With inotify, when a watch is set on a directory and on its child, an
event on the child is reported twice, once with wd of the parent watch
and once with wd of the child watch without the filename.
With fanotify, when a watch is set on a directory and on its child, an
event on the child is reported twice, but it has the exact same
information - either an open file descriptor of the child or an encoded
fid of the child.
The reason that the two identical events are not merged is because the
object id used for merging events in the queue is the child inode in one
event and parent inode in the other.
For events with path or dentry data, use the victim inode instead of the
watched inode as the object id for event merging, so that the event
reported on parent will be merged with the event reported on the child.
The event inode field is used only for comparison in queue merges and
cannot be dereferenced after handle_event(), because it does not hold a
refcount on the inode.
Replace it with an abstract id to do the same thing.
If an operation's flag `needs_file` is set, the function
io_req_set_file() calls io_file_get() to obtain a `struct file*`.
This fails for `O_PATH` file descriptors, because io_file_get() calls
fget(), which rejects `O_PATH` file descriptors. To support `O_PATH`,
fdget_raw() must be used (like path_init() in `fs/namei.c` does).
This rejection causes io_req_set_file() to throw `-EBADF`. This
breaks the operations `openat`, `openat2` and `statx`, where `O_PATH`
file descriptors are commonly used.
This could be solved by adding support for `O_PATH` file descriptors
with another `io_op_def` flag, but since those three operations don't
need the `struct file*` but operate directly on the numeric file
descriptors, the best solution here is to simply remove `needs_file`
(and the accompanying flag `fd_non_reg`).
Cache a copy of the name for the life time of the backing_dev_info
structure so that we can reference it even after unregistering.
Fixes: 68f23b89067f ("memcg: fix a crash in wb_workfn when a device disappears") Reported-by: Yufen Yu <yuyufen@huawei.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Jan Kara <jack@suse.cz> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Sasha Levin <sashal@kernel.org>
bdi_dev_name is not a fast path function, move it out of line. This
prepares for using it from modular callers without having to export
an implementation detail like bdi_unknown_name.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Jan Kara <jack@suse.cz> Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Sasha Levin <sashal@kernel.org>
When I run my memcg testcase which creates lots of memcgs, I found
there're unexpected out of memory logs while there're still enough
available free memory. The error log is
The reason is when we try to create more than MEM_CGROUP_ID_MAX memcgs,
an -ENOMEM errno will be set by mem_cgroup_css_alloc(), but the right
errno should be -ENOSPC "No space left on device", which is an
appropriate errno for userspace's failed mkdir.
As the errno really misled me, we should make it right. After this
patch, the error log will be
mkdir: cannot create directory 'foo.65533': No space left on device
[akpm@linux-foundation.org: s/EBUSY/ENOSPC/, per Michal]
[akpm@linux-foundation.org: s/EBUSY/ENOSPC/, per Michal] Fixes: 73f576c04b94 ("mm: memcontrol: fix cgroup creation failure after many small jobs") Suggested-by: Matthew Wilcox <willy@infradead.org> Signed-off-by: Yafang Shao <laoar.shao@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Acked-by: Michal Hocko <mhocko@kernel.org> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: Vladimir Davydov <vdavydov.dev@gmail.com> Link: http://lkml.kernel.org/r/20200407063621.GA18914@dhcp22.suse.cz Link: http://lkml.kernel.org/r/1586192163-20099-1-git-send-email-laoar.shao@gmail.com Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If the trapping instruction contains a ':', for a memory access through
segment registers for example, the sed substitution will insert the '*'
marker in the middle of the instruction instead of the line address:
I started to think I had forgotten some quirk of the assembly syntax
before noticing that it was actually coming from the script. Fix it to
add the address marker at the right place for these instructions:
Elsewhere in the file, there is a list_for_each_entry with
&vdev->resv_regions as the second argument, suggesting that
&vdev->resv_regions is the list head. So exchange the
arguments on the list_add call to put the list head in the
second argument.
When the current frame address (CFA) is stored on the stack (i.e.,
cfa->base == CFI_SP_INDIRECT), objtool neglects to adjust the stack
offset when there are subsequent pushes or pops. This results in bad
ORC data at the end of the ENTER_IRQ_STACK macro, when it puts the
previous stack pointer on the stack and does a subsequent push.
This fixes the following unwinder warning:
WARNING: can't dereference registers at 00000000f0a6bdba for ip interrupt_entry+0x9f/0xa0
Fixes: 627fce14809b ("objtool: Add ORC unwind table generation") Reported-by: Vince Weaver <vincent.weaver@maine.edu> Reported-by: Dave Jones <dsj@fb.com> Reported-by: Steven Rostedt <rostedt@goodmis.org> Reported-by: Vegard Nossum <vegard.nossum@oracle.com> Reported-by: Joe Mario <jmario@redhat.com> Reviewed-by: Miroslav Benes <mbenes@suse.cz> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Ingo Molnar <mingo@kernel.org> Cc: Andy Lutomirski <luto@kernel.org> Cc: Jann Horn <jannh@google.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: https://lore.kernel.org/r/853d5d691b29e250333332f09b8e27410b2d9924.1587808742.git.jpoimboe@redhat.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The re-entrancy happens because the irq state is the OR of
the interrupt state and the resamplefd state. That is, we don't
want to show the state as 0 until we've had a chance to set the
resamplefd. But if the interrupt has _not_ gone low then
ioapic_set_irq is invoked again, causing an infinite loop.
This can only happen for a level-triggered interrupt, otherwise
irqfd_inject would immediately set the KVM_USERSPACE_IRQ_SOURCE_ID high
and then low. Fortunately, in the case of level-triggered interrupts the VMEXIT already happens because
TMR is set. Thus, fix the bug by restricting the lazy invocation
of the ack notifier to edge-triggered interrupts, the only ones that
need it.
gcc-10 points out that a code path exists where a pointer to a stack
variable may be passed back to the caller:
net/netfilter/nfnetlink_osf.c: In function 'nf_osf_hdr_ctx_init':
cc1: warning: function may return address of local variable [-Wreturn-local-addr]
net/netfilter/nfnetlink_osf.c:171:16: note: declared here
171 | struct tcphdr _tcph;
| ^~~~~
I am not sure whether this can happen in practice, but moving the
variable declaration into the callers avoids the problem.
If the UDP header of a local VXLAN endpoint is NAT-ed, and the VXLAN
device has disabled UDP checksums and enabled Tx checksum offloading,
then the skb passed to udp_manip_pkt() has hdr->check == 0 (outer
checksum disabled) and skb->ip_summed == CHECKSUM_PARTIAL (inner packet
checksum offloaded).
Because of the ->ip_summed value, udp_manip_pkt() tries to update the
outer checksum with the new address and port, leading to an invalid
checksum sent on the wire, as the original null checksum obviously
didn't take the old address and port into account.
So, we can't take ->ip_summed into account in udp_manip_pkt(), as it
might not refer to the checksum we're acting on. Instead, we can base
the decision to update the UDP checksum entirely on the value of
hdr->check, because it's null if and only if checksum is disabled:
* A fully computed checksum can't be 0, since a 0 checksum is
represented by the CSUM_MANGLED_0 value instead.
* A partial checksum can't be 0, since the pseudo-header always adds
at least one non-zero value (the UDP protocol type 0x11) and adding
more values to the sum can't make it wrap to 0 as the carry is then
added to the wrapped number.
* A disabled checksum uses the special value 0.
The problem seems to be there from day one, although it was probably
not visible before UDP tunnels were implemented.
Fixes: 5b1158e909ec ("[NETFILTER]: Add NAT support for nf_conntrack") Signed-off-by: Guillaume Nault <gnault@redhat.com> Reviewed-by: Florian Westphal <fw@strlen.de> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When trying to lock read-only pages, sev_pin_memory() fails because
FOLL_WRITE is used as the flag for get_user_pages_fast().
Commit 73b0140bf0fe ("mm/gup: change GUP fast to use flags rather than a
write 'bool'") updated the get_user_pages_fast() call sites to use
flags, but incorrectly updated the call in sev_pin_memory(). As the
original coding of this call was correct, revert the change made by that
commit.
Fixes: 73b0140bf0fe ("mm/gup: change GUP fast to use flags rather than a write 'bool'") Signed-off-by: Janakarajan Natarajan <Janakarajan.Natarajan@amd.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Reviewed-by: Ira Weiny <ira.weiny@intel.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Sean Christopherson <sean.j.christopherson@intel.com> Cc: Vitaly Kuznetsov <vkuznets@redhat.com> Cc: Wanpeng Li <wanpengli@tencent.com> Cc: Jim Mattson <jmattson@google.com> Cc: Joerg Roedel <joro@8bytes.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Ingo Molnar <mingo@redhat.com> Cc: Borislav Petkov <bp@alien8.de> Cc: "H . Peter Anvin" <hpa@zytor.com> Cc: Mike Marshall <hubcap@omnibond.com> Cc: Brijesh Singh <brijesh.singh@amd.com> Link: http://lkml.kernel.org/r/20200423152419.87202-1-Janakarajan.Natarajan@amd.com Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Current logic incorrectly uses the enum ioapic_irq_destination_types
to check the posted interrupt destination types. However, the value was
set using APIC_DM_XXX macros, which are left-shifted by 8 bits.
Fixes by using the APIC_DM_FIXED and APIC_DM_LOWEST instead.
fsnotify()
[ realign the stack and store previous SP in R10 ]
<IRQ>
[ only IRET regs saved ]
common_interrupt()
interrupt_entry()
<NMI>
[ full pt_regs saved ]
...
[ unwind stack ]
When the unwinder goes through the NMI and the IRQ on the stack, and
then sees fsnotify(), it doesn't have access to the value of R10,
because it only has the five IRET registers. So the unwind stops
prematurely.
However, because the interrupt_entry() code is careful not to clobber
R10 before saving the full regs, the unwinder should be able to read R10
from the previously saved full pt_regs associated with the NMI.
Handle this case properly. When encountering an IRET regs frame
immediately after a full pt_regs frame, use the pt_regs as a backup
which can be used to get the C register values.
Also, note that a call frame resets the 'prev_regs' value, because a
function is free to clobber the registers. For this fix to work, the
IRET and full regs frames must be adjacent, with no FUNC frames in
between. So replace the FUNC hint in interrupt_entry() with an
IRET_REGS hint.
Fixes: ee9f8fce9964 ("x86/unwind: Add the ORC unwinder") Reviewed-by: Miroslav Benes <mbenes@suse.cz> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Ingo Molnar <mingo@kernel.org> Cc: Andy Lutomirski <luto@kernel.org> Cc: Dave Jones <dsj@fb.com> Cc: Jann Horn <jannh@google.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Link: https://lore.kernel.org/r/97a408167cc09f1cfa0de31a7b70dd88868d743f.1587808742.git.jpoimboe@redhat.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If the ORC entry type is unknown, nothing else can be done other than
reporting an error. Exit the function instead of breaking out of the
switch statement.
Fixes: ee9f8fce9964 ("x86/unwind: Add the ORC unwinder") Reviewed-by: Miroslav Benes <mbenes@suse.cz> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Ingo Molnar <mingo@kernel.org> Cc: Andy Lutomirski <luto@kernel.org> Cc: Dave Jones <dsj@fb.com> Cc: Jann Horn <jannh@google.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Link: https://lore.kernel.org/r/a7fa668ca6eabbe81ab18b2424f15adbbfdc810a.1587808742.git.jpoimboe@redhat.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If the unwinder is called before the ORC data has been initialized,
orc_find() returns NULL, and it tries to fall back to using frame
pointers. This can cause some unexpected warnings during boot.
Move the 'orc_init' check from orc_find() to __unwind_init(), so that it
doesn't even try to unwind from an uninitialized state.
Fixes: ee9f8fce9964 ("x86/unwind: Add the ORC unwinder") Reviewed-by: Miroslav Benes <mbenes@suse.cz> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Ingo Molnar <mingo@kernel.org> Cc: Andy Lutomirski <luto@kernel.org> Cc: Dave Jones <dsj@fb.com> Cc: Jann Horn <jannh@google.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Link: https://lore.kernel.org/r/069d1499ad606d85532eb32ce39b2441679667d5.1587808742.git.jpoimboe@redhat.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When unwinding an inactive task, the ORC unwinder skips the first frame
by default. If both the 'regs' and 'first_frame' parameters of
unwind_start() are NULL, 'state->sp' and 'first_frame' are later
initialized to the same value for an inactive task. Given there is a
"less than or equal to" comparison used at the end of __unwind_start()
for skipping stack frames, the first frame is skipped.
Drop the equal part of the comparison and make the behavior equivalent
to the frame pointer unwinder.
Fixes: ee9f8fce9964 ("x86/unwind: Add the ORC unwinder") Reviewed-by: Miroslav Benes <mbenes@suse.cz> Signed-off-by: Miroslav Benes <mbenes@suse.cz> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Ingo Molnar <mingo@kernel.org> Cc: Andy Lutomirski <luto@kernel.org> Cc: Dave Jones <dsj@fb.com> Cc: Jann Horn <jannh@google.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Link: https://lore.kernel.org/r/7f08db872ab59e807016910acdbe82f744de7065.1587808742.git.jpoimboe@redhat.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The LEAQ instruction in rewind_stack_do_exit() moves the stack pointer
directly below the pt_regs at the top of the task stack before calling
do_exit(). Tell the unwinder to expect pt_regs.
Fixes: 8c1f75587a18 ("x86/entry/64: Add unwind hint annotations") Reviewed-by: Miroslav Benes <mbenes@suse.cz> Signed-off-by: Jann Horn <jannh@google.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Ingo Molnar <mingo@kernel.org> Cc: Andy Lutomirski <luto@kernel.org> Cc: Dave Jones <dsj@fb.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Link: https://lore.kernel.org/r/68c33e17ae5963854916a46f522624f8e1d264f2.1587808742.git.jpoimboe@redhat.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
UNWIND_HINT_FUNC has some limitations: specifically, it doesn't reset
all the registers to undefined. This causes objtool to get confused
about the RBP push in __switch_to_asm(), resulting in bad ORC data.
While __switch_to_asm() does do some stack magic, it's otherwise a
normal callable-from-C function, so just annotate it as a function,
which makes objtool happy and allows it to produces the correct hints
automatically.
Fixes: 8c1f75587a18 ("x86/entry/64: Add unwind hint annotations") Reviewed-by: Miroslav Benes <mbenes@suse.cz> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Ingo Molnar <mingo@kernel.org> Cc: Andy Lutomirski <luto@kernel.org> Cc: Dave Jones <dsj@fb.com> Cc: Jann Horn <jannh@google.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Link: https://lore.kernel.org/r/03d0411920d10f7418f2e909210d8e9a3b2ab081.1587808742.git.jpoimboe@redhat.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In swapgs_restore_regs_and_return_to_usermode, after the stack is
switched to the trampoline stack, the existing UNWIND_HINT_REGS hint is
no longer valid, which can result in the following ORC unwinder warning:
WARNING: can't dereference registers at 000000003aeb0cdd for ip swapgs_restore_regs_and_return_to_usermode+0x93/0xa0
For full correctness, we could try to add complicated unwind hints so
the unwinder could continue to find the registers, but when when it's
this close to kernel exit, unwind hints aren't really needed anymore and
it's fine to just use an empty hint which tells the unwinder to stop.
For consistency, also move the UNWIND_HINT_EMPTY in
entry_SYSCALL_64_after_hwframe to a similar location.
Fixes: 3e3b9293d392 ("x86/entry/64: Return to userspace from the trampoline stack") Reported-by: Vince Weaver <vincent.weaver@maine.edu> Reported-by: Dave Jones <dsj@fb.com> Reported-by: Dr. David Alan Gilbert <dgilbert@redhat.com> Reported-by: Joe Mario <jmario@redhat.com> Reported-by: Jann Horn <jannh@google.com> Reported-by: Linus Torvalds <torvalds@linux-foundation.org> Reviewed-by: Miroslav Benes <mbenes@suse.cz> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Ingo Molnar <mingo@kernel.org> Cc: Andy Lutomirski <luto@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: https://lore.kernel.org/r/60ea8f562987ed2d9ace2977502fe481c0d7c9a0.1587808742.git.jpoimboe@redhat.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The PUSH_AND_CLEAR_REGS macro zeroes each register immediately after
pushing it. If an NMI or exception hits after a register is cleared,
but before the UNWIND_HINT_REGS annotation, the ORC unwinder will
wrongly think the previous value of the register was zero. This can
confuse the unwinding process and cause it to exit early.
Because ORC is simpler than DWARF, there are a limited number of unwind
annotation states, so it's not possible to add an individual unwind hint
after each push/clear combination. Instead, the register clearing
instructions need to be consolidated and moved to after the
UNWIND_HINT_REGS annotation.
Fixes: 3f01daecd545 ("x86/entry/64: Introduce the PUSH_AND_CLEAN_REGS macro") Reviewed-by: Miroslav Benes <mbenes@suse.cz> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Ingo Molnar <mingo@kernel.org> Cc: Andy Lutomirski <luto@kernel.org> Cc: Dave Jones <dsj@fb.com> Cc: Jann Horn <jannh@google.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Link: https://lore.kernel.org/r/68fd3d0bc92ae2d62ff7879d15d3684217d51f08.1587808742.git.jpoimboe@redhat.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
As an optimization, cpa_flush() was changed to optionally only flush
the range in @cpa if it was small enough. However, this range does
not include any direct map aliases changed in cpa_process_alias(). So
small set_memory_() calls that touch that alias don't get the direct
map changes flushed. This situation can happen when the virtual
address taking variants are passed an address in vmalloc or modules
space.
In these cases, force a full TLB flush.
Note this issue does not extend to cases where the set_memory_() calls are
passed a direct map address, or page array, etc, as the primary target. In
those cases the direct map would be flushed.
batadv_v_ogm_process() invokes batadv_hardif_neigh_get(), which returns
a reference of the neighbor object to "hardif_neigh" with increased
refcount.
When batadv_v_ogm_process() returns, "hardif_neigh" becomes invalid, so
the refcount should be decreased to keep refcount balanced.
The reference counting issue happens in one exception handling paths of
batadv_v_ogm_process(). When batadv_v_ogm_orig_get() fails to get the
orig node and returns NULL, the refcnt increased by
batadv_hardif_neigh_get() is not decreased, causing a refcnt leak.
Fix this issue by jumping to "out" label when batadv_v_ogm_orig_get()
fails to get the orig node.
Fixes: 9323158ef9f4 ("batman-adv: OGMv2 - implement originators logic") Signed-off-by: Xiyu Yang <xiyuyang19@fudan.edu.cn> Signed-off-by: Xin Tan <tanxin.ctf@gmail.com> Signed-off-by: Sven Eckelmann <sven@narfation.org> Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
batadv_show_throughput_override() invokes batadv_hardif_get_by_netdev(),
which gets a batadv_hard_iface object from net_dev with increased refcnt
and its reference is assigned to a local pointer 'hard_iface'.
When batadv_store_throughput_override() returns, "hard_iface" becomes
invalid, so the refcount should be decreased to keep refcount balanced.
The issue happens in one error path of
batadv_store_throughput_override(). When batadv_parse_throughput()
returns NULL, the refcnt increased by batadv_hardif_get_by_netdev() is
not decreased, causing a refcnt leak.
Fix this issue by jumping to "out" label when batadv_parse_throughput()
returns NULL.
Fixes: 0b5ecc6811bd ("batman-adv: add throughput override attribute to hard_ifaces") Signed-off-by: Xiyu Yang <xiyuyang19@fudan.edu.cn> Signed-off-by: Xin Tan <tanxin.ctf@gmail.com> Signed-off-by: Sven Eckelmann <sven@narfation.org> Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
batadv_show_throughput_override() invokes batadv_hardif_get_by_netdev(),
which gets a batadv_hard_iface object from net_dev with increased refcnt
and its reference is assigned to a local pointer 'hard_iface'.
When batadv_show_throughput_override() returns, "hard_iface" becomes
invalid, so the refcount should be decreased to keep refcount balanced.
The issue happens in the normal path of
batadv_show_throughput_override(), which forgets to decrease the refcnt
increased by batadv_hardif_get_by_netdev() before the function returns,
causing a refcnt leak.
Fix this issue by calling batadv_hardif_put() before the
batadv_show_throughput_override() returns in the normal path.
Fixes: 0b5ecc6811bd ("batman-adv: add throughput override attribute to hard_ifaces") Signed-off-by: Xiyu Yang <xiyuyang19@fudan.edu.cn> Signed-off-by: Xin Tan <tanxin.ctf@gmail.com> Signed-off-by: Sven Eckelmann <sven@narfation.org> Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
and change to pseudorandom numbers, as this is a traffic dithering
operation that doesn't need crypto-grade.
The previous code operated in 4 steps:
1. Generate a random byte 0 <= rand_tq <= 255
2. Multiply it by BATADV_TQ_MAX_VALUE - tq
3. Divide by 255 (= BATADV_TQ_MAX_VALUE)
4. Return BATADV_TQ_MAX_VALUE - rand_tq
This would apperar to scale (BATADV_TQ_MAX_VALUE - tq) by a random
value between 0/255 and 255/255.
But! The intermediate value between steps 3 and 4 is stored in a u8
variable. So it's truncated, and most of the time, is less than 255, after
which the division produces 0. Specifically, if tq is odd, the product is
always even, and can never be 255. If tq is even, there's exactly one
random byte value that will produce a product byte of 255.
Thus, the return value is 255 (511/512 of the time) or 254 (1/512
of the time).
If we assume that the truncation is a bug, and the code is meant to scale
the input, a simpler way of looking at it is that it's returning a random
value between tq and BATADV_TQ_MAX_VALUE, inclusive.
Well, we have an optimized function for doing just that.
Fixes: 3c12de9a5c75 ("batman-adv: network coding - code and transmit packets if possible") Signed-off-by: George Spelvin <lkml@sdf.org> Signed-off-by: Sven Eckelmann <sven@narfation.org> Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
abs_vdebt is an atomic_64 which tracks how much over budget a given cgroup
is and controls the activation of use_delay mechanism. Once a cgroup goes
over budget from forced IOs, it has to pay it back with its future budget.
The progress guarantee on debt paying comes from the iocg being active -
active iocgs are processed by the periodic timer, which ensures that as time
passes the debts dissipate and the iocg returns to normal operation.
However, both iocg activation and vdebt handling are asynchronous and a
sequence like the following may happen.
1. The iocg is in the process of being deactivated by the periodic timer.
2. A bio enters ioc_rqos_throttle(), calls iocg_activate() which returns
without anything because it still sees that the iocg is already active.
3. The iocg is deactivated.
4. The bio from #2 is over budget but needs to be forced. It increases
abs_vdebt and goes over the threshold and enables use_delay.
5. IO control is enabled for the iocg's subtree and now IOs are attributed
to the descendant cgroups and the iocg itself no longer issues IOs.
This leaves the iocg with stuck abs_vdebt - it has debt but inactive and no
further IOs which can activate it. This can end up unduly punishing all the
descendants cgroups.
The usual throttling path has the same issue - the iocg must be active while
throttled to ensure that future event will wake it up - and solves the
problem by synchronizing the throttling path with a spinlock. abs_vdebt
handling is another form of overage handling and shares a lot of
characteristics including the fact that it isn't in the hottest path.
This patch fixes the above and other possible races by strictly
synchronizing abs_vdebt and use_delay handling with iocg->waitq.lock.
The current max_pfn equals to zero. In this case, I found it caused users
cannot get some page information through /proc such as kpagecount in v5.6
kernel because of new sanity checks. The following message is displayed by
stress-ng test suite with the command "stress-ng --verbose --physpage 1 -t
1" on HiFive unleashed board.
# stress-ng --verbose --physpage 1 -t 1
stress-ng: debug: [109] 4 processors online, 4 processors configured
stress-ng: info: [109] dispatching hogs: 1 physpage
stress-ng: debug: [109] cache allocate: reducing cache level from L3 (too high) to L0
stress-ng: debug: [109] get_cpu_cache: invalid cache_level: 0
stress-ng: info: [109] cache allocate: using built-in defaults as no suitable cache found
stress-ng: debug: [109] cache allocate: default cache size: 2048K
stress-ng: debug: [109] starting stressors
stress-ng: debug: [109] 1 stressor spawned
stress-ng: debug: [110] stress-ng-physpage: started [110] (instance 0)
stress-ng: error: [110] stress-ng-physpage: cannot read page count for address 0x3fd34de000 in /proc/kpagecount, errno=0 (Success)
stress-ng: error: [110] stress-ng-physpage: cannot read page count for address 0x3fd32db078 in /proc/kpagecount, errno=0 (Success)
...
stress-ng: error: [110] stress-ng-physpage: cannot read page count for address 0x3fd32db078 in /proc/kpagecount, errno=0 (Success)
stress-ng: debug: [110] stress-ng-physpage: exited [110] (instance 0)
stress-ng: debug: [109] process [110] terminated
stress-ng: info: [109] successful run completed in 1.00s
#
After applying this patch, the kernel can pass the test.
Commit 64e90a8acb859 ("Introduce STATIC_USERMODEHELPER to mediate
call_usermodehelper()") added the optiont to disable all
call_usermodehelper() calls by setting STATIC_USERMODEHELPER_PATH to
an empty string. When this is done, and crashdump is triggered, it
will crash on null pointer dereference, since we make assumptions
over what call_usermodehelper_exec() did.
This has been reported by Sergey when one triggers a a coredump
with the following configuration:
The way disabling the umh was designed was that call_usermodehelper_exec()
would just return early, without an error. But coredump assumes
certain variables are set up for us when this happens, and calls
ile_start_write(cprm.file) with a NULL file.
Check the return value of gasket_get_bar_index function as it can return
a negative one (-EINVAL). If this happens, a negative index is used in
the "gasket_dev->bar_data" array.
Addresses-Coverity-ID: 1438542 ("Negative array index read") Fixes: 9a69f5087ccc2 ("drivers/staging: Gasket driver framework + Apex driver") Signed-off-by: Oscar Carter <oscar.carter@gmx.com> Cc: stable <stable@vger.kernel.org> Reviewed-by: Richard Yeh <rcy@google.com> Link: https://lore.kernel.org/r/20200501155118.13380-1-oscar.carter@gmx.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Eduard reported a problem mounting cephfs on s390 arch. The feature
mask sent by the MDS is little-endian, so we need to convert it
before storing and testing against it.
Commit 1c30844d2dfe ("mm: reclaim small amounts of memory when an
external fragmentation event occurs") adds a boost_watermark() function
which increases the min watermark in a zone by at least
pageblock_nr_pages or the number of pages in a page block.
On Arm64, with 64K pages and 512M huge pages, this is 8192 pages or
512M. It does this regardless of the number of managed pages managed in
the zone or the likelihood of success.
This can put the zone immediately under water in terms of allocating
pages from the zone, and can cause a small machine to fail immediately
due to OoM. Unlike set_recommended_min_free_kbytes(), which
substantially increases min_free_kbytes and is tied to THP,
boost_watermark() can be called even if THP is not active.
The problem is most likely to appear on architectures such as Arm64
where pageblock_nr_pages is very large.
It is desirable to run the kdump capture kernel in as small a space as
possible to avoid wasting memory. In some architectures, such as Arm64,
there are restrictions on where the capture kernel can run, and
therefore, the space available. A capture kernel running in 768M can
fail due to OoM immediately after boost_watermark() sets the min in zone
DMA32, where most of the memory is, to 512M. It fails even though there
is over 500M of free memory. With boost_watermark() suppressed, the
capture kernel can run successfully in 448M.
This patch limits boost_watermark() to boosting a zone's min watermark
only when there are enough pages that the boost will produce positive
results. In this case that is estimated to be four times as many pages
as pageblock_nr_pages.
Mel said:
: There is no harm in marking it stable. Clearly it does not happen very
: often but it's not impossible. 32-bit x86 is a lot less common now
: which would previously have been vulnerable to triggering this easily.
: ppc64 has a larger base page size but typically only has one zone.
: arm64 is likely the most vulnerable, particularly when CMA is
: configured with a small movable zone.
Fixes: 1c30844d2dfe ("mm: reclaim small amounts of memory when an external fragmentation event occurs") Signed-off-by: Henry Willard <henry.willard@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Reviewed-by: David Hildenbrand <david@redhat.com> Acked-by: Mel Gorman <mgorman@techsingularity.net> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: <stable@vger.kernel.org> Link: http://lkml.kernel.org/r/1588294148-6586-1-git-send-email-henry.willard@oracle.com Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Without CONFIG_PREEMPT, it can happen that we get soft lockups detected,
e.g., while booting up.
watchdog: BUG: soft lockup - CPU#0 stuck for 22s! [swapper/0:1]
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.6.0-next-20200331+ #4
Hardware name: Red Hat KVM, BIOS 1.11.1-4.module+el8.1.0+4066+0f1aadab 04/01/2014
RIP: __pageblock_pfn_to_page+0x134/0x1c0
Call Trace:
set_zone_contiguous+0x56/0x70
page_alloc_init_late+0x166/0x176
kernel_init_freeable+0xfa/0x255
kernel_init+0xa/0x106
ret_from_fork+0x35/0x40
The issue becomes visible when having a lot of memory (e.g., 4TB)
assigned to a single NUMA node - a system that can easily be created
using QEMU. Inside VMs on a hypervisor with quite some memory
overcommit, this is fairly easy to trigger.
Signed-off-by: David Hildenbrand <david@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Reviewed-by: Pavel Tatashin <pasha.tatashin@soleen.com> Reviewed-by: Pankaj Gupta <pankaj.gupta.linux@gmail.com> Reviewed-by: Baoquan He <bhe@redhat.com> Reviewed-by: Shile Zhang <shile.zhang@linux.alibaba.com> Acked-by: Michal Hocko <mhocko@suse.com> Cc: Kirill Tkhai <ktkhai@virtuozzo.com> Cc: Shile Zhang <shile.zhang@linux.alibaba.com> Cc: Pavel Tatashin <pasha.tatashin@soleen.com> Cc: Daniel Jordan <daniel.m.jordan@oracle.com> Cc: Michal Hocko <mhocko@kernel.org> Cc: Alexander Duyck <alexander.duyck@gmail.com> Cc: Baoquan He <bhe@redhat.com> Cc: Oscar Salvador <osalvador@suse.de> Cc: <stable@vger.kernel.org> Link: http://lkml.kernel.org/r/20200416073417.5003-1-david@redhat.com Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In the event that we add to ovflist, before commit 339ddb53d373
("fs/epoll: remove unnecessary wakeups of nested epoll") we would be
woken up by ep_scan_ready_list, and did no wakeup in ep_poll_callback.
With that wakeup removed, if we add to ovflist here, we may never wake
up. Rather than adding back the ep_scan_ready_list wakeup - which was
resulting in unnecessary wakeups, trigger a wake-up in ep_poll_callback.
We noticed that one of our workloads was missing wakeups starting with 339ddb53d373 and upon manual inspection, this wakeup seemed missing to me.
With this patch added, we no longer see missing wakeups. I haven't yet
tried to make a small reproducer, but the existing kselftests in
filesystem/epoll passed for me with this patch.
[khazhy@google.com: use if/elif instead of goto + cleanup suggested by Roman] Link: http://lkml.kernel.org/r/20200424190039.192373-1-khazhy@google.com Fixes: 339ddb53d373 ("fs/epoll: remove unnecessary wakeups of nested epoll") Signed-off-by: Khazhismel Kumykov <khazhy@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Reviewed-by: Roman Penyaev <rpenyaev@suse.de> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Roman Penyaev <rpenyaev@suse.de> Cc: Heiher <r@hev.cc> Cc: Jason Baron <jbaron@akamai.com> Cc: <stable@vger.kernel.org> Link: http://lkml.kernel.org/r/20200424025057.118641-1-khazhy@google.com Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
- fixes a lost wakeup introduced by commit 339ddb53d373 ("fs/epoll:
remove unnecessary wakeups of nested epoll")
- improves performance for events delivery.
The description of the problem is the following: if N (>1) threads are
waiting on ep->wq for new events and M (>1) events come, it is quite
likely that >1 wakeups hit the same wait queue entry, because there is
quite a big window between __add_wait_queue_exclusive() and the
following __remove_wait_queue() calls in ep_poll() function.
This can lead to lost wakeups, because thread, which was woken up, can
handle not all the events in ->rdllist. (in better words the problem is
described here: https://lkml.org/lkml/2019/10/7/905)
The idea of the current patch is to use init_wait() instead of
init_waitqueue_entry().
Internally init_wait() sets autoremove_wake_function as a callback,
which removes the wait entry atomically (under the wq locks) from the
list, thus the next coming wakeup hits the next wait entry in the wait
queue, thus preventing lost wakeups.
Problem is very well reproduced by the epoll60 test case [1].
Wait entry removal on wakeup has also performance benefits, because
there is no need to take a ep->lock and remove wait entry from the queue
after the successful wakeup. Here is the timing output of the epoll60
test case:
With explicit wakeup from ep_scan_ready_list() (the state of the
code prior 339ddb53d373):
real 0m6.970s
user 0m49.786s
sys 0m0.113s
After this patch:
real 0m5.220s
user 0m36.879s
sys 0m0.019s
The other testcase is the stress-epoll [2], where one thread consumes
all the events and other threads produce many events:
With explicit wakeup from ep_scan_ready_list() (the state of the
code prior 339ddb53d373):
(number of "events/ms" represents event bandwidth, thus higher is
better; number of "run-time ms" represents overall time spent
doing the benchmark, thus lower is better)
Commit cc731525f26a ("signal: Remove kernel interal si_code magic")
changed the value of SI_FROMUSER(SI_MESGQ), this means that mq_notify() no
longer works if the sender doesn't have rights to send a signal.
Change __do_notify() to use do_send_sig_info() instead of kill_pid_info()
to avoid check_kill_permission().
This needs the additional notify.sigev_signo != 0 check, shouldn't we
change do_mq_notify() to deny sigev_signo == 0 ?
The dcn20_validate_bandwidth function would have code touching the
incorrect registers emitted outside of the boundaries of the
DC_FP_START/END macros, at least on ppc64le. Work around the
problem by wrapping the whole function instead.
Signed-off-by: Daniel Kolesa <daniel@octaforge.org> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Cc: stable@vger.kernel.org # 5.6.x Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Disable the MEI driver on LBG SPS (server) platforms, some corner
flows such as recovery mode does not work, and the driver
doesn't have working use cases.
It's currently the amba driver's responsibility to initialize the pointer,
dma_parms, for its corresponding struct device. The benefit with this
approach allows us to avoid the initialization and to not waste memory for
the struct device_dma_parameters, as this can be decided on a case by case
basis.
However, it has turned out that this approach is not very practical. Not
only does it lead to open coding, but also to real errors. In principle
callers of dma_set_max_seg_size() doesn't check the error code, but just
assumes it succeeds.
For these reasons, let's do the initialization from the common amba bus at
the device registration point. This also follows the way the PCI devices
are being managed, see pci_device_add().
Suggested-by: Christoph Hellwig <hch@lst.de> Cc: Russell King <linux@armlinux.org.uk> Cc: <stable@vger.kernel.org> Tested-by: Haibo Chen <haibo.chen@nxp.com> Reviewed-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> Reviewed-by: Christoph Hellwig <hch@lst.de> Link: https://lore.kernel.org/r/20200422101013.31267-1-ulf.hansson@linaro.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It's currently the platform driver's responsibility to initialize the
pointer, dma_parms, for its corresponding struct device. The benefit with
this approach allows us to avoid the initialization and to not waste memory
for the struct device_dma_parameters, as this can be decided on a case by
case basis.
However, it has turned out that this approach is not very practical. Not
only does it lead to open coding, but also to real errors. In principle
callers of dma_set_max_seg_size() doesn't check the error code, but just
assumes it succeeds.
For these reasons, let's do the initialization from the common platform bus
at the device registration point. This also follows the way the PCI devices
are being managed, see pci_device_add().
The static analyzer in GCC 10 spotted that in huge_pte_alloc() we may
pass a NULL pmdp into pte_alloc_map() when pmd_alloc() returns NULL:
| CC arch/arm64/mm/pageattr.o
| CC arch/arm64/mm/hugetlbpage.o
| from arch/arm64/mm/hugetlbpage.c:10:
| arch/arm64/mm/hugetlbpage.c: In function ‘huge_pte_alloc’:
| ./arch/arm64/include/asm/pgtable-types.h:28:24: warning: dereference of NULL ‘pmdp’ [CWE-690] [-Wanalyzer-null-dereference]
| ./arch/arm64/include/asm/pgtable.h:436:26: note: in expansion of macro ‘pmd_val’
| arch/arm64/mm/hugetlbpage.c:242:10: note: in expansion of macro ‘pte_alloc_map’
| |arch/arm64/mm/hugetlbpage.c:232:10:
| |./arch/arm64/include/asm/pgtable-types.h:28:24:
| ./arch/arm64/include/asm/pgtable.h:436:26: note: in expansion of macro ‘pmd_val’
| arch/arm64/mm/hugetlbpage.c:242:10: note: in expansion of macro ‘pte_alloc_map’
This can only occur when the kernel cannot allocate a page, and so is
unlikely to happen in practice before other systems start failing.
We can avoid this by bailing out if pmd_alloc() fails, as we do earlier
in the function if pud_alloc() fails.
Fixes: 66b3923a1a0f ("arm64: hugetlb: add support for PTE contiguous bit") Signed-off-by: Mark Rutland <mark.rutland@arm.com> Reported-by: Kyrill Tkachov <kyrylo.tkachov@arm.com> Cc: <stable@vger.kernel.org> # 4.5.x- Cc: Will Deacon <will@kernel.org> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In the unlikely event that a 32bit vcpu traps into the hypervisor
on an instruction that is located right at the end of the 32bit
range, the emulation of that instruction is going to increment
PC past the 32bit range. This isn't great, as userspace can then
observe this value and get a bit confused.
Conversly, userspace can do things like (in the context of a 64bit
guest that is capable of 32bit EL0) setting PSTATE to AArch64-EL0,
set PC to a 64bit value, change PSTATE to AArch32-USR, and observe
that PC hasn't been truncated. More confusion.
Fix both by:
- truncating PC increments for 32bit guests
- sanitizing all 32bit regs every time a core reg is changed by
userspace, and that PSTATE indicates a 32bit mode.
Cc: stable@vger.kernel.org Acked-by: Will Deacon <will@kernel.org> Signed-off-by: Marc Zyngier <maz@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When deciding whether a guest has to be stopped we check whether this
is a private interrupt or not. Unfortunately, there's an off-by-one bug
here, and we fail to recognize a whole range of interrupts as being
global (GICv2 SPIs 32-63).
Fix the condition from > to be >=.
Cc: stable@vger.kernel.org Fixes: abd7229626b93 ("KVM: arm/arm64: Simplify active_change_prepare and plug race") Reported-by: André Przywara <andre.przywara@arm.com> Signed-off-by: Marc Zyngier <maz@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Clear CF and ZF in the VM-Exit path after doing __FILL_RETURN_BUFFER so
that KVM doesn't interpret clobbered RFLAGS as a VM-Fail. Filling the
RSB has always clobbered RFLAGS, its current incarnation just happens
clear CF and ZF in the processs. Relying on the macro to clear CF and
ZF is extremely fragile, e.g. commit 089dd8e53126e ("x86/speculation:
Change FILL_RETURN_BUFFER to work with objtool") tweaks the loop such
that the ZF flag is always set.
Reported-by: Qian Cai <cai@lca.pw> Cc: Rick Edgecombe <rick.p.edgecombe@intel.com> Cc: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Josh Poimboeuf <jpoimboe@redhat.com> Cc: stable@vger.kernel.org Fixes: f2fde6a5bcfcf ("KVM: VMX: Move RSB stuffing to before the first RET after VM-Exit") Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Message-Id: <20200506035355.2242-1-sean.j.christopherson@intel.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In LPAR we will only get an intercept for FC==3 for the PQAP
instruction. Running nested under z/VM can result in other intercepts as
well as ECA_APIE is an effective bit: If one hypervisor layer has
turned this bit off, the end result will be that we will get intercepts for
all function codes. Usually the first one will be a query like PQAP(QCI).
So the WARN_ON_ONCE is not right. Let us simply remove it.
Cc: Pierre Morel <pmorel@linux.ibm.com> Cc: Tony Krowiak <akrowiak@linux.ibm.com> Cc: stable@vger.kernel.org # v5.3+ Fixes: e5282de93105 ("s390: ap: kvm: add PQAP interception for AQIC") Link: https://lore.kernel.org/kvm/20200505083515.2720-1-borntraeger@de.ibm.com Reported-by: Qian Cai <cailca@icloud.com> Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com> Reviewed-by: David Hildenbrand <david@redhat.com> Reviewed-by: Cornelia Huck <cohuck@redhat.com> Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The initial Zinc patchset, after some mailing list discussion, contained
code to ensure that kernel_fpu_enable would not be kept on for more than
a 4k chunk, since it disables preemption. The choice of 4k isn't totally
scientific, but it's not a bad guess either, and it's what's used in
both the x86 poly1305, blake2s, and nhpoly1305 code already (in the form
of PAGE_SIZE, which this commit corrects to be explicitly 4k for the
former two).
Ard did some back of the envelope calculations and found that
at 5 cycles/byte (overestimate) on a 1ghz processor (pretty slow), 4k
means we have a maximum preemption disabling of 20us, which Sebastian
confirmed was probably a good limit.
Unfortunately the chunking appears to have been left out of the final
patchset that added the glue code. So, this commit adds it back in.
Fixes: 84e03fa39fbe ("crypto: x86/chacha - expose SIMD ChaCha routine as library function") Fixes: b3aad5bad26a ("crypto: arm64/chacha - expose arm64 ChaCha routine as library function") Fixes: a44a3430d71b ("crypto: arm/chacha - expose ARM ChaCha routine as library function") Fixes: d7d7b8535662 ("crypto: x86/poly1305 - wire up faster implementations for kernel") Fixes: f569ca164751 ("crypto: arm64/poly1305 - incorporate OpenSSL/CRYPTOGAMS NEON implementation") Fixes: a6b803b3ddc7 ("crypto: arm/poly1305 - incorporate OpenSSL/CRYPTOGAMS NEON implementation") Fixes: ed0356eda153 ("crypto: blake2s - x86_64 SIMD implementation") Cc: Eric Biggers <ebiggers@google.com> Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Cc: stable@vger.kernel.org Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Reviewed-by: Ard Biesheuvel <ardb@kernel.org> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Rather than chunking via PAGE_SIZE, this commit changes the arch
implementations to chunk in explicit 4k parts, so that calculations on
maximum acceptable latency don't suddenly become invalid on platforms
where PAGE_SIZE isn't 4k, such as arm64.
x86_64 lazily maps in the vmalloc pages, and the way this works with per_cpu
areas can be complex, to say the least. Mappings may happen at boot up, and
if nothing synchronizes the page tables, those page mappings may not be
synced till they are used. This causes issues for anything that might touch
one of those mappings in the path of the page fault handler. When one of
those unmapped mappings is touched in the page fault handler, it will cause
another page fault, which in turn will cause a page fault, and leave us in
a loop of page faults.
Commit 763802b53a42 ("x86/mm: split vmalloc_sync_all()") split
vmalloc_sync_all() into vmalloc_sync_unmappings() and
vmalloc_sync_mappings(), as on system exit, it did not need to do a full
sync on x86_64 (although it still needed to be done on x86_32). By chance,
the vmalloc_sync_all() would synchronize the page mappings done at boot up
and prevent the per cpu area from being a problem for tracing in the page
fault handler. But when that synchronization in the exit of a task became a
nop, it caused the problem to appear.
Running on a slower machine, it is possible that the preempt delay kernel
thread may still be executing if the module was immediately removed after
added, and this can cause the kernel to crash as the kernel thread might be
executing after its code has been removed.
There's no reason that the caller of the code shouldn't just wait for the
delay thread to finish, as the thread can also be created by a trigger in
the sysfs code, which also has the same issues.
Link: http://lore.kernel.org/r/5EA2B0C8.2080706@cn.fujitsu.com Cc: stable@vger.kernel.org Fixes: 793937236d1ee ("lib: Add module for testing preemptoff/irqsoff latency tracers") Reported-by: Xiao Yang <yangx.jy@cn.fujitsu.com> Reviewed-by: Xiao Yang <yangx.jy@cn.fujitsu.com> Reviewed-by: Joel Fernandes <joel@joelfernandes.org> Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently we check to make sure there is no error state on the extcon
handle for VBUS when writing to the HS_PHY_GENCONFIG_2 register. When using
the USB role-switch API we still need to write to this register absent an
extcon handle.
This patch makes the appropriate update to ensure the write happens if
role-switching is true.
Fixes: 05559f10ed79 ("usb: chipidea: add role switch class support") Cc: stable <stable@vger.kernel.org> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Philipp Zabel <p.zabel@pengutronix.de> Cc: linux-usb@vger.kernel.org Cc: linux-kernel@vger.kernel.org Cc: Stephen Boyd <swboyd@chromium.org> Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org> Signed-off-by: Peter Chen <peter.chen@nxp.com> Link: https://lore.kernel.org/r/20200507004918.25975-2-peter.chen@kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The state of the center button was not reported to userspace for the
2nd-gen Intuos Pro S when used over Bluetooth due to the pad handling
code not being updated to support its reduced number of buttons. This
patch uses the actual number of buttons present on the tablet to
assemble a button state bitmap.
Link: https://github.com/linuxwacom/xf86-input-wacom/issues/112 Fixes: cd47de45b855 ("HID: wacom: Add 2nd gen Intuos Pro Small support") Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> Cc: stable@vger.kernel.org # v5.3+ Signed-off-by: Jiri Kosina <jkosina@suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The syzbot fuzzer discovered a bad race between in the usbhid driver
between usbhid_stop() and usbhid_close(). In particular,
usbhid_stop() does:
usb_free_urb(usbhid->urbin);
...
usbhid->urbin = NULL; /* don't mess up next start */
and usbhid_close() does:
usb_kill_urb(usbhid->urbin);
with no mutual exclusion. If the two routines happen to run
concurrently so that usb_kill_urb() is called in between the
usb_free_urb() and the NULL assignment, it will access the
deallocated urb structure -- a use-after-free bug.
This patch adds a mutex to the usbhid private structure and uses it to
enforce mutual exclusion of the usbhid_start(), usbhid_stop(),
usbhid_open() and usbhid_close() callbacks.
The referenced commit broke pen and touch input for a variety of devices
such as the Cintiq Pro 32. Affected devices may appear to work normally
for a short amount of time, but eventually loose track of actual touch
state and can leave touch arbitration enabled which prevents the pen
from working. The commit is not itself required for any currently-available
Bluetooth device, and so we revert it to correct the behavior of broken
devices.
This breakage occurs due to a mismatch between the order of collections
and the order of usages on some devices. This commit tries to read the
contact count before processing events, but will fail if the contact
count does not occur prior to the first logical finger collection. This
is the case for devices like the Cintiq Pro 32 which place the contact
count at the very end of the report.
Without the contact count set, touches will only be partially processed.
The `wacom_wac_finger_slot` function will not open any slots since the
number of contacts seen is greater than the expectation of 0, but we will
still end up calling `input_mt_sync_frame` for each finger anyway. This
can cause problems for userspace separate from the issue currently taking
place in the kernel. Only once all of the individual finger collections
have been processed do we finally get to the enclosing collection which
contains the contact count. The value ends up being used for the *next*
report, however.
This delayed use of the contact count can cause the driver to loose track
of the actual touch state and believe that there are contacts down when
there aren't. This leaves touch arbitration enabled and prevents the pen
from working. It can also cause userspace to incorrectly treat single-
finger input as gestures.
Link: https://github.com/linuxwacom/input-wacom/issues/146 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> Reviewed-by: Aaron Armstrong Skomra <aaron.skomra@wacom.com> Fixes: 15893fa40109 ("HID: wacom: generic: read the number of expected touches on a per collection basis") Cc: stable@vger.kernel.org # 5.3+ Signed-off-by: Jiri Kosina <jkosina@suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When we start shutdown in sctp_sf_do_dupcook_a(), we want to bundle
the SHUTDOWN with the COOKIE-ACK to ensure that the peer receives them
at the same time and in the correct order. This bundling was broken by
commit 4ff40b86262b ("sctp: set chunk transport correctly when it's a
new asoc"), which assigns a transport for the COOKIE-ACK, but not for
the SHUTDOWN.
Fix this by passing a reference to the COOKIE-ACK chunk as an argument
to sctp_sf_do_9_2_start_shutdown() and onward to
sctp_make_shutdown(). This way the SHUTDOWN chunk is assigned the same
transport as the COOKIE-ACK chunk, which allows them to be bundled.
In sctp_sf_do_9_2_start_shutdown(), the void *arg parameter was
previously unused. Now that we're taking it into use, it must be a
valid pointer to a chunk, or NULL. There is only one call site where
it's not, in sctp_sf_autoclose_timer_expire(). Fix that too.
Fixes: 4ff40b86262b ("sctp: set chunk transport correctly when it's a new asoc") Signed-off-by: Jere Leppänen <jere.leppanen@nokia.com> Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We've recently switched from extracting the value of HID_DG_CONTACTMAX
at a fixed offset (which may not be correct for all tablets) to
injecting the report into the driver for the generic codepath to handle.
Unfortunately, this change was made for *all* tablets, even those which
aren't generic. Because `wacom_wac_report` ignores reports from non-
generic devices, the contact count never gets initialized. Ultimately
this results in the touch device itself failing to probe, and thus the
loss of touch input.
This commit adds back the fixed-offset extraction for non-generic devices.
Link: https://github.com/linuxwacom/input-wacom/issues/155 Fixes: 184eccd40389 ("HID: wacom: generic: read HID_DG_CONTACTMAX from any feature report") Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> Reviewed-by: Aaron Armstrong Skomra <aaron.skomra@wacom.com> CC: stable@vger.kernel.org # 5.3+ Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com> Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Users with pathological hardware reported CPU stalls on CONFIG_
PREEMPT_VOLUNTARY=y, because the ringbuffers would stay full, meaning
these workers would never terminate. That turned out not to be okay on
systems without forced preemption, which Sultan observed. This commit
adds a cond_resched() to the bottom of each loop iteration, so that
these workers don't hog the core. Note that we don't need this on the
napi poll worker, since that terminates after its budget is expended.
Suggested-by: Sultan Alsawaf <sultan@kerneltoast.com> Reported-by: Wang Jian <larkwang@gmail.com> Fixes: e7096c131e51 ("net: WireGuard secure network tunnel") Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It's already possible to create two different interfaces and loop
packets between them. This has always been possible with tunnels in the
kernel, and isn't specific to wireguard. Therefore, the networking stack
already needs to deal with that. At the very least, the packet winds up
exceeding the MTU and is discarded at that point. So, since this is
already something that happens, there's no need to forbid the not very
exceptional case of routing a packet back to the same interface; this
loop is no different than others, and we shouldn't special case it, but
rather rely on generic handling of loops in general. This also makes it
easier to do interesting things with wireguard such as onion routing.
At the same time, we add a selftest for this, ensuring that both onion
routing works and infinite routing loops do not crash the kernel. We
also add a test case for wireguard interfaces nesting packets and
sending traffic between each other, as well as the loop in this case
too. We make sure to send some throughput-heavy traffic for this use
case, to stress out any possible recursion issues with the locks around
workqueues.
Fixes: e7096c131e51 ("net: WireGuard secure network tunnel") Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
the related system resources were not released when enetc_hw_alloc()
return error in the enetc_pci_mdio_probe(), add iounmap() for error
handling label "err_hw_alloc" to fix it.
Fixes: 6517798dd3432a ("enetc: Make MDIO accessors more generic and export to include/linux/fsl") Cc: Andy Shevchenko <andy.shevchenko@gmail.com> Signed-off-by: Dejin Zheng <zhengdejin5@gmail.com> Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
WireGuard currently only propagates ECN markings on tunnel decap according
to the old RFC3168 specification. However, the spec has since been updated
in RFC6040 to recommend slightly different decapsulation semantics. This
was implemented in the kernel as a set of common helpers for ECN
decapsulation, so let's just switch over WireGuard to using those, so it
can benefit from this enhancement and any future tweaks. We do not drop
packets with invalid ECN marking combinations, because WireGuard is
frequently used to work around broken ISPs, which could be doing that.
Fixes: e7096c131e51 ("net: WireGuard secure network tunnel") Reported-by: Olivier Tilmans <olivier.tilmans@nokia-bell-labs.com> Cc: Dave Taht <dave.taht@gmail.com> Cc: Rodney W. Grimes <ietf@gndrsh.dnsmgr.net> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Prior, if the alloc_percpu of packet_percpu_multicore_worker_alloc
failed, the previously allocated ptr_ring wouldn't be freed. This commit
adds the missing call to ptr_ring_cleanup in the error case.
Reported-by: Sultan Alsawaf <sultan@kerneltoast.com> Fixes: e7096c131e51 ("net: WireGuard secure network tunnel") Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The "info->fs.location" is a u32 that comes from the user via the
ethtool_set_rxnfc() function. We need to check for invalid values to
prevent a buffer overflow.
I copy and pasted this check from the mvpp2_ethtool_cls_rule_ins()
function.
Fixes: 90b509b39ac9 ("net: mvpp2: cls: Add Classification offload support") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The "rss_context" variable comes from the user via ethtool_get_rxfh().
It can be any u32 value except zero. Eventually it gets passed to
mvpp22_rss_ctx() and if it is over MVPP22_N_RSS_TABLES (8) then it
results in an array overflow.
Fixes: 895586d5dc32 ("net: mvpp2: cls: Use RSS contexts to handle RSS tables") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Processing commands by cmd_work_handler() while already in Internal
Error State will result in entry leak, since the handler process force
completion without doorbell. Forced completion doesn't release the entry
and event completion will never arrive, so entry should be released.
Fixes: 73dd3a4839c1 ("net/mlx5: Avoid using pending command interface slots") Signed-off-by: Moshe Shemesh <moshe@mellanox.com> Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
mlx5_cmd_flush() will trigger forced completions to all valid command
entries. Triggered by an asynch event such as fast teardown it can
happen at any stage of the command, including command initialization.
It will trigger forced completion and that can lead to completion on an
uninitialized command entry.
Setting MLX5_CMD_ENT_STATE_PENDING_COMP only after command entry is
initialized will ensure force completion is treated only if command
entry is initialized.
Fixes: 73dd3a4839c1 ("net/mlx5: Avoid using pending command interface slots") Signed-off-by: Moshe Shemesh <moshe@mellanox.com> Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In polling mode, set arm_db member to a value that will avoid CQ
event recovery by the HW.
Otherwise we might get event without completion function.
In addition,empty completion function to was added to protect from
unexpected events.
Fixes: 297cccebdc5a ("net/mlx5: DR, Expose an internal API to issue RDMA operations") Signed-off-by: Erez Shitrit <erezsh@mellanox.com> Reviewed-by: Tariq Toukan <tariqt@mellanox.com> Reviewed-by: Alex Vesker <valex@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The current logic in bnxt_fix_features() will inadvertently turn on both
CTAG and STAG VLAN offload if the user tries to disable both. Fix it
by checking that the user is trying to enable CTAG or STAG before
enabling both. The logic is supposed to enable or disable both CTAG and
STAG together.
Fixes: 5a9f6b238e59 ("bnxt_en: Enable and disable RX CTAG and RX STAG VLAN acceleration together.") Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
bnxt_alloc_ctx_pg_tbls() should return error when the memory size of the
context memory to set up is zero. By returning success (0), the caller
may proceed normally and may crash later when it tries to set up the
memory.
Fixes: 08fe9d181606 ("bnxt_en: Add Level 2 context memory paging support.") Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Improve the slot reset sequence by disabling the device to prevent bad
DMAs if slot reset fails. Return the proper result instead of always
PCI_ERS_RESULT_RECOVERED to the caller.
Fixes: 6316ea6db93d ("bnxt_en: Enable AER support.") Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Broadcom adapters support only maximum of 512 CQs per PF. If user sets
MSIx vectors more than supported CQs, firmware is setting incorrect value
for msix_vec_per_pf_max parameter. Fix it by reducing the BNXT_MSIX_VEC_MAX
value to 512, even though the maximum # of MSIx vectors supported by adapter
are 1280.
Fixes: f399e8497826 ("bnxt_en: Use msix_vec_per_pf_max and msix_vec_per_pf_min devlink params.") Signed-off-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com> Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fix the logic that sets the enable/disable flag for the source MAC
filter according to firmware spec 1.7.1.
In the original firmware spec. before 1.7.1, the VF spoof check flags
were not latched after making the HWRM_FUNC_CFG call, so there was a
need to keep the func_flags so that subsequent calls would perserve
the VF spoof check setting. A change was made in the 1.7.1 spec
so that the flags became latched. So we now set or clear the anti-
spoof setting directly without retrieving the old settings in the
stored vf->func_flags which are no longer valid. We also remove the
unneeded vf->func_flags.
Fixes: 8eb992e876a8 ("bnxt_en: Update firmware interface spec to 1.7.6.2.") Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
RFC 6040 recommends propagating an ECT(1) mark from an outer tunnel header
to the inner header if that inner header is already marked as ECT(0). When
RFC 6040 decapsulation was implemented, this case of propagation was not
added. This simply appears to be an oversight, so let's fix that.
Fixes: eccc1bb8d4b4 ("tunnel: drop packet if ECN present with not-ECT") Reported-by: Bob Briscoe <ietf@bobbriscoe.net> Reported-by: Olivier Tilmans <olivier.tilmans@nokia-bell-labs.com> Cc: Dave Taht <dave.taht@gmail.com> Cc: Stephen Hemminger <stephen@networkplumber.org> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When an application connects to the TIPC topology server and subscribes
to some services, a new connection is created along with some objects -
'tipc_subscription' to store related data correspondingly...
However, there is one omission in the connection handling that when the
connection or application is orderly shutdown (e.g. via SIGQUIT, etc.),
the connection is not closed in kernel, the 'tipc_subscription' objects
are not freed too.
This results in:
- The maximum number of subscriptions (65535) will be reached soon, new
subscriptions will be rejected;
- TIPC module cannot be removed (unless the objects are somehow forced
to release first);
The commit fixes the issue by closing the connection if the 'recvmsg()'
returns '0' i.e. when the peer is shutdown gracefully. It also includes
the other unexpected cases.
Acked-by: Jon Maloy <jmaloy@redhat.com> Acked-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: Tuong Lien <tuong.t.lien@dektech.com.au> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
syzbot managed to set up sfq so that q->scaled_quantum was zero,
triggering an infinite loop in sfq_dequeue()
More generally, we must only accept quantum between 1 and 2^18 - 7,
meaning scaled_quantum must be in [1, 0x7FFF] range.
Otherwise, we also could have a loop in sfq_dequeue()
if scaled_quantum happens to be 0x8000, since slot->allot
could indefinitely switch between 0 and 0x8000.
Fixes: eeaeb068f139 ("sch_sfq: allow big packets and be fair") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot+0251e883fe39e7a0cb0a@syzkaller.appspotmail.com Cc: Jason A. Donenfeld <Jason@zx2c4.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If choke_init() could not allocate q->tab, we would crash later
in choke_reset().
BUG: KASAN: null-ptr-deref in memset include/linux/string.h:366 [inline]
BUG: KASAN: null-ptr-deref in choke_reset+0x208/0x340 net/sched/sch_choke.c:326
Write of size 8 at addr 0000000000000000 by task syz-executor822/7022
Fixes: 77e62da6e60c ("sch_choke: drop all packets in queue during reset") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Cc: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In function nfp_abm_vnic_set_mac, pointer nsp is allocated by nfp_nsp_open.
But when nfp_nsp_has_hwinfo_lookup fail, the pointer is not released,
which can lead to a memory leak bug. Fix this issue by adding
nfp_nsp_close(nsp) in the error path.
Fixes: f6e71efdf9fb1 ("nfp: abm: look up MAC addresses via management FW") Signed-off-by: Qiushi Wu <wu000273@umn.edu> Acked-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add support for Dell Wireless 5816e to drivers/net/usb/qmi_wwan.c
Signed-off-by: Matt Jolly <Kangie@footclan.ninja> Acked-by: Bjørn Mork <bjorn@mork.no> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tls_data_ready() invokes sk_psock_get(), which returns a reference of
the specified sk_psock object to "psock" with increased refcnt.
When tls_data_ready() returns, local variable "psock" becomes invalid,
so the refcount should be decreased to keep refcount balanced.
The reference counting issue happens in one exception handling path of
tls_data_ready(). When "psock->ingress_msg" is empty but "psock" is not
NULL, the function forgets to decrease the refcnt increased by
sk_psock_get(), causing a refcnt leak.
Fix this issue by calling sk_psock_put() on all paths when "psock" is
not NULL.
Signed-off-by: Xiyu Yang <xiyuyang19@fudan.edu.cn> Signed-off-by: Xin Tan <tanxin.ctf@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
bpf_exec_tx_verdict() invokes sk_psock_get(), which returns a reference
of the specified sk_psock object to "psock" with increased refcnt.
When bpf_exec_tx_verdict() returns, local variable "psock" becomes
invalid, so the refcount should be decreased to keep refcount balanced.
The reference counting issue happens in one exception handling path of
bpf_exec_tx_verdict(). When "policy" equals to NULL but "psock" is not
NULL, the function forgets to decrease the refcnt increased by
sk_psock_get(), causing a refcnt leak.
Fix this issue by calling sk_psock_put() on this error path before
bpf_exec_tx_verdict() returns.
Signed-off-by: Xiyu Yang <xiyuyang19@fudan.edu.cn> Signed-off-by: Xin Tan <tanxin.ctf@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 3c1bcc8614db ("net: ethernet: Convert phydev advertize and
supported from u32 to link mode") updated ethernet drivers to use a
linkmode bitmap. It mistakenly dropped a bitwise negation in the
tc35815 ethernet driver on a bitmask to set the supported/advertising
flags.
Found by Anthony via code inspection, not tested as I do not have the
required hardware.
Fixes: 3c1bcc8614db ("net: ethernet: Convert phydev advertize and supported from u32 to link mode") Signed-off-by: Anthony Felice <tony.felice@timesys.com> Reviewed-by: Akshay Bhat <akshay.bhat@timesys.com> Reviewed-by: Heiner Kallweit <hkallweit1@gmail.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Syzkaller again found a path to a kernel crash through bad gso input:
a packet with transport header extending beyond skb_headlen(skb).
Tighten validation at kernel entry:
- Verify that the transport header lies within the linear section.
To avoid pulling linux/tcp.h, verify just sizeof tcphdr.
tcp_gso_segment will call pskb_may_pull (th->doff * 4) before use.
- Match the gso_type against the ip_proto found by the flow dissector.
Fixes: bfd5f4a3d605 ("packet: Add GSO/csum offload support.") Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Read the temperature sensor register from the correct location for the 88E2110 PHY. There is no enable/disable bit on 2110, so make
mv3310_hwmon_config() run on 88X3310 only.
Fixes: 62d01535474b61 ("net: phy: marvell10g: add support for the 88x2110 PHY") Cc: Maxime Chevallier <maxime.chevallier@bootlin.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: Baruch Siach <baruch@tkos.co.il> Reviewed-by: Russell King <rmk+kernel@armlinux.org.uk> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When ENOSPC is set the idx is still valid and gets set to the global
MLX4_SINK_COUNTER_INDEX. However gcc's static analysis cannot tell that
ENOSPC is impossible from mlx4_cmd_imm() and gives this warning:
drivers/net/ethernet/mellanox/mlx4/main.c:2552:28: warning: 'idx' may be
used uninitialized in this function [-Wmaybe-uninitialized]
2552 | priv->def_counter[port] = idx;
Also, when ENOSPC is returned mlx4_allocate_default_counters should not
fail.
Fixes: 6de5f7f6a1fa ("net/mlx4_core: Allocate default counter per port") Signed-off-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Tariq Toukan <tariqt@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
MACsec decryption always occurs in a softirq context. Since
the FPU may not be usable in the softirq context, the call to
decrypt may be scheduled on the cryptd work queue. The cryptd
work queue does not provide ordering guarantees. Therefore,
preserving order requires masking out ASYNC implementations
of gcm(aes).
For instance, an Intel CPU with AES-NI makes available the
generic-gcm-aesni driver from the aesni_intel module to
implement gcm(aes). However, this implementation requires
the FPU, so it is not always available to use from a softirq
context, and will fallback to the cryptd work queue, which
does not preserve frame ordering. With this change, such a
system would select gcm_base(ctr(aes-aesni),ghash-generic).
While the aes-aesni implementation prefers to use the FPU, it
will fallback to the aes-asm implementation if unavailable.
By using a synchronous version of gcm(aes), the decryption
will complete before returning from crypto_aead_decrypt().
Therefore, the macsec_decrypt_done() callback will be called
before returning from macsec_decrypt(). Thus, the order of
calls to macsec_post_decrypt() for the frames is preserved.
While it's presumable that the pure AES-NI version of gcm(aes)
is more performant, the hybrid solution is capable of gigabit
speeds on modest hardware. Regardless, preserving the order
of frames is paramount for many network protocols (e.g.,
triggering TCP retries). Within the MACsec driver itself, the
replay protection is tripped by the out-of-order frames, and
can cause frames to be dropped.
This bug has been present in this code since it was added in
v4.6, however it may not have been noticed since not all CPUs
have FPU offload available. Additionally, the bug manifests
as occasional out-of-order packets that are easily
misattributed to other network phenomena.
When this code was added in v4.6, the crypto/gcm.c code did
not restrict selection of the ghash function based on the
ASYNC flag. For instance, x86 CPUs with PCLMULQDQ would
select the ghash-clmulni driver instead of ghash-generic,
which submits to the cryptd work queue if the FPU is busy.
However, this bug was was corrected in v4.8 by commit b30bdfa86431afbafe15284a3ad5ac19b49b88e3, and was backported
all the way back to the v3.14 stable branch, so this patch
should be applicable back to the v4.6 stable branch.
Signed-off-by: Scott Dial <scott@scottdial.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A call of the function macb_init() can fail in the function
fu540_c000_init. The related system resources were not released
then. use devm_platform_ioremap_resource() to replace ioremap()
to fix it.
Fixes: c218ad559020ff9 ("macb: Add support for SiFive FU540-C000") Cc: Andy Shevchenko <andy.shevchenko@gmail.com> Reviewed-by: Yash Shah <yash.shah@sifive.com> Suggested-by: Nicolas Ferre <nicolas.ferre@microchip.com> Suggested-by: Andy Shevchenko <andy.shevchenko@gmail.com> Signed-off-by: Dejin Zheng <zhengdejin5@gmail.com> Acked-by: Nicolas Ferre <nicolas.ferre@microchip.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Prior to 1d27732f411d ("net: dsa: setup and teardown ports"), we would
not treat failures to set-up an user port as fatal, but after this
commit we would, which is a regression for some systems where interfaces
may be declared in the Device Tree, but the underlying hardware may not
be present (pluggable daughter cards for instance).
Fixes: 1d27732f411d ("net: dsa: setup and teardown ports") Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When ndo_get_phys_port_name() for the CPU port was added we introduced
an early check for when the DSA master network device in
dsa_master_ndo_setup() already implements ndo_get_phys_port_name(). When
we perform the teardown operation in dsa_master_ndo_teardown() we would
not be checking that cpu_dp->orig_ndo_ops was successfully allocated and
non-NULL initialized.
With network device drivers such as virtio_net, this leads to a NPD as
soon as the DSA switch hanging off of it gets torn down because we are
now assigning the virtio_net device's netdev_ops a NULL pointer.
Fixes: da7b9e9b00d4 ("net: dsa: Add ndo_get_phys_port_name() for CPU port") Reported-by: Allen Pais <allen.pais@oracle.com> Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Tested-by: Allen Pais <allen.pais@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
User space can request to delete a range of VLANs from a bridge slave in
one netlink request. For each deleted VLAN the FDB needs to be traversed
in order to flush all the affected entries.
If a large range of VLANs is deleted and the number of FDB entries is
large or the FDB lock is contented, it is possible for the kernel to
loop through the deleted VLANs for a long time. In case preemption is
disabled, this can result in a soft lockup.
Fix this by adding a schedule point after each VLAN is deleted to yield
the CPU, if needed. This is safe because the VLANs are traversed in
process context.
Fixes: bdced7ef7838 ("bridge: support for multiple vlans and vlan ranges in setlink and dellink requests") Signed-off-by: Ido Schimmel <idosch@mellanox.com> Reported-by: Stefan Priebe - Profihost AG <s.priebe@profihost.ag> Tested-by: Stefan Priebe - Profihost AG <s.priebe@profihost.ag> Acked-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When a new neighbor entry has been added, event is generated but it does not
include protocol, because its value is assigned after the event notification
routine has run, so move protocol assignment code earlier.
Fixes: df9b0e30d44c ("neighbor: Add protocol attribute") Cc: David Ahern <dsahern@gmail.com> Signed-off-by: Roman Mashak <mrv@mojatatu.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>