In xfs_ifree, we reset the data/attr forks to extents format without
bothering to free any inline data buffer that might still be around
after all the blocks have been truncated off the file. Prior to commit 43518812d2 ("xfs: remove support for inlining data/extents into the
inode fork") nobody noticed because the leftover inline data after
truncation was small enough to fit inside the inline buffer inside the
fork itself.
However, now that we've removed the inline buffer, we /always/ have to
free the inline data buffer or else we leak them like crazy. This test
was found by turning on kmemleak for generic/001 or generic/388.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
which testcase tries to setup the processor specific debug
registers and configure vCPU for handling guest debug events through
KVM_SET_GUEST_DEBUG. The KVM_SET_GUEST_DEBUG ioctl will get and set
rflags in order to set TF bit if single step is needed. All regs' caches
are reset to avail and GUEST_RFLAGS vmcs field is reset to 0x2 during vCPU
reset. However, the cache of rflags is not reset during vCPU reset. The
function vmx_get_rflags() returns an unreset rflags cache value since
the cache is marked avail, it is 0 after boot. Vmentry fails if the
rflags reserved bit 1 is 0.
This patch fixes it by resetting both the GUEST_RFLAGS vmcs field and
its cache to 0x2 during vCPU reset.
Reported-by: Dmitry Vyukov <dvyukov@google.com> Tested-by: Dmitry Vyukov <dvyukov@google.com> Reviewed-by: David Hildenbrand <david@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Radim Krčmář <rkrcmar@redhat.com> Cc: Nadav Amit <nadav.amit@gmail.com> Cc: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This can be reproduced when running kvm-unit-tests/hyperv_stimer.flat and
cpu-hotplug stress simultaneously. __this_cpu_read(cpu_tsc_khz) returns 0
(set in kvmclock_cpu_down_prep()) when the pCPU is unhotplug which results
in kvm_get_time_scale() gets into an infinite loop.
This patch fixes it by treating the unhotplug pCPU as not using master clock.
Reviewed-by: Radim Krčmář <rkrcmar@redhat.com> Reviewed-by: David Hildenbrand <david@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Radim Krčmář <rkrcmar@redhat.com> Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If we fail to prepare our pages for whatever reason (out of memory in
our case) we need to make sure to drop the block_group->data_rwsem,
otherwise hilarity ensues.
Signed-off-by: Josef Bacik <jbacik@fb.com> Reviewed-by: Omar Sandoval <osandov@fb.com> Reviewed-by: Liu Bo <bo.li.liu@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com>
[ add label and use existing unlocking code ] Signed-off-by: David Sterba <dsterba@suse.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If we want to add a datapath flow, which has more than 500 vxlan outputs'
action, we will get the following error reports:
openvswitch: netlink: Flow action size 32832 bytes exceeds max
openvswitch: netlink: Flow action size 32832 bytes exceeds max
openvswitch: netlink: Actions may not be safe on all matching packets
... ...
It seems that we can simply enlarge the MAX_ACTIONS_BUFSIZE to fix it, but
this is not the root cause. For example, for a vxlan output action, we need
about 60 bytes for the nlattr, but after it is converted to the flow
action, it only occupies 24 bytes. This means that we can still support
more than 1000 vxlan output actions for a single datapath flow under the
the current 32k max limitation.
So even if the nla_len(attr) is larger than MAX_ACTIONS_BUFSIZE, we
shouldn't report EINVAL and keep it move on, as the judgement can be
done by the reserve_sfa_size.
Signed-off-by: zhangliping <zhangliping02@baidu.com> Acked-by: Pravin B Shelar <pshelar@ovn.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
register_shrinker is now __must_check, so check it to kill a warning.
Caller of bch_btree_cache_alloc in super.c appropriately checks return
value so this is fully plumbed through.
This V2 fixes checkpatch warnings and improves the commit description,
as I was too hasty getting the previous version out.
The MIPS loongson cpufreq drivers don't build unless configured for the
correct machine type, due to dependency on machine specific architecture
headers and symbols in machine specific platform code.
More specifically loongson1-cpufreq.c uses RST_CPU_EN and RST_CPU,
neither of which is defined in asm/mach-loongson32/regs-clk.h unless
CONFIG_LOONGSON1_LS1B=y, and loongson2_cpufreq.c references
loongson2_clockmod_table[], which is only defined in
arch/mips/loongson64/lemote-2f/clock.c, i.e. when
CONFIG_LEMOTE_MACH2F=y.
Add these dependencies to Kconfig to avoid randconfig / allyesconfig
build failures (e.g. when based on BMIPS which also has a cpufreq
driver).
Signed-off-by: James Hogan <jhogan@kernel.org> Acked-by: Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
According to 82093AA (IOAPIC) manual, Remote IRR and Delivery Status are
read-only. QEMU implements the bits as RO in commit 479c2a1cb7fb
("ioapic: keep RO bits for IOAPIC entry").
Most Bay and Cherry Trail devices use a generic DSDT with all possible
peripheral devices present in the DSDT, with their _STA returning 0x00 or
0x0f based on AML variables which describe what is actually present on
the board.
Since ACPI device objects with a 0x00 status (not present) still get an
entry under /sys/bus/acpi/devices, and those entry had an acpi:PNPID
modalias, userspace would end up loading modules for non present hardware.
This commit fixes this by leaving the modalias empty for non present
devices. This results in 10 modules less being loaded with a generic
distro kernel config on my Cherry Trail test-device (a GPD pocket).
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Some OSes (Linux, Xen) use this behavior to clear the Remote IRR bit for
IOAPICs without an EOI register. They simulate the EOI message manually
by changing the trigger mode to edge and then back to level, with the
entry being masked during this.
QEMU implements this feature in commit ed1263c363c9
("ioapic: clear remote irr bit for edge-triggered interrupts")
As a side effect, this commit removes an incorrect behavior where Remote
IRR was cleared when the redirection table entry was rewritten. This is not
consistent with the manual and also opens an opportunity for a strange
behavior when a redirection table entry is modified from an interrupt
handler that handles the same entry: The modification will clear the
Remote IRR bit even though the interrupt handler is still running.
KVM uses ioapic_handled_vectors to track vectors that need to notify the
IOAPIC on EOI. The problem is that IOAPIC can be reconfigured while an
interrupt with old configuration is pending or running and
ioapic_handled_vectors only remembers the newest configuration;
thus EOI from the old interrupt is not delievered to the IOAPIC.
A previous commit db2bdcbbbd32
("KVM: x86: fix edge EOI and IOAPIC reconfig race")
addressed this issue by adding pending edge-triggered interrupts to
ioapic_handled_vectors, fixing this race for edge-triggered interrupts.
The commit explicitly ignored level-triggered interrupts,
but this race applies to them as well:
1) IOAPIC sends a level triggered interrupt vector to VCPU0
2) VCPU0's handler deasserts the irq line and reconfigures the IOAPIC
to route the vector to VCPU1. The reconfiguration rewrites only the
upper 32 bits of the IOREDTBLn register. (Causes KVM to update
ioapic_handled_vectors for VCPU0 and it no longer includes the vector.)
3) VCPU0 sends EOI for the vector, but it's not delievered to the
IOAPIC because the ioapic_handled_vectors doesn't include the vector.
4) New interrupts are not delievered to VCPU1 because remote_irr bit
is set forever.
Therefore, the correct behavior is to add all pending and running
interrupts to ioapic_handled_vectors.
