4.9.104-stable review

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

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

From: Sven Eckelmann <[email protected]>

[ Upstream commit 011c935fceae5252619ef730baa610c655281dda ]

The function batadv_v_gw_dump stops the processing loop when batadv_v_gw_dump_entry returns a non-0 return code. This should only happen when the buffer is full. Otherwise, an empty message may be returned by batadv_gw_dump. This empty message will then stop the netlink dumping of gateway entries. At worst, not a single entry is returned to userspace even when plenty of possible gateways exist.

---
 net/batman-adv/bat_v.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -920,7 +920,7 @@ static int batadv_v_gw_dump_entry(struct
 	struct batadv_neigh_ifinfo *router_ifinfo = NULL;
 	struct batadv_neigh_node *router;
 	struct batadv_gw_node *curr_gw;
-	int ret = -EINVAL;
+	int ret = 0;
 	void *hdr;
 
 	router = batadv_orig_router_get(gw_node->orig_node, BATADV_IF_DEFAULT);

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

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

From: Jianchao Wang <[email protected]>

[ Upstream commit f25a2dfc20e3a3ed8fe6618c331799dd7bd01190 ]

This patch fixes nvme queue cleanup if requesting an IRQ handler for the queue's vector fails. It does this by resetting the cq_vector to the uninitialized value of -1 so it is ignored for a controller reset.

---
 drivers/nvme/host/pci.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

@@ -1120,7 +1120,7 @@ static int nvme_create_queue(struct nvme
 	nvmeq->cq_vector = qid - 1;
 	result = adapter_alloc_cq(dev, qid, nvmeq);
 	if (result < 0)
-		return result;
+		goto release_vector;
 
 	result = adapter_alloc_sq(dev, qid, nvmeq);
 	if (result < 0)

@@ -1134,9 +1134,12 @@ static int nvme_create_queue(struct nvme
 	return result;
 
  release_sq:
+	dev->online_queues--;
 	adapter_delete_sq(dev, qid);
  release_cq:
 	adapter_delete_cq(dev, qid);
+ release_vector:
+	nvmeq->cq_vector = -1;
 	return result;
 }
 

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

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

From: Colin Ian King <[email protected]>

[ Upstream commit f287eb9013ccf199cbfa4eabd80c36fedfc15a73 ]

The error checks on freq for a negative error return always fails because freq is unsigned and can never be negative. Fix this by making freq a signed long.

Detected with Coccinelle:
drivers/clocksource/fsl_ftm_timer.c:287:5-9: WARNING: Unsigned expression compared with zero: freq <= 0 drivers/clocksource/fsl_ftm_timer.c:291:5-9: WARNING: Unsigned expression compared with zero: freq <= 0

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

@@ -281,7 +281,7 @@ static int __init __ftm_clk_init(struct
 
 static unsigned long __init ftm_clk_init(struct device_node *np)
 {
-	unsigned long freq;
+	long freq;
 
 	freq = __ftm_clk_init(np, "ftm-evt-counter-en", "ftm-evt");
 	if (freq <= 0)

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

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

From: Chengguang Xu <[email protected]>

[ Upstream commit 18106734b512664a8541026519ce4b862498b6c3 ]

When failing from ceph_fs_debugfs_init() in ceph_real_mount(), there is lack of dput of root_dentry and it causes slab errors, so change the calling order of ceph_fs_debugfs_init() and open_root_dentry() and do some cleanups to avoid this issue.

---
 fs/ceph/super.c |   16 +++++-----------
 1 file changed, 5 insertions(+), 11 deletions(-)

@@ -816,7 +816,6 @@ static struct dentry *ceph_real_mount(st
 	int err;
 	unsigned long started = jiffies;  /* note the start time */
 	struct dentry *root;
-	int first = 0;   /* first vfsmount for this super_block */
 
 	dout("mount start %p\n", fsc);
 	mutex_lock(&fsc->client->mount_mutex);

@@ -834,17 +833,17 @@ static struct dentry *ceph_real_mount(st
 			path = fsc->mount_options->server_path + 1;
 			dout("mount opening path %s\n", path);
 		}
+
+		err = ceph_fs_debugfs_init(fsc);
+		if (err < 0)
+			goto out;
+
 		root = open_root_dentry(fsc, path, started);
 		if (IS_ERR(root)) {
 			err = PTR_ERR(root);
 			goto out;
 		}
 		fsc->sb->s_root = dget(root);
-		first = 1;
-
-		err = ceph_fs_debugfs_init(fsc);
-		if (err < 0)
-			goto fail;
 	} else {
 		root = dget(fsc->sb->s_root);
 	}

@@ -854,11 +853,6 @@ static struct dentry *ceph_real_mount(st
 	mutex_unlock(&fsc->client->mount_mutex);
 	return root;
 
-fail:
-	if (first) {
-		dput(fsc->sb->s_root);
-		fsc->sb->s_root = NULL;
-	}
 out:
 	mutex_unlock(&fsc->client->mount_mutex);
 	return ERR_PTR(err);

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

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

From: "David S. Miller" <[email protected]>

[ Upstream commit 13a55372b64e00e564a08d785ca87bd9d454ba30 ]

It is not valid for orion5x to use mac_pton().

First of all, the orion5x buffer is not NULL terminated. mac_pton() has no business operating on non-NULL terminated buffers because only the caller can know that this is valid and in what manner it is ok to parse this NULL'less buffer.

Second of all, orion5x operates on an __iomem pointer, which cannot be dereferenced using normal C pointer operations. Accesses to such areas much be performed with the proper iomem accessors.

---
 arch/arm/mach-orion5x/Kconfig        |    3 -
 arch/arm/mach-orion5x/dns323-setup.c |   53 +++++++++++++++++++++++++++++++++--
 arch/arm/mach-orion5x/tsx09-common.c |   49 +++++++++++++++++++++++++++++---
 3 files changed, 95 insertions(+), 10 deletions(-)

@@ -57,7 +57,6 @@ config MACH_KUROBOX_PRO
 
 config MACH_DNS323
 	bool "D-Link DNS-323"
-	select GENERIC_NET_UTILS
 	select I2C_BOARDINFO if I2C
 	help
 	  Say 'Y' here if you want your kernel to support the

@@ -65,7 +64,6 @@ config MACH_DNS323
 
 config MACH_TS209
 	bool "QNAP TS-109/TS-209"
-	select GENERIC_NET_UTILS
 	help
 	  Say 'Y' here if you want your kernel to support the
 	  QNAP TS-109/TS-209 platform.

@@ -107,7 +105,6 @@ config MACH_LINKSTATION_LS_HGL
 
 config MACH_TS409
 	bool "QNAP TS-409"
-	select GENERIC_NET_UTILS
 	help
 	  Say 'Y' here if you want your kernel to support the
 	  QNAP TS-409 platform.

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

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

From: Ramon Fried <[email protected]>

[ Upstream commit c77f5fbbefc04612755117775e8555c2a7006cac ]

Added MODULE_ALIAS("rpmsg:IPCRTR") to ensure qrtr-smd and qrtr will load when IPCRTR channel is detected.

---
 net/qrtr/smd.c |    1 +
 1 file changed, 1 insertion(+)

@@ -116,5 +116,6 @@ static struct qcom_smd_driver qcom_smd_q
 
 module_qcom_smd_driver(qcom_smd_qrtr_driver);
 
+MODULE_ALIAS("rpmsg:IPCRTR");
 MODULE_DESCRIPTION("Qualcomm IPC-Router SMD interface driver");
 MODULE_LICENSE("GPL v2");

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

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

From: Eric Dumazet <[email protected]>

[ Upstream commit 4c27bf3c5b7434ccb9ab962301da661c26b467a4 ]

r8152 driver handles TSO packets (limited to ~16KB) quite well, but pretends each TSO logical packet is a single packet on the wire.

There is also some error since headers are accounted once, but error rate is small enough that we do not care.

---
 drivers/net/usb/r8152.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1693,7 +1693,7 @@ static int r8152_tx_agg_fill(struct r815
 
 		tx_data += len;
 		agg->skb_len += len;
-		agg->skb_num++;
+		agg->skb_num += skb_shinfo(skb)->gso_segs ?: 1;
 
 		dev_kfree_skb_any(skb);
 

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

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

From: Dave Airlie <[email protected]>

[ Upstream commit 9a191b114906457c4b2494c474f58ae4142d4e67 ]

This exposes to mesa that it can use the fixed ioctl for querying later cap sets, cap set 1 is forever frozen in time.

---
 drivers/gpu/drm/virtio/virtgpu_ioctl.c |   17 +++++++++++------
 include/uapi/drm/virtgpu_drm.h         |    1 +
 2 files changed, 12 insertions(+), 6 deletions(-)

@@ -194,6 +194,9 @@ static int virtio_gpu_getparam_ioctl(str
 	case VIRTGPU_PARAM_3D_FEATURES:
 		value = vgdev->has_virgl_3d == true ? 1 : 0;
 		break;
+	case VIRTGPU_PARAM_CAPSET_QUERY_FIX:
+		value = 1;
+		break;
 	default:
 		return -EINVAL;
 	}

@@ -469,7 +472,7 @@ static int virtio_gpu_get_caps_ioctl(str
 {
 	struct virtio_gpu_device *vgdev = dev->dev_private;
 	struct drm_virtgpu_get_caps *args = data;
-	int size;
+	unsigned size, host_caps_size;
 	int i;
 	int found_valid = -1;
 	int ret;

@@ -478,6 +481,10 @@ static int virtio_gpu_get_caps_ioctl(str
 	if (vgdev->num_capsets == 0)
 		return -ENOSYS;
 
+	/* don't allow userspace to pass 0 */
+	if (args->size == 0)
+		return -EINVAL;
+
 	spin_lock(&vgdev->display_info_lock);
 	for (i = 0; i < vgdev->num_capsets; i++) {
 		if (vgdev->capsets[i].id == args->cap_set_id) {

@@ -493,11 +500,9 @@ static int virtio_gpu_get_caps_ioctl(str
 		return -EINVAL;
 	}
 
-	size = vgdev->capsets[found_valid].max_size;
-	if (args->size > size) {
-		spin_unlock(&vgdev->display_info_lock);
-		return -EINVAL;
-	}
+	host_caps_size = vgdev->capsets[found_valid].max_size;
+	/* only copy to user the minimum of the host caps size or the guest caps size */
+	size = min(args->size, host_caps_size);
 
 	list_for_each_entry(cache_ent, &vgdev->cap_cache, head) {
 		if (cache_ent->id == args->cap_set_id &&

@@ -63,6 +63,7 @@ struct drm_virtgpu_execbuffer {
 };
 
 #define VIRTGPU_PARAM_3D_FEATURES 1 /* do we have 3D features in the hw */
+#define VIRTGPU_PARAM_CAPSET_QUERY_FIX 2 /* do we have the capset fix */
 
 struct drm_virtgpu_getparam {
 	__u64 param;

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

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

From: Yoshihiro Shimoda <[email protected]>

[ Upstream commit d716d9b702bb759dd6fb50804f10a174bd156d71 ]

According to R-Car Gen3 Rev.0.80 manual, the DMATCR can be set to 16,777,215 as maximum. So, this patch fixes the max_chunk_size for safety on all of SoCs. Otherwise, a system may hang if the DMATCR is set to 0 on R-Car Gen3.

---
 drivers/dma/sh/rcar-dmac.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -870,7 +870,7 @@ rcar_dmac_chan_prep_sg(struct rcar_dmac_
 
 	rcar_dmac_chan_configure_desc(chan, desc);
 
-	max_chunk_size = (RCAR_DMATCR_MASK + 1) << desc->xfer_shift;
+	max_chunk_size = RCAR_DMATCR_MASK << desc->xfer_shift;
 
 	/*
 	 * Allocate and fill the transfer chunk descriptors. We own the only

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

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

From: Tang Junhui <[email protected]>

[ Upstream commit 60eb34ec5526e264c2bbaea4f7512d714d791caf ]

Kernel crashed when run fio in a RAID5 backend bcache device, the call trace is bellow:

[  440.012034] kernel BUG at block/blk-ioc.c:146!
[  440.012696] invalid opcode: 0000 [#1] SMP NOPTI
[  440.026537] CPU: 2 PID: 2205 Comm: md127_raid5 Not tainted 4.15.0 #8
[  440.027441] Hardware name: HP ProLiant MicroServer Gen8, BIOS J06 07/16
/2015
[  440.028615] RIP: 0010:put_io_context+0x8b/0x90
[  440.029246] RSP: 0018:ffffa8c882b43af8 EFLAGS: 00010246
[  440.029990] RAX: 0000000000000000 RBX: ffffa8c88294fca0 RCX: 0000000000
0f4240
[  440.031006] RDX: 0000000000000004 RSI: 0000000000000286 RDI: ffffa8c882
94fca0
[  440.032030] RBP: ffffa8c882b43b10 R08: 0000000000000003 R09: ffff949cb8
0c1700
[  440.033206] R10: 0000000000000104 R11: 000000000000b71c R12: 00000000000
01000
[  440.034222] R13: 0000000000000000 R14: ffff949cad84db70 R15: ffff949cb11
bd1e0
[  440.035239] FS:  0000000000000000(0000) GS:ffff949cba280000(0000) knlGS:
0000000000000000
[  440.060190] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  440.084967] CR2: 00007ff0493ef000 CR3: 00000002f1e0a002 CR4: 00000000001
606e0
[  440.110498] Call Trace:
[  440.135443]  bio_disassociate_task+0x1b/0x60
[  440.160355]  bio_free+0x1b/0x60
[  440.184666]  bio_put+0x23/0x30
[  440.208272]  search_free+0x23/0x40 [bcache]
[  440.231448]  cached_dev_write_complete+0x31/0x70 [bcache]
[  440.254468]  closure_put+0xb6/0xd0 [bcache]
[  440.277087]  request_endio+0x30/0x40 [bcache]
[  440.298703]  bio_endio+0xa1/0x120
[  440.319644]  handle_stripe+0x418/0x2270 [raid456]
[  440.340614]  ? load_balance+0x17b/0x9c0
[  440.360506]  handle_active_stripes.isra.58+0x387/0x5a0 [raid456]
[  440.380675]  ? __release_stripe+0x15/0x20 [raid456]
[  440.400132]  raid5d+0x3ed/0x5d0 [raid456]
[  440.419193]  ? schedule+0x36/0x80
[  440.437932]  ? schedule_timeout+0x1d2/0x2f0
[  440.456136]  md_thread+0x122/0x150
[  440.473687]  ? wait_woken+0x80/0x80
[  440.491411]  kthread+0x102/0x140
[  440.508636]  ? find_pers+0x70/0x70
[  440.524927]  ? kthread_associate_blkcg+0xa0/0xa0
[  440.541791]  ret_from_fork+0x35/0x40
[  440.558020] Code: c2 48 00 5b 41 5c 41 5d 5d c3 48 89 c6 4c 89 e7 e8 bb c2
48 00 48 8b 3d bc 36 4b 01 48 89 de e8 7c f7 e0 ff 5b 41 5c 41 5d 5d c3 <0f> 0b
0f 1f 00 0f 1f 44 00 00 55 48 8d 47 b8 48 89 e5 41 57 41
[  440.610020] RIP: put_io_context+0x8b/0x90 RSP: ffffa8c882b43af8
[  440.628575] ---[ end trace a1fd79d85643a73e ]--

All the crash issue happened when a bypass IO coming, in such scenario s->iop.bio is pointed to the s->orig_bio. In search_free(), it finishes the s->orig_bio by calling bio_complete(), and after that, s->iop.bio became invalid, then kernel would crash when calling bio_put(). Maybe its upper layer's faulty, since bio should not be freed before we calling bio_put(), but we'd better calling bio_put() first before calling bio_complete() to notify upper layer ending this bio.

This patch moves bio_complete() under bio_put() to avoid kernel crash.

[mlyle: fixed commit subject for character limits]

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

@@ -633,11 +633,11 @@ static void do_bio_hook(struct search *s
 static void search_free(struct closure *cl)
 {
 	struct search *s = container_of(cl, struct search, cl);
-	bio_complete(s);
 
 	if (s->iop.bio)
 		bio_put(s->iop.bio);
 
+	bio_complete(s);
 	closure_debug_destroy(cl);
 	mempool_free(s, s->d->c->search);
 }

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

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

From: Xin Long <[email protected]>

[ Upstream commit a6aa80446234ec0ad38eecdb8efc59e91daae565 ]

Commit 128bb975dc3c ("ip6_gre: init dev->mtu and dev->hard_header_len correctly") fixed IFLA_MTU ignored on NEWLINK for ip6_gre. The same mtu fix is also needed for ip6_tunnel.

Note that dev->hard_header_len setting for ip6_tunnel works fine, no need to fix it.

---
 net/ipv6/ip6_tunnel.c |   12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

@@ -1972,14 +1972,14 @@ static int ip6_tnl_newlink(struct net *s
 {
 	struct net *net = dev_net(dev);
 	struct ip6_tnl_net *ip6n = net_generic(net, ip6_tnl_net_id);
-	struct ip6_tnl *nt, *t;
 	struct ip_tunnel_encap ipencap;
+	struct ip6_tnl *nt, *t;
+	int err;
 
 	nt = netdev_priv(dev);
 
 	if (ip6_tnl_netlink_encap_parms(data, &ipencap)) {
-		int err = ip6_tnl_encap_setup(nt, &ipencap);
-
+		err = ip6_tnl_encap_setup(nt, &ipencap);
 		if (err < 0)
 			return err;
 	}

@@ -1995,7 +1995,11 @@ static int ip6_tnl_newlink(struct net *s
 			return -EEXIST;
 	}
 
-	return ip6_tnl_create2(dev);
+	err = ip6_tnl_create2(dev);
+	if (!err && tb[IFLA_MTU])
+		ip6_tnl_change_mtu(dev, nla_get_u32(tb[IFLA_MTU]));
+
+	return err;
 }
 
 static int ip6_tnl_changelink(struct net_device *dev, struct nlattr *tb[],

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

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

From: Xin Long <[email protected]>

[ Upstream commit 2b3957c34b6d7f03544b12ebbf875eee430745db ]

Commit 128bb975dc3c ("ip6_gre: init dev->mtu and dev->hard_header_len correctly") fixed IFLA_MTU ignored on NEWLINK for ip6_gre. The same mtu fix is also needed for sit.

Note that dev->hard_header_len setting for sit works fine, no need to fix it. sit is actually ipv4 tunnel, it can't call ip6_tnl_change_mtu to set mtu.

---
 net/ipv6/sit.c |    7 +++++++
 1 file changed, 7 insertions(+)

@@ -1572,6 +1572,13 @@ static int ipip6_newlink(struct net *src
 	if (err < 0)
 		return err;
 
+	if (tb[IFLA_MTU]) {
+		u32 mtu = nla_get_u32(tb[IFLA_MTU]);
+
+		if (mtu >= IPV6_MIN_MTU && mtu <= 0xFFF8 - dev->hard_header_len)
+			dev->mtu = mtu;
+	}
+
 #ifdef CONFIG_IPV6_SIT_6RD
 	if (ipip6_netlink_6rd_parms(data, &ip6rd))
 		err = ipip6_tunnel_update_6rd(nt, &ip6rd);

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

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

From: Florian Fainelli <[email protected]>

[ Upstream commit 0a5aff64f20d92c5a6e9aeed7b5950b0b817bcd9 ]

Jon attempted to fix the amount of RAM on the BCM958625HR in commit c53beb47f621 ("ARM: dts: NSP: Correct RAM amount for BCM958625HR board") but it seems like we tripped over some poorly documented schematics.

The top-level page of the schematics says the board has 2GB, but when you end-up scrolling to page 6, you see two chips of 4GBit (512MB) but what the bootloader really initializes only 512MB, any attempt to use more than that results in data aborts. Fix this again back to 512MB.

---
 arch/arm/boot/dts/bcm958625hr.dts |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -49,7 +49,7 @@
 
 	memory {
 		device_type = "memory";
-		reg = <0x60000000 0x80000000>;
+		reg = <0x60000000 0x20000000>;
 	};
 
 	gpio-restart {

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

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

From: Guenter Roeck <[email protected]>

[ Upstream commit 64c3f648c25d108f346fdc96c15180c6b7d250e9 ]

Once in a while I see build errors similar to the following when building images from a clean tree.

Building powerpc:virtex-ml507:44x/virtex5_defconfig ... failed
------------
Error log:
arch/powerpc/boot/treeboot-akebono.c:37:20: fatal error:
libfdt.h: No such file or directory

Building powerpc:bamboo:smpdev:44x/bamboo_defconfig ... failed
------------
Error log:
arch/powerpc/boot/treeboot-akebono.c:37:20: fatal error:
libfdt.h: No such file or directory

arch/powerpc/boot/treeboot-currituck.c:35:20: fatal error:
libfdt.h: No such file or directory

Rebuilds will succeed.

Turns out that several source files in arch/powerpc/boot/ include libfdt.h, but Makefile dependencies are incomplete. Let's fix that.

---
 arch/powerpc/boot/Makefile |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -92,7 +92,8 @@ $(addprefix $(obj)/,$(zlib-y)): \
 libfdt       := fdt.c fdt_ro.c fdt_wip.c fdt_sw.c fdt_rw.c fdt_strerror.c
 libfdtheader := fdt.h libfdt.h libfdt_internal.h
 
-$(addprefix $(obj)/,$(libfdt) libfdt-wrapper.o simpleboot.o epapr.o opal.o): \
+$(addprefix $(obj)/,$(libfdt) libfdt-wrapper.o simpleboot.o epapr.o opal.o \
+	treeboot-akebono.o treeboot-currituck.o treeboot-iss4xx.o): \
 	$(addprefix $(obj)/,$(libfdtheader))
 
 src-wlib-y := string.S crt0.S crtsavres.S stdio.c decompress.c main.c \

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

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

From: Claudiu Manoil <[email protected]>

[ Upstream commit 590399ddf9561f2ed0839311c8ae1be21597ba68 ]

Don't include in the Rx bytecount of the packet sent up the stack:
the FCB (frame control block), and the padding bytes inserted by the controller into the frame payload, nor the FCS. All these are being pulled out of the skb by gfar_process_frame().
This issue is old, likely from the driver's beginnings, however it was amplified by recent:
commit d903ec77118c ("gianfar: simplify FCS handling and fix memory leak") which basically added the FCS to the Rx bytecount, and so brought this to my attention.

---
 drivers/net/ethernet/freescale/gianfar.c |    7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

@@ -3075,9 +3075,6 @@ static void gfar_process_frame(struct ne
 	if (ndev->features & NETIF_F_RXCSUM)
 		gfar_rx_checksum(skb, fcb);
 
-	/* Tell the skb what kind of packet this is */
-	skb->protocol = eth_type_trans(skb, ndev);
-
 	/* There's need to check for NETIF_F_HW_VLAN_CTAG_RX here.
 	 * Even if vlan rx accel is disabled, on some chips
 	 * RXFCB_VLN is pseudo randomly set.

@@ -3148,13 +3145,15 @@ int gfar_clean_rx_ring(struct gfar_priv_
 			continue;
 		}
 
+		gfar_process_frame(ndev, skb);
+
 		/* Increment the number of packets */
 		total_pkts++;
 		total_bytes += skb->len;
 
 		skb_record_rx_queue(skb, rx_queue->qindex);
 
-		gfar_process_frame(ndev, skb);
+		skb->protocol = eth_type_trans(skb, ndev);
 
 		/* Send the packet up the stack */
 		napi_gro_receive(&rx_queue->grp->napi_rx, skb);

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

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

From: Joey Pabalinas <[email protected]>

[ Upstream commit ecc832758a654e375924ebf06a4ac971acb5ce60 ]

The link to the pdf containing the algorithm description is now a dead link; it seems http://www.ifp.illinois.edu/~srikant/ has been moved to https://sites.google.com/a/illinois.edu/srikant/ and none of the original papers can be found there...

I have replaced it with the only working copy I was able to find.

n.b. there is also a copy available at:

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.296.6350&rep=rep1&type=pdf

However, this seems to only be a *cached* version, so I am unsure exactly how reliable that link can be expected to remain over time and have decided against using that one.

 net/ipv4/tcp_illinois.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

---
 net/ipv4/tcp_illinois.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -6,7 +6,7 @@
  * The algorithm is described in:
  * "TCP-Illinois: A Loss and Delay-Based Congestion Control Algorithm
  *  for High-Speed Networks"
- * http://www.ifp.illinois.edu/~srikant/Papers/liubassri06perf.pdf
+ * http://tamerbasar.csl.illinois.edu/LiuBasarSrikantPerfEvalArtJun2008.pdf
  *
  * Implemented from description in paper and ns-2 simulation.
  * Copyright (C) 2007 Stephen Hemminger <[email protected]>

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

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

From: Max Gurtovoy <[email protected]>

[ Upstream commit bffd2b61670feef18d2535e9b53364d270a1c991 ]

PSDT field section according to NVM_Express-1.3:
"This field specifies whether PRPs or SGLs are used for any data transfer associated with the command. PRPs shall be used for all Admin commands for NVMe over PCIe. SGLs shall be used for all Admin and I/O commands for NVMe over Fabrics. This field shall be set to 01b for NVMe over Fabrics 1.0 implementations.

---
 drivers/nvme/target/core.c |    9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

@@ -491,9 +491,12 @@ bool nvmet_req_init(struct nvmet_req *re
 		goto fail;
 	}
 
-	/* either variant of SGLs is fine, as we don't support metadata */
-	if (unlikely((flags & NVME_CMD_SGL_ALL) != NVME_CMD_SGL_METABUF &&
-		     (flags & NVME_CMD_SGL_ALL) != NVME_CMD_SGL_METASEG)) {
+	/*
+	 * For fabrics, PSDT field shall describe metadata pointer (MPTR) that
+	 * contains an address of a single contiguous physical buffer that is
+	 * byte aligned.
+	 */
+	if (unlikely((flags & NVME_CMD_SGL_ALL) != NVME_CMD_SGL_METABUF)) {
 		status = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
 		goto fail;
 	}

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

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

From: Giulio Benetti <[email protected]>

[ Upstream commit e64b6afa98f3629d0c0c46233bbdbe8acdb56f06 ]

Phase value is not shifted before writing.

Shift left of 28 bits to fit right bits

---
 drivers/gpu/drm/sun4i/sun4i_dotclock.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

@@ -129,10 +129,13 @@ static int sun4i_dclk_get_phase(struct c
 static int sun4i_dclk_set_phase(struct clk_hw *hw, int degrees)
 {
 	struct sun4i_dclk *dclk = hw_to_dclk(hw);
+	u32 val = degrees / 120;
+
+	val <<= 28;
 
 	regmap_update_bits(dclk->regmap, SUN4I_TCON0_IO_POL_REG,
 			   GENMASK(29, 28),
-			   degrees / 120);
+			   val);
 
 	return 0;
 }

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

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

From: Filipe Manana <[email protected]>

[ Upstream commit d4dfc0f4d39475ccbbac947880b5464a74c30b99 ]

When doing an incremental send of a filesystem with the no-holes feature enabled, we end up issuing a write operation when using the no data mode send flag, instead of issuing an update extent operation. Fix this by issuing the update extent operation instead.

Trivial reproducer:

  $ mkfs.btrfs -f -O no-holes /dev/sdc
  $ mkfs.btrfs -f /dev/sdd
  $ mount /dev/sdc /mnt/sdc
  $ mount /dev/sdd /mnt/sdd

  $ xfs_io -f -c "pwrite -S 0xab 0 32K" /mnt/sdc/foobar
  $ btrfs subvolume snapshot -r /mnt/sdc /mnt/sdc/snap1

  $ xfs_io -c "fpunch 8K 8K" /mnt/sdc/foobar
  $ btrfs subvolume snapshot -r /mnt/sdc /mnt/sdc/snap2

  $ btrfs send /mnt/sdc/snap1 | btrfs receive /mnt/sdd
  $ btrfs send --no-data -p /mnt/sdc/snap1 /mnt/sdc/snap2 \
       | btrfs receive -vv /mnt/sdd

Before this change the output of the second receive command is:

receiving snapshot snap2 uuid=f6922049-8c22-e544-9ff9-fc6755918447... utimes
write foobar, offset 8192, len 8192
utimes foobar
BTRFS_IOC_SET_RECEIVED_SUBVOL uuid=f6922049-8c22-e544-9ff9-...

After this change it is:

receiving snapshot snap2 uuid=564d36a3-ebc8-7343-aec9-bf6fda278e64... utimes
update_extent foobar: offset=8192, len=8192
utimes foobar
BTRFS_IOC_SET_RECEIVED_SUBVOL uuid=564d36a3-ebc8-7343-aec9-bf6fda278e64...

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

@@ -4822,6 +4822,9 @@ static int send_hole(struct send_ctx *sc
 	u64 len;
 	int ret = 0;
 
+	if (sctx->flags & BTRFS_SEND_FLAG_NO_FILE_DATA)
+		return send_update_extent(sctx, offset, end - offset);
+
 	p = fs_path_alloc();
 	if (!p)
 		return -ENOMEM;

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

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

From: Michael Ellerman <[email protected]>

[ Upstream commit cd4a6f3ab4d80cb919d15897eb3cbc85c2009d4b ]

The subpage_prot syscall is only functional when the system is using the Hash MMU. Since commit 5b2b80714796 ("powerpc/mm: Invalidate subpage_prot() system call on radix platforms") it returns ENOENT when the Radix MMU is active. Currently this just makes the test fail.

Additionally the syscall is not available if the kernel is built with 4K pages, or if CONFIG_PPC_SUBPAGE_PROT=n, in which case it returns ENOSYS because the syscall is missing entirely.

So check explicitly for ENOENT and ENOSYS and skip if we see either of those.

---
 tools/testing/selftests/powerpc/mm/subpage_prot.c |   14 ++++++++++++++
 1 file changed, 14 insertions(+)

@@ -135,6 +135,16 @@ static int run_test(void *addr, unsigned
 	return 0;
 }
 
+static int syscall_available(void)
+{
+	int rc;
+
+	errno = 0;
+	rc = syscall(__NR_subpage_prot, 0, 0, 0);
+
+	return rc == 0 || (errno != ENOENT && errno != ENOSYS);
+}
+
 int test_anon(void)
 {
 	unsigned long align;

@@ -145,6 +155,8 @@ int test_anon(void)
 	void *mallocblock;
 	unsigned long mallocsize;
 
+	SKIP_IF(!syscall_available());
+
 	if (getpagesize() != 0x10000) {
 		fprintf(stderr, "Kernel page size must be 64K!\n");
 		return 1;

@@ -180,6 +192,8 @@ int test_file(void)
 	off_t filesize;
 	int fd;
 
+	SKIP_IF(!syscall_available());
+
 	fd = open(file_name, O_RDWR);
 	if (fd == -1) {
 		perror("failed to open file");

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

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

From: Paul Mackerras <[email protected]>

[ Upstream commit debd574f4195e205ba505b25e19b2b797f4bcd94 ]

The current code for initializing the VRMA (virtual real memory area) for HPT guests requires the page size of the backing memory to be one of 4kB, 64kB or 16MB. With a radix host we have the possibility that the backing memory page size can be 2MB or 1GB. In these cases, if the guest switches to HPT mode, KVM will not initialize the VRMA and the guest will fail to run.

In fact it is not necessary that the VRMA page size is the same as the backing memory page size; any VRMA page size less than or equal to the backing memory page size is acceptable. Therefore we now choose the largest page size out of the set {4k, 64k, 16M} which is not larger than the backing memory page size.

---
 arch/powerpc/kvm/book3s_hv.c |   12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)

@@ -3107,15 +3107,17 @@ static int kvmppc_hv_setup_htab_rma(stru
 		goto up_out;
 
 	psize = vma_kernel_pagesize(vma);
-	porder = __ilog2(psize);
 
 	up_read(&current->mm->mmap_sem);
 
 	/* We can handle 4k, 64k or 16M pages in the VRMA */
-	err = -EINVAL;
-	if (!(psize == 0x1000 || psize == 0x10000 ||
-	      psize == 0x1000000))
-		goto out_srcu;
+	if (psize >= 0x1000000)
+		psize = 0x1000000;
+	else if (psize >= 0x10000)
+		psize = 0x10000;
+	else
+		psize = 0x1000;
+	porder = __ilog2(psize);
 
 	/* Update VRMASD field in the LPCR */
 	senc = slb_pgsize_encoding(psize);

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

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

From: Sara Sharon <[email protected]>

[ Upstream commit de04d4fbf87b769ab18c480e4f020c53e74bbdd2 ]

We don't have enough room in the TX command for a CCMP 256 key, and need to use key from table.

---
 drivers/net/wireless/intel/iwlwifi/mvm/tx.c |    7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

@@ -406,11 +406,11 @@ static void iwl_mvm_set_tx_cmd_crypto(st
 {
 	struct ieee80211_key_conf *keyconf = info->control.hw_key;
 	u8 *crypto_hdr = skb_frag->data + hdrlen;
+	enum iwl_tx_cmd_sec_ctrl type = TX_CMD_SEC_CCM;
 	u64 pn;
 
 	switch (keyconf->cipher) {
 	case WLAN_CIPHER_SUITE_CCMP:
-	case WLAN_CIPHER_SUITE_CCMP_256:
 		iwl_mvm_set_tx_cmd_ccmp(info, tx_cmd);
 		iwl_mvm_set_tx_cmd_pn(info, crypto_hdr);
 		break;

@@ -434,13 +434,16 @@ static void iwl_mvm_set_tx_cmd_crypto(st
 		break;
 	case WLAN_CIPHER_SUITE_GCMP:
 	case WLAN_CIPHER_SUITE_GCMP_256:
+		type = TX_CMD_SEC_GCMP;
+		/* Fall through */
+	case WLAN_CIPHER_SUITE_CCMP_256:
 		/* TODO: Taking the key from the table might introduce a race
 		 * when PTK rekeying is done, having an old packets with a PN
 		 * based on the old key but the message encrypted with a new
 		 * one.
 		 * Need to handle this.
 		 */
-		tx_cmd->sec_ctl |= TX_CMD_SEC_GCMP | TX_CMD_SEC_KEY_FROM_TABLE;
+		tx_cmd->sec_ctl |= type | TX_CMD_SEC_KEY_FROM_TABLE;
 		tx_cmd->key[0] = keyconf->hw_key_idx;
 		iwl_mvm_set_tx_cmd_pn(info, crypto_hdr);
 		break;

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

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

From: Igor Pylypiv <[email protected]>

[ Upstream commit 7bd3e7b743956afbec30fb525bc3c5e22e3d475c ]

Watchdog close is "expected" when any byte is 'V' not just the last one. Writing "V" to the device fails because the last byte is the end of string.

$ echo V > /dev/watchdog
f71808e_wdt: Unexpected close, not stopping watchdog!

---
 drivers/watchdog/f71808e_wdt.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -566,7 +566,8 @@ static ssize_t watchdog_write(struct fil
 				char c;
 				if (get_user(c, buf + i))
 					return -EFAULT;
-				expect_close = (c == 'V');
+				if (c == 'V')
+					expect_close = true;
 			}
 
 			/* Properly order writes across fork()ed processes */

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

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

From: Jayachandran C <[email protected]>

[ Upstream commit 93ac3deb7c220cbcec032a967220a1f109d58431 ]

According to SBSA spec v3.1 section 5.3:
All registers are 32 bits in size and should be accessed using 32-bit reads and writes. If an access size other than 32 bits is used then the results are IMPLEMENTATION DEFINED. [...]
The Generic Watchdog is little-endian

The current code uses readq to read the watchdog compare register which does a 64-bit access. This fails on ThunderX2 which does not implement 64-bit access to this register.

Fix this by using lo_hi_readq() that does two 32-bit reads.

---
 drivers/watchdog/sbsa_gwdt.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -50,6 +50,7 @@
  */
 
 #include <linux/io.h>
+#include <linux/io-64-nonatomic-lo-hi.h>
 #include <linux/interrupt.h>
 #include <linux/module.h>
 #include <linux/moduleparam.h>

@@ -159,7 +160,7 @@ static unsigned int sbsa_gwdt_get_timele
 	    !(readl(gwdt->control_base + SBSA_GWDT_WCS) & SBSA_GWDT_WCS_WS0))
 		timeleft += readl(gwdt->control_base + SBSA_GWDT_WOR);
 
-	timeleft += readq(gwdt->control_base + SBSA_GWDT_WCV) -
+	timeleft += lo_hi_readq(gwdt->control_base + SBSA_GWDT_WCV) -
 		    arch_counter_get_cntvct();
 
 	do_div(timeleft, gwdt->clk);

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

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

From: Benjamin Poirier <[email protected]>

[ Upstream commit 4e7dc08e57c95673d2edaba8983c3de4dd1f65f5 ]

When autoneg is off, the .check_for_link callback functions clear the get_link_status flag and systematically return a "pseudo-error". This means that the link is not detected as up until the next execution of the e1000_watchdog_task() 2 seconds later.

---
 drivers/net/ethernet/intel/e1000e/ich8lan.c |    2 +-
 drivers/net/ethernet/intel/e1000e/mac.c     |    2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

@@ -1600,7 +1600,7 @@ static s32 e1000_check_for_copper_link_i
 	 * we have already determined whether we have link or not.
 	 */
 	if (!mac->autoneg)
-		return -E1000_ERR_CONFIG;
+		return 1;
 
 	/* Auto-Neg is enabled.  Auto Speed Detection takes care
 	 * of MAC speed/duplex configuration.  So we only need to

@@ -450,7 +450,7 @@ s32 e1000e_check_for_copper_link(struct
 	 * we have already determined whether we have link or not.
 	 */
 	if (!mac->autoneg)
-		return -E1000_ERR_CONFIG;
+		return 1;
 
 	/* Auto-Neg is enabled.  Auto Speed Detection takes care
 	 * of MAC speed/duplex configuration.  So we only need to

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

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

From: Pierre-Yves Kerbrat <[email protected]>

[ Upstream commit aea3fca005fb45f80869f2e8d56fd4e64c1d1fdb ]

Descriptor rings were not initialized at zero when allocated When area contained garbage data, it caused skb_over_panic in e1000_clean_rx_irq (if data had E1000_RXD_STAT_DD bit set)

This patch makes use of dma_zalloc_coherent to make sure the ring is memset at 0 to prevent the area from containing garbage.

Following is the signature of the panic:
[email protected]: skbuff: skb_over_panic: text:80407b20 len:64010 put:64010 head:ab46d800 data:ab46d842 tail:0xab47d24c end:0xab46df40 dev:eth0 [email protected]: BUG: failure at net/core/skbuff.c:105/skb_panic()!
[email protected]: Kernel panic - not syncing: BUG!
[email protected]:
[email protected]: Process swapper/0 (pid: 0, threadinfo=81728000, task=8173cc00 ,cpu: 0)
[email protected]: SP = <815a1c0c>
[email protected]: Stack: 00000001
[email protected]: b2d89800 815e33ac
[email protected]: ea73c040 00000001
[email protected]: 60040003 0000fa0a
[email protected]: 00000002
[email protected]:
[email protected]: 804540c0 815a1c70
[email protected]: b2744000 602ac070
[email protected]: 815a1c44 b2d89800
[email protected]: 8173cc00 815a1c08
[email protected]:
[email protected]: 00000006
[email protected]: 815a1b50 00000000
[email protected]: 80079434 00000001
[email protected]: ab46df40 b2744000
[email protected]: b2d89800
[email protected]:
[email protected]: 0000fa0a 8045745c
[email protected]: 815a1c88 0000fa0a
[email protected]: 80407b20 b2789f80
[email protected]: 00000005 80407b20
[email protected]:
[email protected]:
[email protected]: Call Trace:
[email protected]: [<804540bc>] skb_panic+0xa4/0xa8
[email protected]: [<80079430>] console_unlock+0x2f8/0x6d0
[email protected]: [<80457458>] skb_put+0xa0/0xc0
[email protected]: [<80407b1c>] e1000_clean_rx_irq+0x2dc/0x3e8
[email protected]: [<80407b1c>] e1000_clean_rx_irq+0x2dc/0x3e8
[email protected]: [<804079c8>] e1000_clean_rx_irq+0x188/0x3e8
[email protected]: [<80407b1c>] e1000_clean_rx_irq+0x2dc/0x3e8
[email protected]: [<80468b48>] __dev_kfree_skb_any+0x88/0xa8
[email protected]: [<804101ac>] e1000e_poll+0x94/0x288
[email protected]: [<8046e9d4>] net_rx_action+0x19c/0x4e8
[email protected]: ...
[email protected]: Maximum depth to print reached. Use kstack=<maximum_depth_to_print> To specify a custom value (where 0 means to display the full backtrace) [email protected]: ---[ end Kernel panic - not syncing: BUG!

---
 drivers/net/ethernet/intel/e1000e/netdev.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -2331,8 +2331,8 @@ static int e1000_alloc_ring_dma(struct e
 {
 	struct pci_dev *pdev = adapter->pdev;
 
-	ring->desc = dma_alloc_coherent(&pdev->dev, ring->size, &ring->dma,
-					GFP_KERNEL);
+	ring->desc = dma_zalloc_coherent(&pdev->dev, ring->size, &ring->dma,
+					 GFP_KERNEL);
 	if (!ring->desc)
 		return -ENOMEM;
 

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

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

From: Davidlohr Bueso <[email protected]>

[ Upstream commit 69c907022a7d9325cdc5c9dd064571e445df9a47 ]

At the point of sysfs callback, the call to gup is done without mmap_sem (or any lock for that matter). This is racy. As such, use the get_user_pages_fast() alternative and safely avoid taking the lock, if possible.

---
 arch/ia64/kernel/err_inject.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -142,7 +142,7 @@ store_virtual_to_phys(struct device *dev
 	u64 virt_addr=simple_strtoull(buf, NULL, 16);
 	int ret;
 
-	ret = get_user_pages(virt_addr, 1, FOLL_WRITE, NULL, NULL);
+	ret = get_user_pages_fast(virt_addr, 1, FOLL_WRITE, NULL);
 	if (ret<=0) {
 #ifdef ERR_INJ_DEBUG
 		printk("Virtual address %lx is not existing.\n",virt_addr);

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

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

From: "Kalderon, Michal" <[email protected]>

[ Upstream commit e3fd112cbf21d049faf64ba1471d72b93c22109a ]

Race in qedr_poll_cq, lastest_cqe wasn't protected by lock, leading to a case where two context's accessing poll_cq at the same time lead to one of them having a pointer to an old latest_cqe and reading an invalid cqe element

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

@@ -3467,7 +3467,7 @@ int qedr_poll_cq(struct ib_cq *ibcq, int
 {
 	struct qedr_dev *dev = get_qedr_dev(ibcq->device);
 	struct qedr_cq *cq = get_qedr_cq(ibcq);
-	union rdma_cqe *cqe = cq->latest_cqe;
+	union rdma_cqe *cqe;
 	u32 old_cons, new_cons;
 	unsigned long flags;
 	int update = 0;

@@ -3477,6 +3477,7 @@ int qedr_poll_cq(struct ib_cq *ibcq, int
 		return qedr_gsi_poll_cq(ibcq, num_entries, wc);
 
 	spin_lock_irqsave(&cq->cq_lock, flags);
+	cqe = cq->latest_cqe;
 	old_cons = qed_chain_get_cons_idx_u32(&cq->pbl);
 	while (num_entries && is_valid_cqe(cq, cqe)) {
 		struct qedr_qp *qp;

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

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

From: "Kalderon, Michal" <[email protected]>

[ Upstream commit 551e1c67b4207455375a2e7a285dea1c7e8fc361 ]

iWARP does not support RDMA WRITE or SEND with immediate data. Driver should check this before submitting to FW and return an immediate error

---
 drivers/infiniband/hw/qedr/verbs.c |   10 ++++++++++
 1 file changed, 10 insertions(+)

@@ -2807,6 +2807,11 @@ int __qedr_post_send(struct ib_qp *ibqp,
 
 	switch (wr->opcode) {
 	case IB_WR_SEND_WITH_IMM:
+		if (unlikely(rdma_protocol_iwarp(&dev->ibdev, 1))) {
+			rc = -EINVAL;
+			*bad_wr = wr;
+			break;
+		}
 		wqe->req_type = RDMA_SQ_REQ_TYPE_SEND_WITH_IMM;
 		swqe = (struct rdma_sq_send_wqe_1st *)wqe;
 		swqe->wqe_size = 2;

@@ -2848,6 +2853,11 @@ int __qedr_post_send(struct ib_qp *ibqp,
 		break;
 
 	case IB_WR_RDMA_WRITE_WITH_IMM:
+		if (unlikely(rdma_protocol_iwarp(&dev->ibdev, 1))) {
+			rc = -EINVAL;
+			*bad_wr = wr;
+			break;
+		}
 		wqe->req_type = RDMA_SQ_REQ_TYPE_RDMA_WR_WITH_IMM;
 		rwqe = (struct rdma_sq_rdma_wqe_1st *)wqe;
 

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

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

From: Jack Morgenstein <[email protected]>

[ Upstream commit 0077416a3d529baccbe07ab3242e8db541cfadf6 ]

When using IPv4 addresses in RoCEv2, the GID format for the mapped IPv4 address should be: ::ffff:<4-byte IPv4 address>.

In the cited commit, IPv4 mapped IPV6 addresses had the 3 upper dwords zeroed out by memset, which resulted in deleting the ffff field.

However, since procedure ipv6_addr_v4mapped() already verifies that the gid has format ::ffff:<ipv4 address>, no change is needed for the gid, and the memset can simply be removed.

---
 drivers/infiniband/hw/mlx4/main.c |    2 --
 1 file changed, 2 deletions(-)

@@ -216,8 +216,6 @@ static int mlx4_ib_update_gids_v1_v2(str
 			gid_tbl[i].version = 2;
 			if (!ipv6_addr_v4mapped((struct in6_addr *)&gids[i].gid))
 				gid_tbl[i].type = 1;
-			else
-				memset(&gid_tbl[i].gid, 0, 12);
 		}
 	}
 

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

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

From: Jack M <[email protected]>

[ Upstream commit a18177925c252da7801149abe217c05b80884798 ]

The commit cited below added a gid_type field (RoCEv1 or RoCEv2) to GID properties.

When adding GIDs, this gid_type field was copied over to the hardware gid table. However, when deleting GIDs, the gid_type field was not copied over to the hardware gid table.

As a result, when running RoCEv2, all RoCEv2 gids in the hardware gid table were set to type RoCEv1 when any gid was deleted.

This problem would persist until the next gid was added (which would again restore the gid_type field for all the gids in the hardware gid table).

Fix this by copying over the gid_type field to the hardware gid table when deleting gids, so that the gid_type of all remaining gids is preserved when a gid is deleted.

---
 drivers/infiniband/hw/mlx4/main.c |    9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

@@ -361,8 +361,13 @@ static int mlx4_ib_del_gid(struct ib_dev
 		if (!gids) {
 			ret = -ENOMEM;
 		} else {
-			for (i = 0; i < MLX4_MAX_PORT_GIDS; i++)
-				memcpy(&gids[i].gid, &port_gid_table->gids[i].gid, sizeof(union ib_gid));
+			for (i = 0; i < MLX4_MAX_PORT_GIDS; i++) {
+				memcpy(&gids[i].gid,
+				       &port_gid_table->gids[i].gid,
+				       sizeof(union ib_gid));
+				gids[i].gid_type =
+				    port_gid_table->gids[i].gid_type;
+			}
 		}
 	}
 	spin_unlock_bh(&iboe->lock);

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

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

From: Dan Carpenter <[email protected]>

[ Upstream commit 5d414b178e950ce9685c253994cc730893d5d887 ]

"err" is either zero or possibly uninitialized here. It should be -EINVAL.

---
 drivers/infiniband/hw/mlx5/qp.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

@@ -2809,8 +2809,10 @@ static int __mlx5_ib_modify_qp(struct ib
 		mlx5_ib_qp_disable_pagefaults(qp);
 
 	if (mlx5_cur >= MLX5_QP_NUM_STATE || mlx5_new >= MLX5_QP_NUM_STATE ||
-	    !optab[mlx5_cur][mlx5_new])
+	    !optab[mlx5_cur][mlx5_new]) {
+		err = -EINVAL;
 		goto out;
+	}
 
 	op = optab[mlx5_cur][mlx5_new];
 	optpar = ib_mask_to_mlx5_opt(attr_mask);

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

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

From: Peter Malone <[email protected]>

[ Upstream commit 250c6c49e3b68756b14983c076183568636e2bde ]

Fixing arbitrary kernel leak in case FBIOGETCMAP_SPARC in sbusfb_ioctl_helper().

'index' is defined as an int in sbusfb_ioctl_helper(). We retrieve this from the user:
if (get_user(index, &c->index) ||

    __get_user(count, &c->count) ||
    __get_user(ured, &c->red) ||
    __get_user(ugreen, &c->green) ||
    __get_user(ublue, &c->blue))
       return -EFAULT;

and then we use 'index' in the following way:
red = cmap->red[index + i] >> 8;
green = cmap->green[index + i] >> 8;
blue = cmap->blue[index + i] >> 8;

This is a classic information leak vulnerability. 'index' should be an unsigned int, given its usage above.

This patch is straight-forward; it changes 'index' to unsigned int in two switch-cases: FBIOGETCMAP_SPARC && FBIOPUTCMAP_SPARC.

This patch fixes CVE-2018-6412.

---
 drivers/video/fbdev/sbuslib.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -121,7 +121,7 @@ int sbusfb_ioctl_helper(unsigned long cm
 		unsigned char __user *ured;
 		unsigned char __user *ugreen;
 		unsigned char __user *ublue;
-		int index, count, i;
+		unsigned int index, count, i;
 
 		if (get_user(index, &c->index) ||
 		    __get_user(count, &c->count) ||

@@ -160,7 +160,7 @@ int sbusfb_ioctl_helper(unsigned long cm
 		unsigned char __user *ugreen;
 		unsigned char __user *ublue;
 		struct fb_cmap *cmap = &info->cmap;
-		int index, count, i;
+		unsigned int index, count, i;
 		u8 red, green, blue;
 
 		if (get_user(index, &c->index) ||

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

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

From: Denis Kirjanov <[email protected]>

[ Upstream commit 803fafbe0cd522fa6b9e41ca3b96cfb2e2a2222d ]

__dev_mc_add grabs an adress spinlock so use
atomic context in kmalloc.

/ # ifconfig eth0 inet 192.168.0.111

[   89.331622] BUG: sleeping function called from invalid context at mm/slab.h:420
[   89.339002] in_atomic(): 1, irqs_disabled(): 0, pid: 1035, name: ifconfig
[   89.345799] 2 locks held by ifconfig/1035:
[   89.349908]  #0:  (rtnl_mutex){+.+.}, at: [<(ptrval)>] devinet_ioctl+0xc0/0x8a0
[   89.357258]  #1:  (_xmit_ETHER){+...}, at: [<(ptrval)>] __dev_mc_add+0x28/0x80
[   89.364520] CPU: 1 PID: 1035 Comm: ifconfig Not tainted 4.16.0-rc3-dirty #8
[   89.371464] Call Trace:
[   89.373908] [e959db60] [c066f948] dump_stack+0xa4/0xfc (unreliable)
[   89.380177] [e959db80] [c00671d8] ___might_sleep+0x248/0x280
[   89.385833] [e959dba0] [c01aec34] kmem_cache_alloc_trace+0x174/0x320
[   89.392179] [e959dbd0] [c04ab920] dtsec_add_hash_mac_address+0x130/0x240
[   89.398874] [e959dc00] [c04a9d74] set_multi+0x174/0x1b0
[   89.404093] [e959dc30] [c04afb68] dpaa_set_rx_mode+0x68/0xe0
[   89.409745] [e959dc40] [c057baf8] __dev_mc_add+0x58/0x80
[   89.415052] [e959dc60] [c060fd64] igmp_group_added+0x164/0x190
[   89.420878] [e959dca0] [c060ffa8] ip_mc_inc_group+0x218/0x460
[   89.426617] [e959dce0] [c06120fc] ip_mc_up+0x3c/0x190
[   89.431662] [e959dd10] [c0607270] inetdev_event+0x250/0x620
[   89.437227] [e959dd50] [c005f190] notifier_call_chain+0x80/0xf0
[   89.443138] [e959dd80] [c0573a74] __dev_notify_flags+0x54/0xf0
[   89.448964] [e959dda0] [c05743f8] dev_change_flags+0x48/0x60
[   89.454615] [e959ddc0] [c0606744] devinet_ioctl+0x544/0x8a0
[   89.460180] [e959de10] [c060987c] inet_ioctl+0x9c/0x1f0
[   89.465400] [e959de80] [c05479a8] sock_ioctl+0x168/0x460
[   89.470708] [e959ded0] [c01cf3ec] do_vfs_ioctl+0xac/0x8c0
[   89.476099] [e959df20] [c01cfc40] SyS_ioctl+0x40/0xc0
[   89.481147] [e959df40] [c0011318] ret_from_syscall+0x0/0x3c
[   89.486715] --- interrupt: c01 at 0x1006943c
[   89.486715]     LR = 0x100c45ec

---
 drivers/net/ethernet/freescale/fman/fman_dtsec.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1036,7 +1036,7 @@ int dtsec_add_hash_mac_address(struct fm
 	set_bucket(dtsec->regs, bucket, true);
 
 	/* Create element to be added to the driver hash table */
-	hash_entry = kmalloc(sizeof(*hash_entry), GFP_KERNEL);
+	hash_entry = kmalloc(sizeof(*hash_entry), GFP_ATOMIC);
 	if (!hash_entry)
 		return -ENOMEM;
 	hash_entry->addr = addr;

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

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

From: Hemanth Puranik <[email protected]>

[ Upstream commit cc5db3150e87fe7f7e947bf333b6c1c97f848ecb ]

This patch fixes the warning messages/call traces seen if DMA debug is enabled, In case of fragmented skb's memory was allocated using dma_map_page but freed using dma_unmap_single. This patch modifies buffer allocations in TX path to use dma_map_page in all the places and dma_unmap_page while freeing the buffers.

---
 drivers/net/ethernet/qualcomm/emac/emac-mac.c |   23 ++++++++++++++---------
 1 file changed, 14 insertions(+), 9 deletions(-)

@@ -1250,9 +1250,9 @@ void emac_mac_tx_process(struct emac_ada
 	while (tx_q->tpd.consume_idx != hw_consume_idx) {
 		tpbuf = GET_TPD_BUFFER(tx_q, tx_q->tpd.consume_idx);
 		if (tpbuf->dma_addr) {
-			dma_unmap_single(adpt->netdev->dev.parent,
-					 tpbuf->dma_addr, tpbuf->length,
-					 DMA_TO_DEVICE);
+			dma_unmap_page(adpt->netdev->dev.parent,
+				       tpbuf->dma_addr, tpbuf->length,
+				       DMA_TO_DEVICE);
 			tpbuf->dma_addr = 0;
 		}
 

@@ -1409,9 +1409,11 @@ static void emac_tx_fill_tpd(struct emac
 
 		tpbuf = GET_TPD_BUFFER(tx_q, tx_q->tpd.produce_idx);
 		tpbuf->length = mapped_len;
-		tpbuf->dma_addr = dma_map_single(adpt->netdev->dev.parent,
-						 skb->data, tpbuf->length,
-						 DMA_TO_DEVICE);
+		tpbuf->dma_addr = dma_map_page(adpt->netdev->dev.parent,
+					       virt_to_page(skb->data),
+					       offset_in_page(skb->data),
+					       tpbuf->length,
+					       DMA_TO_DEVICE);
 		ret = dma_mapping_error(adpt->netdev->dev.parent,
 					tpbuf->dma_addr);
 		if (ret)

@@ -1427,9 +1429,12 @@ static void emac_tx_fill_tpd(struct emac
 	if (mapped_len < len) {
 		tpbuf = GET_TPD_BUFFER(tx_q, tx_q->tpd.produce_idx);
 		tpbuf->length = len - mapped_len;
-		tpbuf->dma_addr = dma_map_single(adpt->netdev->dev.parent,
-						 skb->data + mapped_len,
-						 tpbuf->length, DMA_TO_DEVICE);
+		tpbuf->dma_addr = dma_map_page(adpt->netdev->dev.parent,
+					       virt_to_page(skb->data +
+							    mapped_len),
+					       offset_in_page(skb->data +
+							      mapped_len),
+					       tpbuf->length, DMA_TO_DEVICE);
 		ret = dma_mapping_error(adpt->netdev->dev.parent,
 					tpbuf->dma_addr);
 		if (ret)

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

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

From: Jeremy Linton <[email protected]>

[ Upstream commit e06513d78d54e6c7026c9043a39e2c01ee25bdbe ]

The smsc911x driver will crash if it is rmmod'ed while the netdev is up like:

Call trace:
  phy_detach+0x94/0x150
  phy_disconnect+0x40/0x50
  smsc911x_stop+0x104/0x128 [smsc911x]
  __dev_close_many+0xb4/0x138
  dev_close_many+0xbc/0x190
  rollback_registered_many+0x140/0x460
  rollback_registered+0x68/0xb0
  unregister_netdevice_queue+0x100/0x118
  unregister_netdev+0x28/0x38
  smsc911x_drv_remove+0x58/0x130 [smsc911x]
  platform_drv_remove+0x30/0x50
  device_release_driver_internal+0x15c/0x1f8
  driver_detach+0x54/0x98
  bus_remove_driver+0x64/0xe8
  driver_unregister+0x34/0x60
  platform_driver_unregister+0x20/0x30
  smsc911x_cleanup_module+0x14/0xbca8 [smsc911x]
  SyS_delete_module+0x1e8/0x238
  __sys_trace_return+0x0/0x4

This is caused by the mdiobus being unregistered/free'd and the code in phy_detach() attempting to manipulate mdio related structures from unregister_netdev() calling close()

To fix this, we delay the mdiobus teardown until after the netdev is deregistered.

---
 drivers/net/ethernet/smsc/smsc911x.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -2330,14 +2330,14 @@ static int smsc911x_drv_remove(struct pl
 	pdata = netdev_priv(dev);
 	BUG_ON(!pdata);
 	BUG_ON(!pdata->ioaddr);
-	WARN_ON(dev->phydev);
 
 	SMSC_TRACE(pdata, ifdown, "Stopping driver");
 
+	unregister_netdev(dev);
+
 	mdiobus_unregister(pdata->mii_bus);
 	mdiobus_free(pdata->mii_bus);
 
-	unregister_netdev(dev);
 	res = platform_get_resource_byname(pdev, IORESOURCE_MEM,
 					   "smsc911x-memory");
 	if (!res)

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

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

From: Parav Pandit <[email protected]>

[ Upstream commit bb7f8f199c354c4cf155b1d6d55f86eaaed7fa5a ]

resolved_dev returned might be NULL as ifindex is transient number. Ignoring NULL check of resolved_dev might crash the kernel. Therefore perform NULL check before accessing resolved_dev.

Additionally rdma_resolve_ip_route() invokes addr_resolve() which performs check and address translation for loopback ifindex. Therefore, checking it again in rdma_resolve_ip_route() is not helpful. Therefore, the code is simplified to avoid IFF_LOOPBACK check.

---
 drivers/infiniband/core/sa_query.c |    7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

@@ -1137,10 +1137,9 @@ int ib_init_ah_from_path(struct ib_devic
 
 		resolved_dev = dev_get_by_index(dev_addr.net,
 						dev_addr.bound_dev_if);
-		if (resolved_dev->flags & IFF_LOOPBACK) {
-			dev_put(resolved_dev);
-			resolved_dev = idev;
-			dev_hold(resolved_dev);
+		if (!resolved_dev) {
+			dev_put(idev);
+			return -ENODEV;
 		}
 		ndev = ib_get_ndev_from_path(rec);
 		rcu_read_lock();

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

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

From: Arvind Yadav <[email protected]>

[ Upstream commit 351b2bccede1cb673ec7957b35ea997ea24c8884 ]

Never directly free @dev after calling device_register(), even if it returned an error! Always use put_device() to give up the reference initialized.

---
 drivers/xen/xenbus/xenbus_probe.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

@@ -470,8 +470,11 @@ int xenbus_probe_node(struct xen_bus_typ
 
 	/* Register with generic device framework. */
 	err = device_register(&xendev->dev);
-	if (err)
+	if (err) {
+		put_device(&xendev->dev);
+		xendev = NULL;
 		goto fail;
+	}
 
 	return 0;
 fail:

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

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

From: Marc Zyngier <[email protected]>

[ Upstream commit e21da1c992007594d391e7b301779cf30f438691 ]

A recent update to the ARM SMCCC ARCH_WORKAROUND_1 specification allows firmware to return a non zero, positive value to describe that although the mitigation is implemented at the higher exception level, the CPU on which the call is made is not affected.

Let's relax the check on the return value from ARCH_WORKAROUND_1 so that we only error out if the returned value is negative.

---
 arch/arm64/kernel/cpu_errata.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -160,7 +160,7 @@ static int enable_smccc_arch_workaround_
 	case PSCI_CONDUIT_HVC:
 		arm_smccc_1_1_hvc(ARM_SMCCC_ARCH_FEATURES_FUNC_ID,
 				  ARM_SMCCC_ARCH_WORKAROUND_1, &res);
-		if (res.a0)
+		if ((int)res.a0 < 0)
 			return 0;
 		cb = call_hvc_arch_workaround_1;
 		smccc_start = __smccc_workaround_1_hvc_start;

@@ -170,7 +170,7 @@ static int enable_smccc_arch_workaround_
 	case PSCI_CONDUIT_SMC:
 		arm_smccc_1_1_smc(ARM_SMCCC_ARCH_FEATURES_FUNC_ID,
 				  ARM_SMCCC_ARCH_WORKAROUND_1, &res);
-		if (res.a0)
+		if ((int)res.a0 < 0)
 			return 0;
 		cb = call_smc_arch_workaround_1;
 		smccc_start = __smccc_workaround_1_smc_start;

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

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

From: Gregory CLEMENT <[email protected]>

[ Upstream commit 3cd2c313f1d618f92d1294addc6c685c17065761 ]

On the CP110 components which are present on the Armada 7K/8K SoC we need to explicitly enable the clock for the registers. However it is not needed for the AP8xx component, that's why this clock is optional.

With this patch both clock have now a name, but in order to be backward compatible, the name of the first clock is not used. It allows to still use this clock with a device tree using the old binding.

---
 Documentation/devicetree/bindings/dma/mv-xor-v2.txt |    6 ++++
 drivers/dma/mv_xor_v2.c                             |   25 ++++++++++++++++----
 2 files changed, 25 insertions(+), 6 deletions(-)

@@ -11,7 +11,11 @@ Required properties:
   interrupts.
 
 Optional properties:
-- clocks: Optional reference to the clock used by the XOR engine.
+- clocks: Optional reference to the clocks used by the XOR engine.
+- clock-names: mandatory if there is a second clock, in this case the
+   name must be "core" for the first clock and "reg" for the second
+   one
+
 
 Example:
 

@@ -152,6 +152,7 @@ struct mv_xor_v2_device {
 	void __iomem *dma_base;
 	void __iomem *glob_base;
 	struct clk *clk;
+	struct clk *reg_clk;
 	struct tasklet_struct irq_tasklet;
 	struct list_head free_sw_desc;
 	struct dma_device dmadev;

@@ -697,13 +698,26 @@ static int mv_xor_v2_probe(struct platfo
 	if (ret)
 		return ret;
 
+	xor_dev->reg_clk = devm_clk_get(&pdev->dev, "reg");
+	if (PTR_ERR(xor_dev->reg_clk) != -ENOENT) {
+		if (!IS_ERR(xor_dev->reg_clk)) {
+			ret = clk_prepare_enable(xor_dev->reg_clk);
+			if (ret)
+				return ret;
+		} else {
+			return PTR_ERR(xor_dev->reg_clk);
+		}
+	}
+
 	xor_dev->clk = devm_clk_get(&pdev->dev, NULL);
-	if (IS_ERR(xor_dev->clk) && PTR_ERR(xor_dev->clk) == -EPROBE_DEFER)
-		return -EPROBE_DEFER;
+	if (IS_ERR(xor_dev->clk) && PTR_ERR(xor_dev->clk) == -EPROBE_DEFER) {
+		ret = EPROBE_DEFER;
+		goto disable_reg_clk;
+	}
 	if (!IS_ERR(xor_dev->clk)) {
 		ret = clk_prepare_enable(xor_dev->clk);
 		if (ret)
-			return ret;
+			goto disable_reg_clk;
 	}
 
 	ret = platform_msi_domain_alloc_irqs(&pdev->dev, 1,

@@ -812,8 +826,9 @@ free_hw_desq:
 free_msi_irqs:
 	platform_msi_domain_free_irqs(&pdev->dev);
 disable_clk:
-	if (!IS_ERR(xor_dev->clk))
-		clk_disable_unprepare(xor_dev->clk);
+	clk_disable_unprepare(xor_dev->clk);
+disable_reg_clk:
+	clk_disable_unprepare(xor_dev->reg_clk);
 	return ret;
 }
 

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

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

From: Florian Westphal <[email protected]>

[ Upstream commit 932909d9b28d27e807ff8eecb68c7748f6701628 ]

The last rule in the blob has next_entry offset that is same as total size. This made "ebtables32 -A OUTPUT -d de:ad:be:ef:01:02" fail on 64 bit kernel.

---
 net/bridge/netfilter/ebtables.c |    6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

@@ -2097,8 +2097,12 @@ static int size_entry_mwt(struct ebt_ent
 	 * offsets are relative to beginning of struct ebt_entry (i.e., 0).
 	 */
 	for (i = 0; i < 4 ; ++i) {
-		if (offsets[i] >= *total)
+		if (offsets[i] > *total)
 			return -EINVAL;
+
+		if (i < 3 && offsets[i] == *total)
+			return -EINVAL;
+
 		if (i == 0)
 			continue;
 		if (offsets[i-1] > offsets[i])

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

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

From: Michael Chan <[email protected]>

[ Upstream commit 3c4fe80b32c685bdc02b280814d0cfe80d441c72 ]

During initialization, if we encounter errors, there is a code path that calls bnxt_hwrm_vnic_set_tpa() with invalid VNIC ID. This may cause a warning in firmware logs.

---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -3401,6 +3401,9 @@ static int bnxt_hwrm_vnic_set_tpa(struct
 	struct bnxt_vnic_info *vnic = &bp->vnic_info[vnic_id];
 	struct hwrm_vnic_tpa_cfg_input req = {0};
 
+	if (vnic->fw_vnic_id == INVALID_HW_RING_ID)
+		return 0;
+
 	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_VNIC_TPA_CFG, -1, -1);
 
 	if (tpa_flags) {

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

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

From: Arvind Yadav <[email protected]>

[ Upstream commit 537f4146c53c95aac977852b371bafb9c6755ee1 ]

Never directly free @dev after calling device_register(), even if it returned an error! Always use put_device() to give up the reference initialized in this function instead.

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

@@ -5272,7 +5272,7 @@ int workqueue_sysfs_register(struct work
 
 	ret = device_register(&wq_dev->dev);
 	if (ret) {
-		kfree(wq_dev);
+		put_device(&wq_dev->dev);
 		wq->wq_dev = NULL;
 		return ret;
 	}

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

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

From: Sabrina Dubroca <[email protected]>

[ Upstream commit d52e5a7e7ca49457dd31fc8b42fb7c0d58a31221 ]

Prior to the rework of PMTU information storage in commit 2c8cec5c10bc ("ipv4: Cache learned PMTU information in inetpeer."), when a PMTU event advertising a PMTU smaller than net.ipv4.route.min_pmtu was received, we would disable setting the DF flag on packets by locking the MTU metric, and set the PMTU to net.ipv4.route.min_pmtu.

Since then, we don't disable DF, and set PMTU to net.ipv4.route.min_pmtu, so the intermediate router that has this link with a small MTU will have to drop the packets.

This patch reestablishes pre-2.6.39 behavior by splitting rtable->rt_pmtu into a bitfield with rt_mtu_locked and rt_pmtu. rt_mtu_locked indicates that we shouldn't set the DF bit on that path, and is checked in ip_dont_fragment().

One possible workaround is to set net.ipv4.route.min_pmtu to a value low enough to accommodate the lowest MTU encountered.

---
 include/net/ip.h        |   11 +++++++++--
 include/net/ip_fib.h    |    1 +
 include/net/route.h     |    3 ++-
 net/ipv4/route.c        |   26 +++++++++++++++++++-------
 net/ipv4/xfrm4_policy.c |    1 +
 5 files changed, 32 insertions(+), 10 deletions(-)

@@ -304,6 +304,13 @@ int ip_decrease_ttl(struct iphdr *iph)
 	return --iph->ttl;
 }
 
+static inline int ip_mtu_locked(const struct dst_entry *dst)
+{
+	const struct rtable *rt = (const struct rtable *)dst;
+
+	return rt->rt_mtu_locked || dst_metric_locked(dst, RTAX_MTU);
+}
+
 static inline
 int ip_dont_fragment(const struct sock *sk, const struct dst_entry *dst)
 {

@@ -311,7 +318,7 @@ int ip_dont_fragment(const struct sock *
 
 	return  pmtudisc == IP_PMTUDISC_DO ||
 		(pmtudisc == IP_PMTUDISC_WANT &&
-		 !(dst_metric_locked(dst, RTAX_MTU)));
+		 !ip_mtu_locked(dst));
 }
 
 static inline bool ip_sk_accept_pmtu(const struct sock *sk)

@@ -337,7 +344,7 @@ static inline unsigned int ip_dst_mtu_ma
 	struct net *net = dev_net(dst->dev);
 
 	if (net->ipv4.sysctl_ip_fwd_use_pmtu ||
-	    dst_metric_locked(dst, RTAX_MTU) ||
+	    ip_mtu_locked(dst) ||
 	    !forwarding)
 		return dst_mtu(dst);
 

@@ -57,6 +57,7 @@ struct fib_nh_exception {
 	int				fnhe_genid;
 	__be32				fnhe_daddr;
 	u32				fnhe_pmtu;
+	bool				fnhe_mtu_locked;
 	__be32				fnhe_gw;
 	unsigned long			fnhe_expires;
 	struct rtable __rcu		*fnhe_rth_input;

@@ -63,7 +63,8 @@ struct rtable {
 	__be32			rt_gateway;
 
 	/* Miscellaneous cached information */
-	u32			rt_pmtu;
+	u32			rt_mtu_locked:1,
+				rt_pmtu:31;
 
 	u32			rt_table_id;
 

@@ -618,6 +618,7 @@ static inline u32 fnhe_hashfun(__be32 da
 static void fill_route_from_fnhe(struct rtable *rt, struct fib_nh_exception *fnhe)
 {
 	rt->rt_pmtu = fnhe->fnhe_pmtu;
+	rt->rt_mtu_locked = fnhe->fnhe_mtu_locked;
 	rt->dst.expires = fnhe->fnhe_expires;
 
 	if (fnhe->fnhe_gw) {

@@ -628,7 +629,7 @@ static void fill_route_from_fnhe(struct
 }
 
 static void update_or_create_fnhe(struct fib_nh *nh, __be32 daddr, __be32 gw,
-				  u32 pmtu, unsigned long expires)
+				  u32 pmtu, bool lock, unsigned long expires)
 {
 	struct fnhe_hash_bucket *hash;
 	struct fib_nh_exception *fnhe;

@@ -665,8 +666,10 @@ static void update_or_create_fnhe(struct
 			fnhe->fnhe_genid = genid;
 		if (gw)
 			fnhe->fnhe_gw = gw;
-		if (pmtu)
+		if (pmtu) {
 			fnhe->fnhe_pmtu = pmtu;
+			fnhe->fnhe_mtu_locked = lock;
+		}
 		fnhe->fnhe_expires = max(1UL, expires);
 		/* Update all cached dsts too */
 		rt = rcu_dereference(fnhe->fnhe_rth_input);

@@ -690,6 +693,7 @@ static void update_or_create_fnhe(struct
 		fnhe->fnhe_daddr = daddr;
 		fnhe->fnhe_gw = gw;
 		fnhe->fnhe_pmtu = pmtu;
+		fnhe->fnhe_mtu_locked = lock;
 		fnhe->fnhe_expires = expires;
 
 		/* Exception created; mark the cached routes for the nexthop

@@ -771,7 +775,8 @@ static void __ip_do_redirect(struct rtab
 				struct fib_nh *nh = &FIB_RES_NH(res);
 
 				update_or_create_fnhe(nh, fl4->daddr, new_gw,
-						0, jiffies + ip_rt_gc_timeout);
+						0, false,
+						jiffies + ip_rt_gc_timeout);
 			}
 			if (kill_route)
 				rt->dst.obsolete = DST_OBSOLETE_KILL;

@@ -983,15 +988,18 @@ static void __ip_rt_update_pmtu(struct r
 {
 	struct dst_entry *dst = &rt->dst;
 	struct fib_result res;
+	bool lock = false;
 
-	if (dst_metric_locked(dst, RTAX_MTU))
+	if (ip_mtu_locked(dst))
 		return;
 
 	if (ipv4_mtu(dst) < mtu)
 		return;
 
-	if (mtu < ip_rt_min_pmtu)
+	if (mtu < ip_rt_min_pmtu) {
+		lock = true;
 		mtu = ip_rt_min_pmtu;
+	}
 
 	if (rt->rt_pmtu == mtu &&
 	    time_before(jiffies, dst->expires - ip_rt_mtu_expires / 2))

@@ -1001,7 +1009,7 @@ static void __ip_rt_update_pmtu(struct r
 	if (fib_lookup(dev_net(dst->dev), fl4, &res, 0) == 0) {
 		struct fib_nh *nh = &FIB_RES_NH(res);
 
-		update_or_create_fnhe(nh, fl4->daddr, 0, mtu,
+		update_or_create_fnhe(nh, fl4->daddr, 0, mtu, lock,
 				      jiffies + ip_rt_mtu_expires);
 	}
 	rcu_read_unlock();

@@ -1256,7 +1264,7 @@ static unsigned int ipv4_mtu(const struc
 
 	mtu = READ_ONCE(dst->dev->mtu);
 
-	if (unlikely(dst_metric_locked(dst, RTAX_MTU))) {
+	if (unlikely(ip_mtu_locked(dst))) {
 		if (rt->rt_uses_gateway && mtu > 576)
 			mtu = 576;
 	}

@@ -1481,6 +1489,7 @@ struct rtable *rt_dst_alloc(struct net_d
 		rt->rt_is_input = 0;
 		rt->rt_iif = 0;
 		rt->rt_pmtu = 0;
+		rt->rt_mtu_locked = 0;
 		rt->rt_gateway = 0;
 		rt->rt_uses_gateway = 0;
 		rt->rt_table_id = 0;

@@ -2403,6 +2412,7 @@ struct dst_entry *ipv4_blackhole_route(s
 		rt->rt_is_input = ort->rt_is_input;
 		rt->rt_iif = ort->rt_iif;
 		rt->rt_pmtu = ort->rt_pmtu;
+		rt->rt_mtu_locked = ort->rt_mtu_locked;
 
 		rt->rt_genid = rt_genid_ipv4(net);
 		rt->rt_flags = ort->rt_flags;

@@ -2505,6 +2515,8 @@ static int rt_fill_info(struct net *net,
 	memcpy(metrics, dst_metrics_ptr(&rt->dst), sizeof(metrics));
 	if (rt->rt_pmtu && expires)
 		metrics[RTAX_MTU - 1] = rt->rt_pmtu;
+	if (rt->rt_mtu_locked && expires)
+		metrics[RTAX_LOCK - 1] |= BIT(RTAX_MTU);
 	if (rtnetlink_put_metrics(skb, metrics) < 0)
 		goto nla_put_failure;
 

@@ -97,6 +97,7 @@ static int xfrm4_fill_dst(struct xfrm_ds
 	xdst->u.rt.rt_gateway = rt->rt_gateway;
 	xdst->u.rt.rt_uses_gateway = rt->rt_uses_gateway;
 	xdst->u.rt.rt_pmtu = rt->rt_pmtu;
+	xdst->u.rt.rt_mtu_locked = rt->rt_mtu_locked;
 	xdst->u.rt.rt_table_id = rt->rt_table_id;
 	INIT_LIST_HEAD(&xdst->u.rt.rt_uncached);
 

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

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

From: Cathy Zhou <[email protected]>

[ Upstream commit cf55612a945039476abfd73e39064b2e721c3272 ]

The NETIF_F_GSO_SOFTWARE implies support for GSO on SCTP, but the sunvnet driver does not support GSO for sctp. Here we remove the NETIF_F_GSO_SOFTWARE feature flag and only report NETIF_F_ALL_TSO instead.

---
 drivers/net/ethernet/sun/sunvnet.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -198,7 +198,7 @@ static struct vnet *vnet_new(const u64 *
 	dev->ethtool_ops = &vnet_ethtool_ops;
 	dev->watchdog_timeo = VNET_TX_TIMEOUT;
 
-	dev->hw_features = NETIF_F_TSO | NETIF_F_GSO | NETIF_F_GSO_SOFTWARE |
+	dev->hw_features = NETIF_F_TSO | NETIF_F_GSO | NETIF_F_ALL_TSO |
 			   NETIF_F_HW_CSUM | NETIF_F_SG;
 	dev->features = dev->hw_features;
 

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

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

From: Lucas Stach <[email protected]>

[ Upstream commit 6a055b92de15af987b4027826d43aa103c65a3c4 ]

Right now the vblank event completion is racing with the atomic update, which is especially bad when the PRE is in use, as one of the hardware issue workaround might extend the atomic commit for quite some time.

If the vblank IRQ happens to trigger during that time, we will prematurely signal the atomic commit completion to userspace, which causes tearing when userspace re-uses a framebuffer we haven't managed to flip away from yet.

---
 drivers/gpu/drm/imx/ipuv3-crtc.c |    5 +++++
 1 file changed, 5 insertions(+)

@@ -189,7 +189,11 @@ static void ipu_crtc_atomic_begin(struct
 				  struct drm_crtc_state *old_crtc_state)
 {
 	drm_crtc_vblank_on(crtc);
+}
 
+static void ipu_crtc_atomic_flush(struct drm_crtc *crtc,
+				  struct drm_crtc_state *old_crtc_state)
+{
 	spin_lock_irq(&crtc->dev->event_lock);
 	if (crtc->state->event) {
 		WARN_ON(drm_crtc_vblank_get(crtc));

@@ -257,6 +261,7 @@ static const struct drm_crtc_helper_func
 	.mode_set_nofb = ipu_crtc_mode_set_nofb,
 	.atomic_check = ipu_crtc_atomic_check,
 	.atomic_begin = ipu_crtc_atomic_begin,
+	.atomic_flush = ipu_crtc_atomic_flush,
 	.atomic_disable = ipu_crtc_atomic_disable,
 	.enable = ipu_crtc_enable,
 };

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

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

From: Rob Herring <[email protected]>

[ Upstream commit 101646a24a2f9cdb61d7732459fbf068a7bbb542 ]

Microblaze doesn't set CONFIG_NO_BOOTMEM and so memblock_virt_alloc() doesn't work for CONFIG_HAVE_MEMBLOCK && !CONFIG_NO_BOOTMEM.

Similar change was already done by others architectures "ARM: mm: Remove bootmem code and switch to NO_BOOTMEM" (sha1: 84f452b1e8fc73ac0e31254c66e3e2260ce5263d)
or
"openrisc: Consolidate setup to use memblock instead of bootmem" (sha1: 266c7fad157265bb54d17db1c9545f2aaa488643)
or
"parisc: Drop bootmem and switch to memblock"
(sha1: 4fe9e1d957e45ad8eba9885ee860a0e93d13a7c7)
or
"powerpc: Remove bootmem allocator"
(sha1: 10239733ee8617bac3f1c1769af43a88ed979324)
or
"s390/mm: Convert bootmem to memblock"
(sha1: 50be634507284eea38df78154d22615d21200b42)
or
"sparc64: Convert over to NO_BOOTMEM."
(sha1: 625d693e9784f988371e69c2b41a2172c0be6c11)
or
"xtensa: drop sysmem and switch to memblock"
(sha1: 0e46c1115f5816949220d62dd3ff04aa68e7ac6b)

Issue was introduced by:
"of/fdt: use memblock_virt_alloc for early alloc"
(sha1: 0fa1c579349fdd90173381712ad78aa99c09d38b)

---
 arch/microblaze/Kconfig   |    1 
 arch/microblaze/mm/init.c |   56 ++++------------------------------------------
 2 files changed, 7 insertions(+), 50 deletions(-)

@@ -23,6 +23,7 @@ config MICROBLAZE
 	select HAVE_FTRACE_MCOUNT_RECORD
 	select HAVE_FUNCTION_GRAPH_TRACER
 	select HAVE_FUNCTION_TRACER
+	select NO_BOOTMEM
 	select HAVE_MEMBLOCK
 	select HAVE_MEMBLOCK_NODE_MAP
 	select HAVE_OPROFILE

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

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

From: Toshiaki Makita <[email protected]>

[ Upstream commit 4bbb3e0e8239f9079bf1fe20b3c0cb598714ae61 ]

When we have a bridge with vlan_filtering on and a vlan device on top of it, packets would be corrupted in skb_vlan_untag() called from br_dev_xmit().

The problem sits in skb_reorder_vlan_header() used in skb_vlan_untag(), which makes use of skb->mac_len. In this function mac_len is meant for handling rx path with vlan devices with reorder_header disabled, but in tx path mac_len is typically 0 and cannot be used, which is the problem in this case.

The current code even does not properly handle rx path (skb_vlan_untag() called from __netif_receive_skb_core()) with reorder_header off actually.

In rx path single tag case, it works as follows:

- Before skb_reorder_vlan_header()

mac_header data
v v

   +-------------------+-------------+------+----
   |        ETH        |    VLAN     | ETH  |
   |       ADDRS       | TPID | TCI  | TYPE |
   +-------------------+-------------+------+----

<-------- mac_len --------->
<------------->
to be removed

- After skb_reorder_vlan_header()

            mac_header                     data
                 v                          v
                 +-------------------+------+----
                 |        ETH        | ETH  |
                 |       ADDRS       | TYPE |
                 +-------------------+------+----
                 <-------- mac_len --------->

This is ok, but in rx double tag case, it corrupts packets:

- Before skb_reorder_vlan_header()

mac_header data
v v

   +-------------------+-------------+-------------+------+----
   |        ETH        |    VLAN     |    VLAN     | ETH  |
   |       ADDRS       | TPID | TCI  | TPID | TCI  | TYPE |
   +-------------------+-------------+-------------+------+----

<--------------- mac_len ---------------->
<------------->
should be removed
<--------------------------->
actually will be removed

- After skb_reorder_vlan_header()

            mac_header                                   data
                 v                                        v
                               +-------------------+------+----
                               |        ETH        | ETH  |
                               |       ADDRS       | TYPE |
                               +-------------------+------+----
                 <--------------- mac_len ---------------->

So, two of vlan tags are both removed while only inner one should be removed and mac_header (and mac_len) is broken.

skb_vlan_untag() is meant for removing the vlan header at (skb->data - 2), so use skb->data and skb->mac_header to calculate the right offset.

---
 include/uapi/linux/if_ether.h |    1 +
 net/core/skbuff.c             |    7 +++++--
 2 files changed, 6 insertions(+), 2 deletions(-)

@@ -29,6 +29,7 @@
  */
 
 #define ETH_ALEN	6		/* Octets in one ethernet addr	 */
+#define ETH_TLEN	2		/* Octets in ethernet type field */
 #define ETH_HLEN	14		/* Total octets in header.	 */
 #define ETH_ZLEN	60		/* Min. octets in frame sans FCS */
 #define ETH_DATA_LEN	1500		/* Max. octets in payload	 */

@@ -4475,13 +4475,16 @@ EXPORT_SYMBOL_GPL(skb_gso_validate_mtu);
 
 static struct sk_buff *skb_reorder_vlan_header(struct sk_buff *skb)
 {
+	int mac_len;
+
 	if (skb_cow(skb, skb_headroom(skb)) < 0) {
 		kfree_skb(skb);
 		return NULL;
 	}
 
-	memmove(skb->data - ETH_HLEN, skb->data - skb->mac_len - VLAN_HLEN,
-		2 * ETH_ALEN);
+	mac_len = skb->data - skb_mac_header(skb);
+	memmove(skb_mac_header(skb) + VLAN_HLEN, skb_mac_header(skb),
+		mac_len - VLAN_HLEN - ETH_TLEN);
 	skb->mac_header += VLAN_HLEN;
 	return skb;
 }

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

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

From: Matthias Schiffer <[email protected]>

[ Upstream commit 6f27d2c2a8c236d296201c19abb8533ec20d212b ]

Checking for 0 is insufficient: when an SKB without a batadv header, but with a VLAN header is received, hdr_size will be 4, making the following code interpret the Ethernet header as a batadv header.

---
 net/batman-adv/distributed-arp-table.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -391,7 +391,7 @@ static void batadv_dbg_arp(struct batadv
 		   batadv_arp_hw_src(skb, hdr_size), &ip_src,
 		   batadv_arp_hw_dst(skb, hdr_size), &ip_dst);
 
-	if (hdr_size == 0)
+	if (hdr_size < sizeof(struct batadv_unicast_packet))
 		return;
 
 	unicast_4addr_packet = (struct batadv_unicast_4addr_packet *)skb->data;

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

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

From: Sven Eckelmann <[email protected]>

[ Upstream commit fc04fdb2c8a894283259f5621d31d75610701091 ]

batadv_check_unicast_ttvn may redirect a packet to itself or another originator. This involves rewriting the ttvn and the destination address in the batadv unicast header. These field were not yet pulled (with skb rcsum update) and thus any change to them also requires a change in the receive checksum.

---
 net/batman-adv/routing.c |   15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)

@@ -724,6 +724,7 @@ out:
 /**
  * batadv_reroute_unicast_packet - update the unicast header for re-routing
  * @bat_priv: the bat priv with all the soft interface information
+ * @skb: unicast packet to process
  * @unicast_packet: the unicast header to be updated
  * @dst_addr: the payload destination
  * @vid: VLAN identifier

@@ -735,7 +736,7 @@ out:
  * Return: true if the packet header has been updated, false otherwise
  */
 static bool
-batadv_reroute_unicast_packet(struct batadv_priv *bat_priv,
+batadv_reroute_unicast_packet(struct batadv_priv *bat_priv, struct sk_buff *skb,
 			      struct batadv_unicast_packet *unicast_packet,
 			      u8 *dst_addr, unsigned short vid)
 {

@@ -764,8 +765,10 @@ batadv_reroute_unicast_packet(struct bat
 	}
 
 	/* update the packet header */
+	skb_postpull_rcsum(skb, unicast_packet, sizeof(*unicast_packet));
 	ether_addr_copy(unicast_packet->dest, orig_addr);
 	unicast_packet->ttvn = orig_ttvn;
+	skb_postpush_rcsum(skb, unicast_packet, sizeof(*unicast_packet));
 
 	ret = true;
 out:

@@ -806,7 +809,7 @@ static bool batadv_check_unicast_ttvn(st
 	 * the packet to
 	 */
 	if (batadv_tt_local_client_is_roaming(bat_priv, ethhdr->h_dest, vid)) {
-		if (batadv_reroute_unicast_packet(bat_priv, unicast_packet,
+		if (batadv_reroute_unicast_packet(bat_priv, skb, unicast_packet,
 						  ethhdr->h_dest, vid))
 			batadv_dbg_ratelimited(BATADV_DBG_TT,
 					       bat_priv,

@@ -852,7 +855,7 @@ static bool batadv_check_unicast_ttvn(st
 	 * destination can possibly be updated and forwarded towards the new
 	 * target host
 	 */
-	if (batadv_reroute_unicast_packet(bat_priv, unicast_packet,
+	if (batadv_reroute_unicast_packet(bat_priv, skb, unicast_packet,
 					  ethhdr->h_dest, vid)) {
 		batadv_dbg_ratelimited(BATADV_DBG_TT, bat_priv,
 				       "Rerouting unicast packet to %pM (dst=%pM): TTVN mismatch old_ttvn=%u new_ttvn=%u\n",

@@ -875,12 +878,14 @@ static bool batadv_check_unicast_ttvn(st
 	if (!primary_if)
 		return false;
 
+	/* update the packet header */
+	skb_postpull_rcsum(skb, unicast_packet, sizeof(*unicast_packet));
 	ether_addr_copy(unicast_packet->dest, primary_if->net_dev->dev_addr);
+	unicast_packet->ttvn = curr_ttvn;
+	skb_postpush_rcsum(skb, unicast_packet, sizeof(*unicast_packet));
 
 	batadv_hardif_put(primary_if);
 
-	unicast_packet->ttvn = curr_ttvn;
-
 	return true;
 }
 

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

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

From: Stefano Brivio <[email protected]>

[ Upstream commit dd1df24737727e119c263acf1be2a92763938297 ]

This re-introduces the effect of commit a32452366b72 ("vti4:
Don't count header length twice.") which was accidentally reverted by merge commit f895f0cfbb77 ("Merge branch 'master' of
git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec").

The commit message from Steffen Klassert said:

We currently count the size of LL_MAX_HEADER and struct iphdr twice for vti4 devices, this leads to a wrong device mtu. The size of LL_MAX_HEADER and struct iphdr is already counted in ip_tunnel_bind_dev(), so don't do it again in vti_tunnel_init().

And this is still the case now: ip_tunnel_bind_dev() already accounts for the header length of the link layer (not necessarily LL_MAX_HEADER, if the output device is found), plus one IP header.

For example, with a vti device on top of veth, with MTU of 1500, the existing implementation would set the initial vti MTU to 1332, accounting once for LL_MAX_HEADER (128, included in hard_header_len by vti) and twice for the same IP header (once from hard_header_len, once from ip_tunnel_bind_dev()).

It should instead be 1480, because ip_tunnel_bind_dev() is able to figure out that the output device is veth, so no additional link layer header is attached, and will properly count one single IP header.

The existing issue had the side effect of avoiding PMTUD for most xfrm policies, by arbitrarily lowering the initial MTU. However, the only way to get a consistent PMTU value is to let the xfrm PMTU discovery do its course, and commit d6af1a31cc72 ("vti: Add pmtu handling to vti_xmit.") now takes care of local delivery cases where the application ignores local socket notifications.

---
 net/ipv4/ip_vti.c |    1 -
 1 file changed, 1 deletion(-)

@@ -396,7 +396,6 @@ static int vti_tunnel_init(struct net_de
 	memcpy(dev->dev_addr, &iph->saddr, 4);
 	memcpy(dev->broadcast, &iph->daddr, 4);
 
-	dev->hard_header_len	= LL_MAX_HEADER + sizeof(struct iphdr);
 	dev->mtu		= ETH_DATA_LEN;
 	dev->flags		= IFF_NOARP;
 	dev->addr_len		= 4;

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

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

From: Stefano Brivio <[email protected]>

[ Upstream commit 03080e5ec72740c1a62e6730f2a5f3f114f11b19 ]

Don't hardcode a MTU value on vti tunnel initialization, ip_tunnel_newlink() is able to deal with this already. See also commit ffc2b6ee4174 ("ip_gre: fix IFLA_MTU ignored on NEWLINK").

---
 net/ipv4/ip_vti.c |    1 -
 1 file changed, 1 deletion(-)

@@ -396,7 +396,6 @@ static int vti_tunnel_init(struct net_de
 	memcpy(dev->dev_addr, &iph->saddr, 4);
 	memcpy(dev->broadcast, &iph->daddr, 4);
 
-	dev->mtu		= ETH_DATA_LEN;
 	dev->flags		= IFF_NOARP;
 	dev->addr_len		= 4;
 	dev->features		|= NETIF_F_LLTX;

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

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

From: Song Liu <[email protected]>

[ Upstream commit c917e0f259908e75bd2a65877e25f9d90c22c848 ]

When a perf_event is attached to parent cgroup, it should count events for all children cgroups:

parent_group <---- perf_event
\
- child_group <---- process(es)

However, in our tests, we found this perf_event cannot report reliable results. Here is an example case:

  # create cgroups
  mkdir -p /sys/fs/cgroup/p/c
  # start perf for parent group
  perf stat -e instructions -G "p"

  # on another console, run test process in child cgroup:
  stressapptest -s 2 -M 1000 & echo $! > /sys/fs/cgroup/p/c/cgroup.procs

  # after the test process is done, stop perf in the first console shows

<not counted> instructions p

The instruction should not be "not counted" as the process runs in the child cgroup.

We found this is because perf_event->cgrp and cpuctx->cgrp are not identical, thus perf_event->cgrp are not updated properly.

This patch fixes this by updating perf_cgroup properly for ancestor cgroup(s).

---
 kernel/events/core.c |   21 ++++++++++++++++-----
 1 file changed, 16 insertions(+), 5 deletions(-)

@@ -634,9 +634,15 @@ static inline void __update_cgrp_time(st
 
 static inline void update_cgrp_time_from_cpuctx(struct perf_cpu_context *cpuctx)
 {
-	struct perf_cgroup *cgrp_out = cpuctx->cgrp;
-	if (cgrp_out)
-		__update_cgrp_time(cgrp_out);
+	struct perf_cgroup *cgrp = cpuctx->cgrp;
+	struct cgroup_subsys_state *css;
+
+	if (cgrp) {
+		for (css = &cgrp->css; css; css = css->parent) {
+			cgrp = container_of(css, struct perf_cgroup, css);
+			__update_cgrp_time(cgrp);
+		}
+	}
 }
 
 static inline void update_cgrp_time_from_event(struct perf_event *event)

@@ -664,6 +670,7 @@ perf_cgroup_set_timestamp(struct task_st
 {
 	struct perf_cgroup *cgrp;
 	struct perf_cgroup_info *info;
+	struct cgroup_subsys_state *css;
 
 	/*
 	 * ctx->lock held by caller

@@ -674,8 +681,12 @@ perf_cgroup_set_timestamp(struct task_st
 		return;
 
 	cgrp = perf_cgroup_from_task(task, ctx);
-	info = this_cpu_ptr(cgrp->info);
-	info->timestamp = ctx->timestamp;
+
+	for (css = &cgrp->css; css; css = css->parent) {
+		cgrp = container_of(css, struct perf_cgroup, css);
+		info = this_cpu_ptr(cgrp->info);
+		info->timestamp = ctx->timestamp;
+	}
 }
 
 #define PERF_CGROUP_SWOUT	0x1 /* cgroup switch out every event */

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

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

From: Stefan Wahren <[email protected]>

[ Upstream commit 9b9322db5c5a1917a66c71fe47c3848a9a31227e ]

The commit "regulatory: add NUL to request alpha2" increases the length of alpha2 to 3. This causes a regression on brcmfmac, because brcmf_cfg80211_reg_notifier() expect valid ISO3166 codes in the complete array. So fix this accordingly.

---
 drivers/net/wireless/broadcom/brcm80211/brcmfmac/cfg80211.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -6789,7 +6789,7 @@ static void brcmf_cfg80211_reg_notifier(
 	int i;
 
 	/* ignore non-ISO3166 country codes */
-	for (i = 0; i < sizeof(req->alpha2); i++)
+	for (i = 0; i < 2; i++)
 		if (req->alpha2[i] < 'A' || req->alpha2[i] > 'Z') {
 			brcmf_err("not a ISO3166 code (0x%02x 0x%02x)\n",
 				  req->alpha2[0], req->alpha2[1]);

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

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

From: Nicolas Pitre <[email protected]>

[ Upstream commit 825d487583089f9a33d31650c9c41f6474aab7fc ]

Some filesystems have timestamps with coarse precision that may allow for a recently built object file to have the same timestamp as the updated time on one of its dependency files. When that happens, the object file doesn't get rebuilt as it should.

This is especially the case on filesystems that don't have sub-second time precision, such as ext3 or Ext4 with 128B inodes.

Let's prevent that by making sure updated dependency files have a newer timestamp than the first file we created (i.e. autoksyms.h.tmpnew).

---
 scripts/adjust_autoksyms.sh |    7 +++++++
 1 file changed, 7 insertions(+)

@@ -83,6 +83,13 @@ while read sympath; do
 	depfile="include/config/ksym/${sympath}.h"
 	mkdir -p "$(dirname "$depfile")"
 	touch "$depfile"
+	# Filesystems with coarse time precision may create timestamps
+	# equal to the one from a file that was very recently built and that
+	# needs to be rebuild. Let's guard against that by making sure our
+	# dep files are always newer than the first file we created here.
+	while [ ! "$depfile" -nt "$new_ksyms_file" ]; do
+		touch "$depfile"
+	done
 	echo $((count += 1))
 done | tail -1 )
 changed=${changed:-0}

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

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

From: Chien Tin Tung <[email protected]>

[ Upstream commit 5f3e3b85cc0a5eae1c46d72e47d3de7bf208d9e2 ]

The option size check is using optval instead of optlen causing the set option call to fail. Use the correct field, optlen, for size check.

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

@@ -1296,7 +1296,7 @@ static ssize_t ucma_set_option(struct uc
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
 
-	if (unlikely(cmd.optval > KMALLOC_MAX_SIZE))
+	if (unlikely(cmd.optlen > KMALLOC_MAX_SIZE))
 		return -EINVAL;
 
 	optval = memdup_user((void __user *) (unsigned long) cmd.optval,

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

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

From: "Kalderon, Michal" <[email protected]>

[ Upstream commit c3594f22302cca5e924e47ec1cc8edd265708f41 ]

QPs that were configured with ack timeout value lower than 1 msec will not implement re-transmission timeout. This means that if a packet / ACK were dropped, the QP will not retransmit this packet.

This can lead to an application hang.

---
 drivers/infiniband/hw/qedr/verbs.c |   25 +++++++++++++++----------
 1 file changed, 15 insertions(+), 10 deletions(-)

@@ -1888,18 +1888,23 @@ int qedr_modify_qp(struct ib_qp *ibqp, s
 		SET_FIELD(qp_params.modify_flags,
 			  QED_ROCE_MODIFY_QP_VALID_ACK_TIMEOUT, 1);
 
-		qp_params.ack_timeout = attr->timeout;
-		if (attr->timeout) {
-			u32 temp;
-
-			temp = 4096 * (1UL << attr->timeout) / 1000 / 1000;
-			/* FW requires [msec] */
-			qp_params.ack_timeout = temp;
-		} else {
-			/* Infinite */
+		/* The received timeout value is an exponent used like this:
+		 *    "12.7.34 LOCAL ACK TIMEOUT
+		 *    Value representing the transport (ACK) timeout for use by
+		 *    the remote, expressed as: 4.096 * 2^timeout [usec]"
+		 * The FW expects timeout in msec so we need to divide the usec
+		 * result by 1000. We'll approximate 1000~2^10, and 4.096 ~ 2^2,
+		 * so we get: 2^2 * 2^timeout / 2^10 = 2^(timeout - 8).
+		 * The value of zero means infinite so we use a 'max_t' to make
+		 * sure that sub 1 msec values will be configured as 1 msec.
+		 */
+		if (attr->timeout)
+			qp_params.ack_timeout =
+					1 << max_t(int, attr->timeout - 8, 0);
+		else
 			qp_params.ack_timeout = 0;
-		}
 	}
+
 	if (attr_mask & IB_QP_RETRY_CNT) {
 		SET_FIELD(qp_params.modify_flags,
 			  QED_ROCE_MODIFY_QP_VALID_RETRY_CNT, 1);

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

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

From: "Kalderon, Michal" <[email protected]>

[ Upstream commit b15606f47b89b0b09936d7f45b59ba6275527041 ]

Return code wasn't set properly when CNQ allocation failed. This only affect error message logging, currently user will receive an error message that says the qedr driver load failed with rc '0', instead of ENOMEM

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

@@ -762,7 +762,8 @@ static struct qedr_dev *qedr_add(struct
 
 	dev->num_cnq = dev->ops->rdma_get_min_cnq_msix(cdev);
 	if (!dev->num_cnq) {
-		DP_ERR(dev, "not enough CNQ resources.\n");
+		DP_ERR(dev, "Failed. At least one CNQ is required.\n");
+		rc = -ENOMEM;
 		goto init_err;
 	}
 

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

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

From: Dan Carpenter <[email protected]>

[ Upstream commit 5dcd8400884cc4a043a6d4617e042489e5d566a9 ]

We moved the dev_hold(real_dev); call earlier in the function but forgot to update the error paths.

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

@@ -3244,7 +3244,7 @@ static int macsec_newlink(struct net *ne
 
 	err = netdev_upper_dev_link(real_dev, dev);
 	if (err < 0)
-		goto unregister;
+		goto put_dev;
 
 	/* need to be already registered so that ->init has run and
 	 * the MAC addr is set

@@ -3280,7 +3280,8 @@ del_dev:
 	macsec_del_dev(macsec);
 unlink:
 	netdev_upper_dev_unlink(real_dev, dev);
-unregister:
+put_dev:
+	dev_put(real_dev);
 	unregister_netdevice(dev);
 	return err;
 }

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

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

From: Yisheng Xie <[email protected]>

[ Upstream commit 8970a63e965b43288c4f5f40efbc2bbf80de7f16 ]

Alexander reported a use of uninitialized memory in __mpol_equal(), which is caused by incorrect use of preferred_node.

When mempolicy in mode MPOL_PREFERRED with flags MPOL_F_LOCAL, it uses numa_node_id() instead of preferred_node, however, __mpol_equal() uses preferred_node without checking whether it is MPOL_F_LOCAL or not.

---
 mm/mempolicy.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -2159,6 +2159,9 @@ bool __mpol_equal(struct mempolicy *a, s
 	case MPOL_INTERLEAVE:
 		return !!nodes_equal(a->v.nodes, b->v.nodes);
 	case MPOL_PREFERRED:
+		/* a's ->flags is the same as b's */
+		if (a->flags & MPOL_F_LOCAL)
+			return true;
 		return a->v.preferred_node == b->v.preferred_node;
 	default:
 		BUG();

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

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

From: David Rientjes <[email protected]>

[ Upstream commit 9d3c3354bb85bab4d865fe95039443f09a4c8394 ]

Commit 2516035499b9 ("mm, thp: remove __GFP_NORETRY from khugepaged and madvised allocations") changed the page allocator to no longer detect thp allocations based on __GFP_NORETRY.

It did not, however, modify the mem cgroup try_charge() path to avoid oom kill for either khugepaged collapsing or thp faulting. It is never expected to oom kill a process to allocate a hugepage for thp; reclaim is governed by the thp defrag mode and MADV_HUGEPAGE, but allocations (and charging) should fallback instead of oom killing processes.

---
 mm/huge_memory.c |    5 +++--
 mm/khugepaged.c  |    8 ++++++--
 2 files changed, 9 insertions(+), 4 deletions(-)

@@ -542,7 +542,8 @@ static int __do_huge_pmd_anonymous_page(
 
 	VM_BUG_ON_PAGE(!PageCompound(page), page);
 
-	if (mem_cgroup_try_charge(page, vma->vm_mm, gfp, &memcg, true)) {
+	if (mem_cgroup_try_charge(page, vma->vm_mm, gfp | __GFP_NORETRY, &memcg,
+				  true)) {
 		put_page(page);
 		count_vm_event(THP_FAULT_FALLBACK);
 		return VM_FAULT_FALLBACK;

@@ -1060,7 +1061,7 @@ alloc:
 	}
 
 	if (unlikely(mem_cgroup_try_charge(new_page, vma->vm_mm,
-					huge_gfp, &memcg, true))) {
+				huge_gfp | __GFP_NORETRY, &memcg, true))) {
 		put_page(new_page);
 		split_huge_pmd(vma, fe->pmd, fe->address);
 		if (page)

@@ -963,7 +963,9 @@ static void collapse_huge_page(struct mm
 		goto out_nolock;
 	}
 
-	if (unlikely(mem_cgroup_try_charge(new_page, mm, gfp, &memcg, true))) {
+	/* Do not oom kill for khugepaged charges */
+	if (unlikely(mem_cgroup_try_charge(new_page, mm, gfp | __GFP_NORETRY,
+					   &memcg, true))) {
 		result = SCAN_CGROUP_CHARGE_FAIL;
 		goto out_nolock;
 	}

@@ -1323,7 +1325,9 @@ static void collapse_shmem(struct mm_str
 		goto out;
 	}
 
-	if (unlikely(mem_cgroup_try_charge(new_page, mm, gfp, &memcg, true))) {
+	/* Do not oom kill for khugepaged charges */
+	if (unlikely(mem_cgroup_try_charge(new_page, mm, gfp | __GFP_NORETRY,
+					   &memcg, true))) {
 		result = SCAN_CGROUP_CHARGE_FAIL;
 		goto out;
 	}

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

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

From: Masami Hiramatsu <[email protected]>

[ Upstream commit 871bef2000968c312a4000b2f56d370dcedbc93c ]

Add a testcase for probe event argument syntax which ensures the kprobe_events interface correctly parses given event arguments.

---
 tools/testing/selftests/ftrace/test.d/kprobe/kprobe_args_syntax.tc |   97 ++++++++++
 1 file changed, 97 insertions(+)
 create mode 100644 tools/testing/selftests/ftrace/test.d/kprobe/kprobe_args_syntax.tc

@@ -0,0 +1,97 @@
+#!/bin/sh
+# SPDX-License-Identifier: GPL-2.0
+# description: Kprobe event argument syntax
+
+[ -f kprobe_events ] || exit_unsupported # this is configurable
+
+grep "x8/16/32/64" README > /dev/null || exit_unsupported # version issue
+
+echo 0 > events/enable
+echo > kprobe_events
+
+PROBEFUNC="vfs_read"
+GOODREG=
+BADREG=
+GOODSYM="_sdata"
+if ! grep -qw ${GOODSYM} /proc/kallsyms ; then
+  GOODSYM=$PROBEFUNC
+fi
+BADSYM="deaqswdefr"
+SYMADDR=0x`grep -w ${GOODSYM} /proc/kallsyms | cut -f 1 -d " "`
+GOODTYPE="x16"
+BADTYPE="y16"
+
+case `uname -m` in
+x86_64|i[3456]86)
+  GOODREG=%ax
+  BADREG=%ex
+;;
+aarch64)
+  GOODREG=%x0
+  BADREG=%ax
+;;
+arm*)
+  GOODREG=%r0
+  BADREG=%ax
+;;
+esac
+
+test_goodarg() # Good-args
+{
+  while [ "$1" ]; do
+    echo "p ${PROBEFUNC} $1" > kprobe_events
+    shift 1
+  done;
+}
+
+test_badarg() # Bad-args
+{
+  while [ "$1" ]; do
+    ! echo "p ${PROBEFUNC} $1" > kprobe_events
+    shift 1
+  done;
+}
+
+echo > kprobe_events
+
+: "Register access"
+test_goodarg ${GOODREG}
+test_badarg ${BADREG}
+
+: "Symbol access"
+test_goodarg "@${GOODSYM}" "@${SYMADDR}" "@${GOODSYM}+10" "@${GOODSYM}-10"
+test_badarg "@" "@${BADSYM}" "@${GOODSYM}*10" "@${GOODSYM}/10" \
+	    "@${GOODSYM}%10" "@${GOODSYM}&10" "@${GOODSYM}|10"
+
+: "Stack access"
+test_goodarg "\$stack" "\$stack0" "\$stack1"
+test_badarg "\$stackp" "\$stack0+10" "\$stack1-10"
+
+: "Retval access"
+echo "r ${PROBEFUNC} \$retval" > kprobe_events
+! echo "p ${PROBEFUNC} \$retval" > kprobe_events
+
+: "Comm access"
+test_goodarg "\$comm"
+
+: "Indirect memory access"
+test_goodarg "+0(${GOODREG})" "-0(${GOODREG})" "+10(\$stack)" \
+	"+0(\$stack1)" "+10(@${GOODSYM}-10)" "+0(+10(+20(\$stack)))"
+test_badarg "+(${GOODREG})" "(${GOODREG}+10)" "-(${GOODREG})" "(${GOODREG})" \
+	"+10(\$comm)" "+0(${GOODREG})+10"
+
+: "Name assignment"
+test_goodarg "varname=${GOODREG}"
+test_badarg "varname=varname2=${GOODREG}"
+
+: "Type syntax"
+test_goodarg "${GOODREG}:${GOODTYPE}"
+test_badarg "${GOODREG}::${GOODTYPE}" "${GOODREG}:${BADTYPE}" \
+	"${GOODTYPE}:${GOODREG}"
+
+: "Combination check"
+
+test_goodarg "\$comm:string" "+0(\$stack):string"
+test_badarg "\$comm:x64" "\$stack:string" "${GOODREG}:string"
+
+echo > kprobe_events

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

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

From: Masami Hiramatsu <[email protected]>

[ Upstream commit 5fbdbed797b6d12d043a5121fdbc8d8b49d10e80 ]

Add a testcase for string type with kprobe event. This tests good/bad syntax combinations and also the traced data is correct in several way.

---
 tools/testing/selftests/ftrace/test.d/kprobe/kprobe_args_string.tc |   46 ++++++++++
 1 file changed, 46 insertions(+)
 create mode 100644 tools/testing/selftests/ftrace/test.d/kprobe/kprobe_args_string.tc

@@ -0,0 +1,46 @@
+#!/bin/sh
+# SPDX-License-Identifier: GPL-2.0
+# description: Kprobe event string type argument
+
+[ -f kprobe_events ] || exit_unsupported # this is configurable
+
+echo 0 > events/enable
+echo > kprobe_events
+
+case `uname -m` in
+x86_64)
+  ARG2=%si
+  OFFS=8
+;;
+i[3456]86)
+  ARG2=%cx
+  OFFS=4
+;;
+aarch64)
+  ARG2=%x1
+  OFFS=8
+;;
+arm*)
+  ARG2=%r1
+  OFFS=4
+;;
+*)
+  echo "Please implement other architecture here"
+  exit_untested
+esac
+
+: "Test get argument (1)"
+echo "p:testprobe create_trace_kprobe arg1=+0(+0(${ARG2})):string" > kprobe_events
+echo 1 > events/kprobes/testprobe/enable
+! echo test >> kprobe_events
+tail -n 1 trace | grep -qe "testprobe.* arg1=\"test\""
+
+echo 0 > events/kprobes/testprobe/enable
+: "Test get argument (2)"
+echo "p:testprobe create_trace_kprobe arg1=+0(+0(${ARG2})):string arg2=+0(+${OFFS}(${ARG2})):string" > kprobe_events
+echo 1 > events/kprobes/testprobe/enable
+! echo test1 test2 >> kprobe_events
+tail -n 1 trace | grep -qe "testprobe.* arg1=\"test1\" arg2=\"test2\""
+
+echo 0 > events/enable
+echo > kprobe_events

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

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

From: Masami Hiramatsu <[email protected]>

[ Upstream commit dfa453bc90eca0febff33c8d292a656e53702158 ]

Add a testcase for probe point definition. This tests symbol, address and symbol+offset syntax. The offset must be positive and smaller than UINT_MAX.

---
 tools/testing/selftests/ftrace/test.d/kprobe/probepoint.tc |   43 +++++++++++++
 1 file changed, 43 insertions(+)
 create mode 100644 tools/testing/selftests/ftrace/test.d/kprobe/probepoint.tc

@@ -0,0 +1,43 @@
+#!/bin/sh
+# SPDX-License-Identifier: GPL-2.0
+# description: Kprobe events - probe points
+
+[ -f kprobe_events ] || exit_unsupported # this is configurable
+
+TARGET_FUNC=create_trace_kprobe
+
+dec_addr() { # hexaddr
+  printf "%d" "0x"`echo $1 | tail -c 8`
+}
+
+set_offs() { # prev target next
+  A1=`dec_addr $1`
+  A2=`dec_addr $2`
+  A3=`dec_addr $3`
+  TARGET="0x$2" # an address
+  PREV=`expr $A1 - $A2` # offset to previous symbol
+  NEXT=+`expr $A3 - $A2` # offset to next symbol
+  OVERFLOW=+`printf "0x%x" ${PREV}` # overflow offset to previous symbol
+}
+
+# We have to decode symbol addresses to get correct offsets.
+# If the offset is not an instruction boundary, it cause -EILSEQ.
+set_offs `grep -A1 -B1 ${TARGET_FUNC} /proc/kallsyms | cut -f 1 -d " " | xargs`
+
+UINT_TEST=no
+# printf "%x" -1 returns (unsigned long)-1.
+if [ `printf "%x" -1 | wc -c` != 9 ]; then
+  UINT_TEST=yes
+fi
+
+echo 0 > events/enable
+echo > kprobe_events
+echo "p:testprobe ${TARGET_FUNC}" > kprobe_events
+echo "p:testprobe ${TARGET}" > kprobe_events
+echo "p:testprobe ${TARGET_FUNC}${NEXT}" > kprobe_events
+! echo "p:testprobe ${TARGET_FUNC}${PREV}" > kprobe_events
+if [ "${UINT_TEST}" = yes ]; then
+! echo "p:testprobe ${TARGET_FUNC}${OVERFLOW}" > kprobe_events
+fi
+echo > kprobe_events
+clear_trace

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

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

From: Jinbum Park <[email protected]>

[ Upstream commit 73b9160d0dfe44dfdaffd6465dc1224c38a4a73c ]

Define vdso_start, vdso_end as array to avoid compile-time analysis error for the case of built with CONFIG_FORTIFY_SOURCE.

and, since vdso_start, vdso_end are used in vdso.c only, move extern-declaration from vdso.h to vdso.c.

If kernel is built with CONFIG_FORTIFY_SOURCE,
compile-time error happens at this code.
- if (memcmp(&vdso_start, "177ELF", 4))

The size of "&vdso_start" is recognized as 1 byte, but n is 4, So that compile-time error is reported.

---
 arch/arm/include/asm/vdso.h |    2 --
 arch/arm/kernel/vdso.c      |   12 +++++++-----
 2 files changed, 7 insertions(+), 7 deletions(-)

@@ -11,8 +11,6 @@ struct mm_struct;
 
 void arm_install_vdso(struct mm_struct *mm, unsigned long addr);
 
-extern char vdso_start, vdso_end;
-
 extern unsigned int vdso_total_pages;
 
 #else /* CONFIG_VDSO */

@@ -39,6 +39,8 @@
 
 static struct page **vdso_text_pagelist;
 
+extern char vdso_start[], vdso_end[];
+
 /* Total number of pages needed for the data and text portions of the VDSO. */
 unsigned int vdso_total_pages __ro_after_init;
 

@@ -179,13 +181,13 @@ static int __init vdso_init(void)
 	unsigned int text_pages;
 	int i;
 
-	if (memcmp(&vdso_start, "\177ELF", 4)) {
+	if (memcmp(vdso_start, "\177ELF", 4)) {
 		pr_err("VDSO is not a valid ELF object!\n");
 		return -ENOEXEC;
 	}
 
-	text_pages = (&vdso_end - &vdso_start) >> PAGE_SHIFT;
-	pr_debug("vdso: %i text pages at base %p\n", text_pages, &vdso_start);
+	text_pages = (vdso_end - vdso_start) >> PAGE_SHIFT;
+	pr_debug("vdso: %i text pages at base %p\n", text_pages, vdso_start);
 
 	/* Allocate the VDSO text pagelist */
 	vdso_text_pagelist = kcalloc(text_pages, sizeof(struct page *),

@@ -200,7 +202,7 @@ static int __init vdso_init(void)
 	for (i = 0; i < text_pages; i++) {
 		struct page *page;
 
-		page = virt_to_page(&vdso_start + i * PAGE_SIZE);
+		page = virt_to_page(vdso_start + i * PAGE_SIZE);
 		vdso_text_pagelist[i] = page;
 	}
 

@@ -211,7 +213,7 @@ static int __init vdso_init(void)
 
 	cntvct_ok = cntvct_functional();
 
-	patch_vdso(&vdso_start);
+	patch_vdso(vdso_start);
 
 	return 0;
 }

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

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

From: Pawel Dembicki <[email protected]>

[ Upstream commit 743989254ea9f132517806d8893ca9b6cf9dc86b ]

BroadMobi BM806U is an Qualcomm MDM9225 based 3G/4G modem. Tested hardware BM806U is mounted on D-Link DWR-921-C3 router. The USB id is added to qmi_wwan.c to allow QMI communication with the BM806U.

Tested on 4.14 kernel and OpenWRT.

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

@@ -889,6 +889,7 @@ static const struct usb_device_id produc
 	{QMI_FIXED_INTF(0x19d2, 0x2002, 4)},	/* ZTE (Vodafone) K3765-Z */
 	{QMI_FIXED_INTF(0x2001, 0x7e19, 4)},	/* D-Link DWM-221 B1 */
 	{QMI_FIXED_INTF(0x2001, 0x7e35, 4)},	/* D-Link DWM-222 */
+	{QMI_FIXED_INTF(0x2020, 0x2033, 4)},	/* BroadMobi BM806U */
 	{QMI_FIXED_INTF(0x0f3d, 0x68a2, 8)},    /* Sierra Wireless MC7700 */
 	{QMI_FIXED_INTF(0x114f, 0x68a2, 8)},    /* Sierra Wireless MC7750 */
 	{QMI_FIXED_INTF(0x1199, 0x68a2, 8)},	/* Sierra Wireless MC7710 in QMI mode */

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

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

From: Stephane Eranian <[email protected]>

[ Upstream commit 71eb9ee9596d8df3d5723c3cfc18774c6235e8b1 ]

this patch fix a bug in how the pebs->real_ip is handled in the PEBS handler. real_ip only exists in Haswell and later processor. It is actually the eventing IP, i.e., where the event occurred. As opposed to the pebs->ip which is the PEBS interrupt IP which is always off by one.

The problem is that the real_ip just like the IP needs to be fixed up because PEBS does not record all the machine state registers, and in particular the code segement (cs). This is why we have the set_linear_ip() function. The problem was that set_linear_ip() was only used on the pebs->ip and not the pebs->real_ip.

We have profiles which ran into invalid callstacks because of this. Here is an example:

 .....  0: ffffffffffffff80 recent entry, marker kernel v
 .....  1: 000000000040044d <= user address in kernel space!
 .....  2: fffffffffffffe00 marker enter user v
 .....  3: 000000000040044d
 .....  4: 00000000004004b6 oldest entry

Debugging output in get_perf_callchain():

 [  857.769909] CALLCHAIN: CPU8 ip=40044d regs->cs=10 user_mode(regs)=0

The problem is that the kernel entry in 1: points to a user level address. How can that be?

The reason is that with PEBS sampling the instruction that caused the event to occur and the instruction where the CPU was when the interrupt was posted may be far apart. And sometime during that time window, the privilege level may change. This happens, for instance, when the PEBS sample is taken close to a kernel entry point. Here PEBS, eventing IP (real_ip) captured a user level instruction. But by the time the PMU interrupt fired, the processor had already entered kernel space. This is why the debug output shows a user address with user_mode() false.

The problem comes from PEBS not recording the code segment (cs) register. The register is used in x86_64 to determine if executing in kernel vs user space. This is okay because the kernel has a software workaround called set_linear_ip(). But the issue in setup_pebs_sample_data() is that set_linear_ip() is never called on the real_ip value when it is available (Haswell and later) and precise_ip > 1.

This patch fixes this problem and eliminates the callchain discrepancy.

The patch restructures the code around set_linear_ip() to minimize the number of times the IP has to be set.

---
 arch/x86/events/intel/ds.c |   25 +++++++++++++++++--------
 1 file changed, 17 insertions(+), 8 deletions(-)

@@ -1110,6 +1110,7 @@ static void setup_pebs_sample_data(struc
 	if (pebs == NULL)
 		return;
 
+	regs->flags &= ~PERF_EFLAGS_EXACT;
 	sample_type = event->attr.sample_type;
 	dsrc = sample_type & PERF_SAMPLE_DATA_SRC;
 

@@ -1154,7 +1155,6 @@ static void setup_pebs_sample_data(struc
 	 */
 	*regs = *iregs;
 	regs->flags = pebs->flags;
-	set_linear_ip(regs, pebs->ip);
 
 	if (sample_type & PERF_SAMPLE_REGS_INTR) {
 		regs->ax = pebs->ax;

@@ -1190,13 +1190,22 @@ static void setup_pebs_sample_data(struc
 #endif
 	}
 
-	if (event->attr.precise_ip > 1 && x86_pmu.intel_cap.pebs_format >= 2) {
-		regs->ip = pebs->real_ip;
-		regs->flags |= PERF_EFLAGS_EXACT;
-	} else if (event->attr.precise_ip > 1 && intel_pmu_pebs_fixup_ip(regs))
-		regs->flags |= PERF_EFLAGS_EXACT;
-	else
-		regs->flags &= ~PERF_EFLAGS_EXACT;
+	if (event->attr.precise_ip > 1) {
+		/* Haswell and later have the eventing IP, so use it: */
+		if (x86_pmu.intel_cap.pebs_format >= 2) {
+			set_linear_ip(regs, pebs->real_ip);
+			regs->flags |= PERF_EFLAGS_EXACT;
+		} else {
+			/* Otherwise use PEBS off-by-1 IP: */
+			set_linear_ip(regs, pebs->ip);
+
+			/* ... and try to fix it up using the LBR entries: */
+			if (intel_pmu_pebs_fixup_ip(regs))
+				regs->flags |= PERF_EFLAGS_EXACT;
+		}
+	} else
+		set_linear_ip(regs, pebs->ip);
+
 
 	if ((sample_type & PERF_SAMPLE_ADDR) &&
 	    x86_pmu.intel_cap.pebs_format >= 1)