel/sched_rr_timeslice_ms 10

Fix this to be milliseconds all around.

---
 include/linux/sched/sysctl.h | 1 +
 kernel/sched/core.c          | 5 +++--
 kernel/sched/rt.c            | 1 +
 kernel/sysctl.c              | 2 +-
 4 files changed, 6 insertions(+), 3 deletions(-)

index 05e8b6e4edcb6..debcd4bf22956 100644
@@ -60,6 +60,7 @@ extern unsigned int sysctl_sched_cfs_bandwidth_slice;
 extern unsigned int sysctl_sched_autogroup_enabled;
 #endif
 
+extern int sysctl_sched_rr_timeslice;
 extern int sched_rr_timeslice;
 
 extern int sched_rr_handler(struct ctl_table *table, int write,

index 870d802c46f90..c7c7ba8807f83 100644
@@ -8389,8 +8389,9 @@ int sched_rr_handler(struct ctl_table *table, int write,
 	/* make sure that internally we keep jiffies */
 	/* also, writing zero resets timeslice to default */
 	if (!ret && write) {
-		sched_rr_timeslice = sched_rr_timeslice <= 0 ?
-			RR_TIMESLICE : msecs_to_jiffies(sched_rr_timeslice);
+		sched_rr_timeslice =
+			sysctl_sched_rr_timeslice <= 0 ? RR_TIMESLICE :
+			msecs_to_jiffies(sysctl_sched_rr_timeslice);
 	}
 	mutex_unlock(&mutex);
 	return ret;

index 9ab4d73e9cc95..5034c41a53130 100644
@@ -9,6 +9,7 @@
 #include <linux/irq_work.h>
 
 int sched_rr_timeslice = RR_TIMESLICE;
+int sysctl_sched_rr_timeslice = (MSEC_PER_SEC / HZ) * RR_TIMESLICE;
 
 static int do_sched_rt_period_timer(struct rt_bandwidth *rt_b, int overrun);
 

index 34449ec0689d0..513e6da318c47 100644
@@ -426,7 +426,7 @@ static struct ctl_table kern_table[] = {
 	},
 	{
 		.procname	= "sched_rr_timeslice_ms",
-		.data		= &sched_rr_timeslice,
+		.data		= &sysctl_sched_rr_timeslice,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
 		.proc_handler	= sched_rr_handler,

2.25.1

From: Rahul Lakkireddy <[email protected]>

[ Upstream commit 27f78cb245abdb86735529c13b0a579f57829e71 ]

TC-U32 passes all keys values and masks in __be32 format. The parser already expects this and hence pass the value and masks in __be32 natively to the parser.

Fixes following sparse warnings in several places:
cxgb4_tc_u32.c:57:21: warning: incorrect type in assignment (different base types)
cxgb4_tc_u32.c:57:21: expected unsigned int [usertype] val
cxgb4_tc_u32.c:57:21: got restricted __be32 [usertype] val
cxgb4_tc_u32_parse.h:48:24: warning: cast to restricted __be32

---
 .../net/ethernet/chelsio/cxgb4/cxgb4_tc_u32.c |  18 +--
 .../chelsio/cxgb4/cxgb4_tc_u32_parse.h        | 122 ++++++++++++------
 2 files changed, 91 insertions(+), 49 deletions(-)

index 49d2debb334e8..6afad4d3385a9 100644
@@ -47,7 +47,7 @@ static int fill_match_fields(struct adapter *adap,
 			     bool next_header)
 {
 	unsigned int i, j;
-	u32 val, mask;
+	__be32 val, mask;
 	int off, err;
 	bool found;
 

@@ -217,7 +217,7 @@ int cxgb4_config_knode(struct net_device *dev, __be16 protocol,
 		const struct cxgb4_next_header *next;
 		bool found = false;
 		unsigned int i, j;
-		u32 val, mask;
+		__be32 val, mask;
 		int off;
 
 		if (t->table[link_uhtid - 1].link_handle) {

@@ -231,10 +231,10 @@ int cxgb4_config_knode(struct net_device *dev, __be16 protocol,
 