This commit introduces a slight performance hit similar to
commit db2bdcbbbd32 ("KVM: x86: fix edge EOI and IOAPIC reconfig race")
for the rare case that the vector is reused by a non-IOAPIC source on
VCPU0. We prefer to keep solution simple and not handle this case just
as the original commit does.
Pedro reported:
During tests that we conducted on KVM, we noticed that executing a "PUSH %ES"
instruction under KVM produces different results on both memory and the SP
register depending on whether EPT support is enabled. With EPT the SP is
reduced by 4 bytes (and the written value is 0-padded) but without EPT support
it is only reduced by 2 bytes. The difference can be observed when the CS.DB
field is 1 (32-bit) but not when it's 0 (16-bit).
The internal segment descriptor cache exist even in real/vm8096 mode. The CS.D
also should be respected instead of just default operand/address-size/66H
prefix/67H prefix during instruction decoding. This patch fixes it by also
adjusting operand/address-size according to CS.D.
Reported-by: Pedro Fonseca <pfonseca@cs.washington.edu> Tested-by: Pedro Fonseca <pfonseca@cs.washington.edu> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Radim Krčmář <rkrcmar@redhat.com> Cc: Nadav Amit <nadav.amit@gmail.com> Cc: Pedro Fonseca <pfonseca@cs.washington.edu> Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Radim Krčmář <rkrcmar@redhat.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In case of instruction-decode failure or emulation failure,
x86_emulate_instruction() will call reexecute_instruction() which will
attempt to use the cr2 value passed to x86_emulate_instruction().
However, when x86_emulate_instruction() is called from
emulate_instruction(), cr2 is not passed (passed as 0) and therefore
it doesn't make sense to execute reexecute_instruction() logic at all.
On this case, handle_emulation_failure() fills kvm_run with
internal-error information which it expects to be delivered
to user-mode for further processing.
However, the code reports a wrong return-value which makes KVM to never
return to user-mode on this scenario.
Fixes: 6d77dbfc88e3 ("KVM: inject #UD if instruction emulation fails and exit to
userspace")
cpuidle_monitor used to assume that cpu0 is always online which is not
a valid assumption on POWER machines. This patch fixes this by getting
the cpu on which the current thread is running, instead of always using
cpu0 for monitoring which may not be online.
Synchronously disable the IRQ to make the following cancel_work_sync
invocation effective.
An interrupt in flight could enqueue further overflow mem work. As we
free the binner BO immediately following vc4_irq_uninstall this caused
a NULL pointer dereference in the work callback vc4_overflow_mem_work.
As it turns out, normally the freeing of IRQs that would fix this is called
inside of the scope of __igb_close(). However, since the device is
already gone by the point we try to unregister the netdevice from the
driver due to a hotplug we end up seeing that the netif isn't present
and thus, forget to free any of the device IRQs.
So: make sure that if we're in the process of dismantling the netdev, we
always allow __igb_close() to be called so that IRQs may be freed
normally. Additionally, only allow igb_close() to be called from
__igb_close() if it hasn't already been called for the given adapter.
Signed-off-by: Lyude Paul <lyude@redhat.com> Fixes: 9474933caf21 ("igb: close/suspend race in netif_device_detach") Cc: Todd Fujinaka <todd.fujinaka@intel.com> Cc: Stephen Hemminger <stephen@networkplumber.org> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Background: ExpressKey Remotes communicate their events via usb dongle.
Each dongle can hold up to 5 pairings at one time and one EKR (identified
by its serial number) can unfortunately be paired with its dongle
more than once. The pairing takes place in a round-robin fashion.
Input devices are only created once per EKR, when a new serial number
is seen in the list of pairings. However, if a device is created for
a "higher" paring index and subsequently a second pairing occurs at a
lower pairing index, unpairing the remote with that serial number from
any pairing index will currently cause a driver crash. This occurs
infrequently, as two remotes are necessary to trigger this bug and most
users have only one remote.
As an illustration, to trigger the bug you need to have two remotes,
and pair them in this order:
1. slot 0 -> remote 1 (input device created for remote 1)
2. slot 1 -> remote 1 (duplicate pairing - no device created)
3. slot 2 -> remote 1 (duplicate pairing - no device created)
4. slot 3 -> remote 1 (duplicate pairing - no device created)
5. slot 4 -> remote 2 (input device created for remote 2)
6. slot 0 -> remote 2 (1 destroyed and recreated at slot 1)
7. slot 1 -> remote 2 (1 destroyed and recreated at slot 2)
8. slot 2 -> remote 2 (1 destroyed and recreated at slot 3)
9. slot 3 -> remote 2 (1 destroyed and not recreated)
10. slot 4 -> remote 2 (2 was already in this slot so no changes)
11. slot 0 -> remote 1 (The current code sees remote 2 was paired over in
one of the dongle slots it occupied and attempts
to remove all information about remote 2 [1]. It
calls wacom_remote_destroy_one for remote 2, but
the destroy function assumes the lowest index is
where the remote's input device was created. The
code "cleans up" the other remote 2 pairings
including the one which the input device was based
on, assuming they were were just duplicate
pairings. However, the cleanup doesn't call the
devres release function for the input device that
was created in slot 4).
This issue is fixed by this commit.
[1] Remote 2 should subsequently be re-created on the next packet from the
EKR at the lowest numbered slot that it occupies (here slot 1).
Fixes: f9036bd43602 ("HID: wacom: EKR: use devres groups to manage resources") Signed-off-by: Aaron Armstrong Skomra <aaron.skomra@wacom.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The user space interface allows specifying the type and mask field used
to allocate the cipher. Only a subset of the possible flags are intended
for user space. Therefore, white-list the allowed flags.
In case the user space caller uses at least one non-allowed flag, EINVAL
is returned.
Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Stephan Mueller <smueller@chronox.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Ensure that the input is byte swabbed before injecting it into the
SHA3 transform. Use the get_unaligned() accessor for this so that
we don't perform unaligned access inadvertently on architectures
that do not support that.
GCM can be invoked with a zero destination buffer. This is possible if
the AAD and the ciphertext have zero lengths and only the tag exists in
the source buffer (i.e. a source buffer cannot be zero). In this case,
the GCM cipher only performs the authentication and no decryption
operation.
When the destination buffer has zero length, it is possible that no page
is mapped to the SG pointing to the destination. In this case,
sg_page(req->dst) is an invalid access. Therefore, page accesses should
only be allowed if the req->dst->length is non-zero which is the
indicator that a page must exist.
This fixes a crash that can be triggered by user space via AF_ALG.
Signed-off-by: Stephan Mueller <smueller@chronox.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
HW can implement UMR wqe re-transmission in various ways.
Thus, add HCA cap to distinguish the needed fence for UMR to make
sure that the wqe wouldn't fail on mkey checks.
Signed-off-by: Max Gurtovoy <maxg@mellanox.com> Acked-by: Leon Romanovsky <leon@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Doug Ledford <dledford@redhat.com> Cc: Marta Rybczynska <mrybczyn@kalray.eu> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
范龙飞 reports that KASAN can report a use-after-free in __lock_acquire.
The reason is due to insufficient serialization in lo_release(), which
will continue to use the loop device even after it has decremented the
lo_refcnt to zero.
In the meantime, another process can come in, open the loop device
again as it is being shut down. Confusion ensues.
Commit bdcf0a423ea1 ("kernel: make groups_sort calling a responsibility
group_info allocators") appears to break nfsd rootsquash in a pretty
major way.
