3.10.37-stable review

This is the start of the stable review cycle for the 3.10.37 release. There are 41 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.

Responses should be made by Sun Apr 13 16:09:00 UTC 2014. Anything received after that time might be too late.

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

	kernel.org/pub/linux/kernel/v3.0/stable-review/patch-3.10.37-rc1.gz

and the diffstat can be found below.

thanks,

greg k-h

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

Greg Kroah-Hartman <[email protected]>

    Linux 3.10.37-rc1

Stephen Boyd <[email protected]>
cpufreq: Fix timer/workqueue corruption due to double queueing

Xiaoguang Chen <[email protected]>
cpufreq: Fix governor start/stop race condition

Ard Biesheuvel <[email protected]>
crypto: ghash-clmulni-intel - use C implementation for setkey()

Finn Thain <[email protected]>
m68k: Skip futex_atomic_cmpxchg_inatomic() test

Heiko Carstens <[email protected]>
futex: Allow architectures to skip futex_atomic_cmpxchg_inatomic() test

Vineet Gupta <[email protected]>
ARC: [nsimosci] Unbork console

Mischa Jonker <[email protected]>
ARC: [nsimosci] Change .dts to use generic 8250 UART

Sasha Levin <[email protected]>
rds: prevent dereference of a NULL device in rds_iw_laddr_check

Dan Carpenter <[email protected]>
isdnloop: several buffer overflows

YOSHIFUJI Hideaki <[email protected]>
isdnloop: Validate NUL-terminated strings from user.

Pablo Neira <[email protected]>
netlink: don't compare the nul-termination in nla_strcmp

Hannes Frederic Sowa <[email protected]>
ipv6: some ipv6 statistic counters failed to disable bh

Paul Durrant <[email protected]>
xen-netback: remove pointless clause from if statement

Michael S. Tsirkin <[email protected]>
vhost: validate vhost_get_vq_desc return value

Michael S. Tsirkin <[email protected]>
vhost: fix total length when packets are too short

Vlad Yasevich <[email protected]>
vlan: Set hard_header_len according to available acceleration

Oliver Neukum <[email protected]>
usbnet: include wait queue head in device structure

Vlad Yasevich <[email protected]>
tg3: Do not include vlan acceleration features in vlan_features

Li RongQing <[email protected]>
netpoll: fix the skb check in pkt_is_ns

Nicolas Dichtel <[email protected]>
ip6mr: fix mfc notification flags

Nicolas Dichtel <[email protected]>
ipmr: fix mfc notification flags

Nicolas Dichtel <[email protected]>
rtnetlink: fix fdb notification flags

David Stevens <[email protected]>
vxlan: fix potential NULL dereference in arp_reduce()

lucien <[email protected]>
ipv6: ip6_append_data_mtu do not handle the mtu of the second fragment properly

Heiner Kallweit <[email protected]>
ipv6: Avoid unnecessary temporary addresses being generated

Matthew Leach <[email protected]>
net: socket: error on a negative msg_namelen

Eric Dumazet <[email protected]>
tcp: tcp_release_cb() should release socket ownership

Peter Boström <[email protected]>
vlan: Set correct source MAC address with TX VLAN offload enabled

Sabrina Dubroca <[email protected]>
ipv6: don't set DST_NOCOUNT for remotely added routes

Anton Nayshtut <[email protected]>
ipv6: Fix exthdrs offload registration.

Eric Dumazet <[email protected]>
net: unix: non blocking recvmsg() should not return -EINTR

Florian Westphal <[email protected]>
inet: frag: make sure forced eviction removes all frags

Linus Lüssing <[email protected]>
bridge: multicast: add sanity check for query source addresses

Daniel Borkmann <[email protected]>

    net: sctp: fix skb leakage in COOKIE ECHO path of chunk->auth_chunk

Nikolay Aleksandrov <[email protected]>
net: fix for a race condition in the inet frag code

Daniel Fu <[email protected]>
cpuidle: Check the result of cpuidle_get_driver() against NULL

Nicolas Dichtel <[email protected]>
kbuild: fix make headers_install when path is too long

Mikulas Patocka <[email protected]>
powernow-k6: reorder frequencies

Mikulas Patocka <[email protected]>
powernow-k6: correctly initialize default parameters

Mikulas Patocka <[email protected]>
powernow-k6: disable cache when changing frequency

Paul Moore <[email protected]>
selinux: correctly label /proc inodes in use before the policy is loaded

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

Diffstat:

 Makefile                                   |   4 +-
 arch/arc/boot/dts/nsimosci.dts             |  12 ++-
 arch/arc/configs/nsimosci_defconfig        |   1 +
 arch/m68k/Kconfig                          |   1 +
 arch/s390/Kconfig                          |   1 +
 arch/x86/crypto/ghash-clmulni-intel_asm.S  |  29 ------
 arch/x86/crypto/ghash-clmulni-intel_glue.c |  14 ++-
 drivers/cpufreq/cpufreq.c                  |  24 +++++
 drivers/cpufreq/cpufreq_governor.c         |   3 +
 drivers/cpufreq/powernow-k6.c              | 147 +++++++++++++++++++++++------
 drivers/cpuidle/driver.c                   |   3 +-
 drivers/isdn/isdnloop/isdnloop.c           |  23 +++--
 drivers/net/ethernet/broadcom/tg3.c        |   5 +-
 drivers/net/usb/usbnet.c                   |  33 ++++---
 drivers/net/vxlan.c                        |   3 +
 drivers/net/xen-netback/netback.c          |   4 +-
 drivers/vhost/net.c                        |  20 +++-
 include/linux/cpufreq.h                    |   1 +
 include/linux/futex.h                      |   4 +
 include/linux/usb/usbnet.h                 |   2 +-
 include/net/sock.h                         |   5 +
 init/Kconfig                               |   7 ++
 kernel/futex.c                             |  14 ++-
 lib/nlattr.c                               |  10 +-
 net/8021q/vlan.c                           |   4 +-
 net/8021q/vlan_dev.c                       |   6 +-
 net/bridge/br_multicast.c                  |   6 ++
 net/core/netpoll.c                         |   2 +-
 net/core/rtnetlink.c                       |  10 +-
 net/core/sock.c                            |   5 +-
 net/ipv4/inet_fragment.c                   |   5 +-
 net/ipv4/ipmr.c                            |  13 ++-
 net/ipv4/tcp_output.c                      |  11 +++
 net/ipv6/addrconf.c                        |   5 +-
 net/ipv6/exthdrs_offload.c                 |   4 +-
 net/ipv6/icmp.c                            |   2 +-
 net/ipv6/ip6_output.c                      |  18 ++--
 net/ipv6/ip6mr.c                           |  13 ++-
 net/ipv6/mcast.c                           |  11 ++-
 net/ipv6/route.c                           |   2 +-
 net/rds/iw.c                               |   3 +-
 net/sctp/sm_make_chunk.c                   |   4 +-
 net/sctp/sm_statefuns.c                    |   5 -
 net/socket.c                               |   4 +
 net/unix/af_unix.c                         |  17 +++-
 scripts/Makefile.headersinst               |  20 ++--
 scripts/headers_install.sh                 |   7 +-
 security/selinux/hooks.c                   |  36 +++++--
 48 files changed, 413 insertions(+), 170 deletions(-)

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

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

From: Paul Moore <[email protected]>

commit f64410ec665479d7b4b77b7519e814253ed0f686 upstream.

This patch is based on an earlier patch by Eric Paris, he describes the problem below:

"If an inode is accessed before policy load it will get placed on a
list of inodes to be initialized after policy load. After policy load we call inode_doinit() which calls inode_doinit_with_dentry() on all inodes accessed before policy load. In the case of inodes in procfs that means we'll end up at the bottom where it does:

     /* Default to the fs superblock SID. */
     isec->sid = sbsec->sid;

     if ((sbsec->flags & SE_SBPROC) && !S_ISLNK(inode->i_mode)) {
             if (opt_dentry) {
                     isec->sclass = inode_mode_to_security_class(...)
                     rc = selinux_proc_get_sid(opt_dentry,
                                               isec->sclass,
                                               &sid);
                     if (rc)
                             goto out_unlock;
                     isec->sid = sid;
             }
     }

Since opt_dentry is null, we'll never call selinux_proc_get_sid() and will leave the inode labeled with the label on the superblock. I believe a fix would be to mimic the behavior of xattrs. Look for an alias of the inode. If it can't be found, just leave the inode uninitialized (and pick it up later) if it can be found, we should be able to call selinux_proc_get_sid() ..."

On a system exhibiting this problem, you will notice a lot of files in /proc with the generic "proc_t" type (at least the ones that were accessed early in the boot), for example:

   # ls -Z /proc/sys/kernel/shmmax | awk '{ print $4 " " $5 }'
   system_u:object_r:proc_t:s0 /proc/sys/kernel/shmmax

However, with this patch in place we see the expected result:

   # ls -Z /proc/sys/kernel/shmmax | awk '{ print $4 " " $5 }'
   system_u:object_r:sysctl_kernel_t:s0 /proc/sys/kernel/shmmax

---
 security/selinux/hooks.c |   36 +++++++++++++++++++++++++++---------
 1 file changed, 27 insertions(+), 9 deletions(-)

@@ -1361,15 +1361,33 @@ static int inode_doinit_with_dentry(stru
 		isec->sid = sbsec->sid;
 
 		if ((sbsec->flags & SE_SBPROC) && !S_ISLNK(inode->i_mode)) {
-			if (opt_dentry) {
-				isec->sclass = inode_mode_to_security_class(inode->i_mode);
-				rc = selinux_proc_get_sid(opt_dentry,
-							  isec->sclass,
-							  &sid);
-				if (rc)
-					goto out_unlock;
-				isec->sid = sid;
-			}
+			/* We must have a dentry to determine the label on
+			 * procfs inodes */
+			if (opt_dentry)
+				/* Called from d_instantiate or
+				 * d_splice_alias. */
+				dentry = dget(opt_dentry);
+			else
+				/* Called from selinux_complete_init, try to
+				 * find a dentry. */
+				dentry = d_find_alias(inode);
+			/*
+			 * This can be hit on boot when a file is accessed
+			 * before the policy is loaded.  When we load policy we
+			 * may find inodes that have no dentry on the
+			 * sbsec->isec_head list.  No reason to complain as
+			 * these will get fixed up the next time we go through
+			 * inode_doinit() with a dentry, before these inodes
+			 * could be used again by userspace.
+			 */
+			if (!dentry)
+				goto out_unlock;
+			isec->sclass = inode_mode_to_security_class(inode->i_mode);
+			rc = selinux_proc_get_sid(dentry, isec->sclass, &sid);
+			dput(dentry);
+			if (rc)
+				goto out_unlock;
+			isec->sid = sid;
 		}
 		break;
 	}

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

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

From: Mikulas Patocka <[email protected]>

commit e20e1d0ac02308e2211306fc67abcd0b2668fb8b upstream.

I found out that a system with k6-3+ processor is unstable during network server load. The system locks up or the network card stops receiving. The reason for the instability is the CPU frequency scaling.

During frequency transition the processor is in "EPM Stop Grant" state. The documentation says that the processor doesn't respond to inquiry requests in this state. Consequently, coherency of processor caches and bus master devices is not maintained, causing the system instability.

This patch flushes the cache during frequency transition. It fixes the instability.

Other minor changes:
* u64 invalue changed to unsigned long because the variable is 32-bit
* move the logic to set the multiplier to a separate function powernow_k6_set_cpu_multiplier
* preserve lower 5 bits of the powernow port instead of 4 (the voltage field has 5 bits)
* mask interrupts when reading the multiplier, so that the port is not open during other activity (running other kernel code with the port open shouldn't cause any misbehavior, but we should better be safe and keep the port closed)

This patch should be backported to all stable kernels. If it doesn't apply cleanly, change it, or ask me to change it.

---
 drivers/cpufreq/powernow-k6.c |   56 +++++++++++++++++++++++++++++-------------
 1 file changed, 39 insertions(+), 17 deletions(-)

@@ -44,23 +44,58 @@ static struct cpufreq_frequency_table cl
 /**
  * powernow_k6_get_cpu_multiplier - returns the current FSB multiplier
  *
- *   Returns the current setting of the frequency multiplier. Core clock
+ * Returns the current setting of the frequency multiplier. Core clock
  * speed is frequency of the Front-Side Bus multiplied with this value.
  */
 static int powernow_k6_get_cpu_multiplier(void)
 {
-	u64 invalue = 0;
+	unsigned long invalue = 0;
 	u32 msrval;
 
+	local_irq_disable();
+
 	msrval = POWERNOW_IOPORT + 0x1;
 	wrmsr(MSR_K6_EPMR, msrval, 0); /* enable the PowerNow port */
 	invalue = inl(POWERNOW_IOPORT + 0x8);
 	msrval = POWERNOW_IOPORT + 0x0;
 	wrmsr(MSR_K6_EPMR, msrval, 0); /* disable it again */
 
+	local_irq_enable();
+
 	return clock_ratio[(invalue >> 5)&7].index;
 }
 
+static void powernow_k6_set_cpu_multiplier(unsigned int best_i)
+{
+	unsigned long outvalue, invalue;
+	unsigned long msrval;
+	unsigned long cr0;
+
+	/* we now need to transform best_i to the BVC format, see AMD#23446 */
+
+	/*
+	 * The processor doesn't respond to inquiry cycles while changing the
+	 * frequency, so we must disable cache.
+	 */
+	local_irq_disable();
+	cr0 = read_cr0();
+	write_cr0(cr0 | X86_CR0_CD);
+	wbinvd();
+
+	outvalue = (1<<12) | (1<<10) | (1<<9) | (best_i<<5);
+
+	msrval = POWERNOW_IOPORT + 0x1;
+	wrmsr(MSR_K6_EPMR, msrval, 0); /* enable the PowerNow port */
+	invalue = inl(POWERNOW_IOPORT + 0x8);
+	invalue = invalue & 0x1f;
+	outvalue = outvalue | invalue;
+	outl(outvalue, (POWERNOW_IOPORT + 0x8));
+	msrval = POWERNOW_IOPORT + 0x0;
+	wrmsr(MSR_K6_EPMR, msrval, 0); /* disable it again */
+
+	write_cr0(cr0);
+	local_irq_enable();
+}
 
 /**
  * powernow_k6_set_state - set the PowerNow! multiplier

@@ -71,8 +106,6 @@ static int powernow_k6_get_cpu_multiplie
 static void powernow_k6_set_state(struct cpufreq_policy *policy,
 		unsigned int best_i)
 {
-	unsigned long outvalue = 0, invalue = 0;
-	unsigned long msrval;
 	struct cpufreq_freqs freqs;
 
 	if (clock_ratio[best_i].index > max_multiplier) {

@@ -85,18 +118,7 @@ static void powernow_k6_set_state(struct
 
 	cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE);
 
-	/* we now need to transform best_i to the BVC format, see AMD#23446 */
-
-	outvalue = (1<<12) | (1<<10) | (1<<9) | (best_i<<5);
-
-	msrval = POWERNOW_IOPORT + 0x1;
-	wrmsr(MSR_K6_EPMR, msrval, 0); /* enable the PowerNow port */
-	invalue = inl(POWERNOW_IOPORT + 0x8);
-	invalue = invalue & 0xf;
-	outvalue = outvalue | invalue;
-	outl(outvalue , (POWERNOW_IOPORT + 0x8));
-	msrval = POWERNOW_IOPORT + 0x0;
-	wrmsr(MSR_K6_EPMR, msrval, 0); /* disable it again */
+	powernow_k6_set_cpu_multiplier(best_i);
 
 	cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE);
 

@@ -164,7 +186,7 @@ static int powernow_k6_cpu_init(struct c
 	}
 
 	/* cpuinfo and default policy values */
-	policy->cpuinfo.transition_latency = 200000;
+	policy->cpuinfo.transition_latency = 500000;
 	policy->cur = busfreq * max_multiplier;
 
 	result = cpufreq_frequency_table_cpuinfo(policy, clock_ratio);

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

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

From: Mikulas Patocka <[email protected]>

commit d82b922a4acc1781d368aceac2f9da43b038cab2 upstream.

The powernow-k6 driver used to read the initial multiplier from the powernow register. However, there is a problem with this:

* If there was a frequency transition before, the multiplier read from the register corresponds to the current multiplier.
* If there was no frequency transition since reset, the field in the register always reads as zero, regardless of the current multiplier that is set using switches on the mainboard and that the CPU is running at.

The zero value corresponds to multiplier 4.5, so as a consequence, the powernow-k6 driver always assumes multiplier 4.5.

For example, if we have 550MHz CPU with bus frequency 100MHz and multiplier 5.5, the powernow-k6 driver thinks that the multiplier is 4.5 and bus frequency is 122MHz. The powernow-k6 driver then sets the multiplier to 4.5, underclocking the CPU to 450MHz, but reports the current frequency as 550MHz.

There is no reliable way how to read the initial multiplier. I modified the driver so that it contains a table of known frequencies (based on parameters of existing CPUs and some common overclocking schemes) and sets the multiplier according to the frequency. If the frequency is unknown (because of unusual overclocking or underclocking), the user must supply the bus speed and maximum multiplier as module parameters.

This patch should be backported to all stable kernels. If it doesn't apply cleanly, change it, or ask me to change it.

---
 drivers/cpufreq/powernow-k6.c |   76 +++++++++++++++++++++++++++++++++++++++---
 1 file changed, 72 insertions(+), 4 deletions(-)

@@ -26,6 +26,14 @@
 static unsigned int                     busfreq;   /* FSB, in 10 kHz */
 static unsigned int                     max_multiplier;
 
+static unsigned int			param_busfreq = 0;
+static unsigned int			param_max_multiplier = 0;
+
+module_param_named(max_multiplier, param_max_multiplier, uint, S_IRUGO);
+MODULE_PARM_DESC(max_multiplier, "Maximum multiplier (allowed values: 20 30 35 40 45 50 55 60)");
+
+module_param_named(bus_frequency, param_busfreq, uint, S_IRUGO);
+MODULE_PARM_DESC(bus_frequency, "Bus frequency in kHz");
 
 /* Clock ratio multiplied by 10 - see table 27 in AMD#23446 */
 static struct cpufreq_frequency_table clock_ratio[] = {

@@ -40,6 +48,27 @@ static struct cpufreq_frequency_table cl
 	{0, CPUFREQ_TABLE_END}
 };
 
+static const struct {
+	unsigned freq;
+	unsigned mult;
+} usual_frequency_table[] = {
+	{ 400000, 40 },	// 100   * 4
+	{ 450000, 45 }, // 100   * 4.5
+	{ 475000, 50 }, //  95   * 5
+	{ 500000, 50 }, // 100   * 5
+	{ 506250, 45 }, // 112.5 * 4.5
+	{ 533500, 55 }, //  97   * 5.5
+	{ 550000, 55 }, // 100   * 5.5
+	{ 562500, 50 }, // 112.5 * 5
+	{ 570000, 60 }, //  95   * 6
+	{ 600000, 60 }, // 100   * 6
+	{ 618750, 55 }, // 112.5 * 5.5
+	{ 660000, 55 }, // 120   * 5.5
+	{ 675000, 60 }, // 112.5 * 6
+	{ 720000, 60 }, // 120   * 6
+};
+
+#define FREQ_RANGE		3000
 
 /**
  * powernow_k6_get_cpu_multiplier - returns the current FSB multiplier

@@ -163,18 +192,57 @@ static int powernow_k6_target(struct cpu
 	return 0;
 }
 
-
 static int powernow_k6_cpu_init(struct cpufreq_policy *policy)
 {
 	unsigned int i, f;
 	int result;
+	unsigned khz;
 
 	if (policy->cpu != 0)
 		return -ENODEV;
 
-	/* get frequencies */
-	max_multiplier = powernow_k6_get_cpu_multiplier();
-	busfreq = cpu_khz / max_multiplier;
+	max_multiplier = 0;
+	khz = cpu_khz;
+	for (i = 0; i < ARRAY_SIZE(usual_frequency_table); i++) {
+		if (khz >= usual_frequency_table[i].freq - FREQ_RANGE &&
+		    khz <= usual_frequency_table[i].freq + FREQ_RANGE) {
+			khz = usual_frequency_table[i].freq;
+			max_multiplier = usual_frequency_table[i].mult;
+			break;
+		}
+	}
+	if (param_max_multiplier) {
+		for (i = 0; (clock_ratio[i].frequency != CPUFREQ_TABLE_END); i++) {
+			if (clock_ratio[i].index == param_max_multiplier) {
+				max_multiplier = param_max_multiplier;
+				goto have_max_multiplier;
+			}
+		}
+		printk(KERN_ERR "powernow-k6: invalid max_multiplier parameter, valid parameters 20, 30, 35, 40, 45, 50, 55, 60\n");
+		return -EINVAL;
+	}
+
+	if (!max_multiplier) {
+		printk(KERN_WARNING "powernow-k6: unknown frequency %u, cannot determine current multiplier\n", khz);
+		printk(KERN_WARNING "powernow-k6: use module parameters max_multiplier and bus_frequency\n");
+		return -EOPNOTSUPP;
+	}
+
+have_max_multiplier:
+	param_max_multiplier = max_multiplier;
+
+	if (param_busfreq) {
+		if (param_busfreq >= 50000 && param_busfreq <= 150000) {
+			busfreq = param_busfreq / 10;
+			goto have_busfreq;
+		}
+		printk(KERN_ERR "powernow-k6: invalid bus_frequency parameter, allowed range 50000 - 150000 kHz\n");
+		return -EINVAL;
+	}
+
+	busfreq = khz / max_multiplier;
+have_busfreq:
+	param_busfreq = busfreq * 10;
 
 	/* table init */
 	for (i = 0; (clock_ratio[i].frequency != CPUFREQ_TABLE_END); i++) {

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

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

From: Mikulas Patocka <[email protected]>

commit 22c73795b101597051924556dce019385a1e2fa0 upstream.

This patch reorders reported frequencies from the highest to the lowest, just like in other frequency drivers.

---
 drivers/cpufreq/powernow-k6.c |   17 ++++++++++-------
 1 file changed, 10 insertions(+), 7 deletions(-)

@@ -37,17 +37,20 @@ MODULE_PARM_DESC(bus_frequency, "Bus fre
 
 /* Clock ratio multiplied by 10 - see table 27 in AMD#23446 */
 static struct cpufreq_frequency_table clock_ratio[] = {
-	{45,  /* 000 -> 4.5x */ 0},
+	{60,  /* 110 -> 6.0x */ 0},
+	{55,  /* 011 -> 5.5x */ 0},
 	{50,  /* 001 -> 5.0x */ 0},
+	{45,  /* 000 -> 4.5x */ 0},
 	{40,  /* 010 -> 4.0x */ 0},
-	{55,  /* 011 -> 5.5x */ 0},
-	{20,  /* 100 -> 2.0x */ 0},
-	{30,  /* 101 -> 3.0x */ 0},
-	{60,  /* 110 -> 6.0x */ 0},
 	{35,  /* 111 -> 3.5x */ 0},
+	{30,  /* 101 -> 3.0x */ 0},
+	{20,  /* 100 -> 2.0x */ 0},
 	{0, CPUFREQ_TABLE_END}
 };
 
+static const u8 index_to_register[8] = { 6, 3, 1, 0, 2, 7, 5, 4 };
+static const u8 register_to_index[8] = { 3, 2, 4, 1, 7, 6, 0, 5 };
+
 static const struct {
 	unsigned freq;
 	unsigned mult;

@@ -91,7 +94,7 @@ static int powernow_k6_get_cpu_multiplie
 
 	local_irq_enable();
 
-	return clock_ratio[(invalue >> 5)&7].index;
+	return clock_ratio[register_to_index[(invalue >> 5)&7]].index;
 }
 
 static void powernow_k6_set_cpu_multiplier(unsigned int best_i)

@@ -111,7 +114,7 @@ static void powernow_k6_set_cpu_multipli
 	write_cr0(cr0 | X86_CR0_CD);
 	wbinvd();
 
-	outvalue = (1<<12) | (1<<10) | (1<<9) | (best_i<<5);
+	outvalue = (1<<12) | (1<<10) | (1<<9) | (index_to_register[best_i]<<5);
 
 	msrval = POWERNOW_IOPORT + 0x1;
 	wrmsr(MSR_K6_EPMR, msrval, 0); /* enable the PowerNow port */

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

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

From: Nicolas Dichtel <[email protected]>

commit c0ff68f1611d6855a06d672989ad5cfea160a4eb upstream.

If headers_install is executed from a deep/long directory structure, the shell's maximum argument length can be execeeded, which breaks the operation with:

| make[2]: execvp: /bin/sh: Argument list too long | make[2]: ***

Instead of passing each files name with the entire path, I give only the file name without the source path and give this path as a new argument to headers_install.pl.

Because there is three possible paths, I have tree input-files list, one per path.

---
 scripts/Makefile.headersinst |   20 ++++++++++++++------
 scripts/headers_install.sh   |    7 +++++--
 2 files changed, 19 insertions(+), 8 deletions(-)

@@ -47,18 +47,24 @@ header-y      := $(filter-out $(generic-
 all-files     := $(header-y) $(genhdr-y) $(wrapper-files)
 output-files  := $(addprefix $(installdir)/, $(all-files))
 
-input-files   := $(foreach hdr, $(header-y), \
+input-files1  := $(foreach hdr, $(header-y), \
 		   $(if $(wildcard $(srcdir)/$(hdr)), \
-			$(wildcard $(srcdir)/$(hdr)), \
+			$(wildcard $(srcdir)/$(hdr))) \
+		   )
+input-files1-name := $(notdir $(input-files1))
+input-files2  := $(foreach hdr, $(header-y), \
+		   $(if  $(wildcard $(srcdir)/$(hdr)),, \
 			$(if $(wildcard $(oldsrcdir)/$(hdr)), \
 				$(wildcard $(oldsrcdir)/$(hdr)), \
 				$(error Missing UAPI file $(srcdir)/$(hdr))) \
-		   )) \
-		 $(foreach hdr, $(genhdr-y), \
+		   ))
+input-files2-name := $(notdir $(input-files2))
+input-files3  := $(foreach hdr, $(genhdr-y), \
 		   $(if	$(wildcard $(gendir)/$(hdr)), \
 			$(wildcard $(gendir)/$(hdr)), \
 			$(error Missing generated UAPI file $(gendir)/$(hdr)) \
 		   ))
+input-files3-name := $(notdir $(input-files3))
 
 # Work out what needs to be removed
 oldheaders    := $(patsubst $(installdir)/%,%,$(wildcard $(installdir)/*.h))

@@ -72,7 +78,9 @@ printdir = $(patsubst $(INSTALL_HDR_PATH
 quiet_cmd_install = INSTALL $(printdir) ($(words $(all-files))\
                             file$(if $(word 2, $(all-files)),s))
       cmd_install = \
-        $(CONFIG_SHELL) $< $(installdir) $(input-files); \
+        $(CONFIG_SHELL) $< $(installdir) $(srcdir) $(input-files1-name); \
+        $(CONFIG_SHELL) $< $(installdir) $(oldsrcdir) $(input-files2-name); \
+        $(CONFIG_SHELL) $< $(installdir) $(gendir) $(input-files3-name); \
         for F in $(wrapper-files); do                                   \
                 echo "\#include <asm-generic/$$F>" > $(installdir)/$$F;    \
         done;                                                           \

@@ -98,7 +106,7 @@ __headersinst: $(subdirs) $(install-file
 	@:
 
 targets += $(install-file)
-$(install-file): scripts/headers_install.sh $(input-files) FORCE
+$(install-file): scripts/headers_install.sh $(input-files1) $(input-files2) $(input-files3) FORCE
 	$(if $(unwanted),$(call cmd,remove),)
 	$(if $(wildcard $(dir $@)),,$(shell mkdir -p $(dir $@)))
 	$(call if_changed,install)

@@ -2,7 +2,7 @@
 
 if [ $# -lt 1 ]
 then
-	echo "Usage: headers_install.sh OUTDIR [FILES...]
+	echo "Usage: headers_install.sh OUTDIR SRCDIR [FILES...]
 	echo
 	echo "Prepares kernel header files for use by user space, by removing"
 	echo "all compiler.h definitions and #includes, removing any"

@@ -10,6 +10,7 @@ then
 	echo "asm/inline/volatile keywords."
 	echo
 	echo "OUTDIR: directory to write each userspace header FILE to."
+	echo "SRCDIR: source directory where files are picked."
 	echo "FILES:  list of header files to operate on."
 
 	exit 1

@@ -19,6 +20,8 @@ fi
 
 OUTDIR="$1"
 shift
+SRCDIR="$1"
+shift
 
 # Iterate through files listed on command line
 

@@ -34,7 +37,7 @@ do
 		-e 's/(^|[^a-zA-Z0-9])__packed([^a-zA-Z0-9_]|$)/\1__attribute__((packed))\2/g' \
 		-e 's/(^|[ \t(])(inline|asm|volatile)([ \t(]|$)/\1__\2__\3/g' \
 		-e 's@#(ifndef|define|endif[ \t]*/[*])[ \t]*_UAPI@#\1 @' \
-		"$i" > "$OUTDIR/$FILE.sed" || exit 1
+		"$SRCDIR/$i" > "$OUTDIR/$FILE.sed" || exit 1
 	scripts/unifdef -U__KERNEL__ -D__EXPORTED_HEADERS__ "$OUTDIR/$FILE.sed" \
 		> "$OUTDIR/$FILE"
 	[ $? -gt 1 ] && exit 1

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

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

From: Daniel Fu <[email protected]>

commit 3b9c10e98021e1f92e6f8c7ce1778b86ba68db10 upstream.

If the current CPU has no cpuidle driver, drv will be NULL in cpuidle_driver_ref(). Check if that is the case before trying to bump up the driver's refcount to prevent the kernel from crashing.

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

@@ -251,7 +251,8 @@ struct cpuidle_driver *cpuidle_driver_re
 	spin_lock(&cpuidle_driver_lock);
 
 	drv = cpuidle_get_driver();
-	drv->refcnt++;
+	if (drv)
+		drv->refcnt++;
 
 	spin_unlock(&cpuidle_driver_lock);
 	return drv;

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

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

From: Nikolay Aleksandrov <[email protected]>

[ Upstream commit 24b9bf43e93e0edd89072da51cf1fab95fc69dec ]

I stumbled upon this very serious bug while hunting for another one, it's a very subtle race condition between inet_frag_evictor, inet_frag_intern and the IPv4/6 frag_queue and expire functions (basically the users of inet_frag_kill/inet_frag_put).

What happens is that after a fragment has been added to the hash chain but before it's been added to the lru_list (inet_frag_lru_add) in inet_frag_intern, it may get deleted (either by an expired timer if the system load is high or the timer sufficiently low, or by the fraq_queue function for different reasons) before it's added to the lru_list, then after it gets added it's a matter of time for the evictor to get to a piece of memory which has been freed leading to a number of different bugs depending on what's left there.

I've been able to trigger this on both IPv4 and IPv6 (which is normal as the frag code is the same), but it's been much more difficult to trigger on IPv4 due to the protocol differences about how fragments are treated.

The setup I used to reproduce this is: 2 machines with 4 x 10G bonded in a RR bond, so the same flow can be seen on multiple cards at the same time. Then I used multiple instances of ping/ping6 to generate fragmented packets and flood the machines with them while running other processes to load the attacked machine.

*It is very important to have the _same flow_ coming in on multiple CPUs concurrently. Usually the attacked machine would die in less than 30 minutes, if configured properly to have many evictor calls and timeouts it could happen in 10 minutes or so.

An important point to make is that any caller (frag_queue or timer) of inet_frag_kill will remove both the timer refcount and the original/guarding refcount thus removing everything that's keeping the frag from being freed at the next inet_frag_put. All of this could happen before the frag was ever added to the LRU list, then it gets added and the evictor uses a freed fragment.

An example for IPv6 would be if a fragment is being added and is at the stage of being inserted in the hash after the hash lock is released, but before inet_frag_lru_add executes (or is able to obtain the lru lock) another overlapping fragment for the same flow arrives at a different CPU which finds it in the hash, but since it's overlapping it drops it invoking inet_frag_kill and thus removing all guarding refcounts, and afterwards freeing it by invoking inet_frag_put which removes the last refcount added previously by inet_frag_find, then inet_frag_lru_add gets executed by inet_frag_intern and we have a freed fragment in the lru_list.

The fix is simple, just move the lru_add under the hash chain locked region so when a removing function is called it'll have to wait for the fragment to be added to the lru_list, and then it'll remove it (it works because the hash chain removal is done before the lru_list one and there's no window between the two list adds when the frag can get dropped). With this fix applied I couldn't kill the same machine in 24 hours with the same setup.

rwlock")

---
 net/ipv4/inet_fragment.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -283,9 +283,10 @@ static struct inet_frag_queue *inet_frag
 
 	atomic_inc(&qp->refcnt);
 	hlist_add_head(&qp->list, &hb->chain);
+	inet_frag_lru_add(nf, qp);
 	spin_unlock(&hb->chain_lock);
 	read_unlock(&f->lock);
-	inet_frag_lru_add(nf, qp);
+
 	return qp;
 }
 

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

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

From: Daniel Borkmann <[email protected]>

[ Upstream commit c485658bae87faccd7aed540fd2ca3ab37992310 ]

While working on ec0223ec48a9 ("net: sctp: fix sctp_sf_do_5_1D_ce to verify if we/peer is AUTH capable"), we noticed that there's a skb memory leakage in the error path.

Running the same reproducer as in ec0223ec48a9 and by unconditionally jumping to the error label (to simulate an error condition) in sctp_sf_do_5_1D_ce() receive path lets kmemleak detector bark about the unfreed chunk->auth_chunk skb clone:

Unreferenced object 0xffff8800b8f3a000 (size 256):
comm "softirq", pid 0, jiffies 4294769856 (age 110.757s) hex dump (first 32 bytes):

    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
    89 ab 75 5e d4 01 58 13 00 00 00 00 00 00 00 00  ..u^..X.........
  backtrace:
    [<ffffffff816660be>] kmemleak_alloc+0x4e/0xb0
    [<ffffffff8119f328>] kmem_cache_alloc+0xc8/0x210
    [<ffffffff81566929>] skb_clone+0x49/0xb0
    [<ffffffffa0467459>] sctp_endpoint_bh_rcv+0x1d9/0x230 [sctp]
    [<ffffffffa046fdbc>] sctp_inq_push+0x4c/0x70 [sctp]
    [<ffffffffa047e8de>] sctp_rcv+0x82e/0x9a0 [sctp]
    [<ffffffff815abd38>] ip_local_deliver_finish+0xa8/0x210
    [<ffffffff815a64af>] nf_reinject+0xbf/0x180
    [<ffffffffa04b4762>] nfqnl_recv_verdict+0x1d2/0x2b0 [nfnetlink_queue]
    [<ffffffffa04aa40b>] nfnetlink_rcv_msg+0x14b/0x250 [nfnetlink]
    [<ffffffff815a3269>] netlink_rcv_skb+0xa9/0xc0
    [<ffffffffa04aa7cf>] nfnetlink_rcv+0x23f/0x408 [nfnetlink]
    [<ffffffff815a2bd8>] netlink_unicast+0x168/0x250
    [<ffffffff815a2fa1>] netlink_sendmsg+0x2e1/0x3f0
    [<ffffffff8155cc6b>] sock_sendmsg+0x8b/0xc0
    [<ffffffff8155d449>] ___sys_sendmsg+0x369/0x380

What happens is that commit bbd0d59809f9 clones the skb containing the AUTH chunk in sctp_endpoint_bh_rcv() when having the edge case that an endpoint requires COOKIE-ECHO chunks to be authenticated:

---------- INIT[RANDOM; CHUNKS; HMAC-ALGO] ----------> <------- INIT-ACK[RANDOM; CHUNKS; HMAC-ALGO] --------- ------------------ AUTH; COOKIE-ECHO ----------------> <-------------------- COOKIE-ACK ---------------------

When we enter sctp_sf_do_5_1D_ce() and before we actually get to the point where we process (and subsequently free) a non-NULL chunk->auth_chunk, we could hit the "goto nomem_init" path from an error condition and thus leave the cloned skb around w/o freeing it.

The fix is to centrally free such clones in sctp_chunk_destroy() handler that is invoked from sctp_chunk_free() after all refs have dropped; and also move both kfree_skb(chunk->auth_chunk) there, so that chunk->auth_chunk is either NULL (since sctp_chunkify() allocs new chunks through kmem_cache_zalloc()) or non-NULL with a valid skb pointer. chunk->skb and chunk->auth_chunk are the only skbs in the sctp_chunk structure that need to be handeled.

While at it, we should use consume_skb() for both. It is the same as dev_kfree_skb() but more appropriately named as we are not a device but a protocol. Also, this effectively replaces the kfree_skb() from both invocations into consume_skb(). Functions are the same only that kfree_skb() assumes that the frame was being dropped after a failure (e.g. for tools like drop monitor), usage of consume_skb() seems more appropriate in function sctp_chunk_destroy() though.

---
 net/sctp/sm_make_chunk.c |    4 ++--
 net/sctp/sm_statefuns.c  |    5 -----
 2 files changed, 2 insertions(+), 7 deletions(-)

@@ -1403,8 +1403,8 @@ static void sctp_chunk_destroy(struct sc
 	BUG_ON(!list_empty(&chunk->list));
 	list_del_init(&chunk->transmitted_list);
 
-	/* Free the chunk skb data and the SCTP_chunk stub itself. */
-	dev_kfree_skb(chunk->skb);
+	consume_skb(chunk->skb);
+	consume_skb(chunk->auth_chunk);
 
 	SCTP_DBG_OBJCNT_DEC(chunk);
 	kmem_cache_free(sctp_chunk_cachep, chunk);

@@ -767,7 +767,6 @@ sctp_disposition_t sctp_sf_do_5_1D_ce(st
 
 		/* Make sure that we and the peer are AUTH capable */
 		if (!net->sctp.auth_enable || !new_asoc->peer.auth_capable) {
-			kfree_skb(chunk->auth_chunk);
 			sctp_association_free(new_asoc);
 			return sctp_sf_pdiscard(net, ep, asoc, type, arg, commands);
 		}

@@ -782,10 +781,6 @@ sctp_disposition_t sctp_sf_do_5_1D_ce(st
 		auth.transport = chunk->transport;
 
 		ret = sctp_sf_authenticate(net, ep, new_asoc, type, &auth);
-
-		/* We can now safely free the auth_chunk clone */
-		kfree_skb(chunk->auth_chunk);
-
 		if (ret != SCTP_IERROR_NO_ERROR) {
 			sctp_association_free(new_asoc);
 			return sctp_sf_pdiscard(net, ep, asoc, type, arg, commands);

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

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

From: =?UTF-8?q?Linus=20L=C3=BCssing?= <[email protected]>

[ Upstream commit 6565b9eeef194afbb3beec80d6dd2447f4091f8c ]

MLD queries are supposed to have an IPv6 link-local source address according to RFC2710, section 4 and RFC3810, section 5.1.14. This patch adds a sanity check to ignore such broken MLD queries.

Without this check, such malformed MLD queries can result in a denial of service: The queries are ignored by any MLD listener therefore they will not respond with an MLD report. However, without this patch these malformed MLD queries would enable the snooping part in the bridge code, potentially shutting down the according ports towards these hosts for multicast traffic as the bridge did not learn about these listeners.

---
 net/bridge/br_multicast.c |    6 ++++++
 1 file changed, 6 insertions(+)

@@ -1176,6 +1176,12 @@ static int br_ip6_multicast_query(struct
 
 	br_multicast_query_received(br, port, !ipv6_addr_any(&ip6h->saddr));
 
+	/* RFC2710+RFC3810 (MLDv1+MLDv2) require link-local source addresses */
+	if (!(ipv6_addr_type(&ip6h->saddr) & IPV6_ADDR_LINKLOCAL)) {
+		err = -EINVAL;
+		goto out;
+	}
+
 	if (skb->len == sizeof(*mld)) {
 		if (!pskb_may_pull(skb, sizeof(*mld))) {
 			err = -EINVAL;

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

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

From: Florian Westphal <[email protected]>

[ Upstream commit e588e2f286ed7da011ed357c24c5b9a554e26595 ]

Quoting Alexander Aring:
While fragmentation and unloading of 6lowpan module I got this kernel Oops after few seconds:

  BUG: unable to handle kernel paging request at f88bbc30
  [..]
  Modules linked in: ipv6 [last unloaded: 6lowpan]
  Call Trace:
   [<c012af4c>] ? call_timer_fn+0x54/0xb3
   [<c012aef8>] ? process_timeout+0xa/0xa
   [<c012b66b>] run_timer_softirq+0x140/0x15f

Problem is that incomplete frags are still around after unload; when their frag expire timer fires, we get crash.

When a netns is removed (also done when unloading module), inet_frag calls the evictor with 'force' argument to purge remaining frags.

The evictor loop terminates when accounted memory ('work') drops to 0 or the lru-list becomes empty. However, the mem accounting is done via percpu counters and may not be accurate, i.e. loop may terminate prematurely.

Alter evictor to only stop once the lru list is empty when force is requested.

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

@@ -211,7 +211,7 @@ int inet_frag_evictor(struct netns_frags
 	}
 
 	work = frag_mem_limit(nf) - nf->low_thresh;
-	while (work > 0) {
+	while (work > 0 || force) {
 		spin_lock(&nf->lru_lock);
 
 		if (list_empty(&nf->lru_list)) {

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

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

From: Eric Dumazet <[email protected]>

[ Upstream commit de1443916791d75fdd26becb116898277bb0273f ]

Some applications didn't expect recvmsg() on a non blocking socket could return -EINTR. This possibility was added as a side effect of commit b3ca9b02b00704 ("net: fix multithreaded signal handling in unix recv routines").

To hit this bug, you need to be a bit unlucky, as the u->readlock mutex is usually held for very small periods.

---
 net/unix/af_unix.c |   17 ++++++++++++-----
 1 file changed, 12 insertions(+), 5 deletions(-)

@@ -1792,8 +1792,11 @@ static int unix_dgram_recvmsg(struct kio
 		goto out;
 
 	err = mutex_lock_interruptible(&u->readlock);
-	if (err) {
-		err = sock_intr_errno(sock_rcvtimeo(sk, noblock));
+	if (unlikely(err)) {
+		/* recvmsg() in non blocking mode is supposed to return -EAGAIN
+		 * sk_rcvtimeo is not honored by mutex_lock_interruptible()
+		 */
+		err = noblock ? -EAGAIN : -ERESTARTSYS;
 		goto out;
 	}
 

@@ -1913,6 +1916,7 @@ static int unix_stream_recvmsg(struct ki
 	struct unix_sock *u = unix_sk(sk);
 	struct sockaddr_un *sunaddr = msg->msg_name;
 	int copied = 0;
+	int noblock = flags & MSG_DONTWAIT;
 	int check_creds = 0;
 	int target;
 	int err = 0;

@@ -1928,7 +1932,7 @@ static int unix_stream_recvmsg(struct ki
 		goto out;
 
 	target = sock_rcvlowat(sk, flags&MSG_WAITALL, size);
-	timeo = sock_rcvtimeo(sk, flags&MSG_DONTWAIT);
+	timeo = sock_rcvtimeo(sk, noblock);
 
 	/* Lock the socket to prevent queue disordering
 	 * while sleeps in memcpy_tomsg

@@ -1940,8 +1944,11 @@ static int unix_stream_recvmsg(struct ki
 	}
 
 	err = mutex_lock_interruptible(&u->readlock);
-	if (err) {
-		err = sock_intr_errno(timeo);
+	if (unlikely(err)) {
+		/* recvmsg() in non blocking mode is supposed to return -EAGAIN
+		 * sk_rcvtimeo is not honored by mutex_lock_interruptible()
+		 */
+		err = noblock ? -EAGAIN : -ERESTARTSYS;
 		goto out;
 	}
 

Since there's apparently little hope that you kindly stop spamming me with this any time soon: The objection to this is still that the non-blocking call shouldn't ever block (and hence, maintain the undocumented property whose loss apparently wasn't noticed by anyone in the last three years(!) as a side effect). That's arguably at least partially my fault because I didn't think about the implications for non-blocking case in 2011. For an example how this should be implemented, have a look at pipe.c (summary: lock uninterruptibly, check state, unlock, go to sleep or return EAGAIN, relock after sleep [if applicable]).

However, if there are actually applications depending on this behaviour, this workaround is surely sensible for dealing with them.

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

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

From: Anton Nayshtut <[email protected]>

[ Upstream commit d2d273ffabd315eecefce21a4391d44b6e156b73 ]

Without this fix, ipv6_exthdrs_offload_init doesn't register IPPROTO_DSTOPTS offload, but returns 0 (as the IPPROTO_ROUTING registration actually succeeds).

This then causes the ipv6_gso_segment to drop IPv6 packets with IPPROTO_DSTOPTS header.

The issue detected and the fix verified by running MS HCK Offload LSO test on top of QEMU Windows guests, as this test sends IPv6 packets with IPPROTO_DSTOPTS.

---
 net/ipv6/exthdrs_offload.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -25,11 +25,11 @@ int __init ipv6_exthdrs_offload_init(voi
 	int ret;
 
 	ret = inet6_add_offload(&rthdr_offload, IPPROTO_ROUTING);
-	if (!ret)
+	if (ret)
 		goto out;
 
 	ret = inet6_add_offload(&dstopt_offload, IPPROTO_DSTOPTS);
-	if (!ret)
+	if (ret)
 		goto out_rt;
 
 out:

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

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

From: Sabrina Dubroca <[email protected]>

[ Upstream commit c88507fbad8055297c1d1e21e599f46960cbee39 ]

DST_NOCOUNT should only be used if an authorized user adds routes locally. In case of routes which are added on behalf of router advertisments this flag must not get used as it allows an unlimited number of routes getting added remotely.

---
 net/ipv6/route.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1428,7 +1428,7 @@ int ip6_route_add(struct fib6_config *cf
 	if (!table)
 		goto out;
 
-	rt = ip6_dst_alloc(net, NULL, DST_NOCOUNT, table);
+	rt = ip6_dst_alloc(net, NULL, (cfg->fc_flags & RTF_ADDRCONF) ? 0 : DST_NOCOUNT, table);
 
 	if (!rt) {
 		err = -ENOMEM;

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

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

From: Peter Boström <[email protected]>

[ Upstream commit dd38743b4cc2f86be250eaf156cf113ba3dd531a ]

With TX VLAN offload enabled the source MAC address for frames sent using the VLAN interface is currently set to the address of the real interface. This is wrong since the VLAN interface may be configured with a different address.

The bug was introduced in commit 2205369a314e12fcec4781cc73ac9c08fc2b47de ("vlan: Fix header ops passthru when doing TX VLAN offload.").

This patch sets the source address before calling the create function of the real interface.

---
 net/8021q/vlan_dev.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -557,6 +557,9 @@ static int vlan_passthru_hard_header(str
 	struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
 	struct net_device *real_dev = vlan->real_dev;
 
+	if (saddr == NULL)
+		saddr = dev->dev_addr;
+
 	return dev_hard_header(skb, real_dev, type, daddr, saddr, len);
 }
 

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

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

From: Eric Dumazet <[email protected]>

[ Upstream commit c3f9b01849ef3bc69024990092b9f42e20df7797 ]

Lars Persson reported following deadlock :

-000 |M:0x0:0x802B6AF8(asm) <-- arch_spin_lock
-001 |tcp_v4_rcv(skb = 0x8BD527A0) <-- sk = 0x8BE6B2A0
-002 |ip_local_deliver_finish(skb = 0x8BD527A0)
-003 |__netif_receive_skb_core(skb = 0x8BD527A0, ?)
-004 |netif_receive_skb(skb = 0x8BD527A0)
-005 |elk_poll(napi = 0x8C770500, budget = 64)
-006 |net_rx_action(?)
-007 |__do_softirq()
-008 |do_softirq()
-009 |local_bh_enable()
-010 |tcp_rcv_established(sk = 0x8BE6B2A0, skb = 0x87D3A9E0, th = 0x814EBE14, ?)
-011 |tcp_v4_do_rcv(sk = 0x8BE6B2A0, skb = 0x87D3A9E0)
-012 |tcp_delack_timer_handler(sk = 0x8BE6B2A0)
-013 |tcp_release_cb(sk = 0x8BE6B2A0)
-014 |release_sock(sk = 0x8BE6B2A0)
-015 |tcp_sendmsg(?, sk = 0x8BE6B2A0, ?, ?)
-016 |sock_sendmsg(sock = 0x8518C4C0, msg = 0x87D8DAA8, size = 4096)
-017 |kernel_sendmsg(?, ?, ?, ?, size = 4096)
-018 |smb_send_kvec()
-019 |smb_send_rqst(server = 0x87C4D400, rqst = 0x87D8DBA0)
-020 |cifs_call_async()
-021 |cifs_async_writev(wdata = 0x87FD6580)
-022 |cifs_writepages(mapping = 0x852096E4, wbc = 0x87D8DC88)
-023 |__writeback_single_inode(inode = 0x852095D0, wbc = 0x87D8DC88)
-024 |writeback_sb_inodes(sb = 0x87D6D800, wb = 0x87E4A9C0, work = 0x87D8DD88)
-025 |__writeback_inodes_wb(wb = 0x87E4A9C0, work = 0x87D8DD88)
-026 |wb_writeback(wb = 0x87E4A9C0, work = 0x87D8DD88)
-027 |wb_do_writeback(wb = 0x87E4A9C0, force_wait = 0)
-028 |bdi_writeback_workfn(work = 0x87E4A9CC)
-029 |process_one_work(worker = 0x8B045880, work = 0x87E4A9CC)
-030 |worker_thread(__worker = 0x8B045880)
-031 |kthread(_create = 0x87CADD90)
-032 |ret_from_kernel_thread(asm)

Bug occurs because __tcp_checksum_complete_user() enables BH, assuming it is running from softirq context.

Lars trace involved a NIC without RX checksum support but other points are problematic as well, like the prequeue stuff.

Problem is triggered by a timer, that found socket being owned by user.

tcp_release_cb() should call tcp_write_timer_handler() or tcp_delack_timer_handler() in the appropriate context :

BH disabled and socket lock held, but 'owned' field cleared, as if they were running from timer handlers.

---
 include/net/sock.h    |    5 +++++
 net/core/sock.c       |    5 ++++-
 net/ipv4/tcp_output.c |   11 +++++++++++
 3 files changed, 20 insertions(+), 1 deletion(-)

@@ -1437,6 +1437,11 @@ static inline void sk_wmem_free_skb(stru
  */
 #define sock_owned_by_user(sk)	((sk)->sk_lock.owned)
 
+static inline void sock_release_ownership(struct sock *sk)
+{
+	sk->sk_lock.owned = 0;
+}
+
 /*
  * Macro so as to not evaluate some arguments when
  * lockdep is not enabled.

@@ -2309,10 +2309,13 @@ void release_sock(struct sock *sk)
 	if (sk->sk_backlog.tail)
 		__release_sock(sk);
 
+	/* Warning : release_cb() might need to release sk ownership,
+	 * ie call sock_release_ownership(sk) before us.
+	 */
 	if (sk->sk_prot->release_cb)
 		sk->sk_prot->release_cb(sk);
 
-	sk->sk_lock.owned = 0;
+	sock_release_ownership(sk);
 	if (waitqueue_active(&sk->sk_lock.wq))
 		wake_up(&sk->sk_lock.wq);
 	spin_unlock_bh(&sk->sk_lock.slock);

@@ -755,6 +755,17 @@ void tcp_release_cb(struct sock *sk)
 	if (flags & (1UL << TCP_TSQ_DEFERRED))
 		tcp_tsq_handler(sk);
 
+	/* Here begins the tricky part :
+	 * We are called from release_sock() with :
+	 * 1) BH disabled
+	 * 2) sk_lock.slock spinlock held
+	 * 3) socket owned by us (sk->sk_lock.owned == 1)
+	 *
+	 * But following code is meant to be called from BH handlers,
+	 * so we should keep BH disabled, but early release socket ownership
+	 */
+	sock_release_ownership(sk);
+
 	if (flags & (1UL << TCP_WRITE_TIMER_DEFERRED)) {
 		tcp_write_timer_handler(sk);
 		__sock_put(sk);

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

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

From: Matthew Leach <[email protected]>

[ Upstream commit dbb490b96584d4e958533fb637f08b557f505657 ]

When copying in a struct msghdr from the user, if the user has set the msg_namelen parameter to a negative value it gets clamped to a valid size due to a comparison between signed and unsigned values.

Ensure the syscall errors when the user passes in a negative value.

---
 net/socket.c |    4 ++++
 1 file changed, 4 insertions(+)

@@ -1964,6 +1964,10 @@ static int copy_msghdr_from_user(struct
 {
 	if (copy_from_user(kmsg, umsg, sizeof(struct msghdr)))
 		return -EFAULT;
+
+	if (kmsg->msg_namelen < 0)
+		return -EINVAL;
+
 	if (kmsg->msg_namelen > sizeof(struct sockaddr_storage))
 		kmsg->msg_namelen = sizeof(struct sockaddr_storage);
 	return 0;

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

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

From: Heiner Kallweit <[email protected]>

[ Upstream commit ecab67015ef6e3f3635551dcc9971cf363cc1cd5 ]

tmp_prefered_lft is an offset to ifp->tstamp, not now. Therefore age needs to be added to the condition.

Age calculation in ipv6_create_tempaddr is different from the one in addrconf_verify and doesn't consider ADDRCONF_TIMER_FUZZ_MINUS. This can cause age in ipv6_create_tempaddr to be less than the one in addrconf_verify and therefore unnecessary temporary address to be generated.
Use age calculation as in addrconf_modify to avoid this.

---
 net/ipv6/addrconf.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

@@ -1111,8 +1111,11 @@ retry:
 	 * Lifetime is greater than REGEN_ADVANCE time units.  In particular,
 	 * an implementation must not create a temporary address with a zero
 	 * Preferred Lifetime.
+	 * Use age calculation as in addrconf_verify to avoid unnecessary
+	 * temporary addresses being generated.
 	 */
-	if (tmp_prefered_lft <= regen_advance) {
+	age = (now - tmp_tstamp + ADDRCONF_TIMER_FUZZ_MINUS) / HZ;
+	if (tmp_prefered_lft <= regen_advance + age) {
 		in6_ifa_put(ifp);
 		in6_dev_put(idev);
 		ret = -1;

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

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

From: lucien <[email protected]>

[ Upstream commit e367c2d03dba4c9bcafad24688fadb79dd95b218 ]

In ip6_append_data_mtu(), when the xfrm mode is not tunnel(such as transport),the ipsec header need to be added in the first fragment, so the mtu will decrease to reserve space for it, then the second fragment come, the mtu should be turn back, as the commit 0c1833797a5a6ec23ea9261d979aa18078720b74 said. however, in the commit a493e60ac4bbe2e977e7129d6d8cbb0dd236be, it use *mtu = min(*mtu, ...) to change the mtu, which lead to the new mtu is alway equal with the first fragment's. and cannot turn back.

when I test through ping6 -c1 -s5000 $ip (mtu=1280):
...frag (0|1232) ESP(spi=0x00002000,seq=0xb), length 1232

...frag (1232|1216)
...frag (2448|1216)
...frag (3664|1216)
...frag (4880|164)

which should be:
...frag (0|1232) ESP(spi=0x00001000,seq=0x1), length 1232

...frag (1232|1232)
...frag (2464|1232)
...frag (3696|1232)
...frag (4928|116)

so delete the min() when change back the mtu.

---
 net/ipv6/ip6_output.c |   14 ++++++--------
 1 file changed, 6 insertions(+), 8 deletions(-)

@@ -1108,21 +1108,19 @@ static void ip6_append_data_mtu(unsigned
 				unsigned int fragheaderlen,
 				struct sk_buff *skb,
 				struct rt6_info *rt,
-				bool pmtuprobe)
+				unsigned int orig_mtu)
 {
 	if (!(rt->dst.flags & DST_XFRM_TUNNEL)) {
 		if (skb == NULL) {
 			/* first fragment, reserve header_len */
-			*mtu = *mtu - rt->dst.header_len;
+			*mtu = orig_mtu - rt->dst.header_len;
 
 		} else {
 			/*
 			 * this fragment is not first, the headers
 			 * space is regarded as data space.
 			 */
-			*mtu = min(*mtu, pmtuprobe ?
-				   rt->dst.dev->mtu :
-				   dst_mtu(rt->dst.path));
+			*mtu = orig_mtu;
 		}
 		*maxfraglen = ((*mtu - fragheaderlen) & ~7)
 			      + fragheaderlen - sizeof(struct frag_hdr);

@@ -1139,7 +1137,7 @@ int ip6_append_data(struct sock *sk, int
 	struct ipv6_pinfo *np = inet6_sk(sk);
 	struct inet_cork *cork;
 	struct sk_buff *skb, *skb_prev = NULL;
-	unsigned int maxfraglen, fragheaderlen, mtu;
+	unsigned int maxfraglen, fragheaderlen, mtu, orig_mtu;
 	int exthdrlen;
 	int dst_exthdrlen;
 	int hh_len;

@@ -1221,6 +1219,7 @@ int ip6_append_data(struct sock *sk, int
 		dst_exthdrlen = 0;
 		mtu = cork->fragsize;
 	}
+	orig_mtu = mtu;
 
 	hh_len = LL_RESERVED_SPACE(rt->dst.dev);
 

@@ -1300,8 +1299,7 @@ alloc_new_skb:
 			if (skb == NULL || skb_prev == NULL)
 				ip6_append_data_mtu(&mtu, &maxfraglen,
 						    fragheaderlen, skb, rt,
-						    np->pmtudisc ==
-						    IPV6_PMTUDISC_PROBE);
+						    orig_mtu);
 
 			skb_prev = skb;
 

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

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

From: David Stevens <[email protected]>

[ Upstream commit 7346135dcd3f9b57f30a5512094848c678d7143e ]

This patch fixes a NULL pointer dereference in the event of an skb allocation failure in arp_reduce().

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

@@ -845,6 +845,9 @@ static int arp_reduce(struct net_device
 
 		neigh_release(n);
 
+		if (reply == NULL)
+			goto out;
+
 		skb_reset_mac_header(reply);
 		__skb_pull(reply, skb_network_offset(reply));
 		reply->ip_summed = CHECKSUM_UNNECESSARY;

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

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

From: Nicolas Dichtel <[email protected]>

[ Upstream commit 1c104a6bebf3c16b6248408b84f91d09ac8a26b6 ]

Commit 3ff661c38c84 ("net: rtnetlink notify events for FDB NTF_SELF adds and deletes") reuses the function nlmsg_populate_fdb_fill() to notify fdb events. But this function was used only for dump and thus was always setting the flag NLM_F_MULTI, which is wrong in case of a single notification.

Libraries like libnl will wait forever for NLMSG_DONE.

---
 net/core/rtnetlink.c |   10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

@@ -1973,12 +1973,13 @@ EXPORT_SYMBOL(rtmsg_ifinfo);
 static int nlmsg_populate_fdb_fill(struct sk_buff *skb,
 				   struct net_device *dev,
 				   u8 *addr, u32 pid, u32 seq,
-				   int type, unsigned int flags)
+				   int type, unsigned int flags,
+				   int nlflags)
 {
 	struct nlmsghdr *nlh;
 	struct ndmsg *ndm;
 
-	nlh = nlmsg_put(skb, pid, seq, type, sizeof(*ndm), NLM_F_MULTI);
+	nlh = nlmsg_put(skb, pid, seq, type, sizeof(*ndm), nlflags);
 	if (!nlh)
 		return -EMSGSIZE;
 

@@ -2016,7 +2017,7 @@ static void rtnl_fdb_notify(struct net_d
 	if (!skb)
 		goto errout;
 
-	err = nlmsg_populate_fdb_fill(skb, dev, addr, 0, 0, type, NTF_SELF);
+	err = nlmsg_populate_fdb_fill(skb, dev, addr, 0, 0, type, NTF_SELF, 0);
 	if (err < 0) {
 		kfree_skb(skb);
 		goto errout;

@@ -2249,7 +2250,8 @@ static int nlmsg_populate_fdb(struct sk_
 
 		err = nlmsg_populate_fdb_fill(skb, dev, ha->addr,
 					      portid, seq,
-					      RTM_NEWNEIGH, NTF_SELF);
+					      RTM_NEWNEIGH, NTF_SELF,
+					      NLM_F_MULTI);
 		if (err < 0)
 			return err;
 skip:

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

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

From: Nicolas Dichtel <[email protected]>

[ Upstream commit 65886f439ab0fdc2dff20d1fa87afb98c6717472 ]

Commit 8cd3ac9f9b7b ("ipmr: advertise new mfc entries via rtnl") reuses the function ipmr_fill_mroute() to notify mfc events.
But this function was used only for dump and thus was always setting the flag NLM_F_MULTI, which is wrong in case of a single notification.

Libraries like libnl will wait forever for NLMSG_DONE.

---
 net/ipv4/ipmr.c |   13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)

@@ -2255,13 +2255,14 @@ int ipmr_get_route(struct net *net, stru
 }
 
 static int ipmr_fill_mroute(struct mr_table *mrt, struct sk_buff *skb,
-			    u32 portid, u32 seq, struct mfc_cache *c, int cmd)
+			    u32 portid, u32 seq, struct mfc_cache *c, int cmd,
+			    int flags)
 {
 	struct nlmsghdr *nlh;
 	struct rtmsg *rtm;
 	int err;
 
-	nlh = nlmsg_put(skb, portid, seq, cmd, sizeof(*rtm), NLM_F_MULTI);
+	nlh = nlmsg_put(skb, portid, seq, cmd, sizeof(*rtm), flags);
 	if (nlh == NULL)
 		return -EMSGSIZE;
 

@@ -2329,7 +2330,7 @@ static void mroute_netlink_event(struct
 	if (skb == NULL)
 		goto errout;
 
-	err = ipmr_fill_mroute(mrt, skb, 0, 0, mfc, cmd);
+	err = ipmr_fill_mroute(mrt, skb, 0, 0, mfc, cmd, 0);
 	if (err < 0)
 		goto errout;
 

@@ -2368,7 +2369,8 @@ static int ipmr_rtm_dumproute(struct sk_
 				if (ipmr_fill_mroute(mrt, skb,
 						     NETLINK_CB(cb->skb).portid,
 						     cb->nlh->nlmsg_seq,
-						     mfc, RTM_NEWROUTE) < 0)
+						     mfc, RTM_NEWROUTE,
+						     NLM_F_MULTI) < 0)
 					goto done;
 next_entry:
 				e++;

@@ -2382,7 +2384,8 @@ next_entry:
 			if (ipmr_fill_mroute(mrt, skb,
 					     NETLINK_CB(cb->skb).portid,
 					     cb->nlh->nlmsg_seq,
-					     mfc, RTM_NEWROUTE) < 0) {
+					     mfc, RTM_NEWROUTE,
+					     NLM_F_MULTI) < 0) {
 				spin_unlock_bh(&mfc_unres_lock);
 				goto done;
 			}

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

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

From: Nicolas Dichtel <[email protected]>

[ Upstream commit f518338b16038beeb73e155e60d0f70beb9379f4 ]

Commit 812e44dd1829 ("ip6mr: advertise new mfc entries via rtnl") reuses the function ip6mr_fill_mroute() to notify mfc events.
But this function was used only for dump and thus was always setting the flag NLM_F_MULTI, which is wrong in case of a single notification.

Libraries like libnl will wait forever for NLMSG_DONE.

---
 net/ipv6/ip6mr.c |   13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)

@@ -2351,13 +2351,14 @@ int ip6mr_get_route(struct net *net,
 }
 
 static int ip6mr_fill_mroute(struct mr6_table *mrt, struct sk_buff *skb,
-			     u32 portid, u32 seq, struct mfc6_cache *c, int cmd)
+			     u32 portid, u32 seq, struct mfc6_cache *c, int cmd,
+			     int flags)
 {
 	struct nlmsghdr *nlh;
 	struct rtmsg *rtm;
 	int err;
 
-	nlh = nlmsg_put(skb, portid, seq, cmd, sizeof(*rtm), NLM_F_MULTI);
+	nlh = nlmsg_put(skb, portid, seq, cmd, sizeof(*rtm), flags);
 	if (nlh == NULL)
 		return -EMSGSIZE;
 

@@ -2425,7 +2426,7 @@ static void mr6_netlink_event(struct mr6
 	if (skb == NULL)
 		goto errout;
 
-	err = ip6mr_fill_mroute(mrt, skb, 0, 0, mfc, cmd);
+	err = ip6mr_fill_mroute(mrt, skb, 0, 0, mfc, cmd, 0);
 	if (err < 0)
 		goto errout;
 

@@ -2464,7 +2465,8 @@ static int ip6mr_rtm_dumproute(struct sk
 				if (ip6mr_fill_mroute(mrt, skb,
 						      NETLINK_CB(cb->skb).portid,
 						      cb->nlh->nlmsg_seq,
-						      mfc, RTM_NEWROUTE) < 0)
+						      mfc, RTM_NEWROUTE,
+						      NLM_F_MULTI) < 0)
 					goto done;
 next_entry:
 				e++;

@@ -2478,7 +2480,8 @@ next_entry:
 			if (ip6mr_fill_mroute(mrt, skb,
 					      NETLINK_CB(cb->skb).portid,
 					      cb->nlh->nlmsg_seq,
-					      mfc, RTM_NEWROUTE) < 0) {
+					      mfc, RTM_NEWROUTE,
+					      NLM_F_MULTI) < 0) {
 				spin_unlock_bh(&mfc_unres_lock);
 				goto done;
 			}

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

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

From: Li RongQing <[email protected]>

[ Not applicable upstream commit, the code here has been removed
upstream. ]

Neighbor Solicitation is ipv6 protocol, so we should check skb->protocol with ETH_P_IPV6

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

@@ -745,7 +745,7 @@ static bool pkt_is_ns(struct sk_buff *sk
 	struct nd_msg *msg;
 	struct ipv6hdr *hdr;
 
-	if (skb->protocol != htons(ETH_P_ARP))
+	if (skb->protocol != htons(ETH_P_IPV6))
 		return false;
 	if (!pskb_may_pull(skb, sizeof(struct ipv6hdr) + sizeof(struct nd_msg)))
 		return false;

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

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

From: Vlad Yasevich <[email protected]>

[ Upstream commit 51dfe7b944998eaeb2b34d314f3a6b16a5fd621b ]

Including hardware acceleration features in vlan_features breaks stacked vlans (Q-in-Q) by marking the bottom vlan interface as capable of acceleration. This causes one of the tags to be lost and the packets are sent with a sing vlan header.

---
 drivers/net/ethernet/broadcom/tg3.c |    5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

@@ -17308,8 +17308,6 @@ static int tg3_init_one(struct pci_dev *
 
 	tg3_init_bufmgr_config(tp);
 
-	features |= NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_CTAG_RX;
-
 	/* 5700 B0 chips do not support checksumming correctly due
 	 * to hardware bugs.
 	 */

@@ -17341,7 +17339,8 @@ static int tg3_init_one(struct pci_dev *
 			features |= NETIF_F_TSO_ECN;
 	}
 
-	dev->features |= features;
+	dev->features |= features | NETIF_F_HW_VLAN_CTAG_TX |
+			 NETIF_F_HW_VLAN_CTAG_RX;
 	dev->vlan_features |= features;
 
 	/*

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

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

From: Oliver Neukum <[email protected]>

[ Upstream commit 14a0d635d18d0fb552dcc979d6d25106e6541f2e ]

This fixes a race which happens by freeing an object on the stack. Quoting Julius:

The fix is to just not allocate the data structure on the stack. As dev->wait is abused as a flag it also takes a runtime PM change to fix this bug.

---
 drivers/net/usb/usbnet.c   |   33 +++++++++++++++++++--------------
 include/linux/usb/usbnet.h |    2 +-
 2 files changed, 20 insertions(+), 15 deletions(-)

@@ -727,14 +727,12 @@ EXPORT_SYMBOL_GPL(usbnet_unlink_rx_urbs)
 // precondition: never called in_interrupt
 static void usbnet_terminate_urbs(struct usbnet *dev)
 {
-	DECLARE_WAIT_QUEUE_HEAD_ONSTACK(unlink_wakeup);
 	DECLARE_WAITQUEUE(wait, current);
 	int temp;
 
 	/* ensure there are no more active urbs */
-	add_wait_queue(&unlink_wakeup, &wait);
+	add_wait_queue(&dev->wait, &wait);
 	set_current_state(TASK_UNINTERRUPTIBLE);
-	dev->wait = &unlink_wakeup;
 	temp = unlink_urbs(dev, &dev->txq) +
 		unlink_urbs(dev, &dev->rxq);
 

@@ -748,15 +746,14 @@ static void usbnet_terminate_urbs(struct
 				  "waited for %d urb completions\n", temp);
 	}
 	set_current_state(TASK_RUNNING);
-	dev->wait = NULL;
-	remove_wait_queue(&unlink_wakeup, &wait);
+	remove_wait_queue(&dev->wait, &wait);
 }
 
 int usbnet_stop (struct net_device *net)
 {
 	struct usbnet		*dev = netdev_priv(net);
 	struct driver_info	*info = dev->driver_info;
-	int			retval;
+	int			retval, pm;
 
 	clear_bit(EVENT_DEV_OPEN, &dev->flags);
 	netif_stop_queue (net);

@@ -766,6 +763,8 @@ int usbnet_stop (struct net_device *net)
 		   net->stats.rx_packets, net->stats.tx_packets,
 		   net->stats.rx_errors, net->stats.tx_errors);
 
+	/* to not race resume */
+	pm = usb_autopm_get_interface(dev->intf);
 	/* allow minidriver to stop correctly (wireless devices to turn off
 	 * radio etc) */
 	if (info->stop) {

@@ -792,6 +791,9 @@ int usbnet_stop (struct net_device *net)
 	dev->flags = 0;
 	del_timer_sync (&dev->delay);
 	tasklet_kill (&dev->bh);
+	if (!pm)
+		usb_autopm_put_interface(dev->intf);
+
 	if (info->manage_power &&
 	    !test_and_clear_bit(EVENT_NO_RUNTIME_PM, &dev->flags))
 		info->manage_power(dev, 0);

@@ -1360,11 +1362,12 @@ static void usbnet_bh (unsigned long par
 	/* restart RX again after disabling due to high error rate */
 	clear_bit(EVENT_RX_KILL, &dev->flags);
 
-	// waiting for all pending urbs to complete?
-	if (dev->wait) {
-		if ((dev->txq.qlen + dev->rxq.qlen + dev->done.qlen) == 0) {
-			wake_up (dev->wait);
-		}
+	/* waiting for all pending urbs to complete?
+	 * only then can we forgo submitting anew
+	 */
+	if (waitqueue_active(&dev->wait)) {
+		if (dev->txq.qlen + dev->rxq.qlen + dev->done.qlen == 0)
+			wake_up_all(&dev->wait);
 
 	// or are we maybe short a few urbs?
 	} else if (netif_running (dev->net) &&

@@ -1502,6 +1505,7 @@ usbnet_probe (struct usb_interface *udev
 	dev->driver_name = name;
 	dev->msg_enable = netif_msg_init (msg_level, NETIF_MSG_DRV
 				| NETIF_MSG_PROBE | NETIF_MSG_LINK);
+	init_waitqueue_head(&dev->wait);
 	skb_queue_head_init (&dev->rxq);
 	skb_queue_head_init (&dev->txq);
 	skb_queue_head_init (&dev->done);

@@ -1694,9 +1698,10 @@ int usbnet_resume (struct usb_interface
 		spin_unlock_irq(&dev->txq.lock);
 
 		if (test_bit(EVENT_DEV_OPEN, &dev->flags)) {
-			/* handle remote wakeup ASAP */
-			if (!dev->wait &&
-				netif_device_present(dev->net) &&
+			/* handle remote wakeup ASAP
+			 * we cannot race against stop
+			 */
+			if (netif_device_present(dev->net) &&
 				!timer_pending(&dev->delay) &&
 				!test_bit(EVENT_RX_HALT, &dev->flags))
 					rx_alloc_submit(dev, GFP_NOIO);

@@ -30,7 +30,7 @@ struct usbnet {
 	struct driver_info	*driver_info;
 	const char		*driver_name;
 	void			*driver_priv;
-	wait_queue_head_t	*wait;
+	wait_queue_head_t	wait;
 	struct mutex		phy_mutex;
 	unsigned char		suspend_count;
 	unsigned char		pkt_cnt, pkt_err;

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

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

From: Vlad Yasevich <[email protected]>

[ Upstream commit fc0d48b8fb449ca007b2057328abf736cb516168 ]

Currently, if the card supports CTAG acceleration we do not account for the vlan header even if we are configuring an 8021AD vlan. This may not be best since we'll do software tagging for 8021AD which will cause data copy on skb head expansion Configure the length based on available hw offload capabilities and vlan protocol.

---
 net/8021q/vlan.c     |    4 +++-
 net/8021q/vlan_dev.c |    3 ++-
 2 files changed, 5 insertions(+), 2 deletions(-)

@@ -305,9 +305,11 @@ static void vlan_sync_address(struct net
 static void vlan_transfer_features(struct net_device *dev,
 				   struct net_device *vlandev)
 {
+	struct vlan_dev_priv *vlan = vlan_dev_priv(vlandev);
+
 	vlandev->gso_max_size = dev->gso_max_size;
 
-	if (dev->features & NETIF_F_HW_VLAN_CTAG_TX)
+	if (vlan_hw_offload_capable(dev->features, vlan->vlan_proto))
 		vlandev->hard_header_len = dev->hard_header_len;
 	else
 		vlandev->hard_header_len = dev->hard_header_len + VLAN_HLEN;

@@ -611,7 +611,8 @@ static int vlan_dev_init(struct net_devi
 #endif
 
 	dev->needed_headroom = real_dev->needed_headroom;
-	if (real_dev->features & NETIF_F_HW_VLAN_CTAG_TX) {
+	if (vlan_hw_offload_capable(real_dev->features,
+				    vlan_dev_priv(dev)->vlan_proto)) {
 		dev->header_ops      = &vlan_passthru_header_ops;
 		dev->hard_header_len = real_dev->hard_header_len;
 	} else {

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

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

From: "Michael S. Tsirkin" <[email protected]>

[ Upstream commit d8316f3991d207fe32881a9ac20241be8fa2bad0 ]

When mergeable buffers are disabled, and the incoming packet is too large for the rx buffer, get_rx_bufs returns success.

This was intentional in order for make recvmsg truncate the packet and then handle_rx would detect err != sock_len and drop it.

Unfortunately we pass the original sock_len to recvmsg - which means we use parts of iov not fully validated.

Fix this up by detecting this overrun and doing packet drop immediately.

CVE-2014-0077

---
 drivers/vhost/net.c |   14 ++++++++++++++
 1 file changed, 14 insertions(+)

@@ -540,6 +540,12 @@ static int get_rx_bufs(struct vhost_virt
 	*iovcount = seg;
 	if (unlikely(log))
 		*log_num = nlogs;
+
+	/* Detect overrun */
+	if (unlikely(datalen > 0)) {
+		r = UIO_MAXIOV + 1;
+		goto err;
+	}
 	return headcount;
 err:
 	vhost_discard_vq_desc(vq, headcount);

@@ -595,6 +601,14 @@ static void handle_rx(struct vhost_net *
 		/* On error, stop handling until the next kick. */
 		if (unlikely(headcount < 0))
 			break;
+		/* On overrun, truncate and discard */
+		if (unlikely(headcount > UIO_MAXIOV)) {
+			msg.msg_iovlen = 1;
+			err = sock->ops->recvmsg(NULL, sock, &msg,
+						 1, MSG_DONTWAIT | MSG_TRUNC);
+			pr_debug("Discarded rx packet: len %zd\n", sock_len);
+			continue;
+		}
 		/* OK, now we need to know about added descriptors. */
 		if (!headcount) {
 			if (unlikely(vhost_enable_notify(&net->dev, vq))) {

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

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

From: "Michael S. Tsirkin" <[email protected]>

[ Upstream commit a39ee449f96a2cd44ce056d8a0a112211a9b1a1f ]

vhost fails to validate negative error code from vhost_get_vq_desc causing
a crash: we are using -EFAULT which is 0xfffffff2 as vector size, which exceeds the allocated size.

The code in question was introduced in commit 8dd014adfea6f173c1ef6378f7e5e7924866c923
vhost-net: mergeable buffers support

CVE-2014-0055

---
 drivers/vhost/net.c |    6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

@@ -513,9 +513,13 @@ static int get_rx_bufs(struct vhost_virt
 			r = -ENOBUFS;
 			goto err;
 		}
-		d = vhost_get_vq_desc(vq->dev, vq, vq->iov + seg,
+		r = vhost_get_vq_desc(vq->dev, vq, vq->iov + seg,
 				      ARRAY_SIZE(vq->iov) - seg, &out,
 				      &in, log, log_num);
+		if (unlikely(r < 0))
+			goto err;
+
+		d = r;
 		if (d == vq->num) {
 			r = 0;
 			goto err;

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

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

From: Paul Durrant <[email protected]>

[ Upstream commit 0576eddf24df716d8570ef8ca11452a9f98eaab2 ]

This patch removes a test in start_new_rx_buffer() that checks whether a copy operation is less than MAX_BUFFER_OFFSET in length, since MAX_BUFFER_OFFSET is defined to be PAGE_SIZE and the only caller of start_new_rx_buffer() already limits copy operations to PAGE_SIZE or less.

---
 drivers/net/xen-netback/netback.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -347,8 +347,8 @@ static bool start_new_rx_buffer(int offs
 	 * into multiple copies tend to give large frags their
 	 * own buffers as before.
 	 */
-	if ((offset + size > MAX_BUFFER_OFFSET) &&
-	    (size <= MAX_BUFFER_OFFSET) && offset && !head)
+	BUG_ON(size > MAX_BUFFER_OFFSET);
+	if ((offset + size > MAX_BUFFER_OFFSET) && offset && !head)
 		return true;
 
 	return false;

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

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

From: Hannes Frederic Sowa <[email protected]>

[ Upstream commit 43a43b6040165f7b40b5b489fe61a4cb7f8c4980 ]

After commit c15b1ccadb323ea ("ipv6: move DAD and addrconf_verify processing to workqueue") some counters are now updated in process context and thus need to disable bh before doing so, otherwise deadlocks can happen on 32-bit archs. Fabio Estevam noticed this while while mounting a NFS volume on an ARM board.

As a compensation for missing this I looked after the other *_STATS_BH and found three other calls which need updating:

1) icmp6_send: ip6_fragment -> icmpv6_send -> icmp6_send (error handling)
2) ip6_push_pending_frames: rawv6_sendmsg -> rawv6_push_pending_frames -> ... (only in case of icmp protocol with raw sockets in error handling)
3) ping6_v6_sendmsg (error handling)

---
 net/ipv6/icmp.c       |    2 +-
 net/ipv6/ip6_output.c |    4 ++--
 net/ipv6/mcast.c      |   11 ++++++-----
 3 files changed, 9 insertions(+), 8 deletions(-)

@@ -508,7 +508,7 @@ static void icmp6_send(struct sk_buff *s
 			      np->tclass, NULL, &fl6, (struct rt6_info *)dst,
 			      MSG_DONTWAIT, np->dontfrag);
 	if (err) {
-		ICMP6_INC_STATS_BH(net, idev, ICMP6_MIB_OUTERRORS);
+		ICMP6_INC_STATS(net, idev, ICMP6_MIB_OUTERRORS);
 		ip6_flush_pending_frames(sk);
 	} else {
 		err = icmpv6_push_pending_frames(sk, &fl6, &tmp_hdr,

@@ -1554,8 +1554,8 @@ int ip6_push_pending_frames(struct sock
 	if (proto == IPPROTO_ICMPV6) {
 		struct inet6_dev *idev = ip6_dst_idev(skb_dst(skb));
 
-		ICMP6MSGOUT_INC_STATS_BH(net, idev, icmp6_hdr(skb)->icmp6_type);
-		ICMP6_INC_STATS_BH(net, idev, ICMP6_MIB_OUTMSGS);
+		ICMP6MSGOUT_INC_STATS(net, idev, icmp6_hdr(skb)->icmp6_type);
+		ICMP6_INC_STATS(net, idev, ICMP6_MIB_OUTMSGS);
 	}
 
 	err = ip6_local_out(skb);

@@ -1439,11 +1439,12 @@ static void mld_sendpack(struct sk_buff
 		      dst_output);
 out:
 	if (!err) {
-		ICMP6MSGOUT_INC_STATS_BH(net, idev, ICMPV6_MLD2_REPORT);
-		ICMP6_INC_STATS_BH(net, idev, ICMP6_MIB_OUTMSGS);
-		IP6_UPD_PO_STATS_BH(net, idev, IPSTATS_MIB_OUTMCAST, payload_len);
-	} else
-		IP6_INC_STATS_BH(net, idev, IPSTATS_MIB_OUTDISCARDS);
+		ICMP6MSGOUT_INC_STATS(net, idev, ICMPV6_MLD2_REPORT);
+		ICMP6_INC_STATS(net, idev, ICMP6_MIB_OUTMSGS);
+		IP6_UPD_PO_STATS(net, idev, IPSTATS_MIB_OUTMCAST, payload_len);
+	} else {
+		IP6_INC_STATS(net, idev, IPSTATS_MIB_OUTDISCARDS);
+	}
 
 	rcu_read_unlock();
 	return;

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

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

From: Pablo Neira <[email protected]>

[ Upstream commit 8b7b932434f5eee495b91a2804f5b64ebb2bc835 ]

nla_strcmp compares the string length plus one, so it's implicitly including the nul-termination in the comparison.

int nla_strcmp(const struct nlattr *nla, const char *str)

 {
        int len = strlen(str) + 1;
        ...
                d = memcmp(nla_data(nla), str, len);

However, if NLA_STRING is used, userspace can send us a string without the nul-termination. This is a problem since the string comparison will not match as the last byte may be not the nul-termination.

Fix this by skipping the comparison of the nul-termination if the attribute data is nul-terminated. Suggested by Thomas Graf.

---
 lib/nlattr.c |   10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

@@ -303,9 +303,15 @@ int nla_memcmp(const struct nlattr *nla,
  */
 int nla_strcmp(const struct nlattr *nla, const char *str)
 {
-	int len = strlen(str) + 1;
-	int d = nla_len(nla) - len;
+	int len = strlen(str);
+	char *buf = nla_data(nla);
+	int attrlen = nla_len(nla);
+	int d;
 
+	if (attrlen > 0 && buf[attrlen - 1] == '\0')
+		attrlen--;
+
+	d = attrlen - len;
 	if (d == 0)
 		d = memcmp(nla_data(nla), str, len);
 

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

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

From: YOSHIFUJI Hideaki <[email protected]>

[ Upstream commit 77bc6bed7121936bb2e019a8c336075f4c8eef62 ]

Return -EINVAL unless all of user-given strings are correctly NUL-terminated.

---
 drivers/isdn/isdnloop/isdnloop.c |    6 ++++++
 1 file changed, 6 insertions(+)

@@ -1070,6 +1070,12 @@ isdnloop_start(isdnloop_card *card, isdn
 		return -EBUSY;
 	if (copy_from_user((char *) &sdef, (char *) sdefp, sizeof(sdef)))
 		return -EFAULT;
+
+	for (i = 0; i < 3; i++) {
+		if (!memchr(sdef.num[i], 0, sizeof(sdef.num[i])))
+			return -EINVAL;
+	}
+
 	spin_lock_irqsave(&card->isdnloop_lock, flags);
 	switch (sdef.ptype) {
 	case ISDN_PTYPE_EURO:

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

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

From: Dan Carpenter <[email protected]>

[ Upstream commit 7563487cbf865284dcd35e9ef5a95380da046737 ]

There are three buffer overflows addressed in this patch.

1) In isdnloop_fake_err() we add an 'E' to a 60 character string and then copy it into a 60 character buffer. I have made the destination buffer 64 characters and I'm changed the sprintf() to a snprintf().

2) In isdnloop_parse_cmd(), p points to a 6 characters into a 60 character buffer so we have 54 characters. The ->eazlist[] is 11 characters long. I have modified the code to return if the source buffer is too long.

3) In isdnloop_command() the cbuf[] array was 60 characters long but the max length of the string then can be up to 79 characters. I made the cbuf array 80 characters long and changed the sprintf() to snprintf(). I also removed the temporary "dial" buffer and changed it to use "p" directly.

Unfortunately, we pass the "cbuf" string from isdnloop_command() to isdnloop_writecmd() which truncates anything over 60 characters to make it fit in card->omsg[]. (It can accept values up to 255 characters so long as there is a '\n' character every 60 characters). For now I have just fixed the memory corruption bug and left the other problems in this driver alone.

---
 drivers/isdn/isdnloop/isdnloop.c |   17 +++++++++--------
 1 file changed, 9 insertions(+), 8 deletions(-)

@@ -518,9 +518,9 @@ static isdnloop_stat isdnloop_cmd_table[
 static void
 isdnloop_fake_err(isdnloop_card *card)
 {
-	char buf[60];
+	char buf[64];
 
-	sprintf(buf, "E%s", card->omsg);
+	snprintf(buf, sizeof(buf), "E%s", card->omsg);
 	isdnloop_fake(card, buf, -1);
 	isdnloop_fake(card, "NAK", -1);
 }

@@ -903,6 +903,8 @@ isdnloop_parse_cmd(isdnloop_card *card)
 	case 7:
 		/* 0x;EAZ */
 		p += 3;
+		if (strlen(p) >= sizeof(card->eazlist[0]))
+			break;
 		strcpy(card->eazlist[ch - 1], p);
 		break;
 	case 8:

@@ -1133,7 +1135,7 @@ isdnloop_command(isdn_ctrl *c, isdnloop_
 {
 	ulong a;
 	int i;
-	char cbuf[60];
+	char cbuf[80];
 	isdn_ctrl cmd;
 	isdnloop_cdef cdef;
 

@@ -1198,7 +1200,6 @@ isdnloop_command(isdn_ctrl *c, isdnloop_
 			break;
 		if ((c->arg & 255) < ISDNLOOP_BCH) {
 			char *p;
-			char dial[50];
 			char dcode[4];
 
 			a = c->arg;

@@ -1210,10 +1211,10 @@ isdnloop_command(isdn_ctrl *c, isdnloop_
 			} else
 				/* Normal Dial */
 				strcpy(dcode, "CAL");
-			strcpy(dial, p);
-			sprintf(cbuf, "%02d;D%s_R%s,%02d,%02d,%s\n", (int) (a + 1),
-				dcode, dial, c->parm.setup.si1,
-				c->parm.setup.si2, c->parm.setup.eazmsn);
+			snprintf(cbuf, sizeof(cbuf),
+				 "%02d;D%s_R%s,%02d,%02d,%s\n", (int) (a + 1),
+				 dcode, p, c->parm.setup.si1,
+				 c->parm.setup.si2, c->parm.setup.eazmsn);
 			i = isdnloop_writecmd(cbuf, strlen(cbuf), 0, card);
 		}
 		break;

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

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

From: Sasha Levin <[email protected]>

[ Upstream commit bf39b4247b8799935ea91d90db250ab608a58e50 ]

Binding might result in a NULL device which is later dereferenced without checking.

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

@@ -239,7 +239,8 @@ static int rds_iw_laddr_check(__be32 add
 	ret = rdma_bind_addr(cm_id, (struct sockaddr *)&sin);
 	/* due to this, we will claim to support IB devices unless we
 	   check node_type. */
-	if (ret || cm_id->device->node_type != RDMA_NODE_RNIC)
+	if (ret || !cm_id->device ||
+	    cm_id->device->node_type != RDMA_NODE_RNIC)
 		ret = -EADDRNOTAVAIL;
 
 	rdsdebug("addr %pI4 ret %d node type %d\n",

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

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

From: Mischa Jonker <[email protected]>

commit 6eda477b3c54b8236868c8784e5e042ff14244f0 upstream.

The Synopsys APB DW UART has a couple of special features that are not in the System C model. In 3.8, the 8250_dw driver didn't really use these features, but from 3.9 onwards, the 8250_dw driver has become incompatible with our model.

---
 arch/arc/boot/dts/nsimosci.dts |    7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

@@ -11,7 +11,7 @@
 
 / {
 	compatible = "snps,nsimosci";
-	clock-frequency = <80000000>;	/* 80 MHZ */
+	clock-frequency = <20000000>;	/* 20 MHZ */
 	#address-cells = <1>;
 	#size-cells = <1>;
 	interrupt-parent = <&intc>;

@@ -44,15 +44,14 @@
 		};
 
 		uart0: serial@c0000000 {
-			compatible = "snps,dw-apb-uart";
+			compatible = "ns8250";
 			reg = <0xc0000000 0x2000>;
 			interrupts = <11>;
-			#clock-frequency = <80000000>;
 			clock-frequency = <3686400>;
 			baud = <115200>;
 			reg-shift = <2>;
 			reg-io-width = <4>;
-			status = "okay";
+			no-loopback-test = <1>;
 		};
 
 		pgu0: pgu@c9000000 {

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

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

From: Vineet Gupta <[email protected]>

commit 61fb4bfc010b0d2940f7fd87acbce6a0f03217cb upstream.

Despite the switch to right UART driver (prev patch), serial console still doesn't work due to missing CONFIG_SERIAL_OF_PLATFORM

Also fix the default cmdline in DT to not refer to out-of-tree ARC framebuffer driver for console.

---
 arch/arc/boot/dts/nsimosci.dts      |    5 ++++-
 arch/arc/configs/nsimosci_defconfig |    1 +
 2 files changed, 5 insertions(+), 1 deletion(-)

@@ -17,7 +17,10 @@
 	interrupt-parent = <&intc>;
 
 	chosen {
-		bootargs = "console=tty0 consoleblank=0";
+		/* this is for console on PGU */
+		/* bootargs = "console=tty0 consoleblank=0"; */
+		/* this is for console on serial */
+		bootargs = "earlycon=uart8250,mmio32,0xc0000000,115200n8 console=ttyS0,115200n8 consoleblank=0 debug";
 	};
 
 	aliases {

@@ -54,6 +54,7 @@ CONFIG_SERIO_ARC_PS2=y
 CONFIG_SERIAL_8250=y
 CONFIG_SERIAL_8250_CONSOLE=y
 CONFIG_SERIAL_8250_DW=y
+CONFIG_SERIAL_OF_PLATFORM=y
 CONFIG_SERIAL_ARC=y
 CONFIG_SERIAL_ARC_CONSOLE=y
 # CONFIG_HW_RANDOM is not set

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

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

From: Heiko Carstens <[email protected]>

commit 03b8c7b623c80af264c4c8d6111e5c6289933666 upstream.

If an architecture has futex_atomic_cmpxchg_inatomic() implemented and there is no runtime check necessary, allow to skip the test within futex_init().

This allows to get rid of some code which would always give the same result, and also allows the compiler to optimize a couple of if statements away.

---
 arch/s390/Kconfig     |    1 +
 include/linux/futex.h |    4 ++++
 init/Kconfig          |    7 +++++++
 kernel/futex.c        |   14 ++++++++++++--
 4 files changed, 24 insertions(+), 2 deletions(-)

@@ -116,6 +116,7 @@ config S390
 	select HAVE_FUNCTION_GRAPH_TRACER
 	select HAVE_FUNCTION_TRACER
 	select HAVE_FUNCTION_TRACE_MCOUNT_TEST
+	select HAVE_FUTEX_CMPXCHG if FUTEX
 	select HAVE_KERNEL_BZIP2
 	select HAVE_KERNEL_GZIP
 	select HAVE_KERNEL_LZMA

@@ -55,7 +55,11 @@ union futex_key {
 #ifdef CONFIG_FUTEX
 extern void exit_robust_list(struct task_struct *curr);
 extern void exit_pi_state_list(struct task_struct *curr);
+#ifdef CONFIG_HAVE_FUTEX_CMPXCHG
+#define futex_cmpxchg_enabled 1
+#else
 extern int futex_cmpxchg_enabled;
+#endif
 #else
 static inline void exit_robust_list(struct task_struct *curr)
 {

@@ -1365,6 +1365,13 @@ config FUTEX
 	  support for "fast userspace mutexes".  The resulting kernel may not
 	  run glibc-based applications correctly.
 
+config HAVE_FUTEX_CMPXCHG
+	bool
+	help
+	  Architectures should select this if futex_atomic_cmpxchg_inatomic()
+	  is implemented and always working. This removes a couple of runtime
+	  checks.
+
 config EPOLL
 	bool "Enable eventpoll support" if EXPERT
 	default y

@@ -67,7 +67,9 @@
 
 #include "rtmutex_common.h"
 
+#ifndef CONFIG_HAVE_FUTEX_CMPXCHG
 int __read_mostly futex_cmpxchg_enabled;
+#endif
 
 #define FUTEX_HASHBITS (CONFIG_BASE_SMALL ? 4 : 8)
 

@@ -2729,10 +2731,10 @@ SYSCALL_DEFINE6(futex, u32 __user *, uad
 	return do_futex(uaddr, op, val, tp, uaddr2, val2, val3);
 }
 
-static int __init futex_init(void)
+static void __init futex_detect_cmpxchg(void)
 {
+#ifndef CONFIG_HAVE_FUTEX_CMPXCHG
 	u32 curval;
-	int i;
 
 	/*
 	 * This will fail and we want it. Some arch implementations do

@@ -2746,6 +2748,14 @@ static int __init futex_init(void)
 	 */
 	if (cmpxchg_futex_value_locked(&curval, NULL, 0, 0) == -EFAULT)
 		futex_cmpxchg_enabled = 1;
+#endif
+}
+
+static int __init futex_init(void)
+{
+	int i;
+
+	futex_detect_cmpxchg();
 
 	for (i = 0; i < ARRAY_SIZE(futex_queues); i++) {
 		plist_head_init(&futex_queues[i].chain);

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

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

From: Finn Thain <[email protected]>

commit e571c58f313d35c56e0018470e3375ddd1fd320e upstream.

Skip the futex_atomic_cmpxchg_inatomic() test in futex_init(). It causes a fatal exception on 68030 (and presumably 68020 also).

---
 arch/m68k/Kconfig |    1 +
 1 file changed, 1 insertion(+)

@@ -16,6 +16,7 @@ config M68K
 	select FPU if MMU
 	select ARCH_WANT_IPC_PARSE_VERSION
 	select ARCH_USES_GETTIMEOFFSET if MMU && !COLDFIRE
+	select HAVE_FUTEX_CMPXCHG if MMU && FUTEX
 	select HAVE_MOD_ARCH_SPECIFIC
 	select MODULES_USE_ELF_REL
 	select MODULES_USE_ELF_RELA

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

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

From: Ard Biesheuvel <[email protected]>

commit 8ceee72808d1ae3fb191284afc2257a2be964725 upstream.

The GHASH setkey() function uses SSE registers but fails to call kernel_fpu_begin()/kernel_fpu_end(). Instead of adding these calls, and then having to deal with the restriction that they cannot be called from interrupt context, move the setkey() implementation to the C domain.

Note that setkey() does not use any particular SSE features and is not expected to become a performance bottleneck.

---
 arch/x86/crypto/ghash-clmulni-intel_asm.S  |   29 -----------------------------
 arch/x86/crypto/ghash-clmulni-intel_glue.c |   14 +++++++++++---
 2 files changed, 11 insertions(+), 32 deletions(-)

@@ -24,10 +24,6 @@
 .align 16
 .Lbswap_mask:
 	.octa 0x000102030405060708090a0b0c0d0e0f
-.Lpoly:
-	.octa 0xc2000000000000000000000000000001
-.Ltwo_one:
-	.octa 0x00000001000000000000000000000001
 
 #define DATA	%xmm0
 #define SHASH	%xmm1

@@ -134,28 +130,3 @@ ENTRY(clmul_ghash_update)
 .Lupdate_just_ret:
 	ret
 ENDPROC(clmul_ghash_update)
-
-/*
- * void clmul_ghash_setkey(be128 *shash, const u8 *key);
- *
- * Calculate hash_key << 1 mod poly
- */
-ENTRY(clmul_ghash_setkey)
-	movaps .Lbswap_mask, BSWAP
-	movups (%rsi), %xmm0
-	PSHUFB_XMM BSWAP %xmm0
-	movaps %xmm0, %xmm1
-	psllq $1, %xmm0
-	psrlq $63, %xmm1
-	movaps %xmm1, %xmm2
-	pslldq $8, %xmm1
-	psrldq $8, %xmm2
-	por %xmm1, %xmm0
-	# reduction
-	pshufd $0b00100100, %xmm2, %xmm1
-	pcmpeqd .Ltwo_one, %xmm1
-	pand .Lpoly, %xmm1
-	pxor %xmm1, %xmm0
-	movups %xmm0, (%rdi)
-	ret
-ENDPROC(clmul_ghash_setkey)

@@ -30,8 +30,6 @@ void clmul_ghash_mul(char *dst, const be
 void clmul_ghash_update(char *dst, const char *src, unsigned int srclen,
 			const be128 *shash);
 
-void clmul_ghash_setkey(be128 *shash, const u8 *key);
-
 struct ghash_async_ctx {
 	struct cryptd_ahash *cryptd_tfm;
 };

@@ -58,13 +56,23 @@ static int ghash_setkey(struct crypto_sh
 			const u8 *key, unsigned int keylen)
 {
 	struct ghash_ctx *ctx = crypto_shash_ctx(tfm);
+	be128 *x = (be128 *)key;
+	u64 a, b;
 
 	if (keylen != GHASH_BLOCK_SIZE) {
 		crypto_shash_set_flags(tfm, CRYPTO_TFM_RES_BAD_KEY_LEN);
 		return -EINVAL;
 	}
 
-	clmul_ghash_setkey(&ctx->shash, key);
+	/* perform multiplication by 'x' in GF(2^128) */
+	a = be64_to_cpu(x->a);
+	b = be64_to_cpu(x->b);
+
+	ctx->shash.a = (__be64)((b << 1) | (a >> 63));
+	ctx->shash.b = (__be64)((a << 1) | (b >> 63));
+
+	if (a >> 63)
+		ctx->shash.b ^= cpu_to_be64(0xc2);
 
 	return 0;
 }

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

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

From: Xiaoguang Chen <[email protected]>

commit 95731ebb114c5f0c028459388560fc2a72fe5049 upstream.

Cpufreq governors' stop and start operations should be carried out in sequence. Otherwise, there will be unexpected behavior, like in the example below.

Suppose there are 4 CPUs and policy->cpu=CPU0, CPU1/2/3 are linked to CPU0. The normal sequence is:

1) Current governor is userspace. An application tries to set the governor to ondemand. It will call __cpufreq_set_policy() in which it will stop the userspace governor and then start the ondemand governor.

2) Current governor is userspace. The online of CPU3 runs on CPU0. It will call cpufreq_add_policy_cpu() in which it will first stop the userspace governor, and then start it again.

If the sequence of the above two cases interleaves, it becomes:

1) Application stops userspace governor
2) Hotplug stops userspace governor

which is a problem, because the governor shouldn't be stopped twice in a row. What happens next is:

3) Application starts ondemand governor
4) Hotplug starts a governor

In step 4, the hotplug is supposed to start the userspace governor, but now the governor has been changed by the application to ondemand, so the ondemand governor is started once again, which is incorrect.

The solution is to prevent policy governors from being stopped multiple times in a row. A governor should only be stopped once for one policy. After it has been stopped, no more governor stop operations should be executed.

Also add a mutex to serialize governor operations.

---
 drivers/cpufreq/cpufreq.c |   24 ++++++++++++++++++++++++
 include/linux/cpufreq.h   |    1 +
 2 files changed, 25 insertions(+)

@@ -46,6 +46,7 @@ static DEFINE_PER_CPU(struct cpufreq_pol
 static DEFINE_PER_CPU(char[CPUFREQ_NAME_LEN], cpufreq_cpu_governor);
 #endif
 static DEFINE_RWLOCK(cpufreq_driver_lock);
+static DEFINE_MUTEX(cpufreq_governor_lock);
 
 /*
  * cpu_policy_rwsem is a per CPU reader-writer semaphore designed to cure

@@ -1563,6 +1564,21 @@ static int __cpufreq_governor(struct cpu
 
 	pr_debug("__cpufreq_governor for CPU %u, event %u\n",
 						policy->cpu, event);
+
+	mutex_lock(&cpufreq_governor_lock);
+	if ((!policy->governor_enabled && (event == CPUFREQ_GOV_STOP)) ||
+	    (policy->governor_enabled && (event == CPUFREQ_GOV_START))) {
+		mutex_unlock(&cpufreq_governor_lock);
+		return -EBUSY;
+	}
+
+	if (event == CPUFREQ_GOV_STOP)
+		policy->governor_enabled = false;
+	else if (event == CPUFREQ_GOV_START)
+		policy->governor_enabled = true;
+
+	mutex_unlock(&cpufreq_governor_lock);
+
 	ret = policy->governor->governor(policy, event);
 
 	if (!ret) {

@@ -1570,6 +1586,14 @@ static int __cpufreq_governor(struct cpu
 			policy->governor->initialized++;
 		else if (event == CPUFREQ_GOV_POLICY_EXIT)
 			policy->governor->initialized--;
+	} else {
+		/* Restore original values */
+		mutex_lock(&cpufreq_governor_lock);
+		if (event == CPUFREQ_GOV_STOP)
+			policy->governor_enabled = true;
+		else if (event == CPUFREQ_GOV_START)
+			policy->governor_enabled = false;
+		mutex_unlock(&cpufreq_governor_lock);
 	}
 
 	/* we keep one module reference alive for

@@ -107,6 +107,7 @@ struct cpufreq_policy {
 	unsigned int		policy; /* see above */
 	struct cpufreq_governor	*governor; /* see below */
 	void			*governor_data;
+	bool			governor_enabled; /* governor start/stop flag */
 
 	struct work_struct	update; /* if update_policy() needs to be
 					 * called, but you're in IRQ context */

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

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

From: Stephen Boyd <[email protected]>

commit 3617f2ca6d0eba48114308532945a7f1577816a4 upstream.

When a CPU is hot removed we'll cancel all the delayed work items via gov_cancel_work(). Normally this will just cancels a delayed timer on each CPU that the policy is managing and the work won't run, but if the work is already running the workqueue code will wait for the work to finish before continuing to prevent the work items from re-queuing themselves like they normally do. This scheme will work most of the time, except for the case where the work function determines that it should adjust the delay for all other CPUs that the policy is managing. If this scenario occurs, the canceling CPU will cancel its own work but queue up the other CPUs works to run. For example:

CPU0 CPU1
---- ----
cpu_down()
...
__cpufreq_remove_dev()
cpufreq_governor_dbs()
case CPUFREQ_GOV_STOP:
gov_cancel_work(dbs_data, policy);
cpu0 work is canceled
timer is canceled
cpu1 work is canceled <work runs>
<waits for cpu1> od_dbs_timer()
gov_queue_work(*, *, true);
cpu0 work queued cpu1 work queued cpu2 work queued
...
cpu1 work is canceled
cpu2 work is canceled
...

At the end of the GOV_STOP case cpu0 still has a work queued to run although the code is expecting all of the works to be canceled. __cpufreq_remove_dev() will then proceed to re-initialize all the other CPUs works except for the CPU that is going down. The CPUFREQ_GOV_START case in cpufreq_governor_dbs() will trample over the queued work and debugobjects will spit out a warning:

WARNING: at lib/debugobjects.c:260 debug_print_object+0x94/0xbc()
ODEBUG: init active (active state 0) object type: timer_list hint: delayed_work_timer_fn+0x0/0x10
Modules linked in:
CPU: 0 PID: 1491 Comm: sh Tainted: G        W    3.10.0 #19
[<c010c178>] (unwind_backtrace+0x0/0x11c) from [<c0109dec>] (show_stack+0x10/0x14)
[<c0109dec>] (show_stack+0x10/0x14) from [<c01904cc>] (warn_slowpath_common+0x4c/0x6c)
[<c01904cc>] (warn_slowpath_common+0x4c/0x6c) from [<c019056c>] (warn_slowpath_fmt+0x2c/0x3c)
[<c019056c>] (warn_slowpath_fmt+0x2c/0x3c) from [<c0388a7c>] (debug_print_object+0x94/0xbc)
[<c0388a7c>] (debug_print_object+0x94/0xbc) from [<c0388e34>] (__debug_object_init+0x2d0/0x340)
[<c0388e34>] (__debug_object_init+0x2d0/0x340) from [<c019e3b0>] (init_timer_key+0x14/0xb0)
[<c019e3b0>] (init_timer_key+0x14/0xb0) from [<c0635f78>] (cpufreq_governor_dbs+0x3e8/0x5f8)
[<c0635f78>] (cpufreq_governor_dbs+0x3e8/0x5f8) from [<c06325a0>] (__cpufreq_governor+0xdc/0x1a4)
[<c06325a0>] (__cpufreq_governor+0xdc/0x1a4) from [<c0633704>] (__cpufreq_remove_dev.isra.10+0x3b4/0x434)
[<c0633704>] (__cpufreq_remove_dev.isra.10+0x3b4/0x434) from [<c08989f4>] (cpufreq_cpu_callback+0x60/0x80)
[<c08989f4>] (cpufreq_cpu_callback+0x60/0x80) from [<c08a43c0>] (notifier_call_chain+0x38/0x68)
[<c08a43c0>] (notifier_call_chain+0x38/0x68) from [<c01938e0>] (__cpu_notify+0x28/0x40)
[<c01938e0>] (__cpu_notify+0x28/0x40) from [<c0892ad4>] (_cpu_down+0x7c/0x2c0)
[<c0892ad4>] (_cpu_down+0x7c/0x2c0) from [<c0892d3c>] (cpu_down+0x24/0x40)
[<c0892d3c>] (cpu_down+0x24/0x40) from [<c0893ea8>] (store_online+0x2c/0x74)
[<c0893ea8>] (store_online+0x2c/0x74) from [<c04519d8>] (dev_attr_store+0x18/0x24)
[<c04519d8>] (dev_attr_store+0x18/0x24) from [<c02a69d4>] (sysfs_write_file+0x100/0x148)
[<c02a69d4>] (sysfs_write_file+0x100/0x148) from [<c0255c18>] (vfs_write+0xcc/0x174)
[<c0255c18>] (vfs_write+0xcc/0x174) from [<c0255f70>] (SyS_write+0x38/0x64)
[<c0255f70>] (SyS_write+0x38/0x64) from [<c0106120>] (ret_fast_syscall+0x0/0x30)

---
 drivers/cpufreq/cpufreq_governor.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -177,6 +177,9 @@ void gov_queue_work(struct dbs_data *dbs
 {
 	int i;
 
+	if (!policy->governor_enabled)
+		return;
+
 	if (!all_cpus) {
 		__gov_queue_work(smp_processor_id(), dbs_data, delay);
 	} else {

Build results:

	total: 126 pass: 122 skipped: 4 fail: 0

Qemu tests all passed. Results are as expected.

Details are available at http://server.roeck-us.net:8010/builders.

Guenter

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

-- Shuah

Shuah Khan
Senior Linux Kernel Developer - Open Source Group
Samsung Research America(Silicon Valley)
[email protected] | (970) 672-0658