 		/* Try to find matches that allow jumps to next header. */
 		for (i = 0; next[i].jump; i++) {
-			if (next[i].offoff != cls->knode.sel->offoff ||
-			    next[i].shift != cls->knode.sel->offshift ||
-			    next[i].mask != cls->knode.sel->offmask ||
-			    next[i].offset != cls->knode.sel->off)
+			if (next[i].sel.offoff != cls->knode.sel->offoff ||
+			    next[i].sel.offshift != cls->knode.sel->offshift ||
+			    next[i].sel.offmask != cls->knode.sel->offmask ||
+			    next[i].sel.off != cls->knode.sel->off)
 				continue;
 
 			/* Found a possible candidate.  Find a key that

@@ -246,9 +246,9 @@ int cxgb4_config_knode(struct net_device *dev, __be16 protocol,
 				val = cls->knode.sel->keys[j].val;
 				mask = cls->knode.sel->keys[j].mask;
 
-				if (next[i].match_off == off &&
-				    next[i].match_val == val &&
-				    next[i].match_mask == mask) {
+				if (next[i].key.off == off &&
+				    next[i].key.val == val &&
+				    next[i].key.mask == mask) {
 					found = true;
 					break;
 				}

2.25.1

From: Chu Lin <[email protected]>

[ Upstream commit 016983d138cbe99a5c0aaae0103ee88f5300beb3 ]

Per the datasheet for max6697, OVERT mask and ALERT mask are different. For example, the 7th bit of OVERT is the local channel but for alert mask, the 6th bit is the local channel. Therefore, we can't apply the same mask for both registers. In addition to that, the max6697 driver is supposed to be compatibale with different models. I manually went over all the listed chips and made sure all chip types have the same layout.

Testing;
    mask value of 0x9 should map to 0x44 for ALERT and 0x84 for OVERT.
    I used iotool to read the reg value back to verify. I only tested this
    change on max6581.

Reference:
https://datasheets.maximintegrated.com/en/ds/MAX6581.pdf
https://datasheets.maximintegrated.com/en/ds/MAX6697.pdf
https://datasheets.maximintegrated.com/en/ds/MAX6699.pdf

---
 drivers/hwmon/max6697.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

index f03a71722849a..d4bb3d6aaf18c 100644
@@ -46,8 +46,9 @@ static const u8 MAX6697_REG_CRIT[] = {
  * Map device tree / platform data register bit map to chip bit map.
  * Applies to alert register and over-temperature register.
  */
-#define MAX6697_MAP_BITS(reg)	((((reg) & 0x7e) >> 1) | \
+#define MAX6697_ALERT_MAP_BITS(reg)	((((reg) & 0x7e) >> 1) | \
 				 (((reg) & 0x01) << 6) | ((reg) & 0x80))
+#define MAX6697_OVERT_MAP_BITS(reg) (((reg) >> 1) | (((reg) & 0x01) << 7))
 
 #define MAX6697_REG_STAT(n)		(0x44 + (n))
 

@@ -586,12 +587,12 @@ static int max6697_init_chip(struct max6697_data *data,
 		return ret;
 
 	ret = i2c_smbus_write_byte_data(client, MAX6697_REG_ALERT_MASK,
-					MAX6697_MAP_BITS(pdata->alert_mask));
+				MAX6697_ALERT_MAP_BITS(pdata->alert_mask));
 	if (ret < 0)
 		return ret;
 
 	ret = i2c_smbus_write_byte_data(client, MAX6697_REG_OVERT_MASK,
-				MAX6697_MAP_BITS(pdata->over_temperature_mask));
+			MAX6697_OVERT_MAP_BITS(pdata->over_temperature_mask));
 	if (ret < 0)
 		return ret;
 

2.25.1

From: Misono Tomohiro <[email protected]>

[ Upstream commit 8b97f9922211c44a739c5cbd9502ecbb9f17f6d1 ]

Although it rarely happens, we should call free_capabilities() if error happens after read_capabilities() to free allocated strings.