It adds a call to groups_sort() inside the loop that copies/squashes
gids, which means the valid gids are sorted along with the following
garbage. The net result is that the highest numbered valid gids are
replaced with any lower-valued garbage gids, possibly including 0.
We should sort only once, after filling in all the gids.
Fixes: bdcf0a423ea1 ("kernel: make groups_sort calling a responsibility ...") Signed-off-by: Ben Hutchings <ben.hutchings@codethink.co.uk> Acked-by: J. Bruce Fields <bfields@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Cc: Wolfgang Walter <linux@stwm.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Alexei found that verifier does not reject stores into context
via BPF_ST instead of BPF_STX. And while looking at it, we
also should not allow XADD variant of BPF_STX.
The context rewriter is only assuming either BPF_LDX_MEM- or
BPF_STX_MEM-type operations, thus reject anything other than
that so that assumptions in the rewriter properly hold. Add
test cases as well for BPF selftests.
Fixes: d691f9e8d440 ("bpf: allow programs to write to certain skb fields") Reported-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In addition to commit b2157399cc98 ("bpf: prevent out-of-bounds
speculation") also change the layout of struct bpf_map such that
false sharing of fast-path members like max_entries is avoided
when the maps reference counter is altered. Therefore enforce
them to be placed into separate cachelines.
/* size: 128, cachelines: 2, members: 17 */
/* sum members: 121, holes: 1, sum holes: 7 */
};
Now all entries in the first cacheline are read only throughout
the life time of the map, set up once during map creation. Overall
struct size and number of cachelines doesn't change from the
reordering. struct bpf_map is usually first member and embedded
in map structs in specific map implementations, so also avoid those
members to sit at the end where it could potentially share the
cacheline with first map values e.g. in the array since remote
CPUs could trigger map updates just as well for those (easily
dirtying members like max_entries intentionally as well) while
having subsequent values in cache.
Quoting from Google's Project Zero blog [1]:
Additionally, at least on the Intel machine on which this was
tested, bouncing modified cache lines between cores is slow,
apparently because the MESI protocol is used for cache coherence
[8]. Changing the reference counter of an eBPF array on one
physical CPU core causes the cache line containing the reference
counter to be bounced over to that CPU core, making reads of the
reference counter on all other CPU cores slow until the changed
reference counter has been written back to memory. Because the
length and the reference counter of an eBPF array are stored in
the same cache line, this also means that changing the reference
counter on one physical CPU core causes reads of the eBPF array's
length to be slow on other physical CPU cores (intentional false
sharing).
While this doesn't 'control' the out-of-bounds speculation through
masking the index as in commit b2157399cc98, triggering a manipulation
of the map's reference counter is really trivial, so lets not allow
to easily affect max_entries from it.
Splitting to separate cachelines also generally makes sense from
a performance perspective anyway in that fast-path won't have a
cache miss if the map gets pinned, reused in other progs, etc out
of control path, thus also avoids unintentional false sharing.
Although a number of JITs do support BPF_ALU | BPF_ARSH | BPF_{K,X}
generation, not all of them do and interpreter does neither. We can
leave existing ones and implement it later in bpf-next for the
remaining ones, but reject this properly in verifier for the time
being.
The BPF interpreter has been used as part of the spectre 2 attack CVE-2017-5715.
A quote from goolge project zero blog:
"At this point, it would normally be necessary to locate gadgets in
the host kernel code that can be used to actually leak data by reading
from an attacker-controlled location, shifting and masking the result
appropriately and then using the result of that as offset to an
attacker-controlled address for a load. But piecing gadgets together
and figuring out which ones work in a speculation context seems annoying.
So instead, we decided to use the eBPF interpreter, which is built into
the host kernel - while there is no legitimate way to invoke it from inside
a VM, the presence of the code in the host kernel's text section is sufficient
to make it usable for the attack, just like with ordinary ROP gadgets."
To make attacker job harder introduce BPF_JIT_ALWAYS_ON config
option that removes interpreter from the kernel in favor of JIT-only mode.
So far eBPF JIT is supported by:
x64, arm64, arm32, sparc64, s390, powerpc64, mips64
The start of JITed program is randomized and code page is marked as read-only.
In addition "constant blinding" can be turned on with net.core.bpf_jit_harden
v2->v3:
- move __bpf_prog_ret0 under ifdef (Daniel)
v1->v2:
- fix init order, test_bpf and cBPF (Daniel's feedback)
- fix offloaded bpf (Jakub's feedback)
- add 'return 0' dummy in case something can invoke prog->bpf_func
- retarget bpf tree. For bpf-next the patch would need one extra hunk.
It will be sent when the trees are merged back to net-next
Considered doing:
int bpf_jit_enable __read_mostly = BPF_EBPF_JIT_DEFAULT;
but it seems better to land the patch as-is and in bpf-next remove
bpf_jit_enable global variable from all JITs, consolidate in one place
and remove this jit_init() function.
- bpf prog_array just like all other types of bpf array accepts 32-bit index.
Clarify that in the comment.
- fix x64 JIT of bpf_tail_call which was incorrectly loading 8 instead of 4 bytes
- tighten corresponding check in the interpreter to stay consistent
The JIT bug can be triggered after introduction of BPF_F_NUMA_NODE flag
in commit 96eabe7a40aa in 4.14. Before that the map_flags would stay zero and
though JIT code is wrong it will check bounds correctly.
Hence two fixes tags. All other JITs don't have this problem.
Signed-off-by: Alexei Starovoitov <ast@kernel.org> Fixes: 96eabe7a40aa ("bpf: Allow selecting numa node during map creation") Fixes: b52f00e6a715 ("x86: bpf_jit: implement bpf_tail_call() helper") Acked-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Martin KaFai Lau <kafai@fb.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The hrtimer interrupt code contains a hang detection and mitigation
mechanism, which prevents that a long delayed hrtimer interrupt causes a
continous retriggering of interrupts which prevent the system from making
progress. If a hang is detected then the timer hardware is programmed with
a certain delay into the future and a flag is set in the hrtimer cpu base
which prevents newly enqueued timers from reprogramming the timer hardware
prior to the chosen delay. The subsequent hrtimer interrupt after the delay
clears the flag and resumes normal operation.
If such a hang happens in the last hrtimer interrupt before a CPU is
unplugged then the hang_detected flag is set and stays that way when the
CPU is plugged in again. At that point the timer hardware is not armed and
it cannot be armed because the hang_detected flag is still active, so
nothing clears that flag. As a consequence the CPU does not receive hrtimer
interrupts and no timers expire on that CPU which results in RCU stalls and
other malfunctions.
Clear the flag along with some other less critical members of the hrtimer
cpu base to ensure starting from a clean state when a CPU is plugged in.
Thanks to Paul, Sebastian and Anna-Maria for their help to get down to the
root cause of that hard to reproduce heisenbug. Once understood it's
trivial and certainly justifies a brown paperbag.
Fixes: 41d2e4949377 ("hrtimer: Tune hrtimer_interrupt hang logic") Reported-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Sebastian Sewior <bigeasy@linutronix.de> Cc: Anna-Maria Gleixner <anna-maria@linutronix.de> Link: https://lkml.kernel.org/r/alpine.DEB.2.20.1801261447590.2067@nanos Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit b94b73733171 ("x86/microcode/intel: Extend BDW late-loading with a
revision check") reduced the impact of erratum BDF90 for Broadwell model
79.
