4.16.5-stable review

This is the start of the stable review cycle for the 4.16.5 release. There are 26 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 Fri Apr 27 10:33:04 UTC 2018. 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-4.16.5-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-4.16.y
and the diffstat can be found below.

thanks,

greg k-h

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

Greg Kroah-Hartman <[email protected]>

    Linux 4.16.5-rc1

Sean Christopherson <[email protected]>
Revert "KVM: X86: Fix SMRAM accessing even if VM is shutdown"

Leon Romanovsky <[email protected]>
RDMA/mlx5: Fix NULL dereference while accessing XRC_TGT QPs

Jiri Olsa <[email protected]>
perf: Return proper values for user stack errors

Jiri Olsa <[email protected]>
perf: Fix sample_max_stack maximum check

Florian Westphal <[email protected]>
netfilter: x_tables: limit allocation requests for blob rule heads

Florian Westphal <[email protected]>
netfilter: compat: reject huge allocation requests

Florian Westphal <[email protected]>
netfilter: compat: prepare xt_compat_init_offsets to return errors

Florian Westphal <[email protected]>
netfilter: x_tables: add counters allocation wrapper

Florian Westphal <[email protected]>
netfilter: x_tables: cap allocations at 512 mbyte

Tetsuo Handa <[email protected]>
mm,vmscan: Allow preallocating memory for register_shrinker().

Thomas Gleixner <[email protected]>
alarmtimer: Init nanosleep alarm timer on stack

Imre Deak <[email protected]>
drm/i915: Fix LSPCON TMDS output buffer enabling from low-power state

Xidong Wang <[email protected]>
drm/i915: Do no use kfree() to free a kmem_cache_alloc() return value

Gaurav K Singh <[email protected]>
drm/i915/audio: Fix audio detection issue on GLK

Jani Nikula <[email protected]>
drm/i915/bios: filter out invalid DDC pins from VBT child devices

Tina Zhang <[email protected]>
drm/i915/gvt: Add drm_format_mod update

Gerd Hoffmann <[email protected]>
drm/i915/gvt: throw error on unhandled vfio ioctls

Daniel J Blueman <[email protected]>
drm/vc4: Fix memory leak during BO teardown

Xiaoming Gao <[email protected]>
x86/tsc: Prevent 32bit truncation in calc_hpet_ref()

Laura Abbott <[email protected]>
posix-cpu-timers: Ensure set_process_cpu_timer is always evaluated

Anson Huang <[email protected]>
clocksource/imx-tpm: Correct -ETIME return condition check

Dou Liyang <[email protected]>

    x86/acpi: Prevent X2APIC id 0xffffffff from being accounted

Nikolay Borisov <[email protected]>
btrfs: Fix race condition between delayed refs and blockgroup removal

David Sterba <[email protected]>
btrfs: fix unaligned access in readdir

Steve French <[email protected]>
cifs: do not allow creating sockets except with SMB1 posix exensions

Long Li <[email protected]>
cifs: smbd: Check for iov length on sending the last iov

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

Diffstat:

 Makefile                                   |  4 +--
 arch/x86/kernel/acpi/boot.c                |  4 +++
 arch/x86/kernel/tsc.c                      |  2 +-
 arch/x86/kvm/mmu.c                         |  2 +-
 drivers/clocksource/timer-imx-tpm.c        |  2 +-
 drivers/gpu/drm/drm_dp_dual_mode_helper.c  | 39 +++++++++++++++++++----
 drivers/gpu/drm/i915/gvt/dmabuf.c          |  1 +
 drivers/gpu/drm/i915/gvt/kvmgt.c           |  2 +-
 drivers/gpu/drm/i915/i915_gem_execbuffer.c |  2 +-
 drivers/gpu/drm/i915/intel_audio.c         |  2 +-
 drivers/gpu/drm/i915/intel_bios.c          | 13 +++++---
 drivers/gpu/drm/vc4/vc4_bo.c               |  2 ++
 drivers/gpu/drm/vc4/vc4_validate_shaders.c |  1 +
 drivers/infiniband/hw/mlx5/qp.c            |  3 +-
 fs/btrfs/delayed-ref.c                     | 19 ++++++++---
 fs/btrfs/delayed-ref.h                     |  1 +
 fs/btrfs/extent-tree.c                     | 16 +++++++---
 fs/btrfs/inode.c                           | 20 +++++++-----
 fs/cifs/dir.c                              |  9 +++---
 fs/cifs/smbdirect.c                        |  2 ++
 fs/super.c                                 |  9 +++---
 include/linux/netfilter/x_tables.h         |  3 +-
 include/linux/shrinker.h                   |  7 ++--
 kernel/events/callchain.c                  | 21 ++++++------
 kernel/events/core.c                       |  4 +--
 kernel/time/alarmtimer.c                   | 34 +++++++++++++++-----
 kernel/time/posix-cpu-timers.c             |  4 ++-
 mm/vmscan.c                                | 21 +++++++++++-
 net/bridge/netfilter/ebtables.c            | 10 ++++--
 net/ipv4/netfilter/arp_tables.c            | 12 ++++---
 net/ipv4/netfilter/ip_tables.c             | 10 ++++--
 net/ipv6/netfilter/ip6_tables.c            | 12 ++++---
 net/netfilter/x_tables.c                   | 51 ++++++++++++++++++++++++------
 33 files changed, 250 insertions(+), 94 deletions(-)

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

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

From: Long Li <[email protected]>

commit ab60ee7bf9a84954f50a66a3d835860e80f99b7f upstream.

When sending the last iov that breaks into smaller buffers to fit the transfer size, it's necessary to check if this is the last iov.

If this is the latest iov, stop and proceed to send pages.

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

@@ -2194,6 +2194,8 @@ int smbd_send(struct smbd_connection *in
 						goto done;
 				}
 				i++;
+				if (i == rqst->rq_nvec)
+					break;
 			}
 			start = i;
 			buflen = 0;

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

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

From: Steve French <[email protected]>

commit 1d0cffa674cfa7d185a302c8c6850fc50b893bed upstream.

RHBZ: 1453123

Since at least the 3.10 kernel and likely a lot earlier we have not been able to create unix domain sockets in a cifs share when mounted using the SFU mount option (except when mounted with the cifs unix extensions to Samba e.g.)
Trying to create a socket, for example using the af_unix command from xfstests will cause :

BUG: unable to handle kernel NULL pointer dereference at 00000000

00000040

Since no one uses or depends on being able to create unix domains sockets on a cifs share the easiest fix to stop this vulnerability is to simply not allow creation of any other special files than char or block devices when sfu is used.

Added update to Ronnie's patch to handle a tcon link leak, and to address a buf leak noticed by Gustavo and Colin.

---
 fs/cifs/dir.c |    9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

@@ -684,6 +684,9 @@ int cifs_mknod(struct inode *inode, stru
 		goto mknod_out;
 	}
 
+	if (!S_ISCHR(mode) && !S_ISBLK(mode))
+		goto mknod_out;
+
 	if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL))
 		goto mknod_out;
 

@@ -692,10 +695,8 @@ int cifs_mknod(struct inode *inode, stru
 
 	buf = kmalloc(sizeof(FILE_ALL_INFO), GFP_KERNEL);
 	if (buf == NULL) {
-		kfree(full_path);
 		rc = -ENOMEM;
-		free_xid(xid);
-		return rc;
+		goto mknod_out;
 	}
 
 	if (backup_cred(cifs_sb))

@@ -742,7 +743,7 @@ int cifs_mknod(struct inode *inode, stru
 		pdev->minor = cpu_to_le64(MINOR(device_number));
 		rc = tcon->ses->server->ops->sync_write(xid, &fid, &io_parms,
 							&bytes_written, iov, 1);
-	} /* else if (S_ISFIFO) */
+	}
 	tcon->ses->server->ops->close(xid, tcon, &fid);
 	d_drop(direntry);
 

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

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

From: David Sterba <[email protected]>

commit 92d32170847bfff2dd08af2c016085779f2fd2a1 upstream.

The last update to readdir introduced a temporary buffer to store the emitted readdir data, but as there are file names of variable length, there's a lot of unaligned access.

This was observed on a sparc64 machine:

Kernel unaligned access at TPC[102f3080] btrfs_real_readdir+0x51c/0x718 [btrfs]

---
 fs/btrfs/inode.c |   20 ++++++++++++--------
 1 file changed, 12 insertions(+), 8 deletions(-)

@@ -44,6 +44,7 @@
 #include <linux/uio.h>
 #include <linux/magic.h>
 #include <linux/iversion.h>
+#include <asm/unaligned.h>
 #include "ctree.h"
 #include "disk-io.h"
 #include "transaction.h"

@@ -5951,11 +5952,13 @@ static int btrfs_filldir(void *addr, int
 		struct dir_entry *entry = addr;
 		char *name = (char *)(entry + 1);
 
-		ctx->pos = entry->offset;
-		if (!dir_emit(ctx, name, entry->name_len, entry->ino,
-			      entry->type))
+		ctx->pos = get_unaligned(&entry->offset);
+		if (!dir_emit(ctx, name, get_unaligned(&entry->name_len),
+					 get_unaligned(&entry->ino),
+					 get_unaligned(&entry->type)))
 			return 1;
-		addr += sizeof(struct dir_entry) + entry->name_len;
+		addr += sizeof(struct dir_entry) +
+			get_unaligned(&entry->name_len);
 		ctx->pos++;
 	}
 	return 0;

@@ -6045,14 +6048,15 @@ again:
 		}
 
 		entry = addr;
-		entry->name_len = name_len;
+		put_unaligned(name_len, &entry->name_len);
 		name_ptr = (char *)(entry + 1);
 		read_extent_buffer(leaf, name_ptr, (unsigned long)(di + 1),
 				   name_len);
-		entry->type = btrfs_filetype_table[btrfs_dir_type(leaf, di)];
+		put_unaligned(btrfs_filetype_table[btrfs_dir_type(leaf, di)],
+				&entry->type);
 		btrfs_dir_item_key_to_cpu(leaf, di, &location);
-		entry->ino = location.objectid;
-		entry->offset = found_key.offset;
+		put_unaligned(location.objectid, &entry->ino);
+		put_unaligned(found_key.offset, &entry->offset);
 		entries++;
 		addr += sizeof(struct dir_entry) + name_len;
 		total_len += sizeof(struct dir_entry) + name_len;

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

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

From: Nikolay Borisov <[email protected]>

commit 5e388e95815408c27f3612190d089afc0774b870 upstream.

