Hypercalls submitted by user space tools via the privcmd driver can
take a long time (potentially many 10s of seconds) if the hypercall
has many sub-operations.
A fully preemptible kernel may deschedule such as task in any upcall
called from a hypercall continuation.
However, in a kernel with voluntary or no preemption, hypercall
continuations in Xen allow event handlers to be run but the task
issuing the hypercall will not be descheduled until the hypercall is
complete and the ioctl returns to user space. These long running
tasks may also trigger the kernel's soft lockup detection.
Add xen_preemptible_hcall_begin() and xen_preemptible_hcall_end() to
bracket hypercalls that may be preempted. Use these in the privcmd
driver.
When returning from an upcall, call xen_maybe_preempt_hcall() which
adds a schedule point if if the current task was within a preemptible
hypercall.
Since _cond_resched() can move the task to a different CPU, clear and
set xen_in_preemptible_hcall around the call.
Signed-off-by: David Vrabel <david.vrabel@citrix.com> Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Nicolas Dichtel [Thu, 16 Jul 2015 09:31:43 +0000 (11:31 +0200)]
rtnl: restore notifications for deleted interfaces
The commit 984ff7a3e060 is an upstream backport. In fact, it depends on
commit 395eea6ccf2b ("rtnetlink: delay RTM_DELLINK notification until after ndo_uninit()")
which has not been backported in 3.18.y.
Before commit 395eea6ccf2b, rollback_registered_many() uses rtmsg_ifinfo().
The call to this function is done with dev->reg_state set to
NETREG_UNREGISTERING, thus testing this reg_state in rtmsg_ifinfo() is
wrong.
Commit e4502c63f56aeca88 (ufs: deal with nfsd/iget races) made ufs
create inodes with I_NEW flag set. However ufs_mkdir() never cleared
this flag. Thus if someone ever tried to lookup the directory by inode
number, he would deadlock waiting for I_NEW to be cleared. Luckily this
mostly happens only if the filesystem is exported over NFS since
otherwise we have the inode attached to dentry and don't look it up by
inode number. In rare cases dentry can get freed without inode being
freed and then we'd hit the deadlock even without NFS export.
Fix the problem by clearing I_NEW before instantiating new directory
inode.
Fixes: e4502c63f56aeca887ced37f24e0def1ef11cec8 Reported-by: Fabian Frederick <fabf@skynet.be> Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Commit e4502c63f56aeca88 (ufs: deal with nfsd/iget races) introduced
unlock_new_inode() call into ufs_add_nondir(). However that function
gets called also from ufs_link() which hands it already initialized
inode and thus unlock_new_inode() complains. The problem is harmless but
annoying.
Fix the problem by opencoding necessary stuff in ufs_link()
Fixes: e4502c63f56aeca887ced37f24e0def1ef11cec8 Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Limit the mounts fs_fully_visible considers to locked mounts.
Unlocked can always be unmounted so considering them adds hassle
but no security benefit.
Eric noticed problems with vhost-scsi and virtio-ccw: vhost-scsi
complained about overwriting values in the config space, which
was triggered by a broken implementation of virtio-ccw's config
get/set routines. It was probably sheer luck that we did not hit
this before.
When writing a value to the config space, the WRITE_CONF ccw will
always write from the beginning of the config space up to and
including the value to be set. If the config space up to the value
has not yet been retrieved from the device, however, we'll end up
overwriting values. Keep track of the known config space and update
if needed to avoid this.
Moreover, READ_CONF will only read the number of bytes it has been
instructed to retrieve, so we must not copy more than that to the
buffer, or we might overwrite trailing values.
Reported-by: Eric Farman <farman@linux.vnet.ibm.com> Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com> Reviewed-by: Eric Farman <farman@linux.vnet.ibm.com> Tested-by: Eric Farman <farman@linux.vnet.ibm.com> Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com> Cc: stable@vger.kernel.org Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Before calling into the filesystem, vfs_setxattr calls
security_inode_setxattr, which ends up calling selinux_inode_setxattr in
our case. That returns -EOPNOTSUPP whenever SBLABEL_MNT is not set.
SBLABEL_MNT was supposed to be set by sb_finish_set_opts, which sets it
only if selinux_is_sblabel_mnt returns true.
The selinux_is_sblabel_mnt logic was broken by eadcabc697e9 "SELinux: do
all flags twiddling in one place", which didn't take into the account
the SECURITY_FS_USE_NATIVE behavior that had been introduced for nfs
with eb9ae686507b "SELinux: Add new labeling type native labels".
This caused setxattr's of security labels over NFSv4.2 to fail.
Cc: stable@kernel.org # 3.13 Cc: Eric Paris <eparis@redhat.com> Cc: David Quigley <dpquigl@davequigley.com> Reported-by: Richard Chan <rc556677@outlook.com> Signed-off-by: J. Bruce Fields <bfields@redhat.com> Acked-by: Stephen Smalley <sds@tycho.nsa.gov>
[PM: added the stable dependency] Signed-off-by: Paul Moore <pmoore@redhat.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
ffs_closed can race with configfs_rmdir which will call config_item_release, so
add an extra check to avoid calling the unregister_gadget_item with an null
gadget item.
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org> Signed-off-by: Felipe Balbi <balbi@ti.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Architectural performance monitoring, version 1, doesn't support fixed counters.
Currently, even if a hypervisor advertises support for architectural
performance monitoring version 1, perf may still try to use the fixed
counters, as the constraints are set up based on the CPU model.
This patch ensures that perf honors the architectural performance monitoring
version returned by CPUID, and it only uses the fixed counters for version 2
and above.
(Some of the ideas in this patch came from Peter Zijlstra.)
Signed-off-by: Imre Palik <imrep@amazon.de> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Anthony Liguori <aliguori@amazon.com> Cc: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Borislav Petkov <bp@alien8.de> Cc: Brian Gerst <brgerst@gmail.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/1433767609-1039-1-git-send-email-imrep.amz@gmail.com Signed-off-by: Ingo Molnar <mingo@kernel.org> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
The warning message in prepend_path is unclear and outdated. It was
added as a warning that the mechanism for generating names of pseudo
files had been removed from prepend_path and d_dname should be used
instead. Unfortunately the warning reads like a general warning,
making it unclear what to do with it.
Remove the warning. The transition it was added to warn about is long
over, and I added code several years ago which in rare cases causes
the warning to fire on legitimate code, and the warning is now firing
and scaring people for no good reason.
Cc: stable@vger.kernel.org Reported-by: Ivan Delalande <colona@arista.com> Reported-by: Omar Sandoval <osandov@osandov.com> Fixes: f48cfddc6729e ("vfs: In d_path don't call d_dname on a mount point") Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Writes were a bit racy, but hard to turn into a bug at the same time.
(Particularly because modern Linux doesn't use this feature anymore.)
