netfilter: nf_tables: set NLM_F_DUMP_INTR if netlink dumping is stale
An updater may interfer with the dumping of any of the object lists.
Fix this by using a per-net generation counter and use the
nl_dump_check_consistent() interface so the NLM_F_DUMP_INTR flag is set
to notify userspace that it has to restart the dump since an updater
has interfered.
This patch also replaces the existing consistency checking code in the
rule dumping path since it is broken. Basically, the value that the
dump callback returns is not propagated to userspace via
netlink_dump_start().
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
netfilter: nf_tables: safe RCU iteration on list when dumping
The dump operation through netlink is not protected by the nfnl_lock.
Thus, a reader process can be dumping any of the existing object
lists while another process can be updating the list content.
This patch resolves this situation by protecting all the object
lists with RCU in the netlink dump path which is the reader side.
The updater path is already protected via nfnl_lock, so use list
manipulation RCU-safe operations.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Hangbin Liu [Fri, 27 Jun 2014 01:57:53 +0000 (09:57 +0800)]
ipv6: Fix MLD Query message check
Based on RFC3810 6.2, we also need to check the hop limit and router alert
option besides source address.
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com> Acked-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org> Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
James M Leddy [Wed, 25 Jun 2014 21:38:13 +0000 (17:38 -0400)]
udp: Add MIB counters for rcvbuferrors
Add MIB counters for rcvbuferrors in UDP to help diagnose problems.
Signed-off-by: James M Leddy <james.leddy@redhat.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Linus Torvalds [Thu, 26 Jun 2014 04:47:28 +0000 (21:47 -0700)]
Merge branch 'for-next' of git://git.samba.org/sfrench/cifs-2.6
Pull CIFS fixes from Steve French:
"Small set of misc cifs/smb3 fixes"
* 'for-next' of git://git.samba.org/sfrench/cifs-2.6:
[CIFS] fix mount failure with broken pathnames when smb3 mount with mapchars option
cifs: revalidate mapping prior to satisfying read_iter request with cache=loose
fs/cifs: fix regression in cifs_create_mf_symlink()
Linus Torvalds [Thu, 26 Jun 2014 04:38:45 +0000 (21:38 -0700)]
Merge tag 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging
Pull hwmon fixes from Guenter Roeck:
"Various minor fixes"
* tag 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:
hwmon: (gpio-fan) Change name used in hwmon_device_register_with_groups
hwmon: (emc1403) Fix missing 'select REGMAP_I2C' in Kconfig
hwmon: (ntc_thermistor) Use the manufacturer name properly
devicetree: bindings: Document murata vendor prefix
hwmon: (w83l786ng) Report correct minimum fan speed
1) Fix crash in ipvs tot_stats estimator, from Julian Anastasov.
2) Fix OOPS in nf_nat on netns removal, from Florian Westphal.
3) Really really really fix locking issues in slip and slcan tty write
wakeups, from Tyler Hall.
4) Fix checksum offloading in fec driver, from Fugang Duan.
5) Off by one in BPF instruction limit test, from Kees Cook.
6) Need to clear all TSO capability flags when doing software TSO in
tg3 driver, from Prashant Sreedharan.
7) Fix memory leak in vlan_reorder_header() error path, from Li
RongQing.
8) Fix various bugs in xen-netfront and xen-netback multiqueue support,
from David Vrabel and Wei Liu.
9) Fix deadlock in cxgb4 driver, from Li RongQing.
10) Prevent double free of no-cache DST entries, from Eric Dumazet.
11) Bad csum_start handling in skb_segment() leads to crashes when
forwarding, from Tom Herbert.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (76 commits)
net: fix setting csum_start in skb_segment()
ipv4: fix dst race in sk_dst_get()
net: filter: Use kcalloc/kmalloc_array to allocate arrays
trivial: net: filter: Change kerneldoc parameter order
trivial: net: filter: Fix typo in comment
net: allwinner: emac: Add missing free_irq
cxgb4: use dev_port to identify ports
xen-netback: bookkeep number of active queues in our own module
tg3: Change nvram command timeout value to 50ms
cxgb4: Not need to hold the adap_rcu_lock lock when read adap_rcu_list
be2net: fix qnq mode detection on VFs
of: mdio: fixup of_phy_register_fixed_link parsing of new bindings
at86rf230: fix irq setup
net: phy: at803x: fix coccinelle warnings
net/mlx4_core: Fix the error flow when probing with invalid VF configuration
tulip: Poll link status more frequently for Comet chips
net: huawei_cdc_ncm: increase command buffer size
drivers: net: cpsw: fix dual EMAC stall when connected to same switch
xen-netfront: recreate queues correctly when reconnecting
xen-netfront: fix oops when disconnected from backend
...
It looks like a likely culprit is that SKB_GSO_CB()->csum_start is
not set correctly when doing non-scatter gather. We are using
offset as opposed to doffset.
Reported-by: Dave Jones <davej@redhat.com> Tested-by: Dave Jones <davej@redhat.com> Signed-off-by: Tom Herbert <therbert@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Fixes: 7e2b10c1e52ca ("net: Support for multiple checksums with gso") Acked-by: Tom Herbert <therbert@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Linus Torvalds [Thu, 26 Jun 2014 03:06:06 +0000 (20:06 -0700)]
Merge tag 'nfs-for-3.16-2' of git://git.linux-nfs.org/projects/trondmy/linux-nfs
Pull NFS client fixes from Trond Myklebust:
"Highlights include:
- Stable fix for a data corruption case due to incorrect cache
validation
- Fix a couple of false positive cache invalidations
- Fix NFSv4 security negotiation issues"
* tag 'nfs-for-3.16-2' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
NFSv4: test SECINFO RPC_AUTH_GSS pseudoflavors for support
NFS Return -EPERM if no supported or matching SECINFO flavor
NFS check the return of nfs4_negotiate_security in nfs4_submount
NFS: Don't mark the data cache as invalid if it has been flushed
NFS: Clear NFS_INO_REVAL_PAGECACHE when we update the file size
nfs: Fix cache_validity check in nfs_write_pageuptodate()
Eric Dumazet [Tue, 24 Jun 2014 17:05:11 +0000 (10:05 -0700)]
ipv4: fix dst race in sk_dst_get()
When IP route cache had been removed in linux-3.6, we broke assumption
that dst entries were all freed after rcu grace period. DST_NOCACHE
dst were supposed to be freed from dst_release(). But it appears
we want to keep such dst around, either in UDP sockets or tunnels.
In sk_dst_get() we need to make sure dst refcount is not 0
before incrementing it, or else we might end up freeing a dst
twice.
DST_NOCACHE set on a dst does not mean this dst can not be attached
to a socket or a tunnel.
Then, before actual freeing, we need to observe a rcu grace period
to make sure all other cpus can catch the fact the dst is no longer
usable.
Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Dormando <dormando@rydia.net> Signed-off-by: David S. Miller <davem@davemloft.net>
Tobias Klauser [Tue, 24 Jun 2014 13:33:22 +0000 (15:33 +0200)]
net: filter: Use kcalloc/kmalloc_array to allocate arrays
Use kcalloc/kmalloc_array to make it clear we're allocating arrays. No
integer overflow can actually happen here, since len/flen is guaranteed
to be less than BPF_MAXINSNS (4096). However, this changed makes sure
we're not going to get one if BPF_MAXINSNS were ever increased.
Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Acked-by: Daniel Borkmann <dborkman@redhat.com> Acked-by: Alexei Starovoitov <ast@plumgrid.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Tobias Klauser [Tue, 24 Jun 2014 13:33:21 +0000 (15:33 +0200)]
trivial: net: filter: Change kerneldoc parameter order
Change the order of the parameters to sk_unattached_filter_create() in
the kerneldoc to reflect the order they appear in the actual function.
This fix is only cosmetic, in the generated doc they still appear in the
correct order without the fix.
Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Acked-by: Daniel Borkmann <dborkman@redhat.com> Acked-by: Alexei Starovoitov <ast@plumgrid.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Tobias Klauser [Tue, 24 Jun 2014 13:33:20 +0000 (15:33 +0200)]
trivial: net: filter: Fix typo in comment
Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Acked-by: Daniel Borkmann <dborkman@redhat.com> Acked-by: Alexei Starovoitov <ast@plumgrid.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Maxime Ripard [Mon, 23 Jun 2014 20:49:40 +0000 (22:49 +0200)]
net: allwinner: emac: Add missing free_irq
If the mdio probe function fails in emac_open, the interrupt we just requested
isn't freed. If emac_open is called again, for example because we try to set up
the interface again, the kernel will oops because the interrupt wasn't properly
released.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com> Cc: <stable@vger.kernel.org> # 3.11+ Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 3f85944fe207d0225ef21a2c0951d4946fc9a95d ("net: Add sysfs file
for port number") introduce dev_port to network devices. cxgb4 adapters
have multiple ports on the same PCI function, and used dev_id to
identify those ports. That use was removed by commit 8c367fcbe6549195d2eb11e62bea233f811aad41 ("cxgb4: Do not set
net_device::dev_id to VI index"), since dev_id should be used only when
devices share the same MAC address.