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

index e27f7e12c05bb..9b4ad6c74041e 100644
@@ -895,7 +895,7 @@ static int acpi_power_meter_add(struct acpi_device *device)
 
 	res = setup_attrs(resource);
 	if (res)
-		goto exit_free;
+		goto exit_free_capability;
 
 	resource->hwmon_dev = hwmon_device_register(&device->dev);
 	if (IS_ERR(resource->hwmon_dev)) {

@@ -908,6 +908,8 @@ static int acpi_power_meter_add(struct acpi_device *device)
 
 exit_remove:
 	remove_attrs(resource);
+exit_free_capability:
+	free_capabilities(resource);
 exit_free:
 	kfree(resource);
 exit:

2.25.1

From: Hou Tao <[email protected]>

[ Upstream commit e7eea44eefbdd5f0345a0a8b80a3ca1c21030d06 ]

Else there will be memory leak if alloc_disk() fails.

---
 drivers/block/virtio_blk.c | 1 +
 1 file changed, 1 insertion(+)

index f287eec36b282..e57a1f6e39d5c 100644
@@ -749,6 +749,7 @@ static int virtblk_probe(struct virtio_device *vdev)
 	put_disk(vblk->disk);
 out_free_vq:
 	vdev->config->del_vqs(vdev);
+	kfree(vblk->vqs);
 out_free_vblk:
 	kfree(vblk);
 out_free_index:

2.25.1

From: Chris Packham <[email protected]>

[ Upstream commit cd217f2300793a106b49c7dfcbfb26e348bc7593 ]

The PCA9665 datasheet says that I2CSTA = 78h indicates that SCL is stuck low, this differs to the PCA9564 which uses 90h for this indication. Treat either 0x78 or 0x90 as an indication that the SCL line is stuck.

Based on looking through the PCA9564 and PCA9665 datasheets this should be safe for both chips. The PCA9564 should not return 0x78 for any valid state and the PCA9665 should not return 0x90.

---
 drivers/i2c/algos/i2c-algo-pca.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

index e370804ec8bc6..3a9db4626cb60 100644
@@ -326,7 +326,8 @@ static int pca_xfer(struct i2c_adapter *i2c_adap,
 			DEB2("BUS ERROR - SDA Stuck low\n");
 			pca_reset(adap);
 			goto out;
-		case 0x90: /* Bus error - SCL stuck low */
+		case 0x78: /* Bus error - SCL stuck low (PCA9665) */
+		case 0x90: /* Bus error - SCL stuck low (PCA9564) */
 			DEB2("BUS ERROR - SCL Stuck low\n");
 			pca_reset(adap);
 			goto out;

2.25.1

From: Greg Kroah-Hartman <[email protected]>

This reverts commit 5ef30e443e6d3654cccecec99cf481a69a0a6d3b which is commit f0bd62b64016508938df9babe47f65c2c727d25c upstream.

It causes a number of reported issues and a fix for it has not hit Linus's tree yet. Revert this to resolve those problems.

---
 sound/usb/card.h     |    4 ----
 sound/usb/endpoint.c |   43 +++++--------------------------------------
 sound/usb/endpoint.h |    1 -
 sound/usb/pcm.c      |    2 --
 4 files changed, 5 insertions(+), 45 deletions(-)

@@ -80,10 +80,6 @@ struct snd_usb_endpoint {
 	dma_addr_t sync_dma;		/* DMA address of syncbuf */
 
 	unsigned int pipe;		/* the data i/o pipe */
-	unsigned int framesize[2];	/* small/large frame sizes in samples */
-	unsigned int sample_rem;	/* remainder from division fs/fps */
-	unsigned int sample_accum;	/* sample accumulator */
-	unsigned int fps;		/* frames per second */
 	unsigned int freqn;		/* nominal sampling rate in fs/fps in Q16.16 format */
 	unsigned int freqm;		/* momentary sampling rate in fs/fps in Q16.16 format */
 	int	   freqshift;		/* how much to shift the feedback value to get Q16.16 */