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
[Actually the next patch makes it much, much easier to trigger the race
so I'm including this one for stable@ as well. - Paolo] Cc: stable@vger.kernel.org Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
On VM entry, we disable access to the VFP registers in order to
perform a lazy save/restore of these registers.
On VM exit, we restore access, test if we did enable them before,
and save/restore the guest/host registers if necessary. In this
sequence, the FPEXC register is always accessed, irrespective
of the trapping configuration.
If the guest didn't touch the VFP registers, then the HCPTR access
has now enabled such access, but we're missing a barrier to ensure
architectural execution of the new HCPTR configuration. If the HCPTR
access has been delayed/reordered, the subsequent access to FPEXC
will cause a trap, which we aren't prepared to handle at all.
The same condition exists when trapping to enable VFP for the guest.
The fix is to introduce a barrier after enabling VFP access. In the
vmexit case, it can be relaxed to only takes place if the guest hasn't
accessed its view of the VFP registers, making the access to FPEXC safe.
The set_hcptr macro is modified to deal with both vmenter/vmexit and
vmtrap operations, and now takes an optional label that is branched to
when the guest hasn't touched the VFP registers.
Some of the macros defined in kvm_arm.h are useful in assembly files, but are
not compatible with the assembler. Change any C language integer constant
definitions using appended U, UL, or ULL to the UL() preprocessor macro. Also,
add a preprocessor include of the asm/memory.h file which defines the UL()
macro.
Fixes build errors like these when using kvm_arm.h in assembly
source files:
Error: unexpected characters following instruction at operand 3 -- `and x0,x1,#((1U<<25)-1)'
Acked-by: Mark Rutland <mark.rutland@arm.com> Signed-off-by: Geoff Levand <geoff@infradead.org> Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
If hardware doesn't support DecodeAssist - a feature that provides
more information about the intercept in the VMCB, KVM decodes the
instruction and then updates the next_rip vmcb control field.
However, NRIP support itself depends on cpuid Fn8000_000A_EDX[NRIPS].
Since skip_emulated_instruction() doesn't verify nrip support
before accepting control.next_rip as valid, avoid writing this
field if support isn't present.
Signed-off-by: Bandan Das <bsd@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Replacing a xattr consists of doing a lookup for its existing value, delete
the current value from the respective leaf, release the search path and then
finally insert the new value. This leaves a time window where readers (getxattr,
listxattrs) won't see any value for the xattr. Xattrs are used to store ACLs,
so this has security implications.
This change also fixes 2 other existing issues which were:
*) Deleting the old xattr value without verifying first if the new xattr will
fit in the existing leaf item (in case multiple xattrs are packed in the
same item due to name hash collision);
*) Returning -EEXIST when the flag XATTR_CREATE is given and the xattr doesn't
exist but we have have an existing item that packs muliple xattrs with
the same name hash as the input xattr. In this case we should return ENOSPC.
A test case for xfstests follows soon.
Thanks to Alexandre Oliva for reporting the non-atomicity of the xattr replace
implementation.
Reported-by: Alexandre Oliva <oliva@gnu.org> Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: Chris Mason <clm@fb.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
mc_saved_tmp is a static array allocated on the stack, we need to make
sure mc_saved_count stays within its bounds, otherwise we're overflowing
the stack in _save_mc(). A specially crafted microcode header could lead
to a kernel crash or potentially kernel execution.
nfnl_cthelper_parse_tuple() is called from nfnl_cthelper_new(),
nfnl_cthelper_get() and nfnl_cthelper_del(). In each case they pass
a pointer to an nf_conntrack_tuple data structure local variable:
struct nf_conntrack_tuple tuple;
...
ret = nfnl_cthelper_parse_tuple(&tuple, tb[NFCTH_TUPLE]);
The problem is that this local variable is not initialized, and
nfnl_cthelper_parse_tuple() only initializes two fields: src.l3num and
dst.protonum. This leaves all other fields with undefined values
based on whatever is on the stack:
The symptom observed was that when the rpc and tns helpers were added
then traffic to port 1536 was being sent to user-space.
Signed-off-by: Ian Wilson <iwilson@brocade.com> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Worse yet, reading the error message (the filter again) it says that
there was no error, when there clearly was. The issue is that the
code that checks the input does not check for balanced ops. That is,
having an op between a closed parenthesis and the next token.
This would only cause a warning, and fail out before doing any real
harm, but it should still not caues a warning, and the error reported
should work:
file_remove_suid() could mistakenly set S_NOSEC inode bit when root was
modifying the file. As a result following writes to the file by ordinary
user would avoid clearing suid or sgid bits.
Fix the bug by checking actual mode bits before setting S_NOSEC.
CC: stable@vger.kernel.org Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
commit 514ac99c64b "can: fix multiple delivery of a single CAN frame for
overlapping CAN filters" requires the skb->tstamp to be set to check for
identical CAN skbs.
As net timestamping is influenced by several players (netstamp_needed and
netdev_tstamp_prequeue) Manfred missed a proper timestamp which leads to
CAN frame loss.
As skb timestamping became now mandatory for CAN related skbs this patch
makes sure that received CAN skbs always have a proper timestamp set.
Maybe there's a better solution in the future but this patch fixes the
CAN frame loss so far.
Reported-by: Manfred Schlaegl <manfred.schlaegl@gmx.at> Signed-off-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: Sasha Levin <sasha.levin@oracle.com>
Commit 0244756edc4b98c ("ufs: sb mutex merge + mutex_destroy") generated
deadlocks in read/write mode on mkdir.
This patch partially reverts it keeping fixes by Andrew Morton and
mutex_destroy()
[AV: fixed a missing bit in ufs_remount()]
Signed-off-by: Fabian Frederick <fabf@skynet.be> Reported-by: Ian Campbell <ian.campbell@citrix.com> Suggested-by: Jan Kara <jack@suse.cz> Cc: Ian Campbell <ian.campbell@citrix.com> Cc: Evgeniy Dushistov <dushistov@mail.ru> Cc: Alexey Khoroshilov <khoroshilov@ispras.ru> Cc: Roger Pau Monne <roger.pau@citrix.com> Cc: Ian Jackson <Ian.Jackson@eu.citrix.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
This reverts commit 9ef7db7f38d0 ("ufs: fix deadlocks introduced by sb
mutex merge") That patch tried to solve commit 0244756edc4b98c ("ufs: sb
mutex merge + mutex_destroy") which is itself partially reverted due to
multiple deadlocks.