When the delayed refs for a head are all run, eventually cleanup_ref_head is called which (in case of deletion) obtains a reference for the relevant btrfs_space_info struct by querying the bg for the range. This is problematic because when the last extent of a bg is deleted a race window emerges between removal of that bg and the subsequent invocation of cleanup_ref_head. This can result in cache being null and either a null pointer dereference or assertion failure.

	task: ffff8d04d31ed080 task.stack: ffff9e5dc10cc000
	RIP: 0010:assfail.constprop.78+0x18/0x1a [btrfs]
	RSP: 0018:ffff9e5dc10cfbe8 EFLAGS: 00010292
	RAX: 0000000000000044 RBX: 0000000000000000 RCX: 0000000000000000
	RDX: ffff8d04ffc1f868 RSI: ffff8d04ffc178c8 RDI: ffff8d04ffc178c8
	RBP: ffff8d04d29e5ea0 R08: 00000000000001f0 R09: 0000000000000001
	R10: ffff9e5dc0507d58 R11: 0000000000000001 R12: ffff8d04d29e5ea0
	R13: ffff8d04d29e5f08 R14: ffff8d04efe29b40 R15: ffff8d04efe203e0
	FS:  00007fbf58ead500(0000) GS:ffff8d04ffc00000(0000) knlGS:0000000000000000
	CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
	CR2: 00007fe6c6975648 CR3: 0000000013b2a000 CR4: 00000000000006f0
	DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
	DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
	Call Trace:
	 __btrfs_run_delayed_refs+0x10e7/0x12c0 [btrfs]
	 btrfs_run_delayed_refs+0x68/0x250 [btrfs]
	 btrfs_should_end_transaction+0x42/0x60 [btrfs]
	 btrfs_truncate_inode_items+0xaac/0xfc0 [btrfs]
	 btrfs_evict_inode+0x4c6/0x5c0 [btrfs]
	 evict+0xc6/0x190
	 do_unlinkat+0x19c/0x300
	 do_syscall_64+0x74/0x140
	 entry_SYSCALL_64_after_hwframe+0x3d/0xa2
	RIP: 0033:0x7fbf589c57a7

To fix this, introduce a new flag "is_system" to head_ref structs, which is populated at insertion time. This allows to decouple the querying for the spaceinfo from querying the possibly deleted bg.

---
 fs/btrfs/delayed-ref.c |   19 ++++++++++++++-----
 fs/btrfs/delayed-ref.h |    1 +
 fs/btrfs/extent-tree.c |   16 +++++++++++-----
 3 files changed, 26 insertions(+), 10 deletions(-)

@@ -553,8 +553,10 @@ add_delayed_ref_head(struct btrfs_fs_inf
 		     struct btrfs_delayed_ref_head *head_ref,
 		     struct btrfs_qgroup_extent_record *qrecord,
 		     u64 bytenr, u64 num_bytes, u64 ref_root, u64 reserved,
-		     int action, int is_data, int *qrecord_inserted_ret,
+		     int action, int is_data, int is_system,
+		     int *qrecord_inserted_ret,
 		     int *old_ref_mod, int *new_ref_mod)
+
 {
 	struct btrfs_delayed_ref_head *existing;
 	struct btrfs_delayed_ref_root *delayed_refs;

@@ -598,6 +600,7 @@ add_delayed_ref_head(struct btrfs_fs_inf
 	head_ref->ref_mod = count_mod;
 	head_ref->must_insert_reserved = must_insert_reserved;
 	head_ref->is_data = is_data;
+	head_ref->is_system = is_system;
 	head_ref->ref_tree = RB_ROOT;
 	INIT_LIST_HEAD(&head_ref->ref_add_list);
 	RB_CLEAR_NODE(&head_ref->href_node);

@@ -785,6 +788,7 @@ int btrfs_add_delayed_tree_ref(struct bt
 	struct btrfs_delayed_ref_root *delayed_refs;
 	struct btrfs_qgroup_extent_record *record = NULL;
 	int qrecord_inserted;
+	int is_system = (ref_root == BTRFS_CHUNK_TREE_OBJECTID);
 
 	BUG_ON(extent_op && extent_op->is_data);
 	ref = kmem_cache_alloc(btrfs_delayed_tree_ref_cachep, GFP_NOFS);

@@ -813,8 +817,8 @@ int btrfs_add_delayed_tree_ref(struct bt
 	 */
 	head_ref = add_delayed_ref_head(fs_info, trans, head_ref, record,
 					bytenr, num_bytes, 0, 0, action, 0,
-					&qrecord_inserted, old_ref_mod,
-					new_ref_mod);
+					is_system, &qrecord_inserted,
+					old_ref_mod, new_ref_mod);
 
 	add_delayed_tree_ref(fs_info, trans, head_ref, &ref->node, bytenr,
 			     num_bytes, parent, ref_root, level, action);

@@ -881,7 +885,7 @@ int btrfs_add_delayed_data_ref(struct bt
 	 */
 	head_ref = add_delayed_ref_head(fs_info, trans, head_ref, record,
 					bytenr, num_bytes, ref_root, reserved,
-					action, 1, &qrecord_inserted,
+					action, 1, 0, &qrecord_inserted,
 					old_ref_mod, new_ref_mod);
 
 	add_delayed_data_ref(fs_info, trans, head_ref, &ref->node, bytenr,

@@ -911,9 +915,14 @@ int btrfs_add_delayed_extent_op(struct b
 	delayed_refs = &trans->transaction->delayed_refs;
 	spin_lock(&delayed_refs->lock);
 
+	/*
+	 * extent_ops just modify the flags of an extent and they don't result
+	 * in ref count changes, hence it's safe to pass false/0 for is_system
+	 * argument
+	 */
 	add_delayed_ref_head(fs_info, trans, head_ref, NULL, bytenr,
 			     num_bytes, 0, 0, BTRFS_UPDATE_DELAYED_HEAD,
-			     extent_op->is_data, NULL, NULL, NULL);
+			     extent_op->is_data, 0, NULL, NULL, NULL);
 
 	spin_unlock(&delayed_refs->lock);
 	return 0;

@@ -139,6 +139,7 @@ struct btrfs_delayed_ref_head {
 	 */
 	unsigned int must_insert_reserved:1;
 	unsigned int is_data:1;
+	unsigned int is_system:1;
 	unsigned int processing:1;
 };
 

@@ -2615,13 +2615,19 @@ static int cleanup_ref_head(struct btrfs
 	trace_run_delayed_ref_head(fs_info, head, 0);
 
 	if (head->total_ref_mod < 0) {
-		struct btrfs_block_group_cache *cache;
+		struct btrfs_space_info *space_info;
+		u64 flags;
 
-		cache = btrfs_lookup_block_group(fs_info, head->bytenr);
-		ASSERT(cache);
-		percpu_counter_add(&cache->space_info->total_bytes_pinned,
+		if (head->is_data)
+			flags = BTRFS_BLOCK_GROUP_DATA;
+		else if (head->is_system)
+			flags = BTRFS_BLOCK_GROUP_SYSTEM;
+		else
+			flags = BTRFS_BLOCK_GROUP_METADATA;
+		space_info = __find_space_info(fs_info, flags);
+		ASSERT(space_info);
+		percpu_counter_add(&space_info->total_bytes_pinned,
 				   -head->num_bytes);
-		btrfs_put_block_group(cache);
 
 		if (head->is_data) {
 			spin_lock(&delayed_refs->lock);

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

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

From: Dou Liyang <[email protected]>

commit 10daf10ab154e31237a8c07242be3063fb6a9bf4 upstream.

RongQing reported that there are some X2APIC id 0xffffffff in his machine's ACPI MADT table, which makes the number of possible CPU inaccurate.

The reason is that the ACPI X2APIC parser has no sanity check for APIC ID 0xffffffff, which is an invalid id in all APIC types. See "Intel® 64 Architecture x2APIC Specification", Chapter 2.4.1.

Add a sanity check to acpi_parse_x2apic() which ignores the invalid id.

---
 arch/x86/kernel/acpi/boot.c |    4 ++++
 1 file changed, 4 insertions(+)

@@ -215,6 +215,10 @@ acpi_parse_x2apic(struct acpi_subtable_h
 	apic_id = processor->local_apic_id;
 	enabled = processor->lapic_flags & ACPI_MADT_ENABLED;
 
+	/* Ignore invalid ID */
+	if (apic_id == 0xffffffff)
+		return 0;
+
 	/*
 	 * We need to register disabled CPU as well to permit
 	 * counting disabled CPUs. This allows us to size

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

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

From: Anson Huang <[email protected]>

commit 7407188489c62a7b5694bc75a6db2b82af94c9a5 upstream.

The additional brakects added to tpm_set_next_event's return value computation causes (int) forced type conversion NOT taking effect, and the incorrect value return will cause various system timer issue, like RCU stall etc..

Remove the additional brackets to make sure tpm_set_next_event always returns correct value.

---
 drivers/clocksource/timer-imx-tpm.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -105,7 +105,7 @@ static int tpm_set_next_event(unsigned l
 	 * of writing CNT registers which may cause the min_delta event got
 	 * missed, so we need add a ETIME check here in case it happened.
 	 */
-	return (int)((next - now) <= 0) ? -ETIME : 0;
+	return (int)(next - now) <= 0 ? -ETIME : 0;
 }
 
 static int tpm_set_state_oneshot(struct clock_event_device *evt)

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

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

From: Laura Abbott <[email protected]>

commit c3bca5d450b620dd3d36e14b5e1f43639fd47d6b upstream.

Commit a9445e47d897 ("posix-cpu-timers: Make set_process_cpu_timer() more robust") moved the check into the 'if' statement. Unfortunately, it did so on the right side of an && which means that it may get short circuited and never evaluated. This is easily reproduced with:

$ cat loop.c
void main() {
  struct rlimit res;
  /* set the CPU time limit */
  getrlimit(RLIMIT_CPU,&res);
  res.rlim_cur = 2;
  res.rlim_max = 2;
  setrlimit(RLIMIT_CPU,&res);

while (1);
}

Which will hang forever instead of being killed. Fix this by pulling the evaluation out of the if statement but checking the return value instead.

Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1568337

---
 kernel/time/posix-cpu-timers.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

@@ -1205,10 +1205,12 @@ void set_process_cpu_timer(struct task_s
 			   u64 *newval, u64 *oldval)
 {
 	u64 now;
+	int ret;
 
 	WARN_ON_ONCE(clock_idx == CPUCLOCK_SCHED);
+	ret = cpu_timer_sample_group(clock_idx, tsk, &now);
 
-	if (oldval && cpu_timer_sample_group(clock_idx, tsk, &now) != -EINVAL) {
+	if (oldval && ret != -EINVAL) {
 		/*
 		 * We are setting itimer. The *oldval is absolute and we update
 		 * it to be relative, *newval argument is relative and we update

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

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

From: Xiaoming Gao <[email protected]>

commit d3878e164dcd3925a237a20e879432400e369172 upstream.

The TSC calibration code uses HPET as reference. The conversion normalizes the delta of two HPET timestamps:

    hpetref = ((tshpet1 - tshpet2) * HPET_PERIOD) / 1e6

and then divides the normalized delta of the corresponding TSC timestamps by the result to calulate the TSC frequency.

    tscfreq = ((tstsc1 - tstsc2 ) * 1e6) / hpetref

This uses do_div() which takes an u32 as the divisor, which worked so far because the HPET frequency was low enough that 'hpetref' never exceeded 32bit.

On Skylake machines the HPET frequency increased so 'hpetref' can exceed 32bit. do_div() truncates the divisor, which causes the calibration to fail.

Use div64_u64() to avoid the problem.

[ tglx: Fixes whitespace mangled patch and rewrote changelog ]

---
 arch/x86/kernel/tsc.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -317,7 +317,7 @@ static unsigned long calc_hpet_ref(u64 d
 	hpet2 -= hpet1;
 	tmp = ((u64)hpet2 * hpet_readl(HPET_PERIOD));
 	do_div(tmp, 1000000);
-	do_div(deltatsc, tmp);
+	deltatsc = div64_u64(deltatsc, tmp);
 
 	return (unsigned long) deltatsc;
 }

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

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

From: Daniel J Blueman <[email protected]>

commit c0db1b677e1d584fab5d7ac76a32e1c0157542e0 upstream.

During BO teardown, an indirect list 'uniform_addr_offsets' wasn't being freed leading to leaking many 128B allocations. Fix the memory leak by releasing it at teardown time.

---
 drivers/gpu/drm/vc4/vc4_bo.c               |    2 ++
 drivers/gpu/drm/vc4/vc4_validate_shaders.c |    1 +
 2 files changed, 3 insertions(+)

@@ -195,6 +195,7 @@ static void vc4_bo_destroy(struct vc4_bo
 	vc4_bo_set_label(obj, -1);
 
 	if (bo->validated_shader) {
+		kfree(bo->validated_shader->uniform_addr_offsets);
 		kfree(bo->validated_shader->texture_samples);
 		kfree(bo->validated_shader);
 		bo->validated_shader = NULL;

@@ -591,6 +592,7 @@ void vc4_free_object(struct drm_gem_obje
 	}
 
 	if (bo->validated_shader) {
+		kfree(bo->validated_shader->uniform_addr_offsets);
 		kfree(bo->validated_shader->texture_samples);
 		kfree(bo->validated_shader);
 		bo->validated_shader = NULL;

@@ -942,6 +942,7 @@ vc4_validate_shader(struct drm_gem_cma_o
 fail:
 	kfree(validation_state.branch_targets);
 	if (validated_shader) {
+		kfree(validated_shader->uniform_addr_offsets);
 		kfree(validated_shader->texture_samples);
 		kfree(validated_shader);
 	}

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

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

From: Gerd Hoffmann <[email protected]>

commit 9f591ae60e1be026901398ef99eede91237aa3a1 upstream.

On unknown/unhandled ioctls the driver should return an error, so userspace knows it tried to use something unsupported.

---
 drivers/gpu/drm/i915/gvt/kvmgt.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1284,7 +1284,7 @@ static long intel_vgpu_ioctl(struct mdev
 
 	}
 
-	return 0;
+	return -ENOTTY;
 }
 
 static ssize_t

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

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

From: Tina Zhang <[email protected]>

commit 10996f802109c83421ca30556cfe36ffc3bebae3 upstream.

Add drm_format_mod update, which is omitted.

---
 drivers/gpu/drm/i915/gvt/dmabuf.c |    1 +
 1 file changed, 1 insertion(+)

@@ -323,6 +323,7 @@ static void update_fb_info(struct vfio_d
 		      struct intel_vgpu_fb_info *fb_info)
 {
 	gvt_dmabuf->drm_format = fb_info->drm_format;
+	gvt_dmabuf->drm_format_mod = fb_info->drm_format_mod;
 	gvt_dmabuf->width = fb_info->width;
 	gvt_dmabuf->height = fb_info->height;
 	gvt_dmabuf->stride = fb_info->stride;

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

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

From: Jani Nikula <[email protected]>

commit a3520b8992e57bc94ab6ec9f95f09c6c932555fd upstream.

The VBT contains the DDC pin to use for specific ports. Alas, sometimes the field appears to contain bogus data, and while we check for it later on in intel_gmbus_get_adapter() we fail to check the returned NULL on errors. Oops results.

The simplest approach seems to be to catch and ignore the bogus DDC pins already at the VBT parsing phase, reverting to fixed per port default pins. This doesn't guarantee display working, but at least it prevents the oops. And we continue to be fuzzed by VBT.

One affected machine is Dell Latitude 5590 where a BIOS upgrade added invalid DDC pins.

Typical backtrace:

[   35.461411] WARN_ON(!intel_gmbus_is_valid_pin(dev_priv, pin))
[   35.461432] WARNING: CPU: 6 PID: 411 at drivers/gpu/drm/i915/intel_i2c.c:844 intel_gmbus_get_adapter+0x32/0x37 [i915]
[   35.461437] Modules linked in: i915 ahci libahci dm_snapshot dm_bufio dm_raid raid456 async_raid6_recov async_pq raid6_pq async_xor xor async_memcpy async_tx
[   35.461445] CPU: 6 PID: 411 Comm: kworker/u16:2 Not tainted 4.16.0-rc7.x64-g1cda370ffded #1
[   35.461447] Hardware name: Dell Inc. Latitude 5590/0MM81M, BIOS 1.1.9 03/13/2018
[   35.461450] Workqueue: events_unbound async_run_entry_fn
[   35.461465] RIP: 0010:intel_gmbus_get_adapter+0x32/0x37 [i915]
[   35.461467] RSP: 0018:ffff9b4e43d47c40 EFLAGS: 00010286
[   35.461469] RAX: 0000000000000000 RBX: ffff98f90639f800 RCX: ffffffffae051960
[   35.461471] RDX: 0000000000000001 RSI: 0000000000000092 RDI: 0000000000000246
[   35.461472] RBP: ffff98f905410000 R08: 0000004d062a83f6 R09: 00000000000003bd
[   35.461474] R10: 0000000000000031 R11: ffffffffad4eda58 R12: ffff98f905410000
[   35.461475] R13: ffff98f9064c1000 R14: ffff9b4e43d47cf0 R15: ffff98f905410000
[   35.461477] FS:  0000000000000000(0000) GS:ffff98f92e580000(0000) knlGS:0000000000000000
[   35.461479] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   35.461481] CR2: 00007f5682359008 CR3: 00000001b700c005 CR4: 00000000003606e0
[   35.461483] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[   35.461484] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[   35.461486] Call Trace:
[   35.461501]  intel_hdmi_set_edid+0x37/0x27f [i915]
[   35.461515]  intel_hdmi_detect+0x7c/0x97 [i915]
[   35.461518]  drm_helper_probe_single_connector_modes+0xe1/0x6c0
[   35.461521]  drm_setup_crtcs+0x129/0xa6a
[   35.461523]  ? __switch_to_asm+0x34/0x70
[   35.461525]  ? __switch_to_asm+0x34/0x70
[   35.461527]  ? __switch_to_asm+0x40/0x70
[   35.461528]  ? __switch_to_asm+0x34/0x70
[   35.461529]  ? __switch_to_asm+0x40/0x70
[   35.461531]  ? __switch_to_asm+0x34/0x70
[   35.461532]  ? __switch_to_asm+0x40/0x70
[   35.461534]  ? __switch_to_asm+0x34/0x70
[   35.461536]  __drm_fb_helper_initial_config_and_unlock+0x34/0x46f
[   35.461538]  ? __switch_to_asm+0x40/0x70
[   35.461541]  ? _cond_resched+0x10/0x33
[   35.461557]  intel_fbdev_initial_config+0xf/0x1c [i915]
[   35.461560]  async_run_entry_fn+0x2e/0xf5
[   35.461563]  process_one_work+0x15b/0x364
[   35.461565]  worker_thread+0x2c/0x3a0
[   35.461567]  ? process_one_work+0x364/0x364
[   35.461568]  kthread+0x10c/0x122
[   35.461570]  ? _kthread_create_on_node+0x5d/0x5d
[   35.461572]  ret_from_fork+0x35/0x40
[   35.461574] Code: 74 16 89 f6 48 8d 04 b6 48 c1 e0 05 48 29 f0 48 8d 84 c7 e8 11 00 00 c3 48 c7 c6 b0 19 1e c0 48 c7 c7 64 8a 1c c0 e8 47 88 ed ec <0f> 0b 31 c0 c3 8b 87 a4 04 00 00 80 e4 fc 09 c6 89 b7 a4 04 00
[   35.461604] WARNING: CPU: 6 PID: 411 at drivers/gpu/drm/i915/intel_i2c.c:844 intel_gmbus_get_adapter+0x32/0x37 [i915]
[   35.461606] ---[ end trace 4fe1e63e2dd93373 ]---
[   35.461609] BUG: unable to handle kernel NULL pointer dereference at 0000000000000010
[   35.461613] IP: i2c_transfer+0x4/0x86
[   35.461614] PGD 0 P4D 0
[   35.461616] Oops: 0000 [#1] SMP PTI
[   35.461618] Modules linked in: i915 ahci libahci dm_snapshot dm_bufio dm_raid raid456 async_raid6_recov async_pq raid6_pq async_xor xor async_memcpy async_tx
[   35.461624] CPU: 6 PID: 411 Comm: kworker/u16:2 Tainted: G        W        4.16.0-rc7.x64-g1cda370ffded #1
[   35.461625] Hardware name: Dell Inc. Latitude 5590/0MM81M, BIOS 1.1.9 03/13/2018
[   35.461628] Workqueue: events_unbound async_run_entry_fn
[   35.461630] RIP: 0010:i2c_transfer+0x4/0x86
[   35.461631] RSP: 0018:ffff9b4e43d47b30 EFLAGS: 00010246
[   35.461633] RAX: ffff9b4e43d47b6e RBX: 0000000000000005 RCX: 0000000000000001
[   35.461635] RDX: 0000000000000002 RSI: ffff9b4e43d47b80 RDI: 0000000000000000
[   35.461636] RBP: ffff9b4e43d47bd8 R08: 0000004d062a83f6 R09: 00000000000003bd
[   35.461638] R10: 0000000000000031 R11: ffffffffad4eda58 R12: 0000000000000002
[   35.461639] R13: 0000000000000001 R14: ffff9b4e43d47b6f R15: ffff9b4e43d47c07
[   35.461641] FS:  0000000000000000(0000) GS:ffff98f92e580000(0000) knlGS:0000000000000000
[   35.461643] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   35.461645] CR2: 0000000000000010 CR3: 00000001b700c005 CR4: 00000000003606e0
[   35.461646] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[   35.461647] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[   35.461649] Call Trace:
[   35.461652]  drm_do_probe_ddc_edid+0xb3/0x128
[   35.461654]  drm_get_edid+0xe5/0x38d
[   35.461669]  intel_hdmi_set_edid+0x45/0x27f [i915]
[   35.461684]  intel_hdmi_detect+0x7c/0x97 [i915]
[   35.461687]  drm_helper_probe_single_connector_modes+0xe1/0x6c0
[   35.461689]  drm_setup_crtcs+0x129/0xa6a
[   35.461691]  ? __switch_to_asm+0x34/0x70
[   35.461693]  ? __switch_to_asm+0x34/0x70
[   35.461694]  ? __switch_to_asm+0x40/0x70
[   35.461696]  ? __switch_to_asm+0x34/0x70
[   35.461697]  ? __switch_to_asm+0x40/0x70
[   35.461698]  ? __switch_to_asm+0x34/0x70
[   35.461700]  ? __switch_to_asm+0x40/0x70
[   35.461701]  ? __switch_to_asm+0x34/0x70
[   35.461703]  __drm_fb_helper_initial_config_and_unlock+0x34/0x46f
[   35.461705]  ? __switch_to_asm+0x40/0x70
[   35.461707]  ? _cond_resched+0x10/0x33
[   35.461724]  intel_fbdev_initial_config+0xf/0x1c [i915]
[   35.461727]  async_run_entry_fn+0x2e/0xf5
[   35.461729]  process_one_work+0x15b/0x364
[   35.461731]  worker_thread+0x2c/0x3a0
[   35.461733]  ? process_one_work+0x364/0x364
[   35.461734]  kthread+0x10c/0x122
[   35.461736]  ? _kthread_create_on_node+0x5d/0x5d
[   35.461738]  ret_from_fork+0x35/0x40
[   35.461739] Code: 5c fa e1 ad 48 89 df e8 ea fb ff ff e9 2a ff ff ff 0f 1f 44 00 00 31 c0 e9 43 fd ff ff 31 c0 45 31 e4 e9 c5 fd ff ff 41 54 55 53 <48> 8b 47 10 48 83 78 10 00 74 70 41 89 d4 48 89 f5 48 89 fb 65
[   35.461756] RIP: i2c_transfer+0x4/0x86 RSP: ffff9b4e43d47b30
[   35.461757] CR2: 0000000000000010
[   35.461759] ---[ end trace 4fe1e63e2dd93374 ]---

