Pull power management fixes from Rafael Wysocki:
"These fix a cpufreq regression introduced by recent changes related to
the generic DT driver, an initialization time memory leak in cpuidle
on ARM, a PM core bug that may cause system suspend/resume to fail on
some systems, a request type validation issue in the PM QoS framework
and two documentation-related issues.
Specifics:
- Fix a regression in cpufreq on systems using DT as the source of
CPU configuration information where two different code paths
attempt to create the cpufreq-dt device object (there can be only
one) and fix up the "compatible" matching for some TI platforms on
top of that (Viresh Kumar, Dave Gerlach).
- Fix an initialization time memory leak in cpuidle on ARM which
occurs if the cpuidle driver initialization fails (Stefan Wahren).
- Fix a PM core function that checks whether or not there are any
system suspend/resume callbacks for a device, but forgets to check
legacy callbacks which then may be skipped incorrectly and the
system may crash and/or the device may become unusable after a
suspend-resume cycle (Rafael Wysocki).
- Fix request type validation for latency tolerance PM QoS requests
which may lead to unexpected behavior (Jan Schönherr).
- Fix a broken link to PM documentation from a header file and a typo
in a PM document (Geert Uytterhoeven, Rafael Wysocki)"
* tag 'pm-4.14-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
cpufreq: ti-cpufreq: Support additional am43xx platforms
ARM: cpuidle: Avoid memleak if init fail
cpufreq: dt-platdev: Add some missing platforms to the blacklist
PM: core: Fix device_pm_check_callbacks()
PM: docs: Drop an excess character from devices.rst
PM / QoS: Use the correct variable to check the QoS request type
driver core: Fix link to device power management documentation
Pull input fixes from Dmitry Torokhov:
- fixes for two long standing issues (lock up and a crash) in force
feedback handling in uinput driver
- tweak to firmware update timing in Elan I2C touchpad driver.
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
Input: elan_i2c - extend Flash-Write delay
Input: uinput - avoid crash when sending FF request to device going away
Input: uinput - avoid FF flush when destroying device
Pull seccomp updates from Kees Cook:
"Major additions:
- sysctl and seccomp operation to discover available actions
(tyhicks)
- new per-filter configurable logging infrastructure and sysctl
(tyhicks)
- SECCOMP_RET_LOG to log allowed syscalls (tyhicks)
- SECCOMP_RET_KILL_PROCESS as the new strictest possible action
- self-tests for new behaviors"
[ This is the seccomp part of the security pull request during the merge
window that was nixed due to unrelated problems - Linus ]
* tag 'seccomp-v4.14-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
samples: Unrename SECCOMP_RET_KILL
selftests/seccomp: Test thread vs process killing
seccomp: Implement SECCOMP_RET_KILL_PROCESS action
seccomp: Introduce SECCOMP_RET_KILL_PROCESS
seccomp: Rename SECCOMP_RET_KILL to SECCOMP_RET_KILL_THREAD
seccomp: Action to log before allowing
seccomp: Filter flag to log all actions except SECCOMP_RET_ALLOW
seccomp: Selftest for detection of filter flag support
seccomp: Sysctl to configure actions that are allowed to be logged
seccomp: Operation for checking if an action is available
seccomp: Sysctl to display available actions
seccomp: Provide matching filter for introspection
selftests/seccomp: Refactor RET_ERRNO tests
selftests/seccomp: Add simple seccomp overhead benchmark
selftests/seccomp: Add tests for basic ptrace actions
Pull xen fixes from Juergen Gross:
"A fix for a missing __init annotation and two cleanup patches"
* tag 'for-linus-4.14b-rc2-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
xen, arm64: drop dummy lookup_address()
xen: don't compile pv-specific parts if XEN_PV isn't configured
xen: x86: mark xen_find_pt_base as __init
In linux-4.13, Wei worked hard to convert dst to a traditional
refcounted model, removing GC.
We now want to make sure a dst refcount can not transition from 0 back
to 1.
The problem here is that input path attached a not refcounted dst to an
skb. Then later, because packet is forwarded and hits skb_dst_force()
before exiting RCU section, we might try to take a refcount on one dst
that is about to be freed, if another cpu saw 1 -> 0 transition in
dst_release() and queued the dst for freeing after one RCU grace period.
Lets unify skb_dst_force() and skb_dst_force_safe(), since we should
always perform the complete check against dst refcount, and not assume
it is not zero.
Bugzilla : https://bugzilla.kernel.org/show_bug.cgi?id=197005
[ 989.919496] skb_dst_force+0x32/0x34
[ 989.919498] __dev_queue_xmit+0x1ad/0x482
[ 989.919501] ? eth_header+0x28/0xc6
[ 989.919502] dev_queue_xmit+0xb/0xd
[ 989.919504] neigh_connected_output+0x9b/0xb4
[ 989.919507] ip_finish_output2+0x234/0x294
[ 989.919509] ? ipt_do_table+0x369/0x388
[ 989.919510] ip_finish_output+0x12c/0x13f
[ 989.919512] ip_output+0x53/0x87
[ 989.919513] ip_forward_finish+0x53/0x5a
[ 989.919515] ip_forward+0x2cb/0x3e6
[ 989.919516] ? pskb_trim_rcsum.part.9+0x4b/0x4b
[ 989.919518] ip_rcv_finish+0x2e2/0x321
[ 989.919519] ip_rcv+0x26f/0x2eb
[ 989.919522] ? vlan_do_receive+0x4f/0x289
[ 989.919523] __netif_receive_skb_core+0x467/0x50b
[ 989.919526] ? tcp_gro_receive+0x239/0x239
[ 989.919529] ? inet_gro_receive+0x226/0x238
[ 989.919530] __netif_receive_skb+0x4d/0x5f
[ 989.919532] netif_receive_skb_internal+0x5c/0xaf
[ 989.919533] napi_gro_receive+0x45/0x81
[ 989.919536] ixgbe_poll+0xc8a/0xf09
[ 989.919539] ? kmem_cache_free_bulk+0x1b6/0x1f7
[ 989.919540] net_rx_action+0xf4/0x266
[ 989.919543] __do_softirq+0xa8/0x19d
[ 989.919545] irq_exit+0x5d/0x6b
[ 989.919546] do_IRQ+0x9c/0xb5
[ 989.919548] common_interrupt+0x93/0x93
[ 989.919548] </IRQ>
Similarly dst_clone() can use dst_hold() helper to have additional
debugging, as a follow up to commit 44ebe79149 ("net: add debug
atomic_inc_not_zero() in dst_hold()")
In net-next we will convert dst atomic_t to refcount_t for peace of
mind.