Signed-off-by: Fabian Frederick <fabf@skynet.be> Suggested-by: Jan Kara <jack@suse.cz> Cc: Ian Campbell <ian.campbell@citrix.com> Cc: Evgeniy Dushistov <dushistov@mail.ru> Cc: Alexey Khoroshilov <khoroshilov@ispras.ru> Cc: Roger Pau Monne <roger.pau@citrix.com> Cc: Ian Jackson <Ian.Jackson@eu.citrix.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
The related code can be simplified, and also can avoid related warnings
(with allmodconfig under parisc):
CC [M] net/netfilter/nfnetlink_cthelper.o
net/netfilter/nfnetlink_cthelper.c: In function ‘nfnl_cthelper_from_nlattr’:
net/netfilter/nfnetlink_cthelper.c:97:9: warning: passing argument 1 o ‘memcpy’ discards ‘const’ qualifier from pointer target type [-Wdiscarded-array-qualifiers]
memcpy(&help->data, nla_data(attr), help->helper->data_len);
^
In file included from include/linux/string.h:17:0,
from include/uapi/linux/uuid.h:25,
from include/linux/uuid.h:23,
from include/linux/mod_devicetable.h:12,
from ./arch/parisc/include/asm/hardware.h:4,
from ./arch/parisc/include/asm/processor.h:15,
from ./arch/parisc/include/asm/spinlock.h:6,
from ./arch/parisc/include/asm/atomic.h:21,
from include/linux/atomic.h:4,
from ./arch/parisc/include/asm/bitops.h:12,
from include/linux/bitops.h:36,
from include/linux/kernel.h:10,
from include/linux/list.h:8,
from include/linux/module.h:9,
from net/netfilter/nfnetlink_cthelper.c:11:
./arch/parisc/include/asm/string.h:8:8: note: expected ‘void *’ but argument is of type ‘const char (*)[]’
void * memcpy(void * dest,const void *src,size_t count);
^
Signed-off-by: Chen Gang <gang.chen.5i5j@gmail.com> Signed-off-by: Pablo Neira Ayuso <pablo@soleta.eu> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
A huge amount of NIC drivers use the DMA API, however if
compiled under 32-bit an very important part of the DMA API can
be ommitted leading to the drivers not working at all
(especially if used with 'swiotlb=force iommu=soft').
As Prashant Sreedharan explains it: "the driver [tg3] uses
DEFINE_DMA_UNMAP_ADDR(), dma_unmap_addr_set() to keep a copy of
the dma "mapping" and dma_unmap_addr() to get the "mapping"
value. On most of the platforms this is a no-op, but ... with
"iommu=soft and swiotlb=force" this house keeping is required,
... otherwise we pass 0 while calling pci_unmap_/pci_dma_sync_
instead of the DMA address."
As such enable this even when using 32-bit kernels.
Reported-by: Ian Jackson <Ian.Jackson@eu.citrix.com> Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Acked-by: David S. Miller <davem@davemloft.net> Acked-by: Prashant Sreedharan <prashant@broadcom.com> Cc: Borislav Petkov <bp@alien8.de> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Michael Chan <mchan@broadcom.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: boris.ostrovsky@oracle.com Cc: cascardo@linux.vnet.ibm.com Cc: david.vrabel@citrix.com Cc: sanjeevb@broadcom.com Cc: siva.kallam@broadcom.com Cc: vyasevich@gmail.com Cc: xen-devel@lists.xensource.com Link: http://lkml.kernel.org/r/20150417190448.GA9462@l.oracle.com Signed-off-by: Ingo Molnar <mingo@kernel.org> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
On x86-64, __copy_instruction() always returns 0 (error) if the
instruction uses %rip-relative addressing. This is because
kernel_insn_init() is called the second time for 'insn' instance
in such cases and sets all its fields to 0.
Because of this, trying to place a kprobe on such instruction
will fail, register_kprobe() will return -EINVAL.
Ignore an existing mount if the locked readonly, nodev or atime
attributes are less permissive than the desired attributes
of the new mount.
On success ensure the new mount locks all of the same readonly, nodev and
atime attributes as the old mount.
The nosuid and noexec attributes are not checked here as this change
is destined for stable and enforcing those attributes causes a
regression in lxc and libvirt-lxc where those applications will not
start and there are no known executables on sysfs or proc and no known
way to create exectuables without code modifications
Cc: stable@vger.kernel.org Fixes: e51db73532955 ("userns: Better restrictions on when proc and sysfs can be mounted") Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
In RS485 mode, we may want to set the delay_rts_after_send value to 0.
In the datasheet, the 0 value is said to "disable" the Transmitter Timeguard but
this is exactly the expected behavior if we want no delay...
Moreover, if the value was set to non-zero value by device-tree or earlier
ioctl command, it was impossible to change it back to zero.
Reported-by: Sami Pietikäinen <Sami.Pietikainen@wapice.com> Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: stable@vger.kernel.org # 3.2+ Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Fresh mounts of proc and sysfs are a very special case that works very
much like a bind mount. Unfortunately the current structure can not
preserve the MNT_LOCK... mount flags. Therefore refactor the logic
into a form that can be modified to preserve those lock bits.
Add a new filesystem flag FS_USERNS_VISIBLE that requires some mount
of the filesystem be fully visible in the current mount namespace,
before the filesystem may be mounted.
Move the logic for calling fs_fully_visible from proc and sysfs into
fs/namespace.c where it has greater access to mount namespace state.
In commit 5491ce3f79ee ("mmc: sdhci-pxav3: add support for the Armada
38x SDHCI controller"), the sdhci-pxav3 driver was extended to include
support for the SDHCI controller found in the Armada 38x
processor. This mainly involved adding some MBus window related
configuration.
However, this configuration is currently done too early in ->probe():
it is done before clocks are enabled, while this configuration
involves touching the registers of the controller, which will hang the
SoC if the clock is disabled. It wasn't noticed until now because the
bootloader typically leaves gatable clocks enabled, but in situations
where we have a deferred probe (due to a CD GPIO that cannot be taken,
for example), then the probe will be re-tried later, after a clock
disable has been done in the exit path of the failed probe attempt of
the device. This second probe() will hang the system due to the clock
being disabled.
This can for example be produced on Armada 385 GP, which has a CD GPIO
connected to an I2C PCA9555. If the driver for the PCA9555 is not
compiled into the kernel, then we will have the following sequence of
events:
1. The SDHCI probes
2. It does the MBus configuration (which works, because the clock is
left enabled by the bootloader)
3. It enables the clock
4. It tries to get the CD GPIO, which fails due to the driver being
missing, so -EPROBE_DEFER is returned.
5. Before returning -EPROBE_DEFER, the driver cleans up what was
done, which includes disabling the clock.
6. Later on, the SDHCI probe is tried again.
7. It does the MBus configuration, which hangs because the clock is
no longer enabled.
This commit does the obvious fix of doing the MBus configuration after
the clock has been enabled by the driver.
Fixes: 5491ce3f79ee ("mmc: sdhci-pxav3: add support for the Armada 38x SDHCI controller") Cc: <stable@vger.kernel.org> # v3.15+ Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
[jogo: rebased onto 3.18.17] Signed-off-by: Jonas Gorski <jogo@openwrt.org> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
There is NULL pointer dereference possible during statistics update if the route
used for OOTB responce is removed at unfortunate time. If the route exists when
we receive OOTB packet and we finally jump into sctp_packet_transmit() to send
ABORT, but in the meantime route is removed under our feet, we take "no_route"
path and try to update stats with IP_INC_STATS(sock_net(asoc->base.sk), ...).
