Starting with 4.1 the tracing subsystem has its own filesystem
which is automounted in the tracing subdirectory of debugfs.
Prior to this debugfs could be bind mounted in a cloned mount
namespace, but if tracefs has been mounted under debugfs this
now fails because there is a locked child mount. This creates
a regression for container software which bind mounts debugfs
to satisfy the assumption of some userspace software.
In other pseudo filesystems such as proc and sysfs we're already
creating mountpoints like this in such a way that no dirents can
be created in the directories, allowing them to be exceptions to
some MNT_LOCKED tests. In fact we're already do this for the
tracefs mountpoint in sysfs.
Do the same in debugfs_create_automount(), since the intention
here is clearly to create a mountpoint. This fixes the regression,
as locked child mounts on permanently empty directories do not
cause a bind mount to fail.
Based on Sergey's test patch [1], this fixes zram with lz4 compression
on big endian cpus.
Note that the 64-bit preprocessor test is not a cleanup, it's part of
the fix, since those identifiers are bogus (for example, __ppc64__
isn't defined anywhere else in the kernel, which means we'd fall into
the 32-bit definitions on ppc64).
Commit 9567366fefdd ("dm cache metadata: fix READ_LOCK macros and
cleanup WRITE_LOCK macros") uses down_write() instead of down_read() in
cmd_read_lock(), yet up_read() is used to release the lock in
READ_UNLOCK(). Fix it.
Fixes: 9567366fefdd ("dm cache metadata: fix READ_LOCK macros and cleanup WRITE_LOCK macros") Signed-off-by: Ahmed Samy <f.fallen45@gmail.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The READ_LOCK macro was incorrectly returning -EINVAL if
dm_bm_is_read_only() was true -- it will always be true once the cache
metadata transitions to read-only by dm_cache_metadata_set_read_only().
Wrap READ_LOCK and WRITE_LOCK multi-statement macros in do {} while(0).
Also, all accesses of the 'cmd' argument passed to these related macros
are now encapsulated in parenthesis.
A follow-up patch can be developed to eliminate the use of macros in
favor of pure C code. Avoiding that now given that this needs to apply
to stable@.
Reported-by: Ben Hutchings <ben@decadent.org.uk> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Fixes: d14fcf3dd79 ("dm cache: make sure every metadata function checks fail_io") Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When using asynchronous read or write operations on the USB endpoints the
issuer of the IO request is notified by calling the ki_complete() callback
of the submitted kiocb when the URB has been completed.
Calling this ki_complete() callback will free kiocb. Make sure that the
structure is no longer accessed beyond that point, otherwise undefined
behaviour might occur.
On BXT platform Host Controller and Device Controller figure as
same PCI device but with different device function. HCD should
not pass data to Device Controller but only to Host Controllers.
Checking if companion device is Host Controller, otherwise skip.
Signed-off-by: Robert Dobrowolski <robert.dobrowolski@linux.intel.com> Acked-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Otherwise generic-xhci and xhci-platform which have no data get wrongly
detected as XHCI_PLAT_TYPE_MARVELL_ARMADA by xhci_plat_type_is().
This fixes a regression in v4.5 for STiH407 family SoC's which use the
synopsis dwc3 IP, whereby the disable_clk error path gets taken due to
wrongly being detected as XHCI_PLAT_TYPE_MARVELL_ARMADA and the hcd never
gets added.
I suspect this will also fix other dwc3 DT platforms such as Exynos,
although I've only tested on STih410 SoC.
PCI hotpluggable xhci controllers such as some Alpine Ridge solutions will
remove the xhci controller from the PCI bus when the last USB device is
disconnected.
Add a flag to indicate that the host is being removed to avoid queueing
configure_endpoint commands for the dropped