Using dev_port for cxgb4 allows different ports on the same PCI function
to be identified.
Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Wei Liu [Mon, 23 Jun 2014 09:50:17 +0000 (10:50 +0100)]
xen-netback: bookkeep number of active queues in our own module
The original code uses netdev->real_num_tx_queues to bookkeep number of
queues and invokes netif_set_real_num_tx_queues to set the number of
queues. However, netif_set_real_num_tx_queues doesn't allow
real_num_tx_queues to be smaller than 1, which means setting the number
to 0 will not work and real_num_tx_queues is untouched.
This is bogus when xenvif_free is invoked before any number of queues is
allocated. That function needs to iterate through all queues to free
resources. Using the wrong number of queues results in NULL pointer
dereference.
So we bookkeep the number of queues in xen-netback to solve this
problem. This fixes a regression introduced by multiqueue patchset in
3.16-rc1.
There's another bug in original code that the real number of RX queues
is never set. In current Xen multiqueue design, the number of TX queues
and RX queues are in fact the same. We need to set the numbers of TX and
RX queues to the same value.
Also remove xenvif_select_queue and leave queue selection to core
driver, as suggested by David Miller.
Reported-by: Boris Ostrovsky <boris.ostrovsky@oracle.com> Signed-off-by: Wei Liu <wei.liu2@citrix.com> CC: Ian Campbell <ian.campbell@citrix.com> CC: Paul Durrant <paul.durrant@citrix.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 506724c463fcd63477a5e404728a980b71f80bb7 "tg3: Override clock,
link aware and link idle mode during NVRAM dump" changed the timeout
value for nvram command execution from 100ms to 1ms. But the 1ms
timeout value was only sufficient for nvram read operations but not
write operations for most of the devices supported by tg3 driver.
This patch sets the MAX to 50ms. Also it uses usleep_range instead
of udelay.
Signed-off-by: Prashant Sreedharan <prashant@broadcom.com> Signed-off-by: Michael Chan <mchan@broadcom.com> Suggested-by: David Miller <davem@davemloft.net> Signed-off-by: David S. Miller <davem@davemloft.net>
Linus Torvalds [Wed, 25 Jun 2014 19:19:01 +0000 (12:19 -0700)]
Merge tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc
Pull ARM SoC fixes from Arnd Bergmann:
"A new set of bug fixes for 3.16, containing patches for seven
platforms:
at91:
- drivers/misc fix for Kconfig PWM symbol
- correction of several values in DT after conversion to CCF
- fix at91sam9261/at91sam9261ek mistake in slow crystal vs. slow RC osc
imx:
- Use GPIO for card CD/WP on imx51-babbage and eukrea-mbimxsd51,
because controller base CD/WP is not working in esdhc driver due to
runtime PM support
- A couple of random ventana gw5xxx board fixes
- Add IMX_IPUV3_CORE back to defconfig, which gets lost when moving
IPUv3 driver out of staging tree
- Fix enet/fec clock selection on imx6sl
- Fix display node on imx53-m53evk board
- A couple of Cubox-i updates from Russell, which were omitted from
the merge window due to dependency
integrator:
- fix an OF-related regression against 3.15
mvebu:
- mvebu (v7)
- Fix broken SoC ID detection
- Select ARM_CPU_SUSPEND for v7
- Remove armada38x compatible string (no users yet)
- Enable Dove SoC in mvebu_v7_defconfig
- kirkwood
- Fix phy-connection-type on GuruPlug board
qcom:
- enable gsbi driver in defconfig
- fix section mismatch warning in serial driver
samsung:
- use WFI macro in platform_do_lowpower because exynos cpuhotplug
includes a hardcoded WFI instruction and it causes compile error
in Thumb-2 mode.
- fix GIC reg sizes for exynos4 SoCs
- remove reset timer counter value during boot and resume for mct
to fix a big jump in printk timestamps
- fix pm code to check cortex-A9 for another exynos SoCs
- don't rely on firmware's secondary_cpu_start for mcpm
sti:
- Ethernet clocks were wrongly defined for STiH415/416 platforms
- STiH416 B2020 revision E DTS file name contained uppercase, change to
lowercase"
* tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc: (33 commits)
ARM: at91/dt: sam9261: remove slow RC osc
ARM: at91/dt: define sam9261ek slow crystal frequency
ARM: at91/dt: sam9261: correctly define mainck
ARM: at91/dt: sam9n12: correct PLLA ICPLL and OUT values
ARM: at91/dt: sam9x5: correct PLLA ICPLL and OUT values
misc: atmel_pwm: fix Kconfig symbols
ARM: integrator: fix OF-related regression
ARM: mvebu: Fix the improper use of the compatible string armada38x using a wildcard
ARM: dts: kirkwood: fix phy-connection-type for Guruplug
ARM: EXYNOS: Don't rely on firmware's secondary_cpu_start for mcpm
ARM: dts: imx51-eukrea-mbimxsd51-baseboard: unbreak esdhc.
ARM: dts: imx51-babbage: Fix esdhc setup
ARM: dts: mx5: Move the display out of soc {} node
ARM: dts: mx5: Fix IPU port node placement
ARM: mvebu: select ARM_CPU_SUSPEND for Marvell EBU v7 platforms
ARM: mvebu: Fix broken SoC ID detection
ARM: imx_v6_v7_defconfig: Enable CONFIG_IMX_IPUV3_CORE
ARM: multi_v7_defconfig: Add QCOM GSBI driver
ARM: stih41x: Rename stih416-b2020-revE.dts to stih416-b2020e.dts
tty: serial: msm: Fix section mismatch warning
...
Arnd Bergmann [Wed, 25 Jun 2014 18:27:15 +0000 (20:27 +0200)]
Merge tag 'at91-fixes' of git://github.com/at91linux/linux-at91 into fixes
Merge "First AT91 fixes batch for 3.16" from Nicolas Ferre:
- drivers/misc fix for Kconfig PWM symbol
- correction of several values in DT after conversion to CCF
- fix at91sam9261/at91sam9261ek mistake in slow crystal vs. slow RC osc
* tag 'at91-fixes' of git://github.com/at91linux/linux-at91:
ARM: at91/dt: sam9261: remove slow RC osc
ARM: at91/dt: define sam9261ek slow crystal frequency
ARM: at91/dt: sam9261: correctly define mainck
ARM: at91/dt: sam9n12: correct PLLA ICPLL and OUT values
ARM: at91/dt: sam9x5: correct PLLA ICPLL and OUT values
misc: atmel_pwm: fix Kconfig symbols
Arnd Bergmann [Wed, 25 Jun 2014 13:34:00 +0000 (15:34 +0200)]
Merge tag 'mvebu-fixes-3.16' of git://git.infradead.org/linux-mvebu into fixes
Merge "mvebu fixes for v3.16" from Jason Cooper:
- mvebu
- Fix broken SoC ID detection
- Select ARM_CPU_SUSPEND for v7
- Remove armada38x compatible string (no users yet)
- kirkwood
- Fix phy-connection-type on GuruPlug board
* tag 'mvebu-fixes-3.16' of git://git.infradead.org/linux-mvebu:
ARM: mvebu: Fix the improper use of the compatible string armada38x using a wildcard
ARM: dts: kirkwood: fix phy-connection-type for Guruplug
ARM: mvebu: select ARM_CPU_SUSPEND for Marvell EBU v7 platforms
ARM: mvebu: Fix broken SoC ID detection
hwmon: (ntc_thermistor) Use the manufacturer name properly
Murata Manufacturing Co., Ltd is the vendor for
NTC (Negative Temperature coefficient) based Thermistors.
But, the driver extensively uses "NTC" as the vendor name.
This patch corrects the vendor name also updates the
compatibility strings according to the vendor-prefix.txt
Note: Drivers continue to support the previous compatible strings
but further addition of these compatible strings in device tree
is deprecated.