But sctp_ootb_pkt_new() used to prepare responce packet doesn't call
sctp_transport_set_owner() and therefore there is no asoc associated with this
packet. Probably temporary asoc just for OOTB responces is overkill, so just
introduce a check like in all other places in sctp_packet_transmit(), where
"asoc" is dereferenced.
To reproduce this, one needs to
0. ensure that sctp module is loaded (otherwise ABORT is not generated)
1. remove default route on the machine
2. while true; do
ip route del [interface-specific route]
ip route add [interface-specific route]
done
3. send enough OOTB packets (i.e. HB REQs) from another host to trigger ABORT
responce
As bnx2x_init_ptp() is only called if bp->flags contains PTP_SUPPORTED,
we also need to guard bnx2x_stop_ptp() with same condition, otherwise
ptp_task workqueue is not initialized and kernel barfs on
cancel_work_sync()
Fixes: eeed018cbfa30 ("bnx2x: Add timestamping and PTP hardware clock support") Reported-by: Michel Lespinasse <walken@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Michal Kalderon <Michal.Kalderon@qlogic.com> Cc: Ariel Elior <Ariel.Elior@qlogic.com> Cc: Yuval Mintz <Yuval.Mintz@qlogic.com> Cc: David Decotigny <decot@google.com> Acked-by: Sony Chacko <sony.chacko@qlogic.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
When limiting phy link speed using "max-speed" to 100mbps or less on a
giga bit phy, phy never completes auto negotiation and phy state
machine is held in PHY_AN. Fixing this issue by comparing the giga
bit advertise though phydev->supported doesn't have it but phy has
BMSR_ESTATEN set. So that auto negotiation is restarted as old and
new advertise are different and link comes up fine.
Signed-off-by: Mugunthan V N <mugunthanvnm@ti.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Indication of a single completed packet, marked by txbbs_skipped
being bigger then zero, in not enough in order to wake up a
stopped TX queue. The completed packet may contain a single TXBB,
while next packet to be sent (after the wake up) may have multiple
TXBBs (LSO/TSO packets for example), causing overflow in queue followed
by WQE corruption and TX queue timeout.
Instead, wake the stopped queue only when there's enough room for the
worst case (maximum sized WQE) packet that we should need to handle after
the queue is opened again.
Also created an helper routine - mlx4_en_is_tx_ring_full, which checks
if the current TX ring is full or not. It provides better code readability
and removes code duplication.
Signed-off-by: Ido Shamay <idos@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
tcp_fastopen_reset_cipher really cannot be called from interrupt
context. It allocates the tcp_fastopen_context with GFP_KERNEL and
calls crypto_alloc_cipher, which allocates all kind of stuff with
GFP_KERNEL.
Thus, we might sleep when the key-generation is triggered by an
incoming TFO cookie-request which would then happen in interrupt-
context, as shown by enabling CONFIG_DEBUG_ATOMIC_SLEEP:
This patch moves the call to tcp_fastopen_init_key_once to the places
where a listener socket creates its TFO-state, which always happens in
user-context (either from the setsockopt, or implicitly during the
listen()-call)
Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: Hannes Frederic Sowa <hannes@stressinduktion.org> Fixes: 222e83d2e0ae ("tcp: switch tcp_fastopen key generation to net_get_random_once") Signed-off-by: Christoph Paasch <cpaasch@apple.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
The lockless lookups can return entry that is unlinked.
Sometimes they get reference before last neigh_cleanup_and_release,
sometimes they do not need reference. Later, any
modification attempts may result in the following problems:
1. entry is not destroyed immediately because neigh_update
can start the timer for dead entry, eg. on change to NUD_REACHABLE
state. As result, entry lives for some time but is invisible
and out of control.
2. __neigh_event_send can run in parallel with neigh_destroy
while refcnt=0 but if timer is started and expired refcnt can
reach 0 for second time leading to second neigh_destroy and
possible crash.
Thanks to Eric Dumazet and Ying Xue for their work and analyze
on the __neigh_event_send change.
Fixes: 767e97e1e0db ("neigh: RCU conversion of struct neighbour") Fixes: a263b3093641 ("ipv4: Make neigh lookups directly in output packet path.") Fixes: 6fd6ce2056de ("ipv6: Do not depend on rt->n in ip6_finish_output2().") Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: Ying Xue <ying.xue@windriver.com> Signed-off-by: Julian Anastasov <ja@ssi.bg> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
PACKET_FANOUT_LB computes f->rr_cur such that it is modulo
f->num_members. It returns the old value unconditionally, but
f->num_members may have changed since the last store. Ensure
that the return value is always < num.
When modifying the logic, simplify it further by replacing the loop
with an unconditional atomic increment.
Fixes: dc99f600698d ("packet: Add fanout support.") Suggested-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Willem de Bruijn <willemb@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
We need to tell compiler it must not read f->num_members multiple
times. Otherwise testing if num is not zero is flaky, and we could
attempt an invalid divide by 0 in fanout_demux_cpu()
Note bug was present in packet_rcv_fanout_hash() and
packet_rcv_fanout_lb() but final 3.1 had a simple location
after commit 95ec3eb417115fb ("packet: Add 'cpu' fanout policy.")
Fixes: dc99f600698dc ("packet: Add fanout support.") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
After the ->set() spinlocks were removed br_stp_set_bridge_priority
was left running without any protection when used via sysfs. It can
race with port add/del and could result in use-after-free cases and
corrupted lists. Tested by running port add/del in a loop with stp
enabled while setting priority in a loop, crashes are easily
reproducible.
The spinlocks around sysfs ->set() were removed in commit: 14f98f258f19 ("bridge: range check STP parameters")
There's also a race condition in the netlink priority support that is
fixed by this change, but it was introduced recently and the fixes tag
covers it, just in case it's needed the commit is: af615762e972 ("bridge: add ageing_time, stp_state, priority over netlink")
Signed-off-by: Nikolay Aleksandrov <razor@blackwall.org> Fixes: 14f98f258f19 ("bridge: range check STP parameters") Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
->auto_asconf_splist is per namespace and mangled by functions like
sctp_setsockopt_auto_asconf() which doesn't guarantee any serialization.
Also, the call to inet_sk_copy_descendant() was backuping
->auto_asconf_list through the copy but was not honoring
->do_auto_asconf, which could lead to list corruption if it was
different between both sockets.
This commit thus fixes the list handling by using ->addr_wq_lock
spinlock to protect the list. A special handling is done upon socket
creation and destruction for that. Error handlig on sctp_init_sock()
will never return an error after having initialized asconf, so
sctp_destroy_sock() can be called without addrq_wq_lock. The lock now
will be take on sctp_close_sock(), before locking the socket, so we
don't do it in inverse order compared to sctp_addr_wq_timeout_handler().