Fixes: a4c2fd7f78 ("net: remove DST_NOCACHE flag")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Wei Wang <weiwan@google.com>
Reported-by: Paweł Staszewski <pstaszewski@itcare.pl>
Bisected-by: Paweł Staszewski <pstaszewski@itcare.pl>
Acked-by: Wei Wang <weiwan@google.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Normally, when input device supporting force feedback effects is being
destroyed, we try to "flush" currently playing effects, so that the
physical device does not continue vibrating (or executing other effects).
Unfortunately this does not work well for uinput as flushing of the effects
deadlocks with the destroy action:
- if device is being destroyed because the file descriptor is being closed,
then there is noone to even service FF requests;
- if device is being destroyed because userspace sent UI_DEV_DESTROY,
while theoretically it could be possible to service FF requests,
userspace is unlikely to do so (they'd need to make sure FF handling
happens on a separate thread) even if kernel solves the issue with FF
ioctls deadlocking with UI_DEV_DESTROY ioctl on udev->mutex.
To avoid lockups like the one below, let's install a custom input device
flush handler, and avoid trying to flush force feedback effects when we
destroying the device, and instead rely on uinput to shut off the device
properly.
NMI watchdog: Watchdog detected hard LOCKUP on cpu 3
...
<<EOE>> [<ffffffff817a0307>] _raw_spin_lock_irqsave+0x37/0x40
[<ffffffff810e633d>] complete+0x1d/0x50
[<ffffffffa00ba08c>] uinput_request_done+0x3c/0x40 [uinput]
[<ffffffffa00ba587>] uinput_request_submit.part.7+0x47/0xb0 [uinput]
[<ffffffffa00bb62b>] uinput_dev_erase_effect+0x5b/0x76 [uinput]
[<ffffffff815d91ad>] erase_effect+0xad/0xf0
[<ffffffff815d929d>] flush_effects+0x4d/0x90
[<ffffffff815d4cc0>] input_flush_device+0x40/0x60
[<ffffffff815daf1c>] evdev_cleanup+0xac/0xc0
[<ffffffff815daf5b>] evdev_disconnect+0x2b/0x60
[<ffffffff815d74ac>] __input_unregister_device+0xac/0x150
[<ffffffff815d75f7>] input_unregister_device+0x47/0x70
[<ffffffffa00bac45>] uinput_destroy_device+0xb5/0xc0 [uinput]
[<ffffffffa00bb2de>] uinput_ioctl_handler.isra.9+0x65e/0x740 [uinput]
[<ffffffff811231ab>] ? do_futex+0x12b/0xad0
[<ffffffffa00bb3f8>] uinput_ioctl+0x18/0x20 [uinput]
[<ffffffff81241248>] do_vfs_ioctl+0x298/0x480
[<ffffffff81337553>] ? security_file_ioctl+0x43/0x60
[<ffffffff812414a9>] SyS_ioctl+0x79/0x90
[<ffffffff817a04ee>] entry_SYSCALL_64_fastpath+0x12/0x71
Reported-by: Rodrigo Rivas Costa <rodrigorivascosta@gmail.com>
Reported-by: Clément VUCHENER <clement.vuchener@gmail.com>
Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=193741
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Commit 3f1ac7a700 ("net: ethtool: add new ETHTOOL_xLINKSETTINGS API")
deprecated the ethtool_cmd::transceiver field, which was fine in
premise, except that the PHY library was actually using it to report the
type of transceiver: internal or external.
Use the first word of the reserved field to put this __u8 transceiver
field back in. It is made read-only, and we don't expect the
ETHTOOL_xLINKSETTINGS API to be doing anything with this anyway, so this
is mostly for the legacy path where we do:
ethtool_get_settings()
-> dev->ethtool_ops->get_link_ksettings()
-> convert_link_ksettings_to_legacy_settings()
to have no information loss compared to the legacy get_settings API.
Fixes: 3f1ac7a700 ("net: ethtool: add new ETHTOOL_xLINKSETTINGS API")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This reverts commit 74def747bc.
The change to the helper function is only correct for the /proc/irq/
readout usage, but breaks the existing x86 usage of that function.
Reported-by: Yanko Kaneti <yaneti@declera.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Marc Zyngier <marc.zyngier@arm.com>
This patch fixes a bug exhibited by the following scenario:
1. fd1 = perf_event_open with attr.config = ID1
2. attach bpf program prog1 to fd1
3. fd2 = perf_event_open with attr.config = ID1
<this will be successful>
4. user program closes fd2 and prog1 is detached from the tracepoint.
5. user program with fd1 does not work properly as tracepoint
no output any more.
The issue happens at step 4. Multiple perf_event_open can be called
successfully, but only one bpf prog pointer in the tp_event. In the
current logic, any fd release for the same tp_event will free
the tp_event->prog.
The fix is to free tp_event->prog only when the closing fd
corresponds to the one which registered the program.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
On recent Intel platforms (Haswell, Broadwell, Skylake, ApolloLake,
KabyLake, ...), the IEC Coding Type (ICT) bitfield in the Digital
Converter Control #3 needs to be set explicitly for HDMI/DisplayPort
High Bit Rate (HBR) audio playback to work. This was not required in
earlier platforms when HBR was first introduced. The ICT bits are
defined in Section 7.3.3.9 of the HDaudio 1.0a specification.
Since the ICT bitfield was not specified for HDAudio 1.0 devices
(before 2009), we only program it on machines more recent than
Haswell.
We tested that this fix is not needed on Baytrail-I (MinnowBoard
Turbot) and believe by extension it also does not apply to Braswell.
[ Moved AC_VERB_SET_DIGI_CONVERT_3 definition to the right place
by tiwai ]
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98797
Signed-off-by: Sriram Periyasamy <sriramx.periyasamy@intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Subhransu S. Prusty <subhransu.s.prusty@intel.com>
Acked-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Due to commit db3e50f323 (device property: Get rid of struct
fwnode_handle type field), ACPI_HANDLE() inadvertently became
a GPL-only call. The call path that led to that was:
ACPI_HANDLE()
ACPI_COMPANION()
to_acpi_device_node()
is_acpi_device_node()
acpi_device_fwnode_ops
DECLARE_ACPI_FWNODE_OPS(acpi_device_fwnode_ops);
...and the new DECLARE_ACPI_FWNODE_OPS() includes
EXPORT_SYMBOL_GPL, whereas previously it was a static struct.
In order to avoid changing any of that, let's instead provide ever
so slightly better encapsulation of those struct fwnode_operations
instances. Those do not really need to be directly used in
inline function calls in header files. Simply moving two small
functions (is_acpi_device_node and is_acpi_data_node) out of
acpi_bus.h, and into a .c file, does that.
That leaves the internals of struct fwnode_operations as GPL-only
(which I think was the intent all along), but un-breaks any driver
code out there that relies on the ACPI subsystem's being (historically)
an EXPORT_SYMBOL-usable system. By that, I mean, ACPI_HANDLE() and
other basic ACPI calls were non-GPL-protected.
Also, while I'm there, remove a tiny bit of redundancy that was missed
in the earlier commit, by having is_acpi_node() use the other two
routines, instead of checking fwnode directly.
Fixes: db3e50f323 (device property: Get rid of struct fwnode_handle type field)
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The ipmmu-vmsa driver fails in compile-testing on non-OF platforms:
drivers/iommu/ipmmu-vmsa.o: In function `ipmmu_of_xlate':
ipmmu-vmsa.c:(.text+0x740): undefined reference to `of_find_device_by_node'
It would be reasonable to assume that this interface works but
returns failure on non-OF builds, like it does on machines that
have been booted in another way, so this adds another inline
function helper.
Fixes: 7b2d59611f ("iommu/ipmmu-vmsa: Replace local utlb code with fwspec ids")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Rob Herring <robh@kernel.org>
Andrey Konovalov reported a possible out-of-bounds problem for a USB interface
association descriptor. He writes:
It seems there's no proper size check of a USB_DT_INTERFACE_ASSOCIATION
descriptor. It's only checked that the size is >= 2 in
usb_parse_configuration(), so find_iad() might do out-of-bounds access
to intf_assoc->bInterfaceCount.
And he's right, we don't check for crazy descriptors of this type very well, so
resolve this problem. Yet another issue found by syzkaller...
Reported-by: Andrey Konovalov <andreyknvl@google.com>
Tested-by: Andrey Konovalov <andreyknvl@google.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
DRIVER_ATTR is no longer in use, and driver authors should be using
DRIVER_ATTR_RW() or DRIVER_ATTR_RO() or DRIVER_ATTR_WO() instead in
order to always get the permissions correct. So remove it so that no
one can use it anymore.
Acked-by: Alan Tull <atull@kernel.org>
Reviewed-by: Moritz Fischer <mdf@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Correct location as of commit 2728b2d2e5 (PM / core / docs:
Convert sleep states API document to reST).
Fixes: 2728b2d2e5 (PM / core / docs: Convert sleep states API document to reST)
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
There is no plan yet to do a v2 board. And even if we were to do it only
some IPs would actually change, so it be best to add suffixes at that
point, not now !
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
some trivial amdkfd cleanups
* tag 'drm-amdkfd-next-2017-09-02' of git://people.freedesktop.org/~gabbayo/linux:
drm/amdkfd: pass queue's mqd when destroying mqd
drm/amdkfd: remove memset before memcpy
uapi linux/kfd_ioctl.h: only use __u32 and __u64
Pull networking fixes from David Miller:
1) Fix hotplug deadlock in hv_netvsc, from Stephen Hemminger.
2) Fix double-free in rmnet driver, from Dan Carpenter.
3) INET connection socket layer can double put request sockets, fix
from Eric Dumazet.
4) Don't match collect metadata-mode tunnels if the device is down,
from Haishuang Yan.
5) Do not perform TSO6/GSO on ipv6 packets with extensions headers in
be2net driver, from Suresh Reddy.
6) Fix scaling error in gen_estimator, from Eric Dumazet.
7) Fix 64-bit statistics deadlock in systemport driver, from Florian
Fainelli.
8) Fix use-after-free in sctp_sock_dump, from Xin Long.
9) Reject invalid BPF_END instructions in verifier, from Edward Cree.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (43 commits)
mlxsw: spectrum_router: Only handle IPv4 and IPv6 events
Documentation: link in networking docs
tcp: fix data delivery rate
bpf/verifier: reject BPF_ALU64|BPF_END
sctp: do not mark sk dumped when inet_sctp_diag_fill returns err
sctp: fix an use-after-free issue in sctp_sock_dump
netvsc: increase default receive buffer size
tcp: update skb->skb_mstamp more carefully
net: ipv4: fix l3slave check for index returned in IP_PKTINFO
net: smsc911x: Quieten netif during suspend
net: systemport: Fix 64-bit stats deadlock
net: vrf: avoid gcc-4.6 warning
qed: remove unnecessary call to memset
tg3: clean up redundant initialization of tnapi
tls: make tls_sw_free_resources static
sctp: potential read out of bounds in sctp_ulpevent_type_enabled()
MAINTAINERS: review Renesas DT bindings as well
net_sched: gen_estimator: fix scaling error in bytes/packets samples
nfp: wait for the NSP resource to appear on boot
nfp: wait for board state before talking to the NSP
...
Pull MIPS updates from Ralf Baechle:
"This is the main pull request for 4.14 for MIPS; below a summary of
the non-merge commits:
CM:
- Rename mips_cm_base to mips_gcr_base
- Specify register size when generating accessors
- Use BIT/GENMASK for register fields, order & drop shifts
- Add cluster & block args to mips_cm_lock_other()
CPC:
- Use common CPS accessor generation macros
- Use BIT/GENMASK for register fields, order & drop shifts
- Introduce register modify (set/clear/change) accessors
- Use change_*, set_* & clear_* where appropriate
- Add CM/CPC 3.5 register definitions
- Use GlobalNumber macros rather than magic numbers
- Have asm/mips-cps.h include CM & CPC headers
- Cluster support for topology functions
- Detect CPUs in secondary clusters
CPS:
- Read GIC_VL_IDENT directly, not via irqchip driver
DMA:
- Consolidate coherent and non-coherent dma_alloc code
- Don't use dma_cache_sync to implement fd_cacheflush
FPU emulation / FP assist code:
- Another series of 14 commits fixing corner cases such as NaN
propgagation and other special input values.
- Zero bits 32-63 of the result for a CLASS.D instruction.
- Enhanced statics via debugfs
- Do not use bools for arithmetic. GCC 7.1 moans about this.
- Correct user fault_addr type
Generic MIPS:
- Enhancement of stack backtraces
- Cleanup from non-existing options
- Handle non word sized instructions when examining frame
- Fix detection and decoding of ADDIUSP instruction
- Fix decoding of SWSP16 instruction
- Refactor handling of stack pointer in get_frame_info
- Remove unreachable code from force_fcr31_sig()
- Convert to using %pOF instead of full_name
- Remove the R6000 support.
- Move FP code from *_switch.S to *_fpu.S
- Remove unused ST_OFF from r2300_switch.S
- Allow platform to specify multiple its.S files
- Add #includes to various files to ensure code builds reliable and
without warning..
- Remove __invalidate_kernel_vmap_range
- Remove plat_timer_setup
- Declare various variables & functions static
- Abstract CPU core & VP(E) ID access through accessor functions
- Store core & VP IDs in GlobalNumber-style variable
- Unify checks for sibling CPUs
- Add CPU cluster number accessors
- Prevent direct use of generic_defconfig
- Make CONFIG_MIPS_MT_SMP default y
- Add __ioread64_copy
- Remove unnecessary inclusions of linux/irqchip/mips-gic.h
GIC:
- Introduce asm/mips-gic.h with accessor functions
- Use new GIC accessor functions in mips-gic-timer
- Remove counter access functions from irq-mips-gic.c
- Remove gic_read_local_vp_id() from irq-mips-gic.c
- Simplify shared interrupt pending/mask reads in irq-mips-gic.c
- Simplify gic_local_irq_domain_map() in irq-mips-gic.c
- Drop gic_(re)set_mask() functions in irq-mips-gic.c
- Remove gic_set_polarity(), gic_set_trigger(), gic_set_dual_edge(),
gic_map_to_pin() and gic_map_to_vpe() from irq-mips-gic.c.
- Convert remaining shared reg access, local int mask access and
remaining local reg access to new accessors
- Move GIC_LOCAL_INT_* to asm/mips-gic.h
- Remove GIC_CPU_INT* macros from irq-mips-gic.c
- Move various definitions to the driver
- Remove gic_get_usm_range()
- Remove __gic_irq_dispatch() forward declaration
- Remove gic_init()
- Use mips_gic_present() in place of gic_present and remove
gic_present
- Move gic_get_c0_*_int() to asm/mips-gic.h
- Remove linux/irqchip/mips-gic.h
- Inline __gic_init()
- Inline gic_basic_init()
- Make pcpu_masks a per-cpu variable
- Use pcpu_masks to avoid reading GIC_SH_MASK*
- Clean up mti, reserved-cpu-vectors handling
- Use cpumask_first_and() in gic_set_affinity()
- Let the core set struct irq_common_data affinity
microMIPS:
- Fix microMIPS stack unwinding on big endian systems
MIPS-GIC:
- SYNC after enabling GIC region
NUMA:
- Remove the unused parent_node() macro
R6:
- Constify r2_decoder_tables
- Add accessor & bit definitions for GlobalNumber
SMP:
- Constify smp ops
- Allow boot_secondary SMP op to return errors
VDSO:
- Drop gic_get_usm_range() usage
- Avoid use of linux/irqchip/mips-gic.h
Platform changes:
Alchemy:
- Add devboard machine type to cpuinfo
- update cpu feature overrides
- Threaded carddetect irqs for devboards
AR7:
- allow NULL clock for clk_get_rate
BCM63xx:
- Fix ENETDMA_6345_MAXBURST_REG offset
- Allow NULL clock for clk_get_rate
CI20:
- Enable GPIO and RTC drivers in defconfig
- Add ethernet and fixed-regulator nodes to DTS
Generic platform:
- Move Boston and NI 169445 FIT image source to their own files
- Include asm/bootinfo.h for plat_fdt_relocated()
- Include asm/time.h for get_c0_*_int()
- Include asm/bootinfo.h for plat_fdt_relocated()
- Include asm/time.h for get_c0_*_int()
- Allow filtering enabled boards by requirements
- Don't explicitly disable CONFIG_USB_SUPPORT
- Bump default NR_CPUS to 16
JZ4700:
- Probe the jz4740-rtc driver from devicetree
Lantiq:
- Drop check of boot select from the spi-falcon driver.
- Drop check of boot select from the lantiq-flash MTD driver.
- Access boot cause register in the watchdog driver through regmap
- Add device tree binding documentation for the watchdog driver
- Add docs for the RCU DT bindings.
- Convert the fpi bus driver to a platform_driver
- Remove ltq_reset_cause() and ltq_boot_select(
- Switch to a proper reset driver
- Switch to a new drivers/soc GPHY driver
- Add an USB PHY driver for the Lantiq SoCs using the RCU module
- Use of_platform_default_populate instead of __dt_register_buses
- Enable MFD_SYSCON to be able to use it for the RCU MFD
- Replace ltq_boot_select() with dummy implementation.
Loongson 2F:
- Allow NULL clock for clk_get_rate
Malta:
- Use new GIC accessor functions
NI 169445:
- Add support for NI 169445 board.
- Only include in 32r2el kernels
Octeon:
- Add support for watchdog of 78XX SOCs.
- Add support for watchdog of CN68XX SOCs.
- Expose support for mips32r1, mips32r2 and mips64r1
- Enable more drivers in config file
- Add support for accessing the boot vector.
- Remove old boot vector code from watchdog driver
- Define watchdog registers for 70xx, 73xx, 78xx, F75xx.
- Make CSR functions node aware.
- Allow access to CIU3 IRQ domains.
- Misc cleanups in the watchdog driver
Omega2+:
- New board, add support and defconfig
Pistachio:
- Enable Root FS on NFS in defconfig
Ralink:
- Add Mediatek MT7628A SoC
- Allow NULL clock for clk_get_rate
- Explicitly request exclusive reset control in the pci-mt7620 PCI driver.
SEAD3:
- Only include in 32 bit kernels by default
VoCore:
- Add VoCore as a vendor t0 dt-bindings
- Add defconfig file"
* '4.14-features' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus: (167 commits)
MIPS: Refactor handling of stack pointer in get_frame_info
MIPS: Stacktrace: Fix microMIPS stack unwinding on big endian systems
MIPS: microMIPS: Fix decoding of swsp16 instruction
MIPS: microMIPS: Fix decoding of addiusp instruction
MIPS: microMIPS: Fix detection of addiusp instruction
MIPS: Handle non word sized instructions when examining frame
MIPS: ralink: allow NULL clock for clk_get_rate
MIPS: Loongson 2F: allow NULL clock for clk_get_rate
MIPS: BCM63XX: allow NULL clock for clk_get_rate
MIPS: AR7: allow NULL clock for clk_get_rate
MIPS: BCM63XX: fix ENETDMA_6345_MAXBURST_REG offset
mips: Save all registers when saving the frame
MIPS: Add DWARF unwinding to assembly
MIPS: Make SAVE_SOME more standard
MIPS: Fix issues in backtraces
MIPS: jz4780: DTS: Probe the jz4740-rtc driver from devicetree
MIPS: Ci20: Enable RTC driver
watchdog: octeon-wdt: Add support for 78XX SOCs.
watchdog: octeon-wdt: Add support for cn68XX SOCs.
watchdog: octeon-wdt: File cleaning.
...
Commit 86fdb3448c ("sctp: ensure ep is not destroyed before doing the
dump") tried to fix an use-after-free issue by checking !sctp_sk(sk)->ep
with holding sock and sock lock.
But Paolo noticed that endpoint could be destroyed in sctp_rcv without
sock lock protection. It means the use-after-free issue still could be
triggered when sctp_rcv put and destroy ep after sctp_sock_dump checks
!ep, although it's pretty hard to reproduce.
I could reproduce it by mdelay in sctp_rcv while msleep in sctp_close
and sctp_sock_dump long time.
This patch is to add another param cb_done to sctp_for_each_transport
and dump ep->assocs with holding tsp after jumping out of transport's
traversal in it to avoid this issue.
It can also improve sctp diag dump to make it run faster, as no need
to save sk into cb->args[5] and keep calling sctp_for_each_transport
any more.
This patch is also to use int * instead of int for the pos argument
in sctp_for_each_transport, which could make postion increment only
in sctp_for_each_transport and no need to keep changing cb->args[2]
in sctp_sock_filter and sctp_sock_dump any more.
Fixes: 86fdb3448c ("sctp: ensure ep is not destroyed before doing the dump")
Reported-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Which is the equivalent of what we have in regular waitqueues.
I'm not crazy about the name, but this also helps us get both
apis closer -- which iirc comes originally from the -net folks.
We also duplicate the comments for the lockless swait_active(),
from wait.h. Future users will make use of this interface.
Signed-off-by: Davidlohr Bueso <dbueso@suse.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Pull misc leftovers from Al Viro.
* 'work.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
fix the __user misannotations in asm-generic get_user/put_user
fput: Don't reinvent the wheel but use existing llist API
namespace.c: Don't reinvent the wheel but use existing llist API
Pull nowait read support from Al Viro:
"Support IOCB_NOWAIT for buffered reads and block devices"
* 'work.read_write' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
block_dev: support RFW_NOWAIT on block device nodes
fs: support RWF_NOWAIT for buffered reads
fs: support IOCB_NOWAIT in generic_file_buffered_read
fs: pass iocb to do_generic_file_read
Pull mount flag updates from Al Viro:
"Another chunk of fmount preparations from dhowells; only trivial
conflicts for that part. It separates MS_... bits (very grotty
mount(2) ABI) from the struct super_block ->s_flags (kernel-internal,
only a small subset of MS_... stuff).
This does *not* convert the filesystems to new constants; only the
infrastructure is done here. The next step in that series is where the
conflicts would be; that's the conversion of filesystems. It's purely
mechanical and it's better done after the merge, so if you could run
something like
list=$(for i in MS_RDONLY MS_NOSUID MS_NODEV MS_NOEXEC MS_SYNCHRONOUS MS_MANDLOCK MS_DIRSYNC MS_NOATIME MS_NODIRATIME MS_SILENT MS_POSIXACL MS_KERNMOUNT MS_I_VERSION MS_LAZYTIME; do git grep -l $i fs drivers/staging/lustre drivers/mtd ipc mm include/linux; done|sort|uniq|grep -v '^fs/namespace.c$')
sed -i -e 's/\<MS_RDONLY\>/SB_RDONLY/g' \
-e 's/\<MS_NOSUID\>/SB_NOSUID/g' \
-e 's/\<MS_NODEV\>/SB_NODEV/g' \
-e 's/\<MS_NOEXEC\>/SB_NOEXEC/g' \
-e 's/\<MS_SYNCHRONOUS\>/SB_SYNCHRONOUS/g' \
-e 's/\<MS_MANDLOCK\>/SB_MANDLOCK/g' \
-e 's/\<MS_DIRSYNC\>/SB_DIRSYNC/g' \
-e 's/\<MS_NOATIME\>/SB_NOATIME/g' \
-e 's/\<MS_NODIRATIME\>/SB_NODIRATIME/g' \
-e 's/\<MS_SILENT\>/SB_SILENT/g' \
-e 's/\<MS_POSIXACL\>/SB_POSIXACL/g' \
-e 's/\<MS_KERNMOUNT\>/SB_KERNMOUNT/g' \
-e 's/\<MS_I_VERSION\>/SB_I_VERSION/g' \
-e 's/\<MS_LAZYTIME\>/SB_LAZYTIME/g' \
$list
and commit it with something along the lines of 'convert filesystems
away from use of MS_... constants' as commit message, it would save a
quite a bit of headache next cycle"
* 'work.mount' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
VFS: Differentiate mount flags (MS_*) from internal superblock flags
VFS: Convert sb->s_flags & MS_RDONLY to sb_rdonly(sb)
vfs: Add sb_rdonly(sb) to query the MS_RDONLY flag on s_flags
Pull more set_fs removal from Al Viro:
"Christoph's 'use kernel_read and friends rather than open-coding
set_fs()' series"
* 'work.set_fs' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
fs: unexport vfs_readv and vfs_writev
fs: unexport vfs_read and vfs_write
fs: unexport __vfs_read/__vfs_write
lustre: switch to kernel_write
gadget/f_mass_storage: stop messing with the address limit
mconsole: switch to kernel_read
btrfs: switch write_buf to kernel_write
net/9p: switch p9_fd_read to kernel_write
mm/nommu: switch do_mmap_private to kernel_read
serial2002: switch serial2002_tty_write to kernel_{read/write}
fs: make the buf argument to __kernel_write a void pointer
fs: fix kernel_write prototype
fs: fix kernel_read prototype
fs: move kernel_read to fs/read_write.c
fs: move kernel_write to fs/read_write.c
autofs4: switch autofs4_write to __kernel_write
ashmem: switch to ->read_iter
Pull ipc compat cleanup and 64-bit time_t from Al Viro:
"IPC copyin/copyout sanitizing, including 64bit time_t work from Deepa
Dinamani"
* 'work.ipc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
utimes: Make utimes y2038 safe
ipc: shm: Make shmid_kernel timestamps y2038 safe
ipc: sem: Make sem_array timestamps y2038 safe
ipc: msg: Make msg_queue timestamps y2038 safe
ipc: mqueue: Replace timespec with timespec64
ipc: Make sys_semtimedop() y2038 safe
get rid of SYSVIPC_COMPAT on ia64
semtimedop(): move compat to native
shmat(2): move compat to native
msgrcv(2), msgsnd(2): move compat to native
ipc(2): move compat to native
ipc: make use of compat ipc_perm helpers
semctl(): move compat to native
semctl(): separate all layout-dependent copyin/copyout
msgctl(): move compat to native
msgctl(): split the actual work from copyin/copyout
ipc: move compat shmctl to native
shmctl: split the work from copyin/copyout
Pull zstd support from Chris Mason:
"Nick Terrell's patch series to add zstd support to the kernel has been
floating around for a while. After talking with Dave Sterba, Herbert
and Phillip, we decided to send the whole thing in as one pull
request.
zstd is a big win in speed over zlib and in compression ratio over
lzo, and the compression team here at FB has gotten great results
using it in production. Nick will continue to update the kernel side
with new improvements from the open source zstd userland code.
Nick has a number of benchmarks for the main zstd code in his lib/zstd
commit:
I ran the benchmarks on a Ubuntu 14.04 VM with 2 cores and 4 GiB
of RAM. The VM is running on a MacBook Pro with a 3.1 GHz Intel
Core i7 processor, 16 GB of RAM, and a SSD. I benchmarked using
`silesia.tar` [3], which is 211,988,480 B large. Run the following
commands for the benchmark:
sudo modprobe zstd_compress_test
sudo mknod zstd_compress_test c 245 0
sudo cp silesia.tar zstd_compress_test
The time is reported by the time of the userland `cp`.
The MB/s is computed with
1,536,217,008 B / time(buffer size, hash)
which includes the time to copy from userland.
The Adjusted MB/s is computed with
1,536,217,088 B / (time(buffer size, hash) - time(buffer size, none)).
The memory reported is the amount of memory the compressor
requests.
| Method | Size (B) | Time (s) | Ratio | MB/s | Adj MB/s | Mem (MB) |
|----------|----------|----------|-------|---------|----------|----------|
| none | 11988480 | 0.100 | 1 | 2119.88 | - | - |
| zstd -1 | 73645762 | 1.044 | 2.878 | 203.05 | 224.56 | 1.23 |
| zstd -3 | 66988878 | 1.761 | 3.165 | 120.38 | 127.63 | 2.47 |
| zstd -5 | 65001259 | 2.563 | 3.261 | 82.71 | 86.07 | 2.86 |
| zstd -10 | 60165346 | 13.242 | 3.523 | 16.01 | 16.13 | 13.22 |
| zstd -15 | 58009756 | 47.601 | 3.654 | 4.45 | 4.46 | 21.61 |
| zstd -19 | 54014593 | 102.835 | 3.925 | 2.06 | 2.06 | 60.15 |
| zlib -1 | 77260026 | 2.895 | 2.744 | 73.23 | 75.85 | 0.27 |
| zlib -3 | 72972206 | 4.116 | 2.905 | 51.50 | 52.79 | 0.27 |
| zlib -6 | 68190360 | 9.633 | 3.109 | 22.01 | 22.24 | 0.27 |
| zlib -9 | 67613382 | 22.554 | 3.135 | 9.40 | 9.44 | 0.27 |
I benchmarked zstd decompression using the same method on the same
machine. The benchmark file is located in the upstream zstd repo
under `contrib/linux-kernel/zstd_decompress_test.c` [4]. The
memory reported is the amount of memory required to decompress
data compressed with the given compression level. If you know the
maximum size of your input, you can reduce the memory usage of
decompression irrespective of the compression level.
| Method | Time (s) | MB/s | Adjusted MB/s | Memory (MB) |
|----------|----------|---------|---------------|-------------|
| none | 0.025 | 8479.54 | - | - |
| zstd -1 | 0.358 | 592.15 | 636.60 | 0.84 |
| zstd -3 | 0.396 | 535.32 | 571.40 | 1.46 |
| zstd -5 | 0.396 | 535.32 | 571.40 | 1.46 |
| zstd -10 | 0.374 | 566.81 | 607.42 | 2.51 |
| zstd -15 | 0.379 | 559.34 | 598.84 | 4.61 |
| zstd -19 | 0.412 | 514.54 | 547.77 | 8.80 |
| zlib -1 | 0.940 | 225.52 | 231.68 | 0.04 |
| zlib -3 | 0.883 | 240.08 | 247.07 | 0.04 |
| zlib -6 | 0.844 | 251.17 | 258.84 | 0.04 |
| zlib -9 | 0.837 | 253.27 | 287.64 | 0.04 |
I ran a long series of tests and benchmarks on the btrfs side and the
gains are very similar to the core benchmarks Nick ran"
* 'zstd-minimal' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
squashfs: Add zstd support
btrfs: Add zstd support
lib: Add zstd modules
lib: Add xxhash module
Pull device mapper updates from Mike Snitzer:
- Some request-based DM core and DM multipath fixes and cleanups
- Constify a few variables in DM core and DM integrity
- Add bufio optimization and checksum failure accounting to DM
integrity
- Fix DM integrity to avoid checking integrity of failed reads
- Fix DM integrity to use init_completion
- A couple DM log-writes target fixes
- Simplify DAX flushing by eliminating the unnecessary flush
abstraction that was stood up for DM's use.
* tag 'for-4.14/dm-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm:
dax: remove the pmem_dax_ops->flush abstraction
dm integrity: use init_completion instead of COMPLETION_INITIALIZER_ONSTACK
dm integrity: make blk_integrity_profile structure const
dm integrity: do not check integrity for failed read operations
dm log writes: fix >512b sectorsize support
dm log writes: don't use all the cpu while waiting to log blocks
dm ioctl: constify ioctl lookup table
dm: constify argument arrays
dm integrity: count and display checksum failures
dm integrity: optimize writing dm-bufio buffers that are partially changed
dm rq: do not update rq partially in each ending bio
dm rq: make dm-sq requeuing behavior consistent with dm-mq behavior
dm mpath: complain about unsupported __multipath_map_bio() return values
dm mpath: avoid that building with W=1 causes gcc 7 to complain about fall-through
Pull fbdev updates from Bartlomiej Zolnierkiewicz:
- make fbcon a built-time depency for fbdev (fbcon was tristate option
before, now it is a bool) - this is a first step in preparations for
making console_lock usage saner (currently it acts like the BKL for
all things fbdev/fbcon) (Daniel Vetter)
- add fbcon=margin:<color> command line option to select the fbcon
margin color (David Lechner)
- add DMI quirk table for x86 systems which need fbcon rotation
(devices like Asus T100HA, GPD Pocket, the GPD win and the I.T.Works
TW891) (Hans de Goede)
- fix 1bpp logo support for unusual width (needed by LEGO MINDSTORMS
EV3) (David Lechner)
- enable Xilinx FB driver for ARM ZynqMP platform (Michal Simek)
- fix use after free in the error path of udlfb driver (Anton Vasilyev)
- fix error return code handling in pxa3xx_gcu driver (Gustavo A. R.
Silva)
- fix bootparams.screeninfo arguments checking in vgacon (Jan H.
Schönherr)
- do not leak uninitialized padding in clk to userspace in the debug
code of atyfb driver (Vladis Dronov)
- fix compiler warnings in fbcon code and matroxfb driver (Arnd
Bergmann)
- convert fbdev susbsytem to using %pOF instead of full_name (Rob
Herring)
- structures constifications (Arvind Yadav, Bhumika Goyal, Gustavo A.
R. Silva, Julia Lawall)
- misc cleanups (Gustavo A. R. Silva, Hyun Kwon, Julia Lawall, Kuninori
Morimoto, Lynn Lei)
* tag 'fbdev-v4.14' of git://github.com/bzolnier/linux: (75 commits)
video/console: Update BIOS dates list for GPD win console rotation DMI quirk
video/console: Add rotated LCD-panel DMI quirk for the VIOS LTH17
video: fbdev: sis: fix duplicated code for different branches
video: fbdev: make fb_var_screeninfo const
video: fbdev: aty: do not leak uninitialized padding in clk to userspace
vgacon: Prevent faulty bootparams.screeninfo from causing harm
video: fbdev: make fb_videomode const
video/console: Add new BIOS date for GPD pocket to dmi quirk table
fbcon: remove restriction on margin color
video: ARM CLCD: constify amba_id
video: fm2fb: constify zorro_device_id
video: fbdev: annotate fb_fix_screeninfo with const and __initconst
omapfb: constify omap_video_timings structures
video: fbdev: udlfb: Fix use after free on dlfb_usb_probe error path
fbdev: i810: make fb_ops const
fbdev: matrox: make fb_ops const
video: fbdev: pxa3xx_gcu: fix error return code in pxa3xx_gcu_probe()
video: fbdev: Enable Xilinx FB for ZynqMP
video: fbdev: Fix multiple style issues in xilinxfb
video: fbdev: udlfb: constify usb_device_id.
...
Merge misc fixes from Andrew Morton:
"A few leftovers"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
mm, page_owner: skip unnecessary stack_trace entries
arm64: stacktrace: avoid listing stacktrace functions in stacktrace
mm: treewide: remove GFP_TEMPORARY allocation flag
IB/mlx4: fix sprintf format warning
fscache: fix fscache_objlist_show format processing
lib/test_bitmap.c: use ULL suffix for 64-bit constants
procfs: remove unused variable
drivers/media/cec/cec-adap.c: fix build with gcc-4.4.4
idr: remove WARN_ON_ONCE() when trying to replace negative ID
Now that we have added breaks in the wait queue scan and allow bookmark
on scan position, we put this logic in the wake_up_page_bit function.
We can have very long page wait list in large system where multiple
pages share the same wait list. We break the wake up walk here to allow
other cpus a chance to access the list, and not to disable the interrupts
when traversing the list for too long. This reduces the interrupt and
rescheduling latency, and excessive page wait queue lock hold time.
[ v2: Remove bookmark_wake_function ]
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
We encountered workloads that have very long wake up list on large
systems. A waker takes a long time to traverse the entire wake list and
execute all the wake functions.
We saw page wait list that are up to 3700+ entries long in tests of
large 4 and 8 socket systems. It took 0.8 sec to traverse such list
during wake up. Any other CPU that contends for the list spin lock will
spin for a long time. It is a result of the numa balancing migration of
hot pages that are shared by many threads.
Multiple CPUs waking are queued up behind the lock, and the last one
queued has to wait until all CPUs did all the wakeups.
The page wait list is traversed with interrupt disabled, which caused
various problems. This was the original cause that triggered the NMI
watch dog timer in: https://patchwork.kernel.org/patch/9800303/ . Only
extending the NMI watch dog timer there helped.
This patch bookmarks the waker's scan position in wake list and break
the wake up walk, to allow access to the list before the waker resume
its walk down the rest of the wait list. It lowers the interrupt and
rescheduling latency.
This patch also provides a performance boost when combined with the next
patch to break up page wakeup list walk. We saw 22% improvement in the
will-it-scale file pread2 test on a Xeon Phi system running 256 threads.
[ v2: Merged in Linus' changes to remove the bookmark_wake_function, and
simply access to flags. ]
Reported-by: Kan Liang <kan.liang@intel.com>
Tested-by: Kan Liang <kan.liang@intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Commit:
b94f51183b ("kernel/watchdog: prevent false hardlockup on overloaded system")
tries to fix the following issue:
proc_write()
set_sample_period() <--- New sample period becoms visible
<----- Broken starts
proc_watchdog_update()
watchdog_enable_all_cpus() watchdog_hrtimer_fn()
update_watchdog_all_cpus() restart_timer(sample_period)
watchdog_park_threads()
thread->park()
disable_nmi()
<----- Broken ends
The reason why this is broken is that the update of the watchdog threshold
becomes immediately effective and visible for the hrtimer function which
uses that value to rearm the timer. But the NMI/perf side still uses the
old value up to the point where it is disabled. If the rate has been
lowered then the NMI can run fast enough to 'detect' a hard lockup because
the timer has not fired due to the longer period.
The patch 'fixed' this by adding a variable:
proc_write()
set_sample_period()
<----- Broken starts
proc_watchdog_update()
watchdog_enable_all_cpus() watchdog_hrtimer_fn()
update_watchdog_all_cpus() restart_timer(sample_period)
watchdog_park_threads()
park_in_progress = 1
<----- Broken ends
nmi_watchdog()
if (park_in_progress)
return;
The only effect of this variable was to make the window where the breakage
can hit small enough that it was not longer observable in testing. From a
correctness point of view it is a pointless bandaid which merily papers
over the root cause: the unsychronized update of the variable.
Looking deeper into the related code pathes unearthed similar problems in
the watchdog_start()/stop() functions.
watchdog_start()
perf_nmi_event_start()
hrtimer_start()
watchdog_stop()
hrtimer_cancel()
perf_nmi_event_stop()
In both cases the call order is wrong because if the tasks gets preempted
or the VM gets scheduled out long enough after the first call, then there is
a chance that the next NMI will see a stale hrtimer interrupt count and
trigger a false positive hard lockup splat.
Get rid of park_in_progress so the code can be gradually deobfuscated and
pruned from several layers of duct tape papering over the root cause,
which has been either ignored or not understood at all.
Once this is removed the underlying problem will be fixed by rewriting the
proc interface to do a proper synchronized update.
Address the start/stop() ordering problem as well by reverting the call
order, so this part is at least correct now.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Don Zickus <dzickus@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Chris Metcalf <cmetcalf@mellanox.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Sebastian Siewior <bigeasy@linutronix.de>
Cc: Ulrich Obergfell <uobergfe@redhat.com>
Link: http://lkml.kernel.org/r/alpine.DEB.2.20.1709052038270.2393@nanos
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The following deadlock is possible in the watchdog hotplug code:
cpus_write_lock()
...
takedown_cpu()
smpboot_park_threads()
smpboot_park_thread()
kthread_park()
->park() := watchdog_disable()
watchdog_nmi_disable()
perf_event_release_kernel();
put_event()
_free_event()
->destroy() := hw_perf_event_destroy()
x86_release_hardware()
release_ds_buffers()
get_online_cpus()
when a per cpu watchdog perf event is destroyed which drops the last
reference to the PMU hardware. The cleanup code there invokes
get_online_cpus() which instantly deadlocks because the hotplug percpu
rwsem is write locked.
To solve this add a deferring mechanism:
cpus_write_lock()
kthread_park()
watchdog_nmi_disable(deferred)
perf_event_disable(event);
move_event_to_deferred(event);
....
cpus_write_unlock()
cleaup_deferred_events()
perf_event_release_kernel()
This is still properly serialized against concurrent hotplug via the
cpu_add_remove_lock, which is held by the task which initiated the hotplug
event.
This is also used to handle event destruction when the watchdog threads are
parked via other mechanisms than CPU hotplug.
Analyzed-by: Peter Zijlstra <peterz@infradead.org>
Reported-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Don Zickus <dzickus@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Chris Metcalf <cmetcalf@mellanox.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Sebastian Siewior <bigeasy@linutronix.de>
Cc: Ulrich Obergfell <uobergfe@redhat.com>
Link: http://lkml.kernel.org/r/20170912194146.884469246@linutronix.de
Signed-off-by: Ingo Molnar <mingo@kernel.org>