The impact can be reduced further by checking the size of the last level
cache portion per core.
Tony: "The erratum says the problem only occurs on the large-cache SKUs.
So we only need to avoid the update if we are on a big cache SKU that is
also running old microcode."
For more details, see erratum BDF90 in document #334165 (Intel Xeon
Processor E7-8800/4800 v4 Product Family Specification Update) from
September 2017.
Fixes: b94b73733171 ("x86/microcode/intel: Extend BDW late-loading with a revision check") Signed-off-by: Jia Zhang <zhang.jia@linux.alibaba.com> Signed-off-by: Borislav Petkov <bp@suse.de> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Acked-by: Tony Luck <tony.luck@intel.com> Link: https://lkml.kernel.org/r/1516321542-31161-1-git-send-email-zhang.jia@linux.alibaba.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
syzbot reported yet another crash [1] that is caused by
insufficient validation of DODGY packets.
Two bugs are happening here to trigger the crash.
1) Flow dissection leaves with incorrect thoff field.
2) skb_probe_transport_header() sets transport header to this invalid
thoff, even if pointing after skb valid data.
3) qdisc_pkt_len_init() reads out-of-bound data because it
trusts tcp_hdrlen(skb)
Possible fixes :
- Full flow dissector validation before injecting bad DODGY packets in
the stack.
This approach was attempted here : https://patchwork.ozlabs.org/patch/
861874/
- Have more robust functions in the core.
This might be needed anyway for stable versions.
Fixes: 34fad54c2537 ("net: __skb_flow_dissect() must cap its return value") Fixes: a6e544b0a88b ("flow_dissector: Jump to exit code in __skb_flow_dissect") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Willem de Bruijn <willemb@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Acked-by: Jason Wang <jasowang@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tfile->tun could be detached before we close the tun fd,
via tun_detach_all(), so it should not be used to check for
tfile->tx_array.
As Jason suggested, we probably have to clean it up
unconditionally both in __tun_deatch() and tun_detach_all(),
but this requires to check if it is initialized or not.
Currently skb_array_cleanup() doesn't have such a check,
so I check it in the caller and introduce a helper function,
it is a bit ugly but we can always improve it in net-next.
Reported-by: Dmitry Vyukov <dvyukov@google.com> Fixes: 1576d9860599 ("tun: switch to use skb array for tx") Cc: Jason Wang <jasowang@redhat.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Driver periodically samples all neighbors configured in device
in order to update the kernel regarding their state. When finding
an entry configured in HW that doesn't show in neigh_lookup()
driver logs an error message.
This introduces a race when removing multiple neighbors -
it's possible that a given entry would still be configured in HW
as its removal is still being processed but is already removed
from the kernel's neighbor tables.
Simply remove the error message and gracefully accept such events.
Fixes: c723c735fa6b ("mlxsw: spectrum_router: Periodically update the kernel's neigh table") Fixes: 60f040ca11b9 ("mlxsw: spectrum_router: Periodically dump active IPv6 neighbours") Signed-off-by: Yuval Mintz <yuvalm@mellanox.com> Reviewed-by: Ido Schimmel <idosch@mellanox.com> Signed-off-by: Jiri Pirko <jiri@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Validate gso_type during segmentation as SKB_GSO_DODGY sources
may pass packets where the gso_type does not match the contents.
Syzkaller was able to enter the SCTP gso handler with a packet of
gso_type SKB_GSO_TCPV4.
On entry of transport layer gso handlers, verify that the gso_type
matches the transport protocol.
Fixes: 90017accff61 ("sctp: Add GSO support") Link: http://lkml.kernel.org/r/<001a1137452496ffc305617e5fe0@google.com> Reported-by: syzbot+fee64147a25aecd48055@syzkaller.appspotmail.com Signed-off-by: Willem de Bruijn <willemb@google.com> Acked-by: Jason Wang <jasowang@redhat.com> Reviewed-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit b05229f44228 ("gre6: Cleanup GREv6 transmit path,
call common GRE functions") moved dev->mtu initialization
from ip6gre_tunnel_setup() to ip6gre_tunnel_init(), as a
result, the previously set values, before ndo_init(), are
reset in the following cases:
* rtnl_create_link() can update dev->mtu from IFLA_MTU
parameter.
* ip6gre_tnl_link_config() is invoked before ndo_init() in
netlink and ioctl setup, so ndo_init() can reset MTU
adjustments with the lower device MTU as well, dev->mtu
and dev->hard_header_len.
Not applicable for ip6gretap because it has one more call
to ip6gre_tnl_link_config(tunnel, 1) in ip6gre_tap_init().
Fix the first case by updating dev->mtu with 'tb[IFLA_MTU]'
parameter if a user sets it manually on a device creation,
and fix the second one by moving ip6gre_tnl_link_config()
call after register_netdevice().
Fixes: b05229f44228 ("gre6: Cleanup GREv6 transmit path, call common GRE functions") Fixes: db2ec95d1ba4 ("ip6_gre: Fix MTU setting") Signed-off-by: Alexey Kodanev <alexey.kodanev@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The commit 622190669403 ("be2net: Request RSS capability of Rx interface
depending on number of Rx rings") modified be_update_queues() so the
IFACE (HW representation of the netdevice) is destroyed and then
re-created. This causes a regression because potential promiscuous mode
is not restored properly during be_open() because the driver thinks
that the HW has promiscuous mode already enabled.
Note that Lancer is not affected by this bug because RX-filter flags are
disabled during be_close() for this chipset.
Cc: Sathya Perla <sathya.perla@broadcom.com> Cc: Ajit Khaparde <ajit.khaparde@broadcom.com> Cc: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com> Cc: Somnath Kotur <somnath.kotur@broadcom.com> Fixes: 622190669403 ("be2net: Request RSS capability of Rx interface depending on number of Rx rings") Signed-off-by: Ivan Vecera <ivecera@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
ppp_dev_uninit(), which is the .ndo_uninit() handler of PPP devices,
needs to lock pn->all_ppp_mutex. Therefore we mustn't call
register_netdevice() with pn->all_ppp_mutex already locked, or we'd
deadlock in case register_netdevice() fails and calls .ndo_uninit().
Fortunately, we can unlock pn->all_ppp_mutex before calling
register_netdevice(). This lock protects pn->units_idr, which isn't
used in the device registration process.
However, keeping pn->all_ppp_mutex locked during device registration
did ensure that no device in transient state would be published in
pn->units_idr. In practice, unlocking it before calling
register_netdevice() doesn't change this property: ppp_unit_register()
is called with 'ppp_mutex' locked and all searches done in
pn->units_idr hold this lock too.
Fixes: 8cb775bc0a34 ("ppp: fix device unregistration upon netns deletion") Reported-and-tested-by: syzbot+367889b9c9e279219175@syzkaller.appspotmail.com Signed-off-by: Guillaume Nault <g.nault@alphalink.fr> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Map all lookup neigh keys to INADDR_ANY for loopback/point-to-point devices
to avoid making an entry for every remote ip the device needs to talk to.
This used the be the old behavior but became broken in a263b3093641f
(ipv4: Make neigh lookups directly in output packet path) and later removed
in 0bb4087cbec0 (ipv4: Fix neigh lookup keying over loopback/point-to-point
devices) because it was broken.
Signed-off-by: Jim Westfall <jwestfall@surrealistic.net> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use n->primary_key instead of pkey to account for the possibility that a neigh
constructor function may have modified the primary_key value.