Instead of taking the lock on sctp_sock_migrate() for copying and
restoring the list values, it's preferred to avoid rewritting it by
implementing sctp_copy_descendant().
Issue was found with a test application that kept flipping sysctl
default_auto_asconf on and off, but one could trigger it by issuing
simultaneous setsockopt() calls on multiple sockets or by
creating/destroying sockets fast enough. This is only triggerable
locally.
Fixes: 9f7d653b67ae ("sctp: Add Auto-ASCONF support (core).") Reported-by: Ji Jianwen <jiji@redhat.com> Suggested-by: Neil Horman <nhorman@tuxdriver.com> Suggested-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
We saw excessive direct memory compaction triggered by skb_page_frag_refill.
This causes performance issues and add latency. Commit 5640f7685831e0
introduces the order-3 allocation. According to the changelog, the order-3
allocation isn't a must-have but to improve performance. But direct memory
compaction has high overhead. The benefit of order-3 allocation can't
compensate the overhead of direct memory compaction.
This patch makes the order-3 page allocation atomic. If there is no memory
pressure and memory isn't fragmented, the alloction will still success, so we
don't sacrifice the order-3 benefit here. If the atomic allocation fails,
direct memory compaction will not be triggered, skb_page_frag_refill will
fallback to order-0 immediately, hence the direct memory compaction overhead is
avoided. In the allocation failure case, kswapd is waken up and doing
compaction, so chances are allocation could success next time.
alloc_skb_with_frags is the same.
The mellanox driver does similar thing, if this is accepted, we must fix
the driver too.
V3: fix the same issue in alloc_skb_with_frags as pointed out by Eric
V2: make the changelog clearer
Cc: Eric Dumazet <edumazet@google.com> Cc: Chris Mason <clm@fb.com> Cc: Debabrata Banerjee <dbavatar@gmail.com> Signed-off-by: Shaohua Li <shli@fb.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Since the addition of sysfs multicast router support if one set
multicast_router to "2" more than once, then the port would be added to
the hlist every time and could end up linking to itself and thus causing an
endless loop for rlist walkers.
So to reproduce just do:
echo 2 > multicast_router; echo 2 > multicast_router;
in a bridge port and let some igmp traffic flow, for me it hangs up
in br_multicast_flood().
Fix this by adding a check in br_multicast_add_router() if the port is
already linked.
The reason this didn't happen before the addition of multicast_router
sysfs entries is because there's a !hlist_unhashed check that prevents
it.
Signed-off-by: Nikolay Aleksandrov <razor@blackwall.org> Fixes: 0909e11758bd ("bridge: Add multicast_router sysfs entries") Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Since it is possible for vnet_event_napi to end up doing
vnet_control_pkt_engine -> ... -> vnet_send_attr ->
vnet_port_alloc_tx_ring -> ldc_alloc_exp_dring -> kzalloc()
(i.e., in softirq context), kzalloc() should be called with
GFP_ATOMIC from ldc_alloc_exp_dring.
The mcp3021 scaling code is dividing the VDD (full-scale) value in
millivolts by the A2D resolution to obtain the scaling factor. When VDD
is 3300mV (the standard value) and the resolution is 12-bit (4096
divisions), the result is a scale factor of 3300/4096, which is always
one. Effectively, the raw A2D reading is always being returned because
no scaling is applied.
This patch fixes the issue and simplifies the register-to-volts
calculation, removing the unneeded "output_scale" struct member.
The problem is almost exactly the same as the one fixed by feaff8e5b2cf.
We must take a reference to the struct file when running the LOCKU
compound to prevent the final fput from running until the operation is
complete.
Reported-by: Jean Spector <jean@primarydata.com> Signed-off-by: Jeff Layton <jeff.layton@primarydata.com> Cc: stable@vger.kernel.org Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
"Running locktests with a large number of iterations resulted in a
client crash. The test run took a while and hasn't finished after close
to 2 hours. The crash happened right after I gave up and killed the test
(after 107m) with Ctrl+C."
The crash happened because a NULL inode pointer got passed into
locks_get_lock_context. The call chain indicates that file_inode(filp)
returned NULL, which means that f_inode was NULL. Since that's zeroed
out in __fput, that suggests that this filp pointer outlived the last
reference.
Looking at the code, that seems possible. We copy the struct file_lock
that's passed in, but if the task is signalled at an inopportune time we
can end up trying to use that file_lock in rpciod context after the process
that requested it has already returned (and possibly put its filp
reference).
Fix this by taking an extra reference to the filp when we allocate the
lock info, and put it in nfs4_lock_release.
Reported-by: Jean Spector <jean@primarydata.com> Signed-off-by: Jeff Layton <jeff.layton@primarydata.com> Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
rbd_obj_request_create() is called on the main I/O path, so we need to
use GFP_NOIO to make sure allocation doesn't blow back on us. Not all
callers need this, but I'm still hardcoding the flag inside rather than
making it a parameter because a) this is going to stable, and b) those
callers shouldn't really use rbd_obj_request_create() and will be fixed
in the future.
The Ethernet controller found in the Armada 370, 380 and 385 SoCs don't
support TCP/IP checksumming with frame sizes larger than 1600 bytes.
This patch fixes the issue by disabling the features NETIF_F_IP_CSUM and
NETIF_F_TSO for the Armada 370 and compatibles SoCs when the MTU is set
to a value greater than 1600 bytes.
Signed-off-by: Simon Guinot <simon.guinot@sequanux.org> Fixes: c5aff18204da ("net: mvneta: driver for Marvell Armada 370/XP network unit") Cc: <stable@vger.kernel.org> # v3.8+ Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
This patch updates the Ethernet DT nodes for Armada XP SoCs with the
compatible string "marvell,armada-xp-neta".
Signed-off-by: Simon Guinot <simon.guinot@sequanux.org> Fixes: 77916519cba3 ("arm: mvebu: Armada XP MV78230 has only three Ethernet interfaces") Cc: <stable@vger.kernel.org> # v3.8+ Acked-by: Gregory CLEMENT <gregory.clement@free-electrons.com> Reviewed-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
The mvneta driver supports the Ethernet IP found in the Armada 370, XP,
380 and 385 SoCs. Since at least one more hardware feature is available
for the Armada XP SoCs then a way to identify them is needed.
This patch introduces a new compatible string "marvell,armada-xp-neta".
Signed-off-by: Simon Guinot <simon.guinot@sequanux.org> Fixes: c5aff18204da ("net: mvneta: driver for Marvell Armada 370/XP network unit") Cc: <stable@vger.kernel.org> # v3.8+ Acked-by: Gregory CLEMENT <gregory.clement@free-electrons.com> Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
In order for hibernation to reliably work we need to properly turn
off the SDMA block, sadly after numerous attemps i haven't not found
proper sequence for clean and full shutdown. So simply reset both
SDMA block, this makes hibernation works reliably on sea island GPU
family (CI)
Hibernation and suspend to ram were tested (several times) on :
Bonaire
Hawaii
Mullins
Kaveri
Kabini
Cc: stable@vger.kernel.org Signed-off-by: Jérôme Glisse <jglisse@redhat.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
In order for hibernation to reliably work we need to cleanup more
thoroughly the compute ring. Hibernation is different from suspend
resume as when we resume from hibernation the hardware is first
fully initialize by regular kernel then freeze callback happens
(which correspond to a suspend inside the radeon kernel driver)
and turn off each of the block. It turns out we were not cleanly
shutting down the compute ring. This patch fix that.