@@ -27,7 +27,6 @@ void snd_usb_endpoint_release(struct snd
 void snd_usb_endpoint_free(struct snd_usb_endpoint *ep);
 
 int snd_usb_endpoint_implicit_feedback_sink(struct snd_usb_endpoint *ep);
-int snd_usb_endpoint_slave_next_packet_size(struct snd_usb_endpoint *ep);
 int snd_usb_endpoint_next_packet_size(struct snd_usb_endpoint *ep);
 
 void snd_usb_handle_sync_urb(struct snd_usb_endpoint *ep,

@@ -1483,8 +1483,6 @@ static void prepare_playback_urb(struct
 	for (i = 0; i < ctx->packets; i++) {
 		if (ctx->packet_size[i])
 			counts = ctx->packet_size[i];
-		else if (ep->sync_master)
-			counts = snd_usb_endpoint_slave_next_packet_size(ep);
 		else
 			counts = snd_usb_endpoint_next_packet_size(ep);
 

From: Paul Aurich <[email protected]>

commit cc15461c73d7d044d56c47e869a215e49bd429c8 upstream.

Ensure multiuser SMB3 mounts use encryption for all users' tcons if the mount options are configured to require encryption. Without this, only the primary tcon and IPC tcons are guaranteed to be encrypted. Per-user tcons would only be encrypted if the server was configured to require encryption.

---
 fs/cifs/connect.c |    1 +
 1 file changed, 1 insertion(+)

@@ -4216,6 +4216,7 @@ cifs_construct_tcon(struct cifs_sb_info
 	vol_info->no_linux_ext = !master_tcon->unix_ext;
 	vol_info->sectype = master_tcon->ses->sectype;
 	vol_info->sign = master_tcon->ses->sign;
+	vol_info->seal = master_tcon->seal;
 
 	rc = cifs_set_vol_auth(vol_info, master_tcon->ses);
 	if (rc) {

From: Paul Aurich <[email protected]>

commit 00dfbc2f9c61185a2e662f27c45a0bb29b2a134f upstream.

Without this:

- persistent handles will only be enabled for per-user tcons if the server advertises the 'Continuous Availabity' capability
- resilient handles would never be enabled for per-user tcons

---
 fs/cifs/connect.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -4213,6 +4213,8 @@ cifs_construct_tcon(struct cifs_sb_info
 	vol_info->retry = master_tcon->retry;
 	vol_info->nocase = master_tcon->nocase;
 	vol_info->local_lease = master_tcon->local_lease;
+	vol_info->resilient = master_tcon->use_resilient;
+	vol_info->persistent = master_tcon->use_persistent;
 	vol_info->no_linux_ext = !master_tcon->unix_ext;
 	vol_info->sectype = master_tcon->ses->sectype;
 	vol_info->sign = master_tcon->ses->sign;

From: Zhang Xiaoxu <[email protected]>

commit 9ffad9263b467efd8f8dc7ae1941a0a655a2bab2 upstream.

When xfstest generic/035, we found the target file was deleted if the rename return -EACESS.

In cifs_rename2, we unlink the positive target dentry if rename failed with EACESS or EEXIST, even if the target dentry is positived before rename. Then the existing file was deleted.

We should just delete the target file which created during the rename.

---
 fs/cifs/inode.c |   10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

@@ -1770,6 +1770,7 @@ cifs_rename2(struct inode *source_dir, s
 	FILE_UNIX_BASIC_INFO *info_buf_target;
 	unsigned int xid;
 	int rc, tmprc;
+	bool new_target = d_really_is_negative(target_dentry);
 
 	if (flags & ~RENAME_NOREPLACE)
 		return -EINVAL;

@@ -1846,8 +1847,13 @@ cifs_rename2(struct inode *source_dir, s
 	 */
 
 unlink_target:
-	/* Try unlinking the target dentry if it's not negative */
-	if (d_really_is_positive(target_dentry) && (rc == -EACCES || rc == -EEXIST)) {
+	/*
+	 * If the target dentry was created during the rename, try
+	 * unlinking it if it's not negative
+	 */
+	if (new_target &&
+	    d_really_is_positive(target_dentry) &&
+	    (rc == -EACCES || rc == -EEXIST)) {
 		if (d_is_dir(target_dentry))
 			tmprc = cifs_rmdir(target_dir, target_dentry);
 		else

From: Hauke Mehrtens <[email protected]>

commit fcec538ef8cca0ad0b84432235dccd9059c8e6f8 upstream.