Signed-off-by: Jim Westfall <jwestfall@surrealistic.net> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
with the introduction of commit b0eb57cb97e7837ebb746404c2c58c6f536f23fa, it appears that rq->buf_info
is improperly handled. While it is heap allocated when an rx queue is
setup, and freed when torn down, an old line of code in
vmxnet3_rq_destroy was not properly removed, leading to rq->buf_info[0]
being set to NULL prior to its being freed, causing a memory leak, which
eventually exhausts the system on repeated create/destroy operations
(for example, when the mtu of a vmxnet3 interface is changed
frequently.
Fix is pretty straight forward, just move the NULL set to after the
free.
Tested by myself with successful results
Applies to net, and should likely be queued for stable, please
Signed-off-by: Neil Horman <nhorman@tuxdriver.com> Reported-By: boyang@redhat.com CC: boyang@redhat.com CC: Shrikrishna Khare <skhare@vmware.com> CC: "VMware, Inc." <pv-drivers@vmware.com> CC: David S. Miller <davem@davemloft.net> Acked-by: Shrikrishna Khare <skhare@vmware.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
After commit cea0cc80a677 ("sctp: use the right sk after waking up from
wait_buf sleep"), it may change to lock another sk if the asoc has been
peeled off in sctp_wait_for_sndbuf.
However, the asoc's new sk could be already closed elsewhere, as it's in
the sendmsg context of the old sk that can't avoid the new sk's closing.
If the sk's last one refcnt is held by this asoc, later on after putting
this asoc, the new sk will be freed, while under it's own lock.
This patch is to revert that commit, but fix the old issue by returning
error under the old sk's lock.
Fixes: cea0cc80a677 ("sctp: use the right sk after waking up from wait_buf sleep") Reported-by: syzbot+ac6ea7baa4432811eb50@syzkaller.appspotmail.com Signed-off-by: Xin Long <lucien.xin@gmail.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The check in sctp_sockaddr_af is not robust enough to forbid binding a
v4mapped v6 addr on a v4 socket.
The worse thing is that v4 socket's bind_verify would not convert this
v4mapped v6 addr to a v4 addr. syzbot even reported a crash as the v4
socket bound a v6 addr.
This patch is to fix it by doing the common sa.sa_family check first,
then AF_INET check for v4mapped v6 addrs.
Fixes: 7dab83de50c7 ("sctp: Support ipv6only AF_INET6 sockets.") Reported-by: syzbot+7b7b518b1228d2743963@syzkaller.appspotmail.com Acked-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: Xin Long <lucien.xin@gmail.com> Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Hardware statistics retrieval hurts in tight invocation loops.
Avoid extraneous write and enforce strict ordering of writes targeted to
the tally counters dump area address registers.
Signed-off-by: Francois Romieu <romieu@fr.zoreil.com> Tested-by: Oliver Freyermuth <o.freyermuth@googlemail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In pppoe_sendmsg(), reserving dev->hard_header_len bytes of headroom
was probably fine before the introduction of ->needed_headroom in
commit f5184d267c1a ("net: Allow netdevices to specify needed head/tailroom").
But now, virtual devices typically advertise the size of their overhead
in dev->needed_headroom, so we must also take it into account in
skb_reserve().
Allocation size of skb is also updated to take dev->needed_tailroom
into account and replace the arbitrary 32 bytes with the real size of
a PPPoE header.
This issue was discovered by syzbot, who connected a pppoe socket to a
gre device which had dev->header_ops->create == ipgre_header and
dev->hard_header_len == 0. Therefore, PPPoE didn't reserve any
headroom, and dev_hard_header() crashed when ipgre_header() tried to
prepend its header to skb->data.
Admittedly PPPoE shouldn't be allowed to run on non Ethernet-like
interfaces, but reserving space for ->needed_headroom is a more
fundamental issue that needs to be addressed first.
Same problem exists for __pppoe_xmit(), which also needs to take
dev->needed_headroom into account in skb_cow_head().
Fixes: f5184d267c1a ("net: Allow netdevices to specify needed head/tailroom") Reported-by: syzbot+ed0838d0fa4c4f2b528e20286e6dc63effc7c14d@syzkaller.appspotmail.com Signed-off-by: Guillaume Nault <g.nault@alphalink.fr> Reviewed-by: Xin Long <lucien.xin@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When a tcp socket is closed, if it detects that its net namespace is
exiting, close immediately and do not wait for FIN sequence.
For normal sockets, a reference is taken to their net namespace, so it will
never exit while the socket is open. However, kernel sockets do not take a
reference to their net namespace, so it may begin exiting while the kernel
socket is still open. In this case if the kernel socket is a tcp socket,
it will stay open trying to complete its close sequence. The sock's dst(s)
hold a reference to their interface, which are all transferred to the
namespace's loopback interface when the real interfaces are taken down.
When the namespace tries to take down its loopback interface, it hangs
waiting for all references to the loopback interface to release, which
results in messages like:
unregister_netdevice: waiting for lo to become free. Usage count = 1
These messages continue until the socket finally times out and closes.
Since the net namespace cleanup holds the net_mutex while calling its
registered pernet callbacks, any new net namespace initialization is
blocked until the current net namespace finishes exiting.
After this change, the tcp socket notices the exiting net namespace, and
closes immediately, releasing its dst(s) and their reference to the
loopback interface, which lets the net namespace continue exiting.
Link: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1711407
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=97811 Signed-off-by: Dan Streetman <ddstreet@canonical.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Without proper validation of DODGY packets, we might very well
feed qdisc_pkt_len_init() with invalid GSO packets.
tcp_hdrlen() might access out-of-bound data, so let's use
skb_header_pointer() and proper checks.
Whole story is described in commit d0c081b49137 ("flow_dissector:
properly cap thoff field")
We have the goal of validating DODGY packets earlier in the stack,
so we might very well revert this fix in the future.
Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Willem de Bruijn <willemb@google.com> Cc: Jason Wang <jasowang@redhat.com> Reported-by: syzbot+9da69ebac7dddd804552@syzkaller.appspotmail.com Acked-by: Jason Wang <jasowang@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit "net: igmp: Use correct source address on IGMPv3 reports"
introduced a check to validate the source address of locally generated
IGMPv3 packets.
Instead of checking the local interface address directly, it uses
inet_ifa_match(fl4->saddr, ifa), which checks if the address is on the
local subnet (or equal to the point-to-point address if used).
This breaks for point-to-point interfaces, so check against
ifa->ifa_local directly.
Cc: Kevin Cernekee <cernekee@chromium.org> Fixes: a46182b00290 ("net: igmp: Use correct source address on IGMPv3 reports") Reported-by: Sebastian Gottschall <s.gottschall@dd-wrt.com> Signed-off-by: Felix Fietkau <nbd@nbd.name> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fix initialize the uninitialized tx_qlen to an appropriate value when USB
Full Speed is used.
Fixes: 55d7de9de6c3 ("Microchip's LAN7800 family USB 2/3 to 10/100/1000 Ethernet device driver") Signed-off-by: Yuiko Oshino <yuiko.oshino@microchip.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In my last patch, I missed fact that cork.base.dst was not initialized
in ip6_make_skb() :
If ip6_setup_cork() returns an error, we might attempt a dst_release()
on some random pointer.