Hibernation and suspend to ram were tested (several times) on :
Bonaire
Hawaii
Mullins
Kaveri
Kabini
Changed since v1:
- Factor the ring stop logic into a function taking ring as arg.
Cc: stable@vger.kernel.org Signed-off-by: Jérôme Glisse <jglisse@redhat.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
In needs_ilk_vtd_wa(), we pass in the GPU device but compared it against
the ids for the mobile GPU and the mobile host bridge. That latter is
impossible and so likely was just a typo for the desktop GPU device id
(which is also buggy).
drm/i915: Disable WC PTE updates to w/a buggy IOMMU on ILK
Reported-by: Ting-Wei Lan <lantw44@gmail.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91127
References: https://bugzilla.freedesktop.org/show_bug.cgi?id=60391 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: stable@vger.kernel.org Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Fujitsu Lifebook E780 sets the sequence number 0x0f to only only of
the two headphones, thus the driver tries to assign another as the
line-out, and this results in the inconsistent mapping between the
created jack ctl and the actual I/O. Due to this, PulseAudio doesn't
handle it properly and gets the silent output.
Acer Aspire V5 with ALC282 codec needs the similar quirk like Dell
laptops to support the headset mic. The headset mic pin is 0x19 and
it's not exposed by BIOS, thus we need to fix the pincfg as well.
Whilst testing cpu hotplug events on kernel configured with
DEBUG_PREEMPT and DEBUG_ATOMIC_SLEEP we get following BUG message,
caused by calling request_irq() and free_irq() in the context of
hotplug notification (which is in this case atomic context).
[ 40.785859] CPU1: Software reset
[ 40.786660] BUG: sleeping function called from invalid context at mm/slub.c:1241
[ 40.786668] in_atomic(): 1, irqs_disabled(): 128, pid: 0, name: swapper/1
[ 40.786678] Preemption disabled at:[< (null)>] (null)
[ 40.786681]
[ 40.786692] CPU: 1 PID: 0 Comm: swapper/1 Not tainted 3.19.0-rc4-00024-g7dca860 #36
[ 40.786698] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[ 40.786728] [<c0014a00>] (unwind_backtrace) from [<c0011980>] (show_stack+0x10/0x14)
[ 40.786747] [<c0011980>] (show_stack) from [<c0449ba0>] (dump_stack+0x70/0xbc)
[ 40.786767] [<c0449ba0>] (dump_stack) from [<c00c6124>] (kmem_cache_alloc+0xd8/0x170)
[ 40.786785] [<c00c6124>] (kmem_cache_alloc) from [<c005d6f8>] (request_threaded_irq+0x64/0x128)
[ 40.786804] [<c005d6f8>] (request_threaded_irq) from [<c0350b8c>] (exynos4_local_timer_setup+0xc0/0x13c)
[ 40.786820] [<c0350b8c>] (exynos4_local_timer_setup) from [<c0350ca8>] (exynos4_mct_cpu_notify+0x30/0xa8)
[ 40.786838] [<c0350ca8>] (exynos4_mct_cpu_notify) from [<c003b330>] (notifier_call_chain+0x44/0x84)
[ 40.786857] [<c003b330>] (notifier_call_chain) from [<c0022fd4>] (__cpu_notify+0x28/0x44)
[ 40.786873] [<c0022fd4>] (__cpu_notify) from [<c0013714>] (secondary_start_kernel+0xec/0x150)
[ 40.786886] [<c0013714>] (secondary_start_kernel) from [<40008764>] (0x40008764)
Interrupts cannot be requested/freed in the CPU_STARTING/CPU_DYING
notifications which run on the hotplugged cpu with interrupts and
preemption disabled.
To avoid the issue, request the interrupts for all possible cpus in
the boot code. The interrupts are marked NO_AUTOENABLE to avoid a racy
request_irq/disable_irq() sequence. The flag prevents the
request_irq() code from enabling the interrupt immediately.
The interrupt is then enabled in the CPU_STARTING notifier of the
hotplugged cpu and again disabled with disable_irq_nosync() in the
CPU_DYING notifier.
[ tglx: Massaged changelog to match the patch ]
Fixes: 7114cd749a12 ("clocksource: exynos_mct: use (request/free)_irq calls for local timer registration") Reported-by: Krzysztof Kozlowski <k.kozlowski@samsung.com> Reviewed-by: Krzysztof Kozlowski <k.kozlowski@samsung.com> Tested-by: Krzysztof Kozlowski <k.kozlowski@samsung.com> Tested-by: Marcin Jabrzyk <m.jabrzyk@samsung.com> Signed-off-by: Damian Eppel <d.eppel@samsung.com> Cc: m.szyprowski@samsung.com Cc: kyungmin.park@samsung.com Cc: daniel.lezcano@linaro.org Cc: kgene@kernel.org Cc: linux-arm-kernel@lists.infradead.org Link: http://lkml.kernel.org/r/1435324984-7328-1-git-send-email-d.eppel@samsung.com Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: <stable@vger.kernel.org> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
The filter_parse() function can call infix_get_op() which calls
infix_advance() that updates the infix filter pointers for the cnt
and tail without checking if the filter is already at the end, which
will put the cnt to zero and the tail beyond the end. The loop then calls
infix_next() that has
The cnt will now be below zero, and the tail that is returned is
already passed the end of the filter string. So far the allocation
of the filter string usually has some buffer that is zeroed out, but
if the filter string is of the exact size of the allocated buffer
there's no guarantee that the charater after the nul terminating
character will be zero.
The REGSET_VX_LOW ELF notes should contain the lower 64 bit halfes of the
first sixteen 128 bit vector registers. Unfortunately currently we copy
the upper halfes.
Fix this and correctly copy the lower halfes.
Fixes: a62bc0739253 ("s390/kdump: add support for vector extension") Cc: stable@vger.kernel.org # 3.18+ Signed-off-by: Michael Holzheu <holzheu@linux.vnet.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Thinkpad X250, when attached to a dock, has two headphone outs but
no line out. Make sure we don't try to turn this into one headphone
and one line out (since that disables the headphone amp on the dock).