This resolves the hazard between the mtc0 in the change_c0_status() and the mfc0 in configure_exception_vector(). Without resolving this hazard configure_exception_vector() could read an old value and would restore this old value again. This would revert the changes change_c0_status() did. I checked this by printing out the read_c0_status() at the end of per_cpu_trap_init() and the ST0_MX is not set without this patch.

The hazard is documented in the MIPS Architecture Reference Manual Vol.
III: MIPS32/microMIPS32 Privileged Resource Architecture (MD00088), rev 6.03 table 8.1 which includes:

   Producer | Consumer | Hazard
  ----------|----------|----------------------------
   mtc0     | mfc0     | any coprocessor 0 register

I saw this hazard on an Atheros AR9344 rev 2 SoC with a MIPS 74Kc CPU. There the change_c0_status() function would activate the DSPen by setting ST0_MX in the c0_status register. This was reverted and then the system got a DSP exception when the DSP registers were saved in save_dsp() in the first process switch. The crash looks like this:

[    0.089999] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.097796] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.107070] Kernel panic - not syncing: Unexpected DSP exception
[    0.113470] Rebooting in 1 seconds..

We saw this problem in OpenWrt only on the MIPS 74Kc based Atheros SoCs, not on the 24Kc based SoCs. We only saw it with kernel 5.4 not with kernel 4.19, in addition we had to use GCC 8.4 or 9.X, with GCC 8.3 it did not happen.

In the kernel I bisected this problem to commit 9012d011660e ("compiler:
allow all arches to enable CONFIG_OPTIMIZE_INLINING"), but when this was reverted it also happened after commit 172dcd935c34b ("MIPS: Always allocate exception vector for MIPSr2+").

Commit 0b24cae4d535 ("MIPS: Add missing EHB in mtc0 -> mfc0 sequence.") does similar changes to a different file. I am not sure if there are more places affected by this problem.

---
 arch/mips/kernel/traps.c |    1 +
 1 file changed, 1 insertion(+)

@@ -2137,6 +2137,7 @@ static void configure_status(void)
 
 	change_c0_status(ST0_CU|ST0_MX|ST0_RE|ST0_FR|ST0_BEV|ST0_TS|ST0_KX|ST0_SX|ST0_UX,
 			 status_set);
+	back_to_back_c0_hazard();
 }
 
 unsigned int hwrena;

From: Vasily Averin <[email protected]>

Could you please push this patch into stable@? it fixes memory corruption in kernels v3.5 .. v4.10

Lost .data_len definition leads to write beyond end of struct nf_ct_h323_master. Usually it corrupts following struct nf_conn_nat, however if nat is not loaded it corrupts following slab object.

In mainline this problem went away in v4.11,
after commit 9f0f3ebeda47 ("netfilter: helpers: remove data_len usage for inkernel helpers") however many stable kernels are still affected.

---
 net/netfilter/nf_conntrack_h323_main.c |    1 +
 1 file changed, 1 insertion(+)

@@ -1225,6 +1225,7 @@ static struct nf_conntrack_helper nf_con
 	{
 		.name			= "Q.931",
 		.me			= THIS_MODULE,
+		.data_len		= sizeof(struct nf_ct_h323_master),
 		.tuple.src.l3num	= AF_INET6,
 		.tuple.src.u.tcp.port	= cpu_to_be16(Q931_PORT),
 		.tuple.dst.protonum	= IPPROTO_TCP,

From: Peter Jones <[email protected]>

commit 435d1a471598752446a72ad1201b3c980526d869 upstream.

In most cases, such as CONFIG_ACPI_CUSTOM_DSDT and CONFIG_ACPI_TABLE_UPGRADE, boot-time modifications to firmware tables are tied to specific Kconfig options. Currently this is not the case for modifying the ACPI SSDT via the efivar_ssdt kernel command line option and associated EFI variable.

This patch adds CONFIG_EFI_CUSTOM_SSDT_OVERLAYS, which defaults disabled, in order to allow enabling or disabling that feature during the build.