Fixes: 862c03ee1deb ("ipv6: fix possible mem leaks in ipv6_make_skb()") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The logic in __ip6_append_data() assumes that the MTU is at least large
enough for the headers. A device's MTU may be adjusted after being
added while sendmsg() is processing data, resulting in
__ip6_append_data() seeing any MTU. For an mtu smaller than the size of
the fragmentation header, the math results in a negative 'maxfraglen',
which causes problems when refragmenting any previous skb in the
skb_write_queue, leaving it possibly malformed.
Instead sendmsg returns EINVAL when the mtu is calculated to be less
than IPV6_MIN_MTU.
Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Mike Maloney <maloney@google.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 513674b5a2c9 ("net: reevalulate autoflowlabel setting after
sysctl setting") removed the initialisation of
ipv6_pinfo::autoflowlabel and added a second flag to indicate
whether this field or the net namespace default should be used.
The getsockopt() handling for this case was not updated, so it
currently returns 0 for all sockets for which IPV6_AUTOFLOWLABEL is
not explicitly enabled. Fix it to return the effective value, whether
that has been set at the socket or net namespace level.
Fixes: 513674b5a2c9 ("net: reevalulate autoflowlabel setting after sysctl ...") Signed-off-by: Ben Hutchings <ben.hutchings@codethink.co.uk> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
ccid2_hc_tx_rto_expire() timer callback always restarts the timer
again and can run indefinitely (unless it is stopped outside), and after
commit 120e9dabaf55 ("dccp: defer ccid_hc_tx_delete() at dismantle time"),
which moved ccid_hc_tx_delete() (also includes sk_stop_timer()) from
dccp_destroy_sock() to sk_destruct(), this started to happen quite often.
The timer prevents releasing the socket, as a result, sk_destruct() won't
be called.
Found with LTP/dccp_ipsec tests running on the bonding device,
which later couldn't be unloaded after the tests were completed:
unregister_netdevice: waiting for bond0 to become free. Usage count = 148
Fixes: 2a91aa396739 ("[DCCP] CCID2: Initial CCID2 (TCP-Like) implementation") Signed-off-by: Alexey Kodanev <alexey.kodanev@oracle.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For some reason these values are not in the uapi header file, so any
libc has to define it themselves. To prevent them from needing to do
this, just have the kernel provide the correct values.
Ben Hutchings [Fri, 26 Jan 2018 16:23:02 +0000 (16:23 +0000)]
vsyscall: Fix permissions for emulate mode with KAISER/PTI
The backport of KAISER to 4.4 turned vsyscall emulate mode into native
mode. Add a vsyscall_pgprot variable to hold the correct page
protections, like Borislav and Hugh did for 3.2 and 3.18.
Debian's gcc defaults to pie. The global Makefile already defines the -fno-pie option.
Link UML dynamic kernel image also with -no-pie to fix the build.
Signed-off-by: Thomas Meyer <thomas@m3y3r.de> Signed-off-by: Richard Weinberger <richard@nod.at> Cc: Bernie Innocenti <codewiz@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
After do_readv_writev, the inode cache is invalidated anyway, so i_size
will never be read. It will be fetched from the server which will also
know about updates from other machines.
Fixes deadlock on 32-bit SMP.
See https://marc.info/?l=linux-fsdevel&m=151268557427760&w=2
Signed-off-by: Martin Brandenburg <martin@omnibond.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: Mike Marshall <hubcap@omnibond.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Lenovo introduced trackpoint compatible sticks with minimum PS/2 commands.
They supposed to reply with 0x02, 0x03, or 0x04 in response to the
"Read Extended ID" command, so we would know not to try certain extended
commands. Unfortunately even some trackpoints reporting the original IBM
version (0x01 firmware 0x0e) now respond with incorrect data to the "Get
Extended Buttons" command:
Patch series "mm: kswapd spinning on unreclaimable nodes - fixes and
cleanups".
Jia reported a scenario in which the kswapd of a node indefinitely spins
at 100% CPU usage. We have seen similar cases at Facebook.
The kernel's current method of judging its ability to reclaim a node (or
whether to back off and sleep) is based on the amount of scanned pages
in proportion to the amount of reclaimable pages. In Jia's and our
scenarios, there are no reclaimable pages in the node, however, and the
condition for backing off is never met. Kswapd busyloops in an attempt
to restore the watermarks while having nothing to work with.
This series reworks the definition of an unreclaimable node based not on
scanning but on whether kswapd is able to actually reclaim pages in
MAX_RECLAIM_RETRIES (16) consecutive runs. This is the same criteria
the page allocator uses for giving up on direct reclaim and invoking the
OOM killer. If it cannot free any pages, kswapd will go to sleep and
leave further attempts to direct reclaim invocations, which will either
make progress and re-enable kswapd, or invoke the OOM killer.
Patch #1 fixes the immediate problem Jia reported, the remainder are
smaller fixlets, cleanups, and overall phasing out of the old method.
Patch #6 is the odd one out. It's a nice cleanup to get_scan_count(),
and directly related to #5, but in itself not relevant to the series.
If the whole series is too ambitious for 4.11, I would consider the
first three patches fixes, the rest cleanups.
This patch (of 9):
Jia He reports a problem with kswapd spinning at 100% CPU when
requesting more hugepages than memory available in the system:
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
76 root 20 0 0 0 0 R 100.0 0.000 217:17.29 kswapd3
At that time, there are no reclaimable pages left in the node, but as
kswapd fails to restore the high watermarks it refuses to go to sleep.
Kswapd needs to back away from nodes that fail to balance. Up until
commit 1d82de618ddd ("mm, vmscan: make kswapd reclaim in terms of
nodes") kswapd had such a mechanism. It considered zones whose
theoretically reclaimable pages it had reclaimed six times over as
unreclaimable and backed away from them. This guard was erroneously
removed as the patch changed the definition of a balanced node.
However, simply restoring this code wouldn't help in the case reported
here: there *are* no reclaimable pages that could be scanned until the
threshold is met. Kswapd would stay awake anyway.
Introduce a new and much simpler way of backing off. If kswapd runs
through MAX_RECLAIM_RETRIES (16) cycles without reclaiming a single
page, make it back off from the node. This is the same number of shots
direct reclaim takes before declaring OOM. Kswapd will go to sleep on
that node until a direct reclaimer manages to reclaim some pages, thus
proving the node reclaimable again.
[hannes@cmpxchg.org: check kswapd failure against the cumulative nr_reclaimed count] Link: http://lkml.kernel.org/r/20170306162410.GB2090@cmpxchg.org
[shakeelb@google.com: fix condition for throttle_direct_reclaim] Link: http://lkml.kernel.org/r/20170314183228.20152-1-shakeelb@google.com Link: http://lkml.kernel.org/r/20170228214007.5621-2-hannes@cmpxchg.org Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Signed-off-by: Shakeel Butt <shakeelb@google.com> Reported-by: Jia He <hejianet@gmail.com> Tested-by: Jia He <hejianet@gmail.com> Acked-by: Michal Hocko <mhocko@suse.com> Acked-by: Hillf Danton <hillf.zj@alibaba-inc.com> Acked-by: Minchan Kim <minchan@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Cc: Dmitry Shmidt <dimitrysh@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Turns out distros do not want to make retpoline as part of their "ABI",
so this patch should not have been merged. Sorry Andi, this was my
fault, I suggested it when your original patch was the "correct" way of
doing this instead.
Reported-by: Jiri Kosina <jikos@kernel.org> Fixes: 6cfb521ac0d5 ("module: Add retpoline tag to VERMAGIC") Acked-by: Andi Kleen <ak@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: David Woodhouse <dwmw@amazon.co.uk> Cc: rusty@rustcorp.com.au Cc: arjan.van.de.ven@intel.com Cc: jeyu@kernel.org Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The SCSI host byte should be shifted left by 16 in order to have
scsi_decide_disposition() do the right thing (.i.e. requeue the
command).
Signed-off-by: Johannes Thumshirn <jthumshirn@suse.de> Fixes: 661134ad3765 ("[SCSI] libiscsi, bnx2i: make bound ep check common") Cc: Lee Duncan <lduncan@suse.com> Cc: Hannes Reinecke <hare@suse.de> Cc: Bart Van Assche <Bart.VanAssche@sandisk.com> Cc: Chris Leech <cleech@redhat.com> Acked-by: Lee Duncan <lduncan@suse.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
fcntl(0, F_SETOWN, 0x80000000) triggers:
UBSAN: Undefined behaviour in fs/fcntl.c:118:7
negation of -2147483648 cannot be represented in type 'int':
CPU: 1 PID: 18261 Comm: syz-executor Not tainted 4.8.1-0-syzkaller #1
...
Call Trace:
...
[<ffffffffad8f0868>] ? f_setown+0x1d8/0x200
[<ffffffffad8f19a9>] ? SyS_fcntl+0x999/0xf30
[<ffffffffaed1fb00>] ? entry_SYSCALL_64_fastpath+0x23/0xc1
Fix that by checking the arg parameter properly (against INT_MAX) before
"who = -who". And return immediatelly with -EINVAL in case it is wrong.
Note that according to POSIX we can return EINVAL:
http://pubs.opengroup.org/onlinepubs/9699919799/functions/fcntl.html
[EINVAL]
The cmd argument is F_SETOWN and the value of the argument
is not valid as a process or process group identifier.
[v2] returns an error, v1 used to fail silently
[v3] implement proper check for the bad value INT_MIN
Signed-off-by: Jiri Slaby <jslaby@suse.cz> Cc: Jeff Layton <jlayton@poochiereds.net> Cc: "J. Bruce Fields" <bfields@fieldses.org> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: linux-fsdevel@vger.kernel.org Signed-off-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Most extended attributes will fit in a single block. More importantly,
we drop the reference to the inode while holding the transaction open
so the preallocated blocks aren't released. As a result, the inode
may be evicted before it's removed from the transaction's prealloc list
which can cause memory corruption.
Signed-off-by: Jeff Mahoney <jeffm@suse.com> Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The main loop in __discard_prealloc is protected by the reiserfs write lock
which is dropped across schedules like the BKL it replaced. The problem is
that it checks the value, calls a routine that schedules, and then adjusts
the state. As a result, two threads that are calling
reiserfs_prealloc_discard at the same time can race when one calls
reiserfs_free_prealloc_block, the lock is dropped, and the other calls
reiserfs_free_prealloc_block with the same block number. In the right
circumstances, it can cause the prealloc count to go negative.
Signed-off-by: Jeff Mahoney <jeffm@suse.com> Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The capability check in nfnetlink_rcv() verifies that the caller
has CAP_NET_ADMIN in the namespace that "owns" the netlink socket.
However, xt_osf_fingers is shared by all net namespaces on the
system. An unprivileged user can create user and net namespaces
in which he holds CAP_NET_ADMIN to bypass the netlink_net_capable()
check:
vpnns -- nfnl_osf -f /tmp/pf.os
vpnns -- nfnl_osf -f /tmp/pf.os -d
These non-root operations successfully modify the systemwide OS
fingerprint list. Add new capable() checks so that they can't.
Signed-off-by: Kevin Cernekee <cernekee@chromium.org> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Acked-by: Michal Kubecek <mkubecek@suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The capability check in nfnetlink_rcv() verifies that the caller
has CAP_NET_ADMIN in the namespace that "owns" the netlink socket.
However, nfnl_cthelper_list is shared by all net namespaces on the
system. An unprivileged user can create user and net namespaces
in which he holds CAP_NET_ADMIN to bypass the netlink_net_capable()
check:
When early abort is occurred due to invalid ACPI information, Linux kernel
terminates ACPI by calling acpi_terminate() function. The function calls
acpi_ns_terminate() function to delete namespace data and ACPI operand cache
(acpi_gbl_module_code_list).
But the deletion code in acpi_ns_terminate() function is wrapped in
ACPI_EXEC_APP definition, therefore the code is only executed when the
definition exists. If the define doesn't exist, ACPI operand cache
(acpi_gbl_module_code_list) is leaked, and stack dump is shown in kernel log.
This causes a security threat because the old kernel (<= 4.9) shows memory
locations of kernel functions in stack dump, therefore kernel ASLR can be
neutralized.
To fix ACPI operand leak for enhancing security, I made a patch which
removes the ACPI_EXEC_APP define in acpi_ns_terminate() function for
executing the deletion code unconditionally.
Link: https://github.com/acpica/acpica/commit/a23325b2 Signed-off-by: Seunghun Han <kkamagui@gmail.com> Signed-off-by: Lv Zheng <lv.zheng@intel.com> Signed-off-by: Bob Moore <robert.moore@intel.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Acked-by: Lee, Chun-Yi <jlee@suse.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The way acpi_find_child_device() works currently is that, if there
are two (or more) devices with the same _ADR value in the same
namespace scope (which is not specifically allowed by the spec and
the OS behavior in that case is not defined), the first one of them
found to be present (with the help of _STA) will be returned.
This covers the majority of cases, but is not sufficient if some of
the devices in question have a _HID (or _CID) returning some valid
ACPI/PNP device IDs (which is disallowed by the spec) and the
ASL writers' expectation appears to be that the OS will match
devices without a valid ACPI/PNP device ID against a given bus
address first.
To cover this special case as well, modify find_child_checks()
to prefer devices without ACPI/PNP device IDs over devices that
have them.
Suggested-by: Mika Westerberg <mika.westerberg@linux.intel.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Tested-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Jiri Slaby <jslaby@suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When LONG_MIN is passed to msgrcv, one would expect to recieve any
message. But convert_mode does *msgtyp = -*msgtyp and -LONG_MIN is
undefined. In particular, with my gcc -LONG_MIN produces -LONG_MIN
again.
So handle this case properly by assigning LONG_MAX to *msgtyp if
LONG_MIN was specified as msgtyp to msgrcv.
This code:
long msg[] = { 100, 200 };
int m = msgget(IPC_PRIVATE, IPC_CREAT | 0644);
msgsnd(m, &msg, sizeof(msg), 0);
msgrcv(m, &msg, sizeof(msg), LONG_MIN, 0);
Since commit 97a16fc82a7c ("mm, page_alloc: only enforce watermarks for
order-0 allocations"), __zone_watermark_ok() check for high-order
allocations will shortcut per-migratetype free list checks for
ALLOC_HARDER allocations, and return true as long as there's free page
of any migratetype. The intention is that ALLOC_HARDER can allocate
from MIGRATE_HIGHATOMIC free lists, while normal allocations can't.
However, as a side effect, the watermark check will then also return
true when there are pages only on the MIGRATE_ISOLATE list, or (prior to
CMA conversion to ZONE_MOVABLE) on the MIGRATE_CMA list. Since the
allocation cannot actually obtain isolated pages, and might not be able
to obtain CMA pages, this can result in a false positive.
The condition should be rare and perhaps the outcome is not a fatal one.
Still, it's better if the watermark check is correct. There also
shouldn't be a performance tradeoff here.
Link: http://lkml.kernel.org/r/20171102125001.23708-1-vbabka@suse.cz Fixes: 97a16fc82a7c ("mm, page_alloc: only enforce watermarks for order-0 allocations") Signed-off-by: Vlastimil Babka <vbabka@suse.cz> Acked-by: Mel Gorman <mgorman@techsingularity.net> Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com> Cc: Rik van Riel <riel@redhat.com> Cc: David Rientjes <rientjes@google.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The align_offset parameter is used by bitmap_find_next_zero_area_off()
to represent the offset of map's base from the previous alignment
boundary; the function ensures that the returned index, plus the
align_offset, honors the specified align_mask.
The logic introduced by commit b5be83e308f7 ("mm: cma: align to physical
address, not CMA region position") has the cma driver calculate the
offset to the *next* alignment boundary. In most cases, the base
alignment is greater than that specified when making allocations,
resulting in a zero offset whether we align up or down. In the example
given with the commit, the base alignment (8MB) was half the requested
alignment (16MB) so the math also happened to work since the offset is
8MB in both directions. However, when requesting allocations with an
alignment greater than twice that of the base, the returned index would
not be correctly aligned.
Also, the align_order arguments of cma_bitmap_aligned_mask() and
cma_bitmap_aligned_offset() should not be negative so the argument type
was made unsigned.
Fixes: b5be83e308f7 ("mm: cma: align to physical address, not CMA region position") Link: http://lkml.kernel.org/r/20170628170742.2895-1-opendmb@gmail.com Signed-off-by: Angus Clark <angus@angusclark.org> Signed-off-by: Doug Berger <opendmb@gmail.com> Acked-by: Gregory Fong <gregory.0xf0@gmail.com> Cc: Doug Berger <opendmb@gmail.com> Cc: Angus Clark <angus@angusclark.org> Cc: Laura Abbott <labbott@redhat.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Lucas Stach <l.stach@pengutronix.de> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Shiraz Hashim <shashim@codeaurora.org> Cc: Jaewon Kim <jaewon31.kim@samsung.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Vlastimil Babka <vbabka@suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Laurent Dufour has noticed that hwpoinsoned pages are kept charged. In
his particular case he has hit a bad_page("page still charged to
cgroup") when onlining a hwpoison page. While this looks like something
that shouldn't happen in the first place because onlining hwpages and
returning them to the page allocator makes only little sense it shows a
real problem.
hwpoison pages do not get freed usually so we do not uncharge them (at
least not since commit 0a31bc97c80c ("mm: memcontrol: rewrite uncharge
API")). Each charge pins memcg (since e8ea14cc6ead ("mm: memcontrol:
take a css reference for each charged page")) as well and so the
mem_cgroup and the associated state will never go away. Fix this leak
by forcibly uncharging a LRU hwpoisoned page in delete_from_lru_cache().
We also have to tweak uncharge_list because it cannot rely on zero ref
count for these pages.
Commit 1be7107fbe18 ("mm: larger stack guard gap, between vmas") has
introduced a regression in some rust and Java environments which are
trying to implement their own stack guard page. They are punching a new
MAP_FIXED mapping inside the existing stack Vma.
This will confuse expand_{downwards,upwards} into thinking that the
stack expansion would in fact get us too close to an existing non-stack
vma which is a correct behavior wrt safety. It is a real regression on
the other hand.
Let's work around the problem by considering PROT_NONE mapping as a part
of the stack. This is a gros hack but overflowing to such a mapping
would trap anyway an we only can hope that usespace knows what it is
doing and handle it propely.
ARM64 enables both CONFIG_OF and CONFIG_ACPI and the firmware can pass
both ACPI tables and the device tree. Based on the kernel parameter, one
of the two will be chosen. If acpi is enabled, then device tree is not
unflattened.
Currently ARM64 platforms report:
"
Failed to find cpu0 device node
Unable to detect cache hierarchy from DT for CPU 0
"
which is incorrect when booting with ACPI. Also latest ACPI v6.1 has no
support for cache properties/hierarchy.
This patch adds check for unflattened device tree and also returns as
"not supported" if ACPI is runtime enabled.
It also removes the reference to DT from the error message as the cache
hierarchy can be detected from the firmware(OF/DT/ACPI)
With CONFIG_OF enabled on x86, we get the following error on boot:
"
Failed to find cpu0 device node
Unable to detect cache hierarchy from DT for CPU 0
"
and the cacheinfo fails to get populated in the corresponding sysfs
entries. This is because cache_setup_of_node looks for of_node for
setting up the shared cpu_map without checking that it's already
populated in the architecture specific callback.
In order to indicate that the shared cpu_map is already populated, this
patch introduces a boolean `cpu_map_populated` in struct cpu_cacheinfo
that can be used by the generic code to skip cache_shared_cpu_map_setup.
Newer hardware has uncovered a bug in the software implementation of
using MWAITX for the delay function. A value of 0 for the timer is meant
to indicate that a timeout will not be used to exit MWAITX. On newer
hardware this can result in MWAITX never returning, resulting in NMI
soft lockup messages being printed. On older hardware, some of the other
conditions under which MWAITX can exit masked this issue. The AMD APM
does not currently document this and will be updated.
Please refer to http://marc.info/?l=kvm&m=148950623231140 for
information regarding NMI soft lockup messages on an AMD Ryzen 1800X.
This has been root-caused as a 0 passed to MWAITX causing it to wait
indefinitely.
This change has the added benefit of avoiding the unnecessary setup of
MONITORX/MWAITX when the delay value is zero.
KVM only supports PMD hugepages at stage 2 but doesn't actually check
that the provided hugepage memory pagesize is PMD_SIZE before populating
stage 2 entries.
In cases where the backing hugepage size is smaller than PMD_SIZE (such
as when using contiguous hugepages), KVM can end up creating stage 2
mappings that extend beyond the supplied memory.
Fix this by checking for the pagesize of userspace vma before creating
PMD hugepage at stage 2.
Fixes: 66b3923a1a0f77a ("arm64: hugetlb: add support for PTE contiguous bit") Signed-off-by: Punit Agrawal <punit.agrawal@arm.com> Cc: Marc Zyngier <marc.zyngier@arm.com> Reviewed-by: Christoffer Dall <christoffer.dall@linaro.org> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If an invalid CANFD frame is received, from a driver or from a tun
interface, a Kernel warning is generated.
This patch replaces the WARN_ONCE by a simple pr_warn_once, so that a
kernel, bootet with panic_on_warn, does not panic. A printk seems to be
more appropriate here.
Reported-by: syzbot+e3b775f40babeff6e68b@syzkaller.appspotmail.com Suggested-by: Dmitry Vyukov <dvyukov@google.com> Acked-by: Oliver Hartkopp <socketcan@hartkopp.net> Cc: linux-stable <stable@vger.kernel.org> Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de> Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If an invalid CAN frame is received, from a driver or from a tun
interface, a Kernel warning is generated.
This patch replaces the WARN_ONCE by a simple pr_warn_once, so that a
kernel, bootet with panic_on_warn, does not panic. A printk seems to be
more appropriate here.
Reported-by: syzbot+4386709c0c1284dca827@syzkaller.appspotmail.com Suggested-by: Dmitry Vyukov <dvyukov@google.com> Acked-by: Oliver Hartkopp <socketcan@hartkopp.net> Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de> Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
projects / thirdparty / kernel / stable.git / log