Linus Torvalds [Wed, 25 Jun 2014 12:44:17 +0000 (05:44 -0700)]
Merge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc
Pull powerpc fixes and cleanups from Ben Herrenschmidt:
"Here are a handful or two of powerpc fixes and simple/trivial
cleanups. A bunch of them fix ftrace with the new ABI v2 in Little
Endian, the rest is a scattering of fairly simple things"
* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc:
powerpc: Don't skip ePAPR spin-table CPUs
powerpc/module: Fix TOC symbol CRC
powerpc/powernv: Remove OPAL v1 takeover
powerpc/kmemleak: Do not scan the DART table
selftests/powerpc: Use the test harness for the TM DSCR test
powerpc/cell: cbe_thermal.c: Cleaning up a variable is of the wrong type
powerpc/kprobes: Fix jprobes on ABI v2 (LE)
powerpc/ftrace: Use pr_fmt() to namespace error messages
powerpc/ftrace: Fix nop of modules on 64bit LE (ABIv2)
powerpc/ftrace: Fix inverted check of create_branch()
powerpc/ftrace: Fix typo in mask of opcode
powerpc: Add ppc_global_function_entry()
powerpc/macintosh/smu.c: Fix closing brace followed by if
powerpc: Remove __arch_swab*
powerpc: Remove ancient DEBUG_SIG code
powerpc/kerenl: Enable EEH for IO accessors
Linus Torvalds [Wed, 25 Jun 2014 12:30:20 +0000 (05:30 -0700)]
Merge tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost
Pull vhost cleanups from Michael S Tsirkin:
"Two cleanup patches removing code duplication that got introduced by
changes in rc1. Not fixing crashes, but I'd rather not carry the
duplicate code until the next merge window"
* tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost:
vhost-scsi: don't open-code kvfree
vhost-net: don't open-code kvfree
Linus Torvalds [Wed, 25 Jun 2014 12:08:09 +0000 (05:08 -0700)]
Merge tag 'trace-fixes-v3.16-rc1-v2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace
Pull tracing cleanups and fixes from Steven Rostedt:
"This includes three patches from Oleg Nesterov. The first is a fix to
a race condition that happens between enabling/disabling syscall
tracepoints and new process creations (the check to go into the ptrace
path for a process can be set when it shouldn't, or not set when it
should). Not a major bug but one that should be fixed and even
applied to stable.
The other two patches are cleanup/fixes that are not that critical,
but for an -rc1 release would be nice to have. They both deal with
syscall tracepoints.
It also includes a patch to introduce a new macro for the
TRACE_EVENT() format called __field_struct(). Originally, __field()
was used to record any variable into a trace event, but with the
addition of setting the "is signed" attribute, the check causes
anything but a primitive variable to fail to compile. That is,
structs and unions can't be used as they once were. When the "is
signed" check was introduce there were only primitive variables being
recorded. But that will change soon and it was reported that
__field() causes build failures.
To solve the __field() issue, __field_struct() is introduced to allow
trace_events to be able to record complex types too"
* tag 'trace-fixes-v3.16-rc1-v2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
tracing: Add __field_struct macro for TRACE_EVENT()
tracing: syscall_regfunc() should not skip kernel threads
tracing: Change syscall_*regfunc() to check PF_KTHREAD and use for_each_process_thread()
tracing: Fix syscall_*regfunc() vs copy_process() race
Nicolas Ferre [Wed, 25 Jun 2014 09:33:44 +0000 (11:33 +0200)]
misc: atmel_pwm: fix Kconfig symbols
AT91 symbols AT91SAM9263, AT91SAM9RL, and AT91SAM9G45 do not exist and this
patch changes them to their correct ARCH_* version.
These symbols are chosen instead of the SOC_* ones because this driver is not
converted to DT.
Anyway, the ATMEL_PWM symbol and the associated driver will be removed soon,
during the move to the PWM sub-system.
Reported-by: Paul Bolle <pebolle@tiscali.nl> Acked-by: Alexandre Belloni <alexandre.belloni@free-electrons.com> Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Scott Wood [Wed, 25 Jun 2014 01:15:51 +0000 (20:15 -0500)]
powerpc: Don't skip ePAPR spin-table CPUs
Commit 59a53afe70fd530040bdc69581f03d880157f15a "powerpc: Don't setup
CPUs with bad status" broke ePAPR SMP booting. ePAPR says that CPUs
that aren't presently running shall have status of disabled, with
enable-method being used to determine whether the CPU can be enabled.
Fix by checking for spin-table, which is currently the only supported
enable-method.
Signed-off-by: Scott Wood <scottwood@freescale.com> Cc: Michael Neuling <mikey@neuling.org> Cc: Emil Medve <Emilian.Medve@Freescale.com> Cc: stable@vger.kernel.org Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Laurent Dufour [Tue, 24 Jun 2014 08:53:59 +0000 (10:53 +0200)]
powerpc/module: Fix TOC symbol CRC
The commit 71ec7c55ed91 introduced the magic symbol ".TOC." for ELFv2 ABI.
This symbol is built manually and has no CRC value computed. A zero value
is put in the CRC section to avoid modpost complaining about a missing CRC.
Unfortunately, this breaks the kernel module loading when the kernel is
relocated (kdump case for instance) because of the relocation applied to
the kcrctab values.
This patch compute a CRC value for the TOC symbol which will match the one
compute by the kernel when it is relocated - aka '0 - relocate_start' done in
maybe_relocated called by check_version (module.c).
Signed-off-by: Laurent Dufour <ldufour@linux.vnet.ibm.com> Cc: Anton Blanchard <anton@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Michael Ellerman [Tue, 24 Jun 2014 07:17:47 +0000 (17:17 +1000)]
powerpc/powernv: Remove OPAL v1 takeover
In commit 27f4488872d9 "Add OPAL takeover from PowerVM" we added support
for "takeover" on OPAL v1 machines.
This was a mode of operation where we would boot under pHyp, and query
for the presence of OPAL. If detected we would then do a special
sequence to take over the machine, and the kernel would end up running
in hypervisor mode.
OPAL v1 was never a supported product, and was never shipped outside
IBM. As far as we know no one is still using it.
Newer versions of OPAL do not use the takeover mechanism. Although the
query for OPAL should be harmless on machines with newer OPAL, we have
seen a machine where it causes a crash in Open Firmware.
The code in early_init_devtree() to copy boot_command_line into cmd_line
was added in commit 817c21ad9a1f "Get kernel command line accross OPAL
takeover", and AFAIK is only used by takeover, so should also be
removed.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Li RongQing [Fri, 20 Jun 2014 09:32:36 +0000 (17:32 +0800)]
cxgb4: Not need to hold the adap_rcu_lock lock when read adap_rcu_list
cxgb4_netdev maybe lead to dead lock, since it uses a spin lock, and be called
in both thread and softirq context, but not disable BH, the lockdep report is
below; In fact, cxgb4_netdev only reads adap_rcu_list with RCU protection, so
not need to hold spin lock again.
=================================
[ INFO: inconsistent lock state ]
3.14.7+ #24 Tainted: G C O
---------------------------------
inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage.
radvd/3794 [HC0[0]:SC1[1]:HE1:SE0] takes:
(adap_rcu_lock){+.?...}, at: [<ffffffffa09989ea>] clip_add+0x2c/0x116 [cxgb4]
{SOFTIRQ-ON-W} state was registered at:
[<ffffffff810fca81>] __lock_acquire+0x34a/0xe48
[<ffffffff810fd98b>] lock_acquire+0x82/0x9d
[<ffffffff815d6ff8>] _raw_spin_lock+0x34/0x43
[<ffffffffa09989ea>] clip_add+0x2c/0x116 [cxgb4]
[<ffffffffa0998beb>] cxgb4_inet6addr_handler+0x117/0x12c [cxgb4]
[<ffffffff815da98b>] notifier_call_chain+0x32/0x5c
[<ffffffff815da9f9>] __atomic_notifier_call_chain+0x44/0x6e
[<ffffffff815daa32>] atomic_notifier_call_chain+0xf/0x11
[<ffffffff815b1356>] inet6addr_notifier_call_chain+0x16/0x18
[<ffffffffa01f72e5>] ipv6_add_addr+0x404/0x46e [ipv6]
[<ffffffffa01f8df0>] addrconf_add_linklocal+0x5f/0x95 [ipv6]
[<ffffffffa01fc3e9>] addrconf_notify+0x632/0x841 [ipv6]
[<ffffffff815da98b>] notifier_call_chain+0x32/0x5c
[<ffffffff810e09a1>] __raw_notifier_call_chain+0x9/0xb
[<ffffffff810e09b2>] raw_notifier_call_chain+0xf/0x11
[<ffffffff8151b3b7>] call_netdevice_notifiers_info+0x4e/0x56
[<ffffffff8151b3d0>] call_netdevice_notifiers+0x11/0x13
[<ffffffff8151c0a6>] netdev_state_change+0x1f/0x38
[<ffffffff8152f004>] linkwatch_do_dev+0x3b/0x49
[<ffffffff8152f184>] __linkwatch_run_queue+0x10b/0x144
[<ffffffff8152f1dd>] linkwatch_event+0x20/0x27
[<ffffffff810d7bc0>] process_one_work+0x1cb/0x2ee
[<ffffffff810d7e3b>] worker_thread+0x12e/0x1fc
[<ffffffff810dd391>] kthread+0xc4/0xcc
[<ffffffff815dc48c>] ret_from_fork+0x7c/0xb0
irq event stamp: 3388
hardirqs last enabled at (3388): [<ffffffff810c6c85>]
__local_bh_enable_ip+0xaa/0xd9
hardirqs last disabled at (3387): [<ffffffff810c6c2d>]
__local_bh_enable_ip+0x52/0xd9
softirqs last enabled at (3288): [<ffffffffa01f1d5b>]
rcu_read_unlock_bh+0x0/0x2f [ipv6]
softirqs last disabled at (3289): [<ffffffff815ddafc>]
do_softirq_own_stack+0x1c/0x30
other info that might help us debug this:
Possible unsafe locking scenario:
Reported-by: Ben Greear <greearb@candelatech.com> Cc: Casey Leedom <leedom@chelsio.com> Cc: Hariprasad Shenai <hariprasad@chelsio.com> Signed-off-by: Li RongQing <roy.qing.li@gmail.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Acked-by: Casey Leedom <leedom@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Andy Adamson [Thu, 12 Jun 2014 19:02:32 +0000 (15:02 -0400)]
NFSv4: test SECINFO RPC_AUTH_GSS pseudoflavors for support
Fix nfs4_negotiate_security to create an rpc_clnt used to test each SECINFO
returned pseudoflavor. Check credential creation (and gss_context creation)
which is important for RPC_AUTH_GSS pseudoflavors which can fail for multiple
reasons including mis-configuration.
Don't call nfs4_negotiate in nfs4_submount as it was just called by
nfs4_proc_lookup_mountpoint (nfs4_proc_lookup_common)
Signed-off-by: Andy Adamson <andros@netapp.com>
[Trond: fix corrupt return value from nfs_find_best_sec()] Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Andy Adamson [Mon, 9 Jun 2014 19:33:20 +0000 (15:33 -0400)]
NFS Return -EPERM if no supported or matching SECINFO flavor
Do not return RPC_AUTH_UNIX if SEINFO reply tests fail. This
prevents an infinite loop of NFS4ERR_WRONGSEC for non RPC_AUTH_UNIX mounts.
Without this patch, a mount with no sec= option to a server
that does not include RPC_AUTH_UNIX in the
SECINFO return can be presented with an attemtp to use RPC_AUTH_UNIX
which will result in an NFS4ERR_WRONG_SEC which will prompt the SECINFO
call which will again try RPC_AUTH_UNIX....
Signed-off-by: Andy Adamson <andros@netapp.com> Tested-By: Steve Dickson <steved@redhat.com> Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Trond Myklebust [Fri, 20 Jun 2014 17:11:01 +0000 (13:11 -0400)]
NFS: Don't mark the data cache as invalid if it has been flushed
Now that we have functions such as nfs_write_pageuptodate() that use
the cache_validity flags to check if the data cache is valid or not,
it is a little more important to keep the flags in sync with the
state of the data cache.
In particular, we'd like to ensure that if the data cache is empty, we
don't start marking it as needing revalidation.
Reported-by: Scott Mayhew <smayhew@redhat.com> Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Trond Myklebust [Fri, 20 Jun 2014 17:16:38 +0000 (13:16 -0400)]
NFS: Clear NFS_INO_REVAL_PAGECACHE when we update the file size
In nfs_update_inode(), if the change attribute is seen to change on
the server, then we set NFS_INO_REVAL_PAGECACHE in order to make
sure that we check the file size.
However, if we also update the file size in the same function, we
don't need to check it again. So make sure that we clear the
NFS_INO_REVAL_PAGECACHE that was set earlier.
Scott Mayhew [Fri, 20 Jun 2014 12:44:42 +0000 (08:44 -0400)]
nfs: Fix cache_validity check in nfs_write_pageuptodate()
NFS_INO_INVALID_DATA cannot be ignored, even if we have a delegation.
We're still having some problems with data corruption when multiple
clients are appending to a file and those clients are being granted
write delegations on open.
To reproduce:
Client A:
vi /mnt/`hostname -s`
while :; do echo "XXXXXXXXXXXXXXX" >>/mnt/file; sleep $(( $RANDOM % 5 )); done
Client B:
vi /mnt/`hostname -s`
while :; do echo "YYYYYYYYYYYYYYY" >>/mnt/file; sleep $(( $RANDOM % 5 )); done
What's happening is that in nfs_update_inode() we're recognizing that
the file size has changed and we're setting NFS_INO_INVALID_DATA
accordingly, but then we ignore the cache_validity flags in
nfs_write_pageuptodate() because we have a delegation. As a result,
in nfs_updatepage() we're extending the write to cover the full page
even though we've not read in the data to begin with.
Linus Torvalds [Tue, 24 Jun 2014 21:00:13 +0000 (14:00 -0700)]
Merge git://git.kvack.org/~bcrl/aio-fixes
Pull aio fixes from Ben LaHaise:
"These fix a kernel memory disclosure issue (arbitrary kmap() &
copy_to_user()) revealed in CVE-2014-0206 by changes that were
introduced in v3.10"
* git://git.kvack.org/~bcrl/aio-fixes:
aio: fix kernel memory disclosure in io_getevents() introduced in v3.10
aio: fix aio request leak when events are reaped by userspace
Linus Torvalds [Tue, 24 Jun 2014 20:59:00 +0000 (13:59 -0700)]
Merge branch 'fixes' of git://ftp.arm.linux.org.uk/~rmk/linux-arm
Pull ARM fixes from Russell King:
"A number of low impact fixes, the most noticable one is the thumb2
frame pointer fix. We also fix a regression caused during this merge
window with ARM925 CPUs running with caches disabled, and fix a number
of warnings"
* 'fixes' of git://ftp.arm.linux.org.uk/~rmk/linux-arm:
ARM: arm925: ensure assembly sets up writethrough mapping
ARM: perf: fix compiler warning with gcc 4.6.4 (and tidy code)
ARM: l2c: fix dependencies on PL310 errata symbols
ARM: 8069/1: Make thread_save_fp macro aware of THUMB2 mode
ARM: 8068/1: scoop: Remove unused variable
Benjamin LaHaise [Tue, 24 Jun 2014 17:32:51 +0000 (13:32 -0400)]
aio: fix kernel memory disclosure in io_getevents() introduced in v3.10
A kernel memory disclosure was introduced in aio_read_events_ring() in v3.10
by commit a31ad380bed817aa25f8830ad23e1a0480fef797. The changes made to
aio_read_events_ring() failed to correctly limit the index into
ctx->ring_pages[], allowing an attacked to cause the subsequent kmap() of
an arbitrary page with a copy_to_user() to copy the contents into userspace.
This vulnerability has been assigned CVE-2014-0206. Thanks to Mateusz and
Petr for disclosing this issue.
This patch applies to v3.12+. A separate backport is needed for 3.10/3.11.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org> Cc: Mateusz Guzik <mguzik@redhat.com> Cc: Petr Matousek <pmatouse@redhat.com> Cc: Kent Overstreet <kmo@daterainc.com> Cc: Jeff Moyer <jmoyer@redhat.com> Cc: stable@vger.kernel.org
Benjamin LaHaise [Tue, 24 Jun 2014 17:12:55 +0000 (13:12 -0400)]
aio: fix aio request leak when events are reaped by userspace
The aio cleanups and optimizations by kmo that were merged into the 3.10
tree added a regression for userspace event reaping. Specifically, the
reference counts are not decremented if the event is reaped in userspace,
leading to the application being unable to submit further aio requests.
This patch applies to 3.12+. A separate backport is required for 3.10/3.11.
This issue was uncovered as part of CVE-2014-0206.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org> Cc: stable@vger.kernel.org Cc: Kent Overstreet <kmo@daterainc.com> Cc: Mateusz Guzik <mguzik@redhat.com> Cc: Petr Matousek <pmatouse@redhat.com>
Steve French [Mon, 23 Jun 2014 01:38:49 +0000 (20:38 -0500)]
[CIFS] fix mount failure with broken pathnames when smb3 mount with mapchars option
When we SMB3 mounted with mapchars (to allow reserved characters : \ / > < * ?
via the Unicode Windows to POSIX remap range) empty paths
(eg when we open "" to query the root of the SMB3 directory on mount) were not
null terminated so we sent garbarge as a path name on empty paths which caused
SMB2/SMB2.1/SMB3 mounts to fail when mapchars was specified. mapchars is
particularly important since Unix Extensions for SMB3 are not supported (yet)
Signed-off-by: Steve French <smfrench@gmail.com> Cc: <stable@vger.kernel.org> Reviewed-by: David Disseldorp <ddiss@suse.de>
Linus Walleij [Tue, 24 Jun 2014 12:08:07 +0000 (14:08 +0200)]
ARM: integrator: fix OF-related regression
Commit 07e461cd7e73a84f0e3757932b93cc80976fd749
"of: Ensure unique names without sacrificing determinism"
caused a boot failure regression on the Integrator machines.
The problem is probably caused by fiddling too much with
the device tree population in the OF init function, such
as passing the SoC bus device as parent when populating
the device tree.
This patch fixes the problem by:
- Avoiding to explicitly look up the tree root
- Look up devices needed before device population from
the match only, passing NULL as root
- Passing NULL as root and parent when calling
of_platform_populate()
After this the Integrators boot again. Tested on
Integrator/AP and Integrator/CP.
Gregory CLEMENT [Mon, 23 Jun 2014 14:16:51 +0000 (16:16 +0200)]
ARM: mvebu: Fix the improper use of the compatible string armada38x using a wildcard
Wildcards in compatible strings should be avoid. "marvell,armada38x"
was recently introduced but was not yet used.
The armada 385 SoC is a superset of the armada 380 SoC (with more CPUs
and more PCIe slots). So this patch replaces the use of
"marvell,armada38x" by the "marvell,armada380" string.
Catalin Marinas [Fri, 13 Jun 2014 08:44:21 +0000 (09:44 +0100)]
powerpc/kmemleak: Do not scan the DART table
The DART table allocation is registered to kmemleak via the
memblock_alloc_base() call. However, the DART table is later unmapped
and dart_tablebase VA no longer accessible. This patch tells kmemleak
not to scan this block and avoid an unhandled paging request.
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Michael Ellerman [Mon, 23 Jun 2014 03:23:31 +0000 (13:23 +1000)]
powerpc/kprobes: Fix jprobes on ABI v2 (LE)
In commit 721aeaa9 "Build little endian ppc64 kernel with ABIv2", we
missed some updates required in the kprobes code to make jprobes work
when the kernel is built with ABI v2.
Firstly update arch_deref_entry_point() to do the right thing. Now that
we have added ppc_global_function_entry() we can just always use that, it
will do the right thing for 32 & 64 bit and ABI v1 & v2.
Secondly we need to update the code that sets up the register state before
calling the jprobe handler. On ABI v1 we setup r2 to hold the TOC, on ABI
v2 we need to populate r12 with the function entry point address.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Michael Ellerman [Tue, 17 Jun 2014 06:15:36 +0000 (16:15 +1000)]
powerpc/ftrace: Use pr_fmt() to namespace error messages
The printks() in our ftrace code have no prefix, so they appear on the
console with very little context, eg:
Branch out of range
Use pr_fmt() & pr_err() to add a prefix. While we're at it, collapse a
few split lines that don't need to be, and add a missing newline to one
message.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Michael Ellerman [Tue, 17 Jun 2014 06:15:34 +0000 (16:15 +1000)]
powerpc/ftrace: Fix inverted check of create_branch()
In commit 24a1bdc35, "Fix ABIv2 issues with __ftrace_make_call", Anton
changed the logic that creates and patches the branch, and added a
thinko in the check of create_branch(). create_branch() returns the
instruction that was generated, so if we get zero then it succeeded.
The result is we can't ftrace modules:
Branch out of range
WARNING: at ../kernel/trace/ftrace.c:1638
ftrace failed to modify [<d000000004ba001c>] fuse_req_init_context+0x1c/0x90 [fuse]
We should probably fix patch_instruction() to do that check and make the
API saner, but that's a separate patch. For now just invert the test.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Michael Ellerman [Tue, 17 Jun 2014 06:15:33 +0000 (16:15 +1000)]
powerpc/ftrace: Fix typo in mask of opcode
In commit 24a1bdc35, "Fix ABIv2 issues with __ftrace_make_call", Anton
changed the logic that checks for the expected code sequence when
patching a module.
We missed the typo in the mask, 0xffff00000 should be 0xffff0000, which
has the effect of making the test always true.
That makes it impossible to ftrace against modules, eg:
Unexpected call sequence: 48000008e8410018
WARNING: at ../kernel/trace/ftrace.c:1638
ftrace failed to modify [<d000000007cf001c>] rng_dev_open+0x1c/0x70 [rng_core]
Reported-by: David Binderman <dcb314@hotmail.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Michael Ellerman [Tue, 17 Jun 2014 06:15:32 +0000 (16:15 +1000)]
powerpc: Add ppc_global_function_entry()
ABIv2 has the concept of a global and local entry point to a function.
In most cases we are interested in the local entry point, and so that is
what ppc_function_entry() returns.
However we have a case in the ftrace code where we want the global entry
point, and there may be other places we need it too. Rather than special
casing each, add an accessor.
For ABIv1 and 32-bit there is only a single entry point, so we return
that. That means it's safe for the caller to use this without also
checking the ABI version.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Michael Ellerman [Mon, 23 Jun 2014 04:17:47 +0000 (14:17 +1000)]
powerpc: Remove ancient DEBUG_SIG code
We have some compile-time disabled debug code in signal_xx.c. It's from
some ancient time BG, almost certainly part of the original port, given
the very similar code on other arches.
The show_unhandled_signal logic, added in d0c3d534a438 (2.6.24) is
cleaner and prints more useful information, so drop the debug code.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Linus Torvalds [Tue, 24 Jun 2014 00:05:28 +0000 (17:05 -0700)]
Merge tag 'compress-3.16-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core
Pull compress bugfixes from Greg KH:
"Here are two bugfixes for some compression functions that resolve some
errors when uncompressing some pathalogical data. Both were found by
Don A Bailey"
* tag 'compress-3.16-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core:
lz4: ensure length does not wrap
lzo: properly check for overruns
Linus Torvalds [Mon, 23 Jun 2014 23:48:14 +0000 (16:48 -0700)]
Merge branch 'akpm' (patches from Andrew Morton)
Merge fixes from Andrew Morton:
"The nmi patch and watchdog patch aren't actually fixes - they're
features which needed a few last-minutes touchups.
Otherwise, a rather large batch of fixes - ocfs2 review takes a while
and I got distracted and missed last week's batch"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>: (31 commits)
ocfs2/dlm: do not purge lockres that is queued for assert master
ocfs2: do not return DLM_MIGRATE_RESPONSE_MASTERY_REF to avoid endless,loop during umount
ocfs2: manually do the iput once ocfs2_add_entry failed in ocfs2_symlink and ocfs2_mknod
ocfs2: fix a tiny race when running dirop_fileop_racer
ocfs2/dlm: fix misuse of list_move_tail() in dlm_run_purge_list()
ocfs2: refcount: take rw_lock in ocfs2_reflink
ocfs2: revert "ocfs2: fix NULL pointer dereference when dismount and ocfs2rec simultaneously"
ocfs2: fix deadlock when two nodes are converting same lock from PR to EX and idletimeout closes conn
ocfs2: should add inode into orphan dir after updating entry in ocfs2_rename()
mm: fix crashes from mbind() merging vmas
checkpatch: reduce false positives when checking void function return statements
ia64: arch/ia64/include/uapi/asm/fcntl.h needs personality.h
DMA, CMA: fix possible memory leak
slab: fix oops when reading /proc/slab_allocators
shmem: fix faulting into a hole while it's punched
mm: let mm_find_pmd fix buggy race with THP fault
mm: thp: fix DEBUG_PAGEALLOC oops in copy_page_rep()
kernel/watchdog.c: print traces for all cpus on lockup detection
nmi: provide the option to issue an NMI back trace to every cpu but current
Documentation/accounting/getdelays.c: add missing null-terminate after strncpy call
...
Xue jiufei [Mon, 23 Jun 2014 20:22:09 +0000 (13:22 -0700)]
ocfs2/dlm: do not purge lockres that is queued for assert master
When workqueue is delayed, it may occur that a lockres is purged while it
is still queued for master assert. it may trigger BUG() as follows.
N1 N2
dlm_get_lockres()
->dlm_do_master_requery
is the master of lockres,
so queue assert_master work
dlm_thread() start running
and purge the lockres
dlm_assert_master_worker()
send assert master message
to other nodes
receiving the assert_master
message, set master to N2
dlmlock_remote() send create_lock message to N2, but receive DLM_IVLOCKID,
if it is RECOVERY lockres, it triggers the BUG().
Another BUG() is triggered when N3 become the new master and send
assert_master to N1, N1 will trigger the BUG() because owner doesn't
match. So we should not purge lockres when it is queued for assert
master.
Signed-off-by: joyce.xue <xuejiufei@huawei.com> Reviewed-by: Mark Fasheh <mfasheh@suse.de> Cc: Joel Becker <jlbec@evilplan.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
jiangyiwen [Mon, 23 Jun 2014 20:22:09 +0000 (13:22 -0700)]
ocfs2: do not return DLM_MIGRATE_RESPONSE_MASTERY_REF to avoid endless,loop during umount
The following case may lead to endless loop during umount.
node A node B node C node D
umount volume,
migrate lockres1
to B
want to lock lockres1,
send
MASTER_REQUEST_MSG
to C
init block mle
send
MIGRATE_REQUEST_MSG
to C
find a block
mle, and then
return
DLM_MIGRATE_RESPONSE_MASTERY_REF
to B
set C in refmap
umount successfully
try to umount, endless
loop occurs when migrate
lockres1 since C is in
refmap
So we can fix this endless loop case by only returning
DLM_MIGRATE_RESPONSE_MASTERY_REF if it has a mastery mle when receiving
MIGRATE_REQUEST_MSG.
[akpm@linux-foundation.org: coding-style fixes] Signed-off-by: jiangyiwen <jiangyiwen@huawei.com> Reviewed-by: Mark Fasheh <mfasheh@suse.de> Cc: Joel Becker <jlbec@evilplan.org> Cc: Xue jiufei <xuejiufei@huawei.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
jiangyiwen [Mon, 23 Jun 2014 20:22:09 +0000 (13:22 -0700)]
ocfs2: manually do the iput once ocfs2_add_entry failed in ocfs2_symlink and ocfs2_mknod
When the call to ocfs2_add_entry() failed in ocfs2_symlink() and
ocfs2_mknod(), iput() will not be called during dput(dentry) because no
d_instantiate(), and this will lead to umount hung.
Signed-off-by: jiangyiwen <jiangyiwen@huawei.com> Cc: Joel Becker <jlbec@evilplan.org> Reviewed-by: Mark Fasheh <mfasheh@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Yiwen Jiang [Mon, 23 Jun 2014 20:22:09 +0000 (13:22 -0700)]
ocfs2: fix a tiny race when running dirop_fileop_racer
When running dirop_fileop_racer we found a dead lock case.
2 nodes, say Node A and Node B, mount the same ocfs2 volume. Create
/race/16/1 in the filesystem, and let the inode number of dir 16 is less
than the inode number of dir race.
Node A Node B
mv /race/16/1 /race/
right after Node A has got the
EX mode of /race/16/, and tries to
get EX mode of /race
ls /race/16/
In this case, Node A has got the EX mode of /race/16/, and wants to get EX
mode of /race/. Node B has got the PR mode of /race/, and wants to get
the PR mode of /race/16/. Since EX and PR are mutually exclusive, dead
lock happens.
This patch fixes this case by locking in ancestor order before trying
inode number order.
Signed-off-by: Yiwen Jiang <jiangyiwen@huawei.com> Signed-off-by: Joseph Qi <joseph.qi@huawei.com> Cc: Joel Becker <jlbec@evilplan.org> Reviewed-by: Mark Fasheh <mfasheh@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Xue jiufei [Mon, 23 Jun 2014 20:22:08 +0000 (13:22 -0700)]
ocfs2/dlm: fix misuse of list_move_tail() in dlm_run_purge_list()
When a lockres in purge list but is still in use, it should be moved to
the tail of purge list. dlm_thread will continue to check next lockres in
purge list. However, code list_move_tail(&dlm->purge_list,
&lockres->purge) will do *no* movements, so dlm_thread will purge the same
lockres in this loop again and again. If it is in use for a long time,
other lockres will not be processed.
Signed-off-by: Yiwen Jiang <jiangyiwen@huawei.com> Signed-off-by: joyce.xue <xuejiufei@huawei.com> Reviewed-by: Mark Fasheh <mfasheh@suse.de> Cc: Joel Becker <jlbec@evilplan.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
It crashes at
BUG_ON(create && (ext_flags & OCFS2_EXT_REFCOUNTED));
in ocfs2_direct_IO_get_blocks.
ocfs2_direct_IO_get_blocks is expecting the OCFS2_EXT_REFCOUNTED be removed in
ocfs2_prepare_inode_for_write() if it was there. But no cluster lock is taken
during the time before (or inside) ocfs2_prepare_inode_for_write() and after
ocfs2_direct_IO_get_blocks().
It can happen in this case:
Node A(which crashes) Node B
------------------------ ---------------------------
ocfs2_file_aio_write
ocfs2_prepare_inode_for_write
ocfs2_inode_lock
...
ocfs2_inode_unlock
#no refcount found
.... ocfs2_reflink
ocfs2_inode_lock
...
ocfs2_inode_unlock
#now, refcount flag set on extent
...
flush change to disk
ocfs2_direct_IO_get_blocks
ocfs2_get_clusters
#extent map miss
#buffer_head miss
read extents from disk
found refcount flag on extent
crash..
Fix:
Take rw_lock in ocfs2_reflink path
Signed-off-by: Wengang Wang <wen.gang.wang@oracle.com> Reviewed-by: Mark Fasheh <mfasheh@suse.de> Cc: Joel Becker <jlbec@evilplan.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Xue jiufei [Mon, 23 Jun 2014 20:22:08 +0000 (13:22 -0700)]
ocfs2: revert "ocfs2: fix NULL pointer dereference when dismount and ocfs2rec simultaneously"
75f82eaa502c ("ocfs2: fix NULL pointer dereference when dismount and
ocfs2rec simultaneously") may cause umount hang while shutting down
truncate log.
The situation is as followes:
ocfs2_dismout_volume
-> ocfs2_recovery_exit
-> free osb->recovery_map
-> ocfs2_truncate_shutdown
-> lock global bitmap inode
-> ocfs2_wait_for_recovery
-> check whether osb->recovery_map->rm_used is zero
Because osb->recovery_map is already freed, rm_used can be any other
values, so it may yield umount hang.
Signed-off-by: joyce.xue <xuejiufei@huawei.com> Reviewed-by: Mark Fasheh <mfasheh@suse.de> Cc: Joel Becker <jlbec@evilplan.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Two node cluster and both nodes hold a lock at PR level and both want to
convert to EX at the same time. Master node 1 has sent BAST and then
closes the connection due to idletime out. Node 0 receives BAST, sends
unlock req with cancel flag but gets error -ENOTCONN. The problem is
this error is ignored in dlm_send_remote_unlock_request() on the
**incorrect** assumption that the master is dead. See NOTE in comment
why it returns DLM_NORMAL. Upon getting DLM_NORMAL, node 0 proceeds to
sends convert (without cancel flg) which fails with -ENOTCONN. waits 5
sec and resends.
This time gets DLM_IVLOCKID from the master since lock not found in
grant, it had been moved to converting queue in response to conv PR->EX
req. No way out.
Node 1 (master) Node 0
============== ======
lock mode PR PR
convert PR -> EX
mv grant -> convert and que BAST
...
<-------- convert PR -> EX
convert que looks like this: ((node 1, PR -> EX) (node 0, PR -> EX))
...
BAST (want PR -> NL)
------------------>
...
idle timout, conn closed
...
In response to BAST,
sends unlock with cancel convert flag
gets -ENOTCONN. Ignores and
sends remote convert request
gets -ENOTCONN, waits 5 Sec, retries
...
reconnects
<----------------- convert req goes through on next try
does not find lock on grant que
status DLM_IVLOCKID
------------------>
...
No way out. Fix is to keep retrying unlock with cancel flag until it
succeeds or the master dies.
Signed-off-by: Tariq Saeed <tariq.x.saeed@oracle.com> Reviewed-by: Mark Fasheh <mfasheh@suse.de> Cc: Joel Becker <jlbec@evilplan.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
alex chen [Mon, 23 Jun 2014 20:22:07 +0000 (13:22 -0700)]
ocfs2: should add inode into orphan dir after updating entry in ocfs2_rename()
There are two files a and b in dir /mnt/ocfs2.
node A node B
mv a b
In ocfs2_rename(), after calling
ocfs2_orphan_add(), the inode of
file b will be added into orphan
dir.
If ocfs2_update_entry() fails,
ocfs2_rename return error and mv
operation fails. But file b still
exists in the parent dir.
ocfs2_queue_orphan_scan
-> ocfs2_queue_recovery_completion
-> ocfs2_complete_recovery
-> ocfs2_recover_orphans
The inode of the file b will be
put with iput().
ocfs2_evict_inode
-> ocfs2_delete_inode
-> ocfs2_wipe_inode
-> ocfs2_remove_inode
OCFS2_VALID_FL in the inode
i_flags will be cleared.
The file b still can be accessed
on node B.
ls /mnt/ocfs2
When first read the file b with
ocfs2_read_inode_block(). It will
validate the inode using
ocfs2_validate_inode_block().
Because OCFS2_VALID_FL not set in
the inode i_flags, so the file
system will be readonly.
So we should add inode into orphan dir after updating entry in
ocfs2_rename().
Signed-off-by: alex.chen <alex.chen@huawei.com> Reviewed-by: Mark Fasheh <mfasheh@suse.de> Cc: Joel Becker <jlbec@evilplan.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Hugh Dickins [Mon, 23 Jun 2014 20:22:07 +0000 (13:22 -0700)]
mm: fix crashes from mbind() merging vmas
In v2.6.34 commit 9d8cebd4bcd7 ("mm: fix mbind vma merge problem")
introduced vma merging to mbind(), but it should have also changed the
convention of passing start vma from queue_pages_range() (formerly
check_range()) to new_vma_page(): vma merging may have already freed
that structure, resulting in BUG at mm/mempolicy.c:1738 and probably
worse crashes.
fs/notify/fanotify/fanotify_user.c: In function 'SYSC_fanotify_init':
fs/notify/fanotify/fanotify_user.c:726: error: implicit declaration of function 'personality'
fs/notify/fanotify/fanotify_user.c:726: error: 'PER_LINUX32' undeclared (first use in this function)
fs/notify/fanotify/fanotify_user.c:726: error: (Each undeclared identifier is reported only once
fs/notify/fanotify/fanotify_user.c:726: error: for each function it appears in.)
Joonsoo Kim [Mon, 23 Jun 2014 20:22:06 +0000 (13:22 -0700)]
slab: fix oops when reading /proc/slab_allocators
Commit b1cb0982bdd6 ("change the management method of free objects of
the slab") introduced a bug on slab leak detector
('/proc/slab_allocators'). This detector works like as following
decription.
1. traverse all objects on all the slabs.
2. determine whether it is active or not.
3. if active, print who allocate this object.
but that commit changed the way how to manage free objects, so the logic
determining whether it is active or not is also changed. In before, we
regard object in cpu caches as inactive one, but, with this commit, we
mistakenly regard object in cpu caches as active one.
This intoduces kernel oops if DEBUG_PAGEALLOC is enabled. If
DEBUG_PAGEALLOC is enabled, kernel_map_pages() is used to detect who
corrupt free memory in the slab. It unmaps page table mapping if object
is free and map it if object is active. When slab leak detector check
object in cpu caches, it mistakenly think this object active so try to
access object memory to retrieve caller of allocation. At this point,
page table mapping to this object doesn't exist, so oops occurs.
Following is oops message reported from Dave.
It blew up when something tried to read /proc/slab_allocators
(Just cat it, and you should see the oops below)
To fix the problem, I introduce an object status buffer on each slab.
With this, we can track object status precisely, so slab leak detector
would not access active object and no kernel oops would occur. Memory
overhead caused by this fix is only imposed to CONFIG_DEBUG_SLAB_LEAK
which is mainly used for debugging, so memory overhead isn't big
problem.
Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com> Reported-by: Dave Jones <davej@redhat.com> Reported-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp> Reviewed-by: Vladimir Davydov <vdavydov@parallels.com> Cc: Christoph Lameter <cl@linux.com> Cc: Pekka Enberg <penberg@kernel.org> Cc: David Rientjes <rientjes@google.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Hugh Dickins [Mon, 23 Jun 2014 20:22:06 +0000 (13:22 -0700)]
shmem: fix faulting into a hole while it's punched
Trinity finds that mmap access to a hole while it's punched from shmem
can prevent the madvise(MADV_REMOVE) or fallocate(FALLOC_FL_PUNCH_HOLE)
from completing, until the reader chooses to stop; with the puncher's
hold on i_mutex locking out all other writers until it can complete.
It appears that the tmpfs fault path is too light in comparison with its
hole-punching path, lacking an i_data_sem to obstruct it; but we don't
want to slow down the common case.
Extend shmem_fallocate()'s existing range notification mechanism, so
shmem_fault() can refrain from faulting pages into the hole while it's
punched, waiting instead on i_mutex (when safe to sleep; or repeatedly
faulting when not).
[akpm@linux-foundation.org: coding-style fixes] Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Dave Jones <davej@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
I believe this comes about because, whereas collapsing and splitting THP
functions take anon_vma lock in write mode (which excludes concurrent
rmap walks), faulting THP functions (write protection and misplaced
NUMA) do not - and mostly they do not need to.
But they do use a pmdp_clear_flush(), set_pmd_at() sequence which, for
an instant (indeed, for a long instant, given the inter-CPU TLB flush in
there), leaves *pmd neither present not trans_huge.
Which can confuse a concurrent rmap walk, as when removing migration
ptes, seen in the dumped trace. Although that rmap walk has a 4k page
to insert, anon_vmas containing THPs are in no way segregated from
4k-page anon_vmas, so the 4k-intent mm_find_pmd() does need to cope with
that instant when a trans_huge pmd is temporarily absent.
I don't think we need strengthen the locking at the THP end: it's easily
handled with an ACCESS_ONCE() before testing both conditions.
And since mm_find_pmd() had only one caller who wanted a THP rather than
a pmd, let's slightly repurpose it to fail when it hits a THP or
non-present pmd, and open code split_huge_page_address() again.
Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Mel Gorman <mgorman@suse.de> Cc: Bob Liu <bob.liu@oracle.com> Cc: Christoph Lameter <cl@gentwo.org> Cc: Dave Jones <davej@redhat.com> Cc: David Rientjes <rientjes@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Hugh Dickins [Mon, 23 Jun 2014 20:22:05 +0000 (13:22 -0700)]
mm: thp: fix DEBUG_PAGEALLOC oops in copy_page_rep()
Trinity has for over a year been reporting a CONFIG_DEBUG_PAGEALLOC oops
in copy_page_rep() called from copy_user_huge_page() called from
do_huge_pmd_wp_page().
I believe this is a DEBUG_PAGEALLOC false positive, due to the source
page being split, and a tail page freed, while copy is in progress; and
not a problem without DEBUG_PAGEALLOC, since the pmd_same() check will
prevent a miscopy from being made visible.
Fix by adding get_user_huge_page() and put_user_huge_page(): reducing to
the usual get_page() and put_page() on head page in the usual config;
but get and put references to all of the tail pages when
DEBUG_PAGEALLOC.
[akpm@linux-foundation.org: coding-style fixes] Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Dave Jones <davej@redhat.com> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: David Rientjes <rientjes@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Aaron Tomlin [Mon, 23 Jun 2014 20:22:05 +0000 (13:22 -0700)]
kernel/watchdog.c: print traces for all cpus on lockup detection
A 'softlockup' is defined as a bug that causes the kernel to loop in
kernel mode for more than a predefined period to time, without giving
other tasks a chance to run.
Currently, upon detection of this condition by the per-cpu watchdog
task, debug information (including a stack trace) is sent to the system
log.
On some occasions, we have observed that the "victim" rather than the
actual "culprit" (i.e. the owner/holder of the contended resource) is
reported to the user. Often this information has proven to be
insufficient to assist debugging efforts.
To avoid loss of useful debug information, for architectures which
support NMI, this patch makes it possible to improve soft lockup
reporting. This is accomplished by issuing an NMI to each cpu to obtain
a stack trace.
If NMI is not supported we just revert back to the old method. A sysctl
and boot-time parameter is available to toggle this feature.
[dzickus@redhat.com: add CONFIG_SMP in certain areas]
[akpm@linux-foundation.org: additional CONFIG_SMP=n optimisations]
[mq@suse.cz: fix warning] Signed-off-by: Aaron Tomlin <atomlin@redhat.com> Signed-off-by: Don Zickus <dzickus@redhat.com> Cc: David S. Miller <davem@davemloft.net> Cc: Mateusz Guzik <mguzik@redhat.com> Cc: Oleg Nesterov <oleg@redhat.com> Signed-off-by: Jan Moskyto Matejka <mq@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Aaron Tomlin [Mon, 23 Jun 2014 20:22:05 +0000 (13:22 -0700)]
nmi: provide the option to issue an NMI back trace to every cpu but current
Sometimes it is preferred not to use the trigger_all_cpu_backtrace()
routine when one wants to avoid capturing a back trace for current. For
instance if one was previously captured recently.
This patch provides a new routine namely
trigger_allbutself_cpu_backtrace() which offers the flexibility to issue
an NMI to every cpu but current and capture a back trace accordingly.
Patch x86 and sparc to support new routine.
[dzickus@redhat.com: add stub in #else clause]
[dzickus@redhat.com: don't print message in single processor case, wrap with get/put_cpu based on Oleg's suggestion]
[sfr@canb.auug.org.au: undo C99ism] Signed-off-by: Aaron Tomlin <atomlin@redhat.com> Signed-off-by: Don Zickus <dzickus@redhat.com> Acked-by: David S. Miller <davem@davemloft.net> Cc: Mateusz Guzik <mguzik@redhat.com> Cc: Oleg Nesterov <oleg@redhat.com> Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
However, if the percpu_pagelist_fraction sysctl is set by the user, it
is also impossible to restore it to the kernel default since the user
cannot write 0 to the sysctl.
This patch allows the user to write 0 to restore the default behavior.
It still requires a fraction equal to or larger than 8, however, as
stated by the documentation for sanity. If a value in the range [1, 7]
is written, the sysctl will return EINVAL.
This successfully solves the divide by zero issue at the same time.
Signed-off-by: David Rientjes <rientjes@google.com> Reported-by: Oleg Drokin <green@linuxhacker.ru> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
What happened is after updating the watchdog_thresh, the lockup detector
is restarted to utilize the new value. Part of this process involved
disabling preemption. Once preemption was disabled, perf tried to
allocate a new event (as part of the restart). This caused the above
BUG_ON as you can't sleep with preemption disabled.
The preemption restriction seemed agressive as we are not doing anything
on that particular cpu, but with all the online cpus (which are
protected by the get_online_cpus lock). Remove the restriction and the
BUG_ON goes away.
Signed-off-by: Don Zickus <dzickus@redhat.com> Acked-by: Michal Hocko <mhocko@suse.cz> Reported-by: Peter Wu <peter@lekensteyn.nl> Tested-by: Peter Wu <peter@lekensteyn.nl> Acked-by: David Rientjes <rientjes@google.com> Cc: <stable@vger.kernel.org> [3.13+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Remove one inactive maintainer, add two new ones and update my email
address. Plus add Andrew. And fix the glob to include files like
mm/slab_common.c
Signed-off-by: Christoph Lameter <cl@linux.com> Acked-by: David Rientjes <rientjes@google.com> Acked-by: Joonsoo Kim <iamjoonsoo.kim@lge.com> Acked-by: Pekka Enberg <penberg@kernel.org> Cc: Matt Mackall <mpm@selenic.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Naoya Horiguchi [Mon, 23 Jun 2014 20:22:03 +0000 (13:22 -0700)]
hugetlb: fix copy_hugetlb_page_range() to handle migration/hwpoisoned entry
There's a race between fork() and hugepage migration, as a result we try
to "dereference" a swap entry as a normal pte, causing kernel panic.
The cause of the problem is that copy_hugetlb_page_range() can't handle
"swap entry" family (migration entry and hwpoisoned entry) so let's fix
it.
Hugh Dickins [Mon, 23 Jun 2014 20:22:03 +0000 (13:22 -0700)]
tmpfs: ZERO_RANGE and COLLAPSE_RANGE not currently supported
I was well aware of FALLOC_FL_ZERO_RANGE and FALLOC_FL_COLLAPSE_RANGE
support being added to fallocate(); but didn't realize until now that I
had been too stupid to future-proof shmem_fallocate() against new
additions. -EOPNOTSUPP instead of going on to ordinary fallocation.
David Rientjes [Mon, 23 Jun 2014 20:22:03 +0000 (13:22 -0700)]
mm, hotplug: probe interface is available on several platforms
Documentation/memory-hotplug.txt incorrectly states that the memory
driver "probe" interface is only supported on powerpc and is vague about
its application on x86. Clarify the platforms that make this interface
available if memory hotplug is enabled.
Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Petr Tesarik [Mon, 23 Jun 2014 20:22:03 +0000 (13:22 -0700)]
kexec: save PG_head_mask in VMCOREINFO
To allow filtering of huge pages, makedumpfile must be able to identify
them in the dump. This can be done by checking the appropriate page
flag, so communicate its value to makedumpfile through the VMCOREINFO
interface.
There's only one small catch. Depending on how many page flags are
available on a given architecture, this bit can be called PG_head or
PG_compound.
I sent a similar patch back in 2012, but Eric Biederman did not like
using an #ifdef. So, this time I'm adding a common symbol
(PG_head_mask) instead.
See https://lkml.org/lkml/2012/11/28/91 for the previous version.
Signed-off-by: Petr Tesarik <ptesarik@suse.cz> Acked-by: Vivek Goyal <vgoyal@redhat.com> Cc: Eric Biederman <ebiederm@xmission.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Fengguang Wu <fengguang.wu@intel.com> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Shaohua Li <shli@kernel.org> Cc: Alexey Kardashevskiy <aik@ozlabs.ru> Cc: Sasha Levin <sasha.levin@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Srivatsa S. Bhat [Mon, 23 Jun 2014 20:22:02 +0000 (13:22 -0700)]
CPU hotplug, smp: flush any pending IPI callbacks before CPU offline
There is a race between the CPU offline code (within stop-machine) and
the smp-call-function code, which can lead to getting IPIs on the
outgoing CPU, *after* it has gone offline.
Specifically, this can happen when using
smp_call_function_single_async() to send the IPI, since this API allows
sending asynchronous IPIs from IRQ disabled contexts. The exact race
condition is described below.
During CPU offline, in stop-machine, we don't enforce any rule in the
_DISABLE_IRQ stage, regarding the order in which the outgoing CPU and
the other CPUs disable their local interrupts. Due to this, we can
encounter a situation in which an IPI is sent by one of the other CPUs
to the outgoing CPU (while it is *still* online), but the outgoing CPU
ends up noticing it only *after* it has gone offline.
CPU 1 CPU 2
(Online CPU) (CPU going offline)
Enter _PREPARE stage Enter _PREPARE stage
Enter _DISABLE_IRQ stage
=
Got a device interrupt, and | Didn't notice the IPI
the interrupt handler sent an | since interrupts were
IPI to CPU 2 using | disabled on this CPU.
smp_call_function_single_async() |
=
Enter _DISABLE_IRQ stage
Enter _RUN stage Enter _RUN stage
=
Busy loop with interrupts | Invoke take_cpu_down()
disabled. | and take CPU 2 offline
=
Enter _EXIT stage Enter _EXIT stage
Re-enable interrupts Re-enable interrupts
The pending IPI is noted
immediately, but alas,
the CPU is offline at
this point.
This of course, makes the smp-call-function IPI handler code running on
CPU 2 unhappy and it complains about "receiving an IPI on an offline
CPU".
One real example of the scenario on CPU 1 is the block layer's
complete-request call-path:
However, if we look closely, the block layer does check that the target
CPU is online before firing the IPI. So in this case, it is actually
the unfortunate ordering/timing of events in the stop-machine phase that
leads to receiving IPIs after the target CPU has gone offline.
In reality, getting a late IPI on an offline CPU is not too bad by
itself (this can happen even due to hardware latencies in IPI
send-receive). It is a bug only if the target CPU really went offline
without executing all the callbacks queued on its list. (Note that a
CPU is free to execute its pending smp-call-function callbacks in a
batch, without waiting for the corresponding IPIs to arrive for each one
of those callbacks).
So, fixing this issue can be broken up into two parts:
1. Ensure that a CPU goes offline only after executing all the
callbacks queued on it.
2. Modify the warning condition in the smp-call-function IPI handler
code such that it warns only if an offline CPU got an IPI *and* that
CPU had gone offline with callbacks still pending in its queue.
Achieving part 1 is straight-forward - just flush (execute) all the
queued callbacks on the outgoing CPU in the CPU_DYING stage[1],
including those callbacks for which the source CPU's IPIs might not have
been received on the outgoing CPU yet. Once we do this, an IPI that
arrives late on the CPU going offline (either due to the race mentioned
above, or due to hardware latencies) will be completely harmless, since
the outgoing CPU would have executed all the queued callbacks before
going offline.
Overall, this fix (parts 1 and 2 put together) additionally guarantees
that we will see a warning only when the *IPI-sender code* is buggy -
that is, if it queues the callback _after_ the target CPU has gone
offline.
[1]. The CPU_DYING part needs a little more explanation: by the time we
execute the CPU_DYING notifier callbacks, the CPU would have already
been marked offline. But we want to flush out the pending callbacks at
this stage, ignoring the fact that the CPU is offline. So restructure
the IPI handler code so that we can by-pass the "is-cpu-offline?" check
in this particular case. (Of course, the right solution here is to fix
CPU hotplug to mark the CPU offline _after_ invoking the CPU_DYING
notifiers, but this requires a lot of audit to ensure that this change
doesn't break any existing code; hence lets go with the solution
proposed above until that is done).
[akpm@linux-foundation.org: coding-style fixes] Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Suggested-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Cc: Borislav Petkov <bp@suse.de> Cc: Christoph Hellwig <hch@infradead.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Gautham R Shenoy <ego@linux.vnet.ibm.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Mel Gorman <mgorman@suse.de> Cc: Mike Galbraith <mgalbraith@suse.de> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Rafael J. Wysocki <rjw@rjwysocki.net> Cc: Rik van Riel <riel@redhat.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Tejun Heo <tj@kernel.org> Cc: Thomas Gleixner <tglx@linutronix.de> Tested-by: Sachin Kamat <sachin.kamat@samsung.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Steven Miao [Mon, 23 Jun 2014 20:22:02 +0000 (13:22 -0700)]
mm: nommu: per-thread vma cache fix
mm could be removed from current task struct, using previous vma->vm_mm
It will crash on blackfin after updated to Linux 3.15. The commit "mm:
per-thread vma caching" caused the crash. mm could be removed from
current task struct before
projects / linux-beck.git / log