---
 drivers/firmware/efi/Kconfig |   11 +++++++++++
 drivers/firmware/efi/efi.c   |    2 +-
 2 files changed, 12 insertions(+), 1 deletion(-)

@@ -129,6 +129,17 @@ config EFI_TEST
 	  Say Y here to enable the runtime services support via /dev/efi_test.
 	  If unsure, say N.
 
+config EFI_CUSTOM_SSDT_OVERLAYS
+	bool "Load custom ACPI SSDT overlay from an EFI variable"
+	depends on EFI_VARS && ACPI
+	default ACPI_TABLE_UPGRADE
+	help
+	  Allow loading of an ACPI SSDT overlay from an EFI variable specified
+	  by a kernel command line option.
+
+	  See Documentation/admin-guide/acpi/ssdt-overlays.rst for more
+	  information.
+
 endmenu
 
 config UEFI_CPER

@@ -198,7 +198,7 @@ static void generic_ops_unregister(void)
 	efivars_unregister(&generic_efivars);
 }
 
-#if IS_ENABLED(CONFIG_ACPI)
+#ifdef CONFIG_EFI_CUSTOM_SSDT_OVERLAYS
 #define EFIVAR_SSDT_NAME_MAX	16
 static char efivar_ssdt[EFIVAR_SSDT_NAME_MAX] __initdata;
 static int __init efivar_ssdt_setup(char *str)

Results from Linaro’s test farm.
No regressions on arm64, arm, x86_64, and i386.

Summary
------------------------------------------------------------------------

kernel: 4.9.230-rc1
git repo: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git
git branch: linux-4.9.y
git commit: e3bed594af79a41cf71d9942b412663a8c6b4fad
git describe: v4.9.228-216-ge3bed594af79
Test details: https://qa-reports.linaro.org/lkft/linux-stable-rc-4.9-oe/build/v4.9.228-216-ge3bed594af79

No regressions (compared to build v4.9.228)

No fixes (compared to build v4.9.228)

Ran 29962 total tests in the following environments and test suites.

Environments
--------------
- dragonboard-410c - arm64
- hi6220-hikey - arm64
- i386
- juno-r2 - arm64
- juno-r2-compat
- juno-r2-kasan
- qemu_arm
- qemu_arm64
- qemu_i386
- qemu_x86_64
- x15 - arm
- x86_64
- x86-kasan

Test Suites
-----------
* build
* install-android-platform-tools-r2600
* install-android-platform-tools-r2800
* kselftest
* kselftest/drivers
* kselftest/filesystems
* libhugetlbfs
* linux-log-parser
* ltp-cap_bounds-tests
* ltp-commands-tests
* ltp-containers-tests
* ltp-controllers-tests
* ltp-cpuhotplug-tests
* ltp-crypto-tests
* ltp-cve-tests
* ltp-dio-tests
* ltp-fcntl-locktests-tests
* ltp-filecaps-tests
* ltp-fs-tests
* ltp-fs_bind-tests
* ltp-fs_perms_simple-tests
* ltp-fsx-tests
* ltp-hugetlb-tests
* ltp-io-tests
* ltp-ipc-tests
* ltp-math-tests
* ltp-mm-tests
* ltp-nptl-tests
* ltp-pty-tests
* ltp-sched-tests
* ltp-securebits-tests
* ltp-syscalls-tests
* perf
* v4l2-compliance
* kvm-unit-tests
* network-basic-tests
* ltp-open-posix-tests
* kselftest/net

Linaro LKFT
https://lkft.linaro.org

All tests are passing for Tegra ...

Test results for stable-v4.9:

    8 builds:	8 pass, 0 fail
    16 boots:	16 pass, 0 fail
    30 tests:	30 pass, 0 fail

Linux version: 4.9.230-rc1-ge3bed594af79
Boards tested: tegra124-jetson-tk1, tegra20-ventana,

                tegra210-p2371-2180, tegra30-cardhu-a04

Cheers
Jon

nvpublic

Compiled and booted on my test system. No dmesg regressions.

thanks,
-- Shuah

Build results:

	total: 171 pass: 171 fail: 0
Qemu test results:
	total: 386 pass: 386 fail: 0

Guenter