Based on a patch by Fei Li.

v2: s/reverting/sticking/ (Chris)

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105549
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105961

---
 drivers/gpu/drm/i915/intel_bios.c |   13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)

@@ -1255,7 +1255,6 @@ static void parse_ddi_port(struct drm_i9
 		return;
 
 	aux_channel = child->aux_channel;
-	ddc_pin = child->ddc_pin;
 
 	is_dvi = child->device_type & DEVICE_TYPE_TMDS_DVI_SIGNALING;
 	is_dp = child->device_type & DEVICE_TYPE_DISPLAYPORT_OUTPUT;

@@ -1302,9 +1301,15 @@ static void parse_ddi_port(struct drm_i9
 		DRM_DEBUG_KMS("Port %c is internal DP\n", port_name(port));
 
 	if (is_dvi) {
-		info->alternate_ddc_pin = map_ddc_pin(dev_priv, ddc_pin);
-
-		sanitize_ddc_pin(dev_priv, port);
+		ddc_pin = map_ddc_pin(dev_priv, child->ddc_pin);
+		if (intel_gmbus_is_valid_pin(dev_priv, ddc_pin)) {
+			info->alternate_ddc_pin = ddc_pin;
+			sanitize_ddc_pin(dev_priv, port);
+		} else {
+			DRM_DEBUG_KMS("Port %c has invalid DDC pin %d, "
+				      "sticking to defaults\n",
+				      port_name(port), ddc_pin);
+		}
 	}
 
 	if (is_dp) {

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

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

From: Gaurav K Singh <[email protected]>

commit b4615730530be85fc45ab4631c2ad6d8e2d0b97d upstream.

On Geminilake, sometimes audio card is not getting detected after reboot. This is a spurious issue happening on Geminilake. HW codec and HD audio controller link was going out of sync for which there was a fix in i915 driver but was not getting invoked for GLK. Extending this fix to GLK as well.

Tested by Du,Wenkai on GLK board.

Bspec: 21829

v2: Instead of checking GEN9_BC, BXT and GLK macros, use IS_GEN9 macro (Jani N)

---
 drivers/gpu/drm/i915/intel_audio.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -729,7 +729,7 @@ static void i915_audio_component_codec_w
 	struct drm_i915_private *dev_priv = kdev_to_i915(kdev);
 	u32 tmp;
 
-	if (!IS_GEN9_BC(dev_priv))
+	if (!IS_GEN9(dev_priv))
 		return;
 
 	i915_audio_component_get_power(kdev);

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

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

From: Xidong Wang <[email protected]>

commit fcf1fadf4c65eea6c519c773d2d9901e8ad94f5f upstream.

Along the eb_lookup_vmas() error path, the return value from kmem_cache_alloc() was freed using kfree(). Fix it to use the proper kmem_cache_free() instead.

---
 drivers/gpu/drm/i915/i915_gem_execbuffer.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -728,7 +728,7 @@ static int eb_lookup_vmas(struct i915_ex
 
 		err = radix_tree_insert(handles_vma, handle, vma);
 		if (unlikely(err)) {
-			kfree(lut);
+			kmem_cache_free(eb->i915->luts, lut);
 			goto err_obj;
 		}
 

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

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

From: Imre Deak <[email protected]>

commit 7eb2c4dd54ff841f2fe509a84973eb25fa20bda2 upstream.

LSPCON adapters in low-power state may ignore the first I2C write during TMDS output buffer enabling, resulting in a blank screen even with an otherwise enabled pipe. Fix this by reading back and validating the written value a few times.

The problem was noticed on GLK machines with an onboard LSPCON adapter after entering/exiting DC5 power state. Doing an I2C read of the adapter ID as the first transaction - instead of the I2C write to enable the TMDS buffers - returns the correct value. Based on this we assume that the transaction itself is sent properly, it's only the adapter that is not ready for some reason to accept this first write after waking from low-power state. In my case the second I2C write attempt always succeeded.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105854

---
 drivers/gpu/drm/drm_dp_dual_mode_helper.c |   39 ++++++++++++++++++++++++------
 1 file changed, 32 insertions(+), 7 deletions(-)

@@ -350,19 +350,44 @@ int drm_dp_dual_mode_set_tmds_output(enu
 {
 	uint8_t tmds_oen = enable ? 0 : DP_DUAL_MODE_TMDS_DISABLE;
 	ssize_t ret;
+	int retry;
 
 	if (type < DRM_DP_DUAL_MODE_TYPE2_DVI)
 		return 0;
 
-	ret = drm_dp_dual_mode_write(adapter, DP_DUAL_MODE_TMDS_OEN,
-				     &tmds_oen, sizeof(tmds_oen));
-	if (ret) {
-		DRM_DEBUG_KMS("Failed to %s TMDS output buffers\n",
-			      enable ? "enable" : "disable");
-		return ret;
+	/*
+	 * LSPCON adapters in low-power state may ignore the first write, so
+	 * read back and verify the written value a few times.
+	 */
+	for (retry = 0; retry < 3; retry++) {
+		uint8_t tmp;
+
+		ret = drm_dp_dual_mode_write(adapter, DP_DUAL_MODE_TMDS_OEN,
+					     &tmds_oen, sizeof(tmds_oen));
+		if (ret) {
+			DRM_DEBUG_KMS("Failed to %s TMDS output buffers (%d attempts)\n",
+				      enable ? "enable" : "disable",
+				      retry + 1);
+			return ret;
+		}
+
+		ret = drm_dp_dual_mode_read(adapter, DP_DUAL_MODE_TMDS_OEN,
+					    &tmp, sizeof(tmp));
+		if (ret) {
+			DRM_DEBUG_KMS("I2C read failed during TMDS output buffer %s (%d attempts)\n",
+				      enable ? "enabling" : "disabling",
+				      retry + 1);
+			return ret;
+		}
+
+		if (tmp == tmds_oen)
+			return 0;
 	}
 
-	return 0;
+	DRM_DEBUG_KMS("I2C write value mismatch during TMDS output buffer %s\n",
+		      enable ? "enabling" : "disabling");
+
+	return -EIO;
 }
 EXPORT_SYMBOL(drm_dp_dual_mode_set_tmds_output);
 

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

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

From: Thomas Gleixner <[email protected]>

commit bd03143007eb9b03a7f2316c677780561b68ba2a upstream.

syszbot reported the following debugobjects splat:

ODEBUG: object is on stack, but not annotated

 WARNING: CPU: 0 PID: 4185 at lib/debugobjects.c:328

 RIP: 0010:debug_object_is_on_stack lib/debugobjects.c:327 [inline]
 debug_object_init+0x17/0x20 lib/debugobjects.c:391
 debug_hrtimer_init kernel/time/hrtimer.c:410 [inline]
 debug_init kernel/time/hrtimer.c:458 [inline]
 hrtimer_init+0x8c/0x410 kernel/time/hrtimer.c:1259
 alarm_init kernel/time/alarmtimer.c:339 [inline]
 alarm_timer_nsleep+0x164/0x4d0 kernel/time/alarmtimer.c:787
 SYSC_clock_nanosleep kernel/time/posix-timers.c:1226 [inline]
 SyS_clock_nanosleep+0x235/0x330 kernel/time/posix-timers.c:1204
 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287
 entry_SYSCALL_64_after_hwframe+0x42/0xb7

This happens because the hrtimer for the alarm nanosleep is on stack, but the code does not use the proper debug objects initialization.