Beginning at commit d52d3997f843 ("ipv6: Create percpu rt6_info"), the
following INFO splat is logged:
===============================
[ INFO: suspicious RCU usage. ]
4.1.0-rc7-next-20150612 #1 Not tainted
-------------------------------
kernel/sched/core.c:7318 Illegal context switch in RCU-bh read-side critical section!
other info that might help us debug this:
rcu_scheduler_active = 1, debug_locks = 0
3 locks held by systemd/1:
#0: (rtnl_mutex){+.+.+.}, at: [<ffffffff815f0c8f>] rtnetlink_rcv+0x1f/0x40
#1: (rcu_read_lock_bh){......}, at: [<ffffffff816a34e2>] ipv6_add_addr+0x62/0x540
#2: (addrconf_hash_lock){+...+.}, at: [<ffffffff816a3604>] ipv6_add_addr+0x184/0x540
stack backtrace:
CPU: 0 PID: 1 Comm: systemd Not tainted 4.1.0-rc7-next-20150612 #1
Hardware name: TOSHIBA TECRA A50-A/TECRA A50-A, BIOS Version 4.20 04/17/2014
Call Trace:
dump_stack+0x4c/0x6e
lockdep_rcu_suspicious+0xe7/0x120
___might_sleep+0x1d5/0x1f0
__might_sleep+0x4d/0x90
kmem_cache_alloc+0x47/0x250
create_object+0x39/0x2e0
kmemleak_alloc_percpu+0x61/0xe0
pcpu_alloc+0x370/0x630
Additional backtrace lines are truncated. In addition, the above splat
is followed by several "BUG: sleeping function called from invalid
context at mm/slub.c:1268" outputs. As suggested by Martin KaFai Lau,
these are the clue to the fix. Routine kmemleak_alloc_percpu() always
uses GFP_KERNEL for its allocations, whereas it should follow the gfp
from its callers.
Reviewed-by: Catalin Marinas <catalin.marinas@arm.com> Reviewed-by: Kamalesh Babulal <kamalesh@linux.vnet.ibm.com> Acked-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net> Cc: Martin KaFai Lau <kafai@fb.com> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Tejun Heo <tj@kernel.org> Cc: Christoph Lameter <cl@linux-foundation.org> Cc: <stable@vger.kernel.org> [3.18+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
The kmemleak scanning thread can run for minutes. Callbacks like
kmemleak_free() are allowed during this time, the race being taken care
of by the object->lock spinlock. Such lock also prevents a memory block
from being freed or unmapped while it is being scanned by blocking the
kmemleak_free() -> ... -> __delete_object() function until the lock is
released in scan_object().
When a kmemleak error occurs (e.g. it fails to allocate its metadata),
kmemleak_enabled is set and __delete_object() is no longer called on
freed objects. If kmemleak_scan is running at the same time,
kmemleak_free() no longer waits for the object scanning to complete,
allowing the corresponding memory block to be freed or unmapped (in the
case of vfree()). This leads to kmemleak_scan potentially triggering a
page fault.
This patch separates the kmemleak_free() enabling/disabling from the
overall kmemleak_enabled nob so that we can defer the disabling of the
object freeing tracking until the scanning thread completed. The
kmemleak_free_part() is deliberately ignored by this patch since this is
only called during boot before the scanning thread started.
- arch_spin_lock/unlock were lacking the ACQUIRE/RELEASE barriers
Since ARCv2 only provides load/load, store/store and all/all, we need
the full barrier
- LLOCK/SCOND based atomics, bitops, cmpxchg, which return modified
values were lacking the explicit smp barriers.
- Non LLOCK/SCOND varaints don't need the explicit barriers since that
is implicity provided by the spin locks used to implement the
critical section (the spin lock barriers in turn are also fixed in
this commit as explained above
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: stable@vger.kernel.org Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Vineet Gupta <vgupta@synopsys.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Many harddisks (mostly WD ones) have firmware problems and take too
long, more than 10 seconds, to resume from suspend. And this often
exceeds the default DPM watchdog timeout (12 seconds), resulting in a
kernel panic out of sudden.
Since most distros just take the default as is, we should give a bit
more safer value. This patch increases the default value from 12
seconds to one minute, which has been confirmed to be long enough for
such problematic disks.
Current implementation of descriptor init procedure only takes
care about setting/clearing ownership flag in "des0"/"des1"
fields while it is perfectly possible to get unexpected bits
set because of the following factors:
[1] On driver probe underlying memory allocated with
dma_alloc_coherent() might not be zeroed and so
it will be filled with garbage.
[2] During driver operation some bits could be set by SD/MMC
controller (for example error flags etc).
And unexpected and/or randomly set flags in "des0"/"des1"
fields may lead to unpredictable behavior of GMAC DMA block.
This change addresses both items above with:
[1] Use of dma_zalloc_coherent() instead of simple
dma_alloc_coherent() to make sure allocated memory is
zeroed. That shouldn't affect performance because
this allocation only happens once on driver probe.
[2] Do explicit zeroing of both "des0" and "des1" fields
of all buffer descriptors during initialization of
DMA transfer.
And while at it fixed identation of dma_free_coherent()
counterpart as well.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com> Cc: Giuseppe Cavallaro <peppe.cavallaro@st.com> Cc: arc-linux-dev@synopsys.com Cc: linux-kernel@vger.kernel.org Cc: stable@vger.kernel.org Cc: David Miller <davem@davemloft.net> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
When kzalloc() is called under spin_lock(), GFP_ATOMIC should be
used to avoid sleeping allocation.
The call tree is:
of_pci_range_to_resource()
--> pci_register_io_range() <-- takes spin_lock(&io_range_lock);
--> kzalloc()
Signed-off-by: Jingoo Han <jingoohan1@gmail.com> Cc: stable@vger.kernel.org # 3.18+ Signed-off-by: Rob Herring <robh@kernel.org> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Add code to nf_unregister_hook to flush the nf_queue when a hook is
unregistered. This guarantees that the pointer that the nf_queue code
retains into the nf_hook list will remain valid while a packet is
queued.
I tested what would happen if we do not flush queued packets and was
trivially able to obtain the oops below. All that was required was
to stop the nf_queue listening process, to delete all of the nf_tables,
and to awaken the nf_queue listening process.
Cc: stable@vger.kernel.org Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
A ROSE socket doesn't necessarily always have a neighbour pointer so check
if the neighbour pointer is valid before dereferencing it.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Tested-by: Bernard Pidoux <f6bvp@free.fr> Cc: stable@vger.kernel.org #2.6.11+ Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
If we are doing an MST transaction and we've gotten HPD and we
lookup the device from the incoming msg, we should take the mgr
lock around it, so that mst_primary and mstb->ports are valid.
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: stable@vger.kernel.org Signed-off-by: Dave Airlie <airlied@redhat.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
This validates the mst_primary under the lock, and then calls
into the check and send function. This makes the code a lot
easier to understand the locking rules in.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: stable@vger.kernel.org Signed-off-by: Dave Airlie <airlied@redhat.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
ext4 isn't willing to map clusters to a non-extent file. Don't signal
this with an out of space error, since the FS will retry the
allocation (which didn't fail) forever. Instead, return EUCLEAN so
that the operation will fail immediately all the way back to userspace.