Split out the code for the allocated use cases and invoke hrtimer_init_on_stack() for the nanosleep related functions.

---
 kernel/time/alarmtimer.c |   34 ++++++++++++++++++++++++++--------
 1 file changed, 26 insertions(+), 8 deletions(-)

@@ -326,6 +326,17 @@ static int alarmtimer_resume(struct devi
 }
 #endif
 
+static void
+__alarm_init(struct alarm *alarm, enum alarmtimer_type type,
+	     enum alarmtimer_restart (*function)(struct alarm *, ktime_t))
+{
+	timerqueue_init(&alarm->node);
+	alarm->timer.function = alarmtimer_fired;
+	alarm->function = function;
+	alarm->type = type;
+	alarm->state = ALARMTIMER_STATE_INACTIVE;
+}
+
 /**
  * alarm_init - Initialize an alarm structure
  * @alarm: ptr to alarm to be initialized

@@ -335,13 +346,9 @@ static int alarmtimer_resume(struct devi
 void alarm_init(struct alarm *alarm, enum alarmtimer_type type,
 		enum alarmtimer_restart (*function)(struct alarm *, ktime_t))
 {
-	timerqueue_init(&alarm->node);
 	hrtimer_init(&alarm->timer, alarm_bases[type].base_clockid,
-			HRTIMER_MODE_ABS);
-	alarm->timer.function = alarmtimer_fired;
-	alarm->function = function;
-	alarm->type = type;
-	alarm->state = ALARMTIMER_STATE_INACTIVE;
+		     HRTIMER_MODE_ABS);
+	__alarm_init(alarm, type, function);
 }
 EXPORT_SYMBOL_GPL(alarm_init);
 

@@ -719,6 +726,8 @@ static int alarmtimer_do_nsleep(struct a
 
 	__set_current_state(TASK_RUNNING);
 
+	destroy_hrtimer_on_stack(&alarm->timer);
+
 	if (!alarm->data)
 		return 0;
 

@@ -740,6 +749,15 @@ static int alarmtimer_do_nsleep(struct a
 	return -ERESTART_RESTARTBLOCK;
 }
 
+static void
+alarm_init_on_stack(struct alarm *alarm, enum alarmtimer_type type,
+		    enum alarmtimer_restart (*function)(struct alarm *, ktime_t))
+{
+	hrtimer_init_on_stack(&alarm->timer, alarm_bases[type].base_clockid,
+			      HRTIMER_MODE_ABS);
+	__alarm_init(alarm, type, function);
+}
+
 /**
  * alarm_timer_nsleep_restart - restartblock alarmtimer nsleep
  * @restart: ptr to restart block

@@ -752,7 +770,7 @@ static long __sched alarm_timer_nsleep_r
 	ktime_t exp = restart->nanosleep.expires;
 	struct alarm alarm;
 
-	alarm_init(&alarm, type, alarmtimer_nsleep_wakeup);
+	alarm_init_on_stack(&alarm, type, alarmtimer_nsleep_wakeup);
 
 	return alarmtimer_do_nsleep(&alarm, exp, type);
 }

@@ -784,7 +802,7 @@ static int alarm_timer_nsleep(const cloc
 	if (!capable(CAP_WAKE_ALARM))
 		return -EPERM;
 
-	alarm_init(&alarm, type, alarmtimer_nsleep_wakeup);
+	alarm_init_on_stack(&alarm, type, alarmtimer_nsleep_wakeup);
 
 	exp = timespec64_to_ktime(*tsreq);
 	/* Convert (if necessary) to absolute time */

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

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

From: Tetsuo Handa <[email protected]>

commit 8e04944f0ea8b838399049bdcda920ab36ae3b04 upstream.

syzbot is catching so many bugs triggered by commit 9ee332d99e4d5a97 ("sget(): handle failures of register_shrinker()"). That commit expected that calling kill_sb() from deactivate_locked_super() without successful fill_super() is safe, but the reality was different; some callers assign attributes which are needed for kill_sb() after sget() succeeds.

For example, [1] is a report where sb->s_mode (which seems to be either FMODE_READ | FMODE_EXCL | FMODE_WRITE or FMODE_READ | FMODE_EXCL) is not assigned unless sget() succeeds. But it does not worth complicate sget() so that register_shrinker() failure path can safely call kill_block_super() via kill_sb(). Making alloc_super() fail if memory allocation for register_shrinker() failed is much simpler. Let's avoid calling deactivate_locked_super() from sget_userns() by preallocating memory for the shrinker and making register_shrinker() in sget_userns() never fail.

[1] https://syzkaller.appspot.com/bug?id=588996a25a2587be2e3a54e8646728fb9cae44e7

---
 fs/super.c               |    9 ++++-----
 include/linux/shrinker.h |    7 +++++--
 mm/vmscan.c              |   21 ++++++++++++++++++++-
 3 files changed, 29 insertions(+), 8 deletions(-)

@@ -166,6 +166,7 @@ static void destroy_unused_super(struct
 	security_sb_free(s);
 	put_user_ns(s->s_user_ns);
 	kfree(s->s_subtype);
+	free_prealloced_shrinker(&s->s_shrink);
 	/* no delays needed */
 	destroy_super_work(&s->destroy_work);
 }

@@ -251,6 +252,8 @@ static struct super_block *alloc_super(s
 	s->s_shrink.count_objects = super_cache_count;
 	s->s_shrink.batch = 1024;
 	s->s_shrink.flags = SHRINKER_NUMA_AWARE | SHRINKER_MEMCG_AWARE;
+	if (prealloc_shrinker(&s->s_shrink))
+		goto fail;
 	return s;
 
 fail:

@@ -517,11 +520,7 @@ retry:
 	hlist_add_head(&s->s_instances, &type->fs_supers);
 	spin_unlock(&sb_lock);
 	get_filesystem(type);
-	err = register_shrinker(&s->s_shrink);
-	if (err) {
-		deactivate_locked_super(s);
-		s = ERR_PTR(err);
-	}
+	register_shrinker_prepared(&s->s_shrink);
 	return s;
 }
 

@@ -75,6 +75,9 @@ struct shrinker {
 #define SHRINKER_NUMA_AWARE	(1 << 0)
 #define SHRINKER_MEMCG_AWARE	(1 << 1)
 
-extern int register_shrinker(struct shrinker *);
-extern void unregister_shrinker(struct shrinker *);
+extern int prealloc_shrinker(struct shrinker *shrinker);
+extern void register_shrinker_prepared(struct shrinker *shrinker);
+extern int register_shrinker(struct shrinker *shrinker);
+extern void unregister_shrinker(struct shrinker *shrinker);
+extern void free_prealloced_shrinker(struct shrinker *shrinker);
 #endif

@@ -258,7 +258,7 @@ unsigned long lruvec_lru_size(struct lru
 /*
  * Add a shrinker callback to be called from the vm.
  */
-int register_shrinker(struct shrinker *shrinker)
+int prealloc_shrinker(struct shrinker *shrinker)
 {
 	size_t size = sizeof(*shrinker->nr_deferred);
 

@@ -268,10 +268,29 @@ int register_shrinker(struct shrinker *s
 	shrinker->nr_deferred = kzalloc(size, GFP_KERNEL);
 	if (!shrinker->nr_deferred)
 		return -ENOMEM;
+	return 0;
+}
+
+void free_prealloced_shrinker(struct shrinker *shrinker)
+{
+	kfree(shrinker->nr_deferred);
+	shrinker->nr_deferred = NULL;
+}
 
+void register_shrinker_prepared(struct shrinker *shrinker)
+{
 	down_write(&shrinker_rwsem);
 	list_add_tail(&shrinker->list, &shrinker_list);
 	up_write(&shrinker_rwsem);
+}
+
+int register_shrinker(struct shrinker *shrinker)
+{
+	int err = prealloc_shrinker(shrinker);
+
+	if (err)
+		return err;
+	register_shrinker_prepared(shrinker);
 	return 0;
 }
 EXPORT_SYMBOL(register_shrinker);

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

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

From: Florian Westphal <[email protected]>

commit 19926968ea86a286aa6fbea16ee3f2e7442f10f0 upstream.

Arbitrary limit, however, this still allows huge rulesets (> 1 million rules). This helps with automated fuzzer as it prevents oom-killer invocation.

---
 net/netfilter/x_tables.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -40,6 +40,7 @@ MODULE_AUTHOR("Harald Welte <laforge@net
 MODULE_DESCRIPTION("{ip,ip6,arp,eb}_tables backend module");
 
 #define XT_PCPU_BLOCK_SIZE 4096
+#define XT_MAX_TABLE_SIZE	(512 * 1024 * 1024)
 
 struct compat_delta {
 	unsigned int offset; /* offset in kernel */

@@ -1029,7 +1030,7 @@ struct xt_table_info *xt_alloc_table_inf
 	struct xt_table_info *info = NULL;
 	size_t sz = sizeof(*info) + size;
 
-	if (sz < sizeof(*info))
+	if (sz < sizeof(*info) || sz >= XT_MAX_TABLE_SIZE)
 		return NULL;
 
 	/* __GFP_NORETRY is not fully supported by kvmalloc but it should

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

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

From: Florian Westphal <[email protected]>

commit c84ca954ac9fa67a6ce27f91f01e4451c74fd8f6 upstream.

allows to have size checks in a single spot.
This is supposed to reduce oom situations when fuzz-testing xtables.

---
 include/linux/netfilter/x_tables.h |    1 +
 net/ipv4/netfilter/arp_tables.c    |    2 +-
 net/ipv4/netfilter/ip_tables.c     |    2 +-
 net/ipv6/netfilter/ip6_tables.c    |    2 +-
 net/netfilter/x_tables.c           |   15 +++++++++++++++
 5 files changed, 19 insertions(+), 3 deletions(-)

@@ -301,6 +301,7 @@ int xt_data_to_user(void __user *dst, co
 
 void *xt_copy_counters_from_user(const void __user *user, unsigned int len,
 				 struct xt_counters_info *info, bool compat);
+struct xt_counters *xt_counters_alloc(unsigned int counters);
 
 struct xt_table *xt_register_table(struct net *net,
 				   const struct xt_table *table,

@@ -895,7 +895,7 @@ static int __do_replace(struct net *net,
 	struct arpt_entry *iter;
 
 	ret = 0;
-	counters = vzalloc(num_counters * sizeof(struct xt_counters));
+	counters = xt_counters_alloc(num_counters);
 	if (!counters) {
 		ret = -ENOMEM;
 		goto out;

@@ -1057,7 +1057,7 @@ __do_replace(struct net *net, const char
 	struct ipt_entry *iter;
 
 	ret = 0;
-	counters = vzalloc(num_counters * sizeof(struct xt_counters));
+	counters = xt_counters_alloc(num_counters);
 	if (!counters) {
 		ret = -ENOMEM;
 		goto out;

@@ -1075,7 +1075,7 @@ __do_replace(struct net *net, const char
 	struct ip6t_entry *iter;
 
 	ret = 0;
-	counters = vzalloc(num_counters * sizeof(struct xt_counters));
+	counters = xt_counters_alloc(num_counters);
 	if (!counters) {
 		ret = -ENOMEM;
 		goto out;

@@ -1199,6 +1199,21 @@ static int xt_jumpstack_alloc(struct xt_
 	return 0;
 }
 
+struct xt_counters *xt_counters_alloc(unsigned int counters)
+{
+	struct xt_counters *mem;
+
+	if (counters == 0 || counters > INT_MAX / sizeof(*mem))
+		return NULL;
+
+	counters *= sizeof(*mem);
+	if (counters > XT_MAX_TABLE_SIZE)
+		return NULL;
+
+	return vzalloc(counters);
+}
+EXPORT_SYMBOL(xt_counters_alloc);
+
 struct xt_table_info *
 xt_replace_table(struct xt_table *table,
 	      unsigned int num_counters,

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

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

From: Florian Westphal <[email protected]>

commit 9782a11efc072faaf91d4aa60e9d23553f918029 upstream.

should have no impact, function still always returns 0. This patch is only to ease review.

---
 include/linux/netfilter/x_tables.h |    2 +-
 net/bridge/netfilter/ebtables.c    |   10 ++++++++--
 net/ipv4/netfilter/arp_tables.c    |   10 +++++++---
 net/ipv4/netfilter/ip_tables.c     |    8 ++++++--
 net/ipv6/netfilter/ip6_tables.c    |   10 +++++++---
 net/netfilter/x_tables.c           |    4 +++-
 6 files changed, 32 insertions(+), 12 deletions(-)

@@ -510,7 +510,7 @@ void xt_compat_unlock(u_int8_t af);
 
 int xt_compat_add_offset(u_int8_t af, unsigned int offset, int delta);
 void xt_compat_flush_offsets(u_int8_t af);
-void xt_compat_init_offsets(u_int8_t af, unsigned int number);
+int xt_compat_init_offsets(u8 af, unsigned int number);
 int xt_compat_calc_jump(u_int8_t af, unsigned int offset);
 
 int xt_compat_match_offset(const struct xt_match *match);

@@ -1821,10 +1821,14 @@ static int compat_table_info(const struc
 {
 	unsigned int size = info->entries_size;
 	const void *entries = info->entries;
+	int ret;
 
 	newinfo->entries_size = size;
 
-	xt_compat_init_offsets(NFPROTO_BRIDGE, info->nentries);
+	ret = xt_compat_init_offsets(NFPROTO_BRIDGE, info->nentries);
+	if (ret)
+		return ret;
+
 	return EBT_ENTRY_ITERATE(entries, size, compat_calc_entry, info,
 							entries, newinfo);
 }

@@ -2268,7 +2272,9 @@ static int compat_do_replace(struct net
 
 	xt_compat_lock(NFPROTO_BRIDGE);
 
-	xt_compat_init_offsets(NFPROTO_BRIDGE, tmp.nentries);
+	ret = xt_compat_init_offsets(NFPROTO_BRIDGE, tmp.nentries);
+	if (ret < 0)
+		goto out_unlock;
 	ret = compat_copy_entries(entries_tmp, tmp.entries_size, &state);
 	if (ret < 0)
 		goto out_unlock;

@@ -781,7 +781,9 @@ static int compat_table_info(const struc
 	memcpy(newinfo, info, offsetof(struct xt_table_info, entries));
 	newinfo->initial_entries = 0;
 	loc_cpu_entry = info->entries;
-	xt_compat_init_offsets(NFPROTO_ARP, info->number);
+	ret = xt_compat_init_offsets(NFPROTO_ARP, info->number);
+	if (ret)
+		return ret;
 	xt_entry_foreach(iter, loc_cpu_entry, info->size) {
 		ret = compat_calc_entry(iter, info, loc_cpu_entry, newinfo);
 		if (ret != 0)

@@ -1167,7 +1169,7 @@ static int translate_compat_table(struct
 	struct compat_arpt_entry *iter0;
 	struct arpt_replace repl;
 	unsigned int size;
-	int ret = 0;
+	int ret;
 
 	info = *pinfo;
 	entry0 = *pentry0;

@@ -1176,7 +1178,9 @@ static int translate_compat_table(struct
 
 	j = 0;
 	xt_compat_lock(NFPROTO_ARP);
-	xt_compat_init_offsets(NFPROTO_ARP, compatr->num_entries);
+	ret = xt_compat_init_offsets(NFPROTO_ARP, compatr->num_entries);
+	if (ret)
+		goto out_unlock;
 	/* Walk through entries, checking offsets. */
 	xt_entry_foreach(iter0, entry0, compatr->size) {
 		ret = check_compat_entry_size_and_hooks(iter0, info, &size,

@@ -945,7 +945,9 @@ static int compat_table_info(const struc
 	memcpy(newinfo, info, offsetof(struct xt_table_info, entries));
 	newinfo->initial_entries = 0;
 	loc_cpu_entry = info->entries;
-	xt_compat_init_offsets(AF_INET, info->number);
+	ret = xt_compat_init_offsets(AF_INET, info->number);
+	if (ret)
+		return ret;
 	xt_entry_foreach(iter, loc_cpu_entry, info->size) {
 		ret = compat_calc_entry(iter, info, loc_cpu_entry, newinfo);
 		if (ret != 0)

@@ -1418,7 +1420,9 @@ translate_compat_table(struct net *net,
 
 	j = 0;
 	xt_compat_lock(AF_INET);
-	xt_compat_init_offsets(AF_INET, compatr->num_entries);
+	ret = xt_compat_init_offsets(AF_INET, compatr->num_entries);
+	if (ret)
+		goto out_unlock;
 	/* Walk through entries, checking offsets. */
 	xt_entry_foreach(iter0, entry0, compatr->size) {
 		ret = check_compat_entry_size_and_hooks(iter0, info, &size,

@@ -962,7 +962,9 @@ static int compat_table_info(const struc
 	memcpy(newinfo, info, offsetof(struct xt_table_info, entries));
 	newinfo->initial_entries = 0;
 	loc_cpu_entry = info->entries;
-	xt_compat_init_offsets(AF_INET6, info->number);
+	ret = xt_compat_init_offsets(AF_INET6, info->number);
+	if (ret)
+		return ret;
 	xt_entry_foreach(iter, loc_cpu_entry, info->size) {
 		ret = compat_calc_entry(iter, info, loc_cpu_entry, newinfo);
 		if (ret != 0)

@@ -1425,7 +1427,7 @@ translate_compat_table(struct net *net,
 	struct compat_ip6t_entry *iter0;
 	struct ip6t_replace repl;
 	unsigned int size;
-	int ret = 0;
+	int ret;
 
 	info = *pinfo;
 	entry0 = *pentry0;

@@ -1434,7 +1436,9 @@ translate_compat_table(struct net *net,
 
 	j = 0;
 	xt_compat_lock(AF_INET6);
-	xt_compat_init_offsets(AF_INET6, compatr->num_entries);
+	ret = xt_compat_init_offsets(AF_INET6, compatr->num_entries);
+	if (ret)
+		goto out_unlock;
 	/* Walk through entries, checking offsets. */
 	xt_entry_foreach(iter0, entry0, compatr->size) {
 		ret = check_compat_entry_size_and_hooks(iter0, info, &size,

@@ -604,10 +604,12 @@ int xt_compat_calc_jump(u_int8_t af, uns
 }
 EXPORT_SYMBOL_GPL(xt_compat_calc_jump);
 
-void xt_compat_init_offsets(u_int8_t af, unsigned int number)
+int xt_compat_init_offsets(u8 af, unsigned int number)
 {
 	xt[af].number = number;
 	xt[af].cur = 0;
+
+	return 0;
 }
 EXPORT_SYMBOL(xt_compat_init_offsets);
 

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

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

From: Florian Westphal <[email protected]>

commit 7d7d7e02111e9a4dc9d0658597f528f815d820fd upstream.

no need to bother even trying to allocating huge compat offset arrays, such ruleset is rejected later on anyway becaus we refuse to allocate overly large rule blobs.

However, compat translation happens before blob allocation, so we should add a check there too.

This is supposed to help with fuzzing by avoiding oom-killer.

---
 net/netfilter/x_tables.c |   26 ++++++++++++++++++--------
 1 file changed, 18 insertions(+), 8 deletions(-)

@@ -554,14 +554,8 @@ int xt_compat_add_offset(u_int8_t af, un
 {
 	struct xt_af *xp = &xt[af];
 
-	if (!xp->compat_tab) {
-		if (!xp->number)
-			return -EINVAL;
-		xp->compat_tab = vmalloc(sizeof(struct compat_delta) * xp->number);
-		if (!xp->compat_tab)
-			return -ENOMEM;
-		xp->cur = 0;
-	}
+	if (WARN_ON(!xp->compat_tab))
+		return -ENOMEM;
 
 	if (xp->cur >= xp->number)
 		return -EINVAL;

@@ -606,6 +600,22 @@ EXPORT_SYMBOL_GPL(xt_compat_calc_jump);
 
 int xt_compat_init_offsets(u8 af, unsigned int number)
 {
+	size_t mem;
+
+	if (!number || number > (INT_MAX / sizeof(struct compat_delta)))
+		return -EINVAL;
+
+	if (WARN_ON(xt[af].compat_tab))
+		return -EINVAL;
+
+	mem = sizeof(struct compat_delta) * number;
+	if (mem > XT_MAX_TABLE_SIZE)
+		return -ENOMEM;
+
+	xt[af].compat_tab = vmalloc(mem);
+	if (!xt[af].compat_tab)
+		return -ENOMEM;
+
 	xt[af].number = number;
 	xt[af].cur = 0;
 

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

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

From: Florian Westphal <[email protected]>

commit 9d5c12a7c08f67999772065afd50fb222072114e upstream.

This is a very conservative limit (134217728 rules), but good enough to not trigger frequent oom from syzkaller.

---
 net/netfilter/x_tables.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -818,6 +818,9 @@ EXPORT_SYMBOL(xt_check_entry_offsets);
  */
 unsigned int *xt_alloc_entry_offsets(unsigned int size)
 {
+	if (size > XT_MAX_TABLE_SIZE / sizeof(unsigned int))
+		return NULL;
+
 	return kvmalloc_array(size, sizeof(unsigned int), GFP_KERNEL | __GFP_ZERO);
 
 }

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

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

From: Jiri Olsa <[email protected]>

commit 5af44ca53d019de47efe6dbc4003dd518e5197ed upstream.

The syzbot hit KASAN bug in perf_callchain_store having the entry stored behind the allocated bounds [1].

We miss the sample_max_stack check for the initial event that allocates callchain buffers. This missing check allows to create an event with sample_max_stack value bigger than the global sysctl maximum:

  # sysctl -a | grep perf_event_max_stack
  kernel.perf_event_max_stack = 127

  # perf record -vv -C 1 -e cycles/max-stack=256/ kill
  ...
  perf_event_attr:
    size                             112
    ...
    sample_max_stack                 256
  ------------------------------------------------------------
  sys_perf_event_open: pid -1  cpu 1  group_fd -1  flags 0x8 = 4

Note the '-C 1', which forces perf record to create just single event. Otherwise it opens event for every cpu, then the sample_max_stack check fails on the second event and all's fine.

The fix is to run the sample_max_stack check also for the first event with callchains.

[1] https://marc.info/?l=linux-kernel&m=152352732920874&w=2

---
 kernel/events/callchain.c |   21 ++++++++++++---------
 1 file changed, 12 insertions(+), 9 deletions(-)

@@ -119,19 +119,22 @@ int get_callchain_buffers(int event_max_
 		goto exit;
 	}
 
+	/*
+	 * If requesting per event more than the global cap,
+	 * return a different error to help userspace figure
+	 * this out.
+	 *
+	 * And also do it here so that we have &callchain_mutex held.
+	 */
+	if (event_max_stack > sysctl_perf_event_max_stack) {
+		err = -EOVERFLOW;
+		goto exit;
+	}
+
 	if (count > 1) {
 		/* If the allocation failed, give up */
 		if (!callchain_cpus_entries)
 			err = -ENOMEM;
-		/*
-		 * If requesting per event more than the global cap,
-		 * return a different error to help userspace figure
-		 * this out.
-		 *
-		 * And also do it here so that we have &callchain_mutex held.
-		 */
-		if (event_max_stack > sysctl_perf_event_max_stack)
-			err = -EOVERFLOW;
 		goto exit;
 	}
 

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

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

From: Jiri Olsa <[email protected]>

commit 78b562fbfa2cf0a9fcb23c3154756b690f4905c1 upstream.

Return immediately when we find issue in the user stack checks. The error value could get overwritten by following check for PERF_SAMPLE_REGS_INTR.

---
 kernel/events/core.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -9730,9 +9730,9 @@ static int perf_copy_attr(struct perf_ev
 		 * __u16 sample size limit.
 		 */
 		if (attr->sample_stack_user >= USHRT_MAX)
-			ret = -EINVAL;
+			return -EINVAL;
 		else if (!IS_ALIGNED(attr->sample_stack_user, sizeof(u64)))
-			ret = -EINVAL;
+			return -EINVAL;
 	}
 
 	if (attr->sample_type & PERF_SAMPLE_REGS_INTR)

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

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

From: Leon Romanovsky <[email protected]>

commit 75a4598209cbe45540baa316c3b51d9db222e96e upstream.

mlx5 modify_qp() relies on FW that the error will be thrown if wrong state is supplied. The missing check in FW causes the following crash while using XRC_TGT QPs.

[   14.769632] BUG: unable to handle kernel NULL pointer dereference at (null)
[   14.771085] IP: mlx5_ib_modify_qp+0xf60/0x13f0
[   14.771894] PGD 800000001472e067 P4D 800000001472e067 PUD 14529067 PMD 0
[   14.773126] Oops: 0002 [#1] SMP PTI
[   14.773763] CPU: 0 PID: 365 Comm: ubsan Not tainted 4.16.0-rc1-00038-g8151138c0793 #119
[   14.775192] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.7.5-0-ge51488c-20140602_164612-nilsson.home.kraxel.org 04/01/2014
[   14.777522] RIP: 0010:mlx5_ib_modify_qp+0xf60/0x13f0
[   14.778417] RSP: 0018:ffffbf48001c7bd8 EFLAGS: 00010246
[   14.779346] RAX: 0000000000000000 RBX: ffff9a8f9447d400 RCX: 0000000000000000
[   14.780643] RDX: 0000000000000000 RSI: 000000000000000a RDI: 0000000000000000
[   14.781930] RBP: 0000000000000000 R08: 00000000000217b0 R09: ffffffffbc9c1504
[   14.783214] R10: fffff4a180519480 R11: ffff9a8f94523600 R12: ffff9a8f9493e240
[   14.784507] R13: ffff9a8f9447d738 R14: 000000000000050a R15: 0000000000000000
[   14.785800] FS:  00007f545b466700(0000) GS:ffff9a8f9fc00000(0000) knlGS:0000000000000000
[   14.787073] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   14.787792] CR2: 0000000000000000 CR3: 00000000144be000 CR4: 00000000000006b0
[   14.788689] Call Trace:
[   14.789007]  _ib_modify_qp+0x71/0x120
[   14.789475]  modify_qp.isra.20+0x207/0x2f0
[   14.790010]  ib_uverbs_modify_qp+0x90/0xe0
[   14.790532]  ib_uverbs_write+0x1d2/0x3c0
[   14.791049]  ? __handle_mm_fault+0x93c/0xe40
[   14.791644]  __vfs_write+0x36/0x180
[   14.792096]  ? handle_mm_fault+0xc1/0x210
[   14.792601]  vfs_write+0xad/0x1e0
[   14.793018]  SyS_write+0x52/0xc0
[   14.793422]  do_syscall_64+0x75/0x180
[   14.793888]  entry_SYSCALL_64_after_hwframe+0x21/0x86
[   14.794527] RIP: 0033:0x7f545ad76099
[   14.794975] RSP: 002b:00007ffd78787468 EFLAGS: 00000287 ORIG_RAX: 0000000000000001
[   14.795958] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f545ad76099
[   14.797075] RDX: 0000000000000078 RSI: 0000000020009000 RDI: 0000000000000003
[   14.798140] RBP: 00007ffd78787470 R08: 00007ffd78787480 R09: 00007ffd78787480
[   14.799207] R10: 00007ffd78787480 R11: 0000000000000287 R12: 00005599ada98760
[   14.800277] R13: 00007ffd78787560 R14: 0000000000000000 R15: 0000000000000000
[   14.801341] Code: 4c 8b 1c 24 48 8b 83 70 02 00 00 48 c7 83 cc 02 00
00 00 00 00 00 48 c7 83 24 03 00 00 00 00 00 00 c7 83 2c 03 00 00 00 00
00 00 <c7> 00 00 00 00 00 48 8b 83 70 02 00 00 c7 40 04 00 00 00 00 4c
[   14.804012] RIP: mlx5_ib_modify_qp+0xf60/0x13f0 RSP: ffffbf48001c7bd8
[   14.804838] CR2: 0000000000000000
[   14.805288] ---[ end trace 3f1da0df5c8b7c37 ]---

---
 drivers/infiniband/hw/mlx5/qp.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -3157,7 +3157,8 @@ static int __mlx5_ib_modify_qp(struct ib
 	 * If we moved a kernel QP to RESET, clean up all old CQ
 	 * entries and reinitialize the QP.
 	 */
-	if (new_state == IB_QPS_RESET && !ibqp->uobject) {
+	if (new_state == IB_QPS_RESET &&
+	    !ibqp->uobject && ibqp->qp_type != IB_QPT_XRC_TGT) {
 		mlx5_ib_cq_clean(recv_cq, base->mqp.qpn,
 				 ibqp->srq ? to_msrq(ibqp->srq) : NULL);
 		if (send_cq != recv_cq)

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

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

From: Sean Christopherson <[email protected]>

commit 2c151b25441ae5c2da66472abd165af785c9ecd2 upstream.

The bug that led to commit 95e057e25892eaa48cad1e2d637b80d0f1a4fac5 was a benign warning (no adverse affects other than the warning itself) that was detected by syzkaller. Further inspection shows that the WARN_ON in question, in handle_ept_misconfig(), is unnecessary and flawed (this was also briefly discussed in the original patch: https://patchwork.kernel.org/patch/10204649).

* The WARN_ON is unnecessary as kvm_mmu_page_fault() will WARN if reserved bits are set in the SPTEs, i.e. it covers the case where an EPT misconfig occurred because of a KVM bug.

* The WARN_ON is flawed because it will fire on any system error code that is hit while handling the fault, e.g. -ENOMEM can be returned by mmu_topup_memory_caches() while handling a legitmate MMIO EPT misconfig.

The original behavior of returning -EFAULT when userspace munmaps an HVA without first removing the memslot is correct and desirable, i.e. KVM is letting userspace know it has generated a bad address. Returning RET_PF_EMULATE masks the WARN_ON in the EPT misconfig path, but does not fix the underlying bug, i.e. the WARN_ON is bogus.

Furthermore, returning RET_PF_EMULATE has the unwanted side effect of causing KVM to attempt to emulate an instruction on any page fault with an invalid HVA translation, e.g. a not-present EPT violation on a VM_PFNMAP VMA whose fault handler failed to insert a PFN.

* There is no guarantee that the fault is directly related to the instruction, i.e. the fault could have been triggered by a side effect memory access in the guest, e.g. while vectoring a #DB or writing a tracing record. This could cause KVM to effectively mask the fault if KVM doesn't model the behavior leading to the fault, i.e. emulation could succeed and resume the guest.

* If emulation does fail, KVM will return EMULATION_FAILED instead of -EFAULT, which is a red herring as the user will either debug a bogus emulation attempt or scratch their head wondering why we were attempting emulation in the first place.

TL;DR: revert to returning -EFAULT and remove the bogus WARN_ON in handle_ept_misconfig in a future patch.

This reverts commit 95e057e25892eaa48cad1e2d637b80d0f1a4fac5.

---
 arch/x86/kvm/mmu.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -3031,7 +3031,7 @@ static int kvm_handle_bad_page(struct kv
 		return RET_PF_RETRY;
 	}
 
-	return RET_PF_EMULATE;
+	return -EFAULT;
 }
 
 static void transparent_hugepage_adjust(struct kvm_vcpu *vcpu,

For v4.16.4-28-gb04d14a:

Build results:

	total: 143 pass: 143 fail: 0
Qemu test results:
	total: 139 pass: 139 fail: 0

Details are available at http://kerneltests.org/builders/.

Guenter

Great! Thanks for testing and letting me know.

greg k-h

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

thanks,
-- Shuah

Thanks for testing both of these and letting me know.

greg k-h

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

We did find a regression in 4.16.4 that is dragonboard-410c specific that we missed the first time around. Discovered using kselftest/printf.sh:

/sbin/modprobe test_printf
[ 22.725551] test_printf: hashing plain 'p' has unexpected format
[ 22.726031] test_printf: failed 1 out of 236 tests
modprobe: ERROR: could not insert 'test_printf': Invalid argument

We'll bisect it and find the issue but since it was already released, I assume you won't want to hold up 4.16.5 for it.

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

kernel: 4.16.5-rc1
git repo:
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git
git branch: linux-4.16.y
git commit: 9a7eb4ea64f3850a0fabfdf2d95f4c8cbffcb324 git describe: v4.16.4-27-g9a7eb4ea64f3
Test details:
https://qa-reports.linaro.org/lkft/linux-stable-rc-4.16-oe/build/v4.16.4-27-g9a7eb4ea64f3

No regressions (compared to build v4.16.4)

Boards, architectures and test suites:
-------------------------------------

dragonboard-410c - arm64
* boot - pass: 18, fail: 2,
* kselftest - pass: 41, fail: 1, skip: 26
* libhugetlbfs - pass: 89, fail: 1, skip: 1
* ltp-containers-tests - pass: 64, skip: 17
* ltp-fcntl-locktests-tests - pass: 2,
* ltp-filecaps-tests - pass: 2,
* ltp-fs-tests - pass: 57, skip: 6
* ltp-fs_bind-tests - pass: 2,
* ltp-fs_perms_simple-tests - pass: 19,
* ltp-fsx-tests - pass: 2,
* ltp-hugetlb-tests - pass: 21, skip: 1
* ltp-io-tests - pass: 3,
* ltp-ipc-tests - pass: 9,
* ltp-nptl-tests - pass: 2,
* ltp-pty-tests - pass: 4,
* ltp-sched-tests - pass: 14,
* ltp-securebits-tests - pass: 4,
* ltp-syscalls-tests - pass: 1017, skip: 133
* ltp-timers-tests - pass: 13,

hi6220-hikey - arm64
* boot - pass: 20,
* kselftest - pass: 45, skip: 22
* libhugetlbfs - pass: 90, skip: 1
* ltp-cap_bounds-tests - pass: 2,
* ltp-containers-tests - pass: 64, skip: 17
* ltp-fcntl-locktests-tests - pass: 2,
* ltp-filecaps-tests - pass: 2,
* ltp-fs-tests - pass: 57, skip: 6
* ltp-fs_bind-tests - pass: 2,
* ltp-fs_perms_simple-tests - pass: 19,
* ltp-fsx-tests - pass: 2,
* ltp-hugetlb-tests - pass: 21, skip: 1
* ltp-io-tests - pass: 3,
* ltp-ipc-tests - pass: 9,
* ltp-math-tests - pass: 11,
* ltp-nptl-tests - pass: 2,
* ltp-pty-tests - pass: 4,
* ltp-sched-tests - pass: 10, skip: 4
* ltp-securebits-tests - pass: 4,
* ltp-syscalls-tests - pass: 1016, skip: 134
* ltp-timers-tests - pass: 13,

juno-r2 - arm64
* boot - pass: 20,
* libhugetlbfs - pass: 90, skip: 1
* ltp-cap_bounds-tests - pass: 2,
* ltp-containers-tests - pass: 64, skip: 17
* ltp-filecaps-tests - pass: 2,
* ltp-fs-tests - pass: 57, skip: 6
* ltp-fs_perms_simple-tests - pass: 19,
* ltp-fsx-tests - pass: 2,
* ltp-hugetlb-tests - pass: 22,
* ltp-io-tests - pass: 3,
* ltp-ipc-tests - pass: 9,
* ltp-math-tests - pass: 11,
* ltp-nptl-tests - pass: 2,
* ltp-pty-tests - pass: 4,
* ltp-sched-tests - pass: 10, skip: 4
* ltp-securebits-tests - pass: 4,
* ltp-syscalls-tests - pass: 1017, skip: 133
* ltp-timers-tests - pass: 13,

qemu_arm
* boot - pass: 15, fail: 5,
* ltp-cap_bounds-tests - pass: 2,
* ltp-containers-tests - pass: 62, fail: 2, skip: 17
* ltp-fcntl-locktests-tests - pass: 2,
* ltp-filecaps-tests - pass: 2,
* ltp-fs-tests - pass: 58, skip: 5
* ltp-fs_bind-tests - pass: 2,
* ltp-fsx-tests - pass: 2,
* ltp-hugetlb-tests - pass: 21, skip: 1
* ltp-io-tests - pass: 3,
* ltp-math-tests - pass: 11,
* ltp-nptl-tests - pass: 2,
* ltp-pty-tests - pass: 4,
* ltp-timers-tests - pass: 13,

qemu_arm64
* boot - pass: 15, fail: 5,
* ltp-cap_bounds-tests - pass: 2,
* ltp-containers-tests - pass: 64, skip: 17
* ltp-fcntl-locktests-tests - pass: 2,
* ltp-filecaps-tests - pass: 2,
* ltp-fs_bind-tests - pass: 2,
* ltp-fs_perms_simple-tests - pass: 19,
* ltp-fsx-tests - pass: 2,
* ltp-hugetlb-tests - pass: 22,
* ltp-io-tests - pass: 3,
* ltp-ipc-tests - pass: 9,
* ltp-nptl-tests - pass: 2,
* ltp-pty-tests - pass: 4,
* ltp-timers-tests - pass: 13,

qemu_x86_64
* boot - pass: 22,
* kselftest - pass: 50, skip: 30
* kselftest-vsyscall-mode-native - pass: 50, skip: 30
* kselftest-vsyscall-mode-none - pass: 50, skip: 30
* libhugetlbfs - pass: 90, skip: 1
* ltp-cap_bounds-tests - pass: 2,
* ltp-containers-tests - pass: 64, skip: 17
* ltp-fcntl-locktests-tests - pass: 2,
* ltp-filecaps-tests - pass: 2,
* ltp-fs-tests - pass: 57, skip: 6
* ltp-fs_bind-tests - pass: 2,
* ltp-fs_perms_simple-tests - pass: 19,
* ltp-fsx-tests - pass: 2,
* ltp-hugetlb-tests - pass: 22,
* ltp-io-tests - pass: 3,
* ltp-ipc-tests - pass: 9,
* ltp-math-tests - pass: 11,
* ltp-nptl-tests - pass: 2,
* ltp-pty-tests - pass: 4,
* ltp-sched-tests - pass: 13, skip: 1
* ltp-securebits-tests - pass: 4,
* ltp-syscalls-tests - pass: 998, skip: 152
* ltp-timers-tests - pass: 13,

x15 - arm
* boot - pass: 20,
* kselftest - pass: 37, skip: 28
* libhugetlbfs - pass: 87, skip: 1
* ltp-cap_bounds-tests - pass: 2,
* ltp-containers-tests - pass: 63, skip: 18
* ltp-fcntl-locktests-tests - pass: 2,
* ltp-filecaps-tests - pass: 2,
* ltp-fs-tests - pass: 58, skip: 5
* ltp-fs_bind-tests - pass: 2,
* ltp-fs_perms_simple-tests - pass: 19,
* ltp-fsx-tests - pass: 2,
* ltp-hugetlb-tests - pass: 20, skip: 2
* ltp-io-tests - pass: 3,
* ltp-ipc-tests - pass: 9,
* ltp-math-tests - pass: 11,
* ltp-nptl-tests - pass: 2,
* ltp-pty-tests - pass: 4,
* ltp-sched-tests - pass: 13, skip: 1
* ltp-securebits-tests - pass: 4,
* ltp-syscalls-tests - pass: 1075, skip: 75
* ltp-timers-tests - pass: 13,

x86_64
* boot - pass: 22,
* kselftest - pass: 55, skip: 20
* kselftest-vsyscall-mode-native - pass: 55, skip: 20
* kselftest-vsyscall-mode-none - pass: 55, skip: 20
* libhugetlbfs - pass: 90, skip: 1
* ltp-cap_bounds-tests - pass: 2,
* ltp-containers-tests - pass: 64, skip: 17
* ltp-fcntl-locktests-tests - pass: 2,
* ltp-filecaps-tests - pass: 2,
* ltp-fs-tests - pass: 58, skip: 5
* ltp-fs_bind-tests - pass: 2,
* ltp-fs_perms_simple-tests - pass: 19,
* ltp-fsx-tests - pass: 2,
* ltp-hugetlb-tests - pass: 22,
* ltp-io-tests - pass: 3,
* ltp-ipc-tests - pass: 9,
* ltp-math-tests - pass: 11,
* ltp-nptl-tests - pass: 2,
* ltp-pty-tests - pass: 4,
* ltp-sched-tests - pass: 9, skip: 5
* ltp-securebits-tests - pass: 4,
* ltp-syscalls-tests - pass: 1034, skip: 116
* ltp-timers-tests - pass: 13,

Linaro QA (BETA)
https://qa-reports.linaro.org

As it's an old issue, no, I'll wait for you all to find it and not hold this release up :)

thanks,

greg k-h