(The fix is either to run e2fsck -E bmap2extent, or to chattr +e the file.)
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Theodore Ts'o <tytso@mit.edu> Cc: stable@vger.kernel.org Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
If we create a CRC filesystem, mount it, and create a symlink with
a path long enough that it can't live in the inode, we get a very
strange result upon remount:
# ls -l mnt
total 4
lrwxrwxrwx. 1 root root 929 Jun 15 16:58 link -> XSLM
XSLM is the V5 symlink block header magic (which happens to be
followed by a NUL, so the string looks terminated).
xfs_readlink_bmap() advanced cur_chunk by the size of the header
for CRC filesystems, but never actually used that pointer; it
kept reading from bp->b_addr, which is the start of the block,
rather than the start of the symlink data after the header.
Looks like this problem goes back to v3.10.
Fixing this gets us reading the proper link target, again.
Cc: stable@vger.kernel.org Signed-off-by: Eric Sandeen <sandeen@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
KVM guest kernels for trap & emulate run in user mode, with a modified
set of kernel memory segments. However the fixmap address is still in
the normal KSeg3 region at 0xfffe0000 regardless, causing problems when
cache alias handling makes use of them when handling copy on write.
Therefore define FIXADDR_TOP as 0x7ffe0000 in the guest kernel mapped
region when CONFIG_KVM_GUEST is defined.
Normally all of the buffers will have been forced out to disk before
we call invalidate_bdev(), but there will be some cases, where a file
system operation was aborted due to an ext4_error(), where there may
still be some dirty buffers in the buffer cache for the device. So
try to force them out to memory before calling invalidate_bdev().
This fixes a warning triggered by generic/081:
WARNING: CPU: 1 PID: 3473 at /usr/projects/linux/ext4/fs/block_dev.c:56 __blkdev_put+0xb5/0x16f()
When building the kernel with a bare-metal (ELF) toolchain, the -shared
option may not be passed down to collect2, resulting in silent corruption
of the vDSO image (in particular, the DYNAMIC section is omitted).
The effect of this corruption is that the dynamic linker fails to find
the vDSO symbols and libc is instead used for the syscalls that we
intended to optimise (e.g. gettimeofday). Functionally, there is no
issue as the sigreturn trampoline is still intact and located by the
kernel.
This patch fixes the problem by explicitly passing -shared to the linker
when building the vDSO.
Cc: <stable@vger.kernel.org> Reported-by: Szabolcs Nagy <Szabolcs.Nagy@arm.com> Reported-by: James Greenlaigh <james.greenhalgh@arm.com> Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
BugLink: https://bugs.launchpad.net/bugs/1394368
This device requires new firmware files
AthrBT_0x11020100.dfu and ramps_0x11020100_40.dfu added to
/lib/firmware/ar3k/ that are not included in linux-firmware yet.
If a write attempt fails, and the write is queued up for resending to
the server, as opposed to being dropped, then we need to set the
appropriate flag so that nfs_file_fsync() does the right thing.
pnfs_do_write() expects the call to pnfs_write_through_mds() to free the
pgio header and to release the layout segment before exiting. The problem
is that nfs_pgio_data_destroy() doesn't actually do this; it only frees
the memory allocated by nfs_generic_pgio().
Ditto for pnfs_do_read()...
Fix in both cases is to add a call to hdr->release(hdr).
We enable _CRS on all systems from 2008 and later. On older systems, we
ignore _CRS and assume the whole physical address space (excluding RAM and
other devices) is available for PCI devices, but on systems that support
physical address spaces larger than 4GB, it's doubtful that the area above
4GB is really available for PCI.
After d56dbf5bab8c ("PCI: Allocate 64-bit BARs above 4G when possible"), we
try to use that space above 4GB *first*, so we're more likely to put a
device there.
On Juan's Toshiba Satellite Pro U200, BIOS left the graphics, sound, 1394,
and card reader devices unassigned (but only after Windows had been
booted). Only the sound device had a 64-bit BAR, so it was the only device
placed above 4GB, and hence the only device that didn't work.
Keep _CRS enabled even on pre-2008 systems if they support physical address
space larger than 4GB.
Commit fd1d0ddf2ae9 (KVM: arm/arm64: check IRQ number on userland
injection) rightly limited the range of interrupts userspace can
inject in a guest, but failed to consider the (unlikely) case where
a guest is configured with 1024 interrupts.
In this case, interrupts ranging from 1020 to 1023 are unuseable,
as they have a special meaning for the GIC CPU interface.
Make sure that these number cannot be used as an IRQ. Also delete
a redundant (and similarily buggy) check in kvm_set_irq.
Reported-by: Peter Maydell <peter.maydell@linaro.org> Cc: Andre Przywara <andre.przywara@arm.com> Cc: <stable@vger.kernel.org> # 4.1, 4.0, 3.19, 3.18 Signed-off-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
The metadata space map has a simplified 'bootstrap' mode that is
operational when extending the space maps. Whilst in this mode it's
possible for some refcount decrement operations to become queued (eg, as
a result of shadowing one of the bitmap indexes). These decrements were
not being applied when switching out of bootstrap mode.
The effect of this bug was the leaking of a 4k metadata block. This is
detected by the latest version of thin_check as a non fatal error.
Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
The memmap freeing code in free_unused_memmap() computes the end of
each memblock by adding the memblock size onto the base. However,
if SPARSEMEM is enabled then the value (start) used for the base
may already have been rounded downwards to work out which memmap
entries to free after the previous memblock.
This may cause memmap entries that are in use to get freed.
In general, you're not likely to hit this problem unless there
are at least 2 memblocks and one of them is not aligned to a
sparsemem section boundary. Note that carve-outs can increase
the number of memblocks by splitting the regions listed in the
device tree.
This problem doesn't occur with SPARSEMEM_VMEMMAP, because the
vmemmap code deals with freeing the unused regions of the memmap
instead of requiring the arch code to do it.
This patch gets the memblock base out of the memblock directly when
computing the block end address to ensure the correct value is used.
Signed-off-by: Dave Martin <Dave.Martin@arm.com> Cc: <stable@vger.kernel.org> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Commit 6c81fe7925cc4c42 ("arm64: enable context tracking") did not
update el0_sp_pc to use ct_user_exit, but this appears to have been
unintentional. In commit 6ab6463aeb5fbc75 ("arm64: adjust el0_sync so
that a function can be called") we made x0 available, and in the return
to userspace we call ct_user_enter in the kernel_exit macro.
Due to this, we currently don't correctly inform RCU of the user->kernel
transition, and may erroneously account for time spent in the kernel as
if we were in an extended quiescent state when CONFIG_CONTEXT_TRACKING
is enabled.
As we do record the kernel->user transition, a userspace application
making accesses from an unaligned stack pointer can demonstrate the
imbalance, provoking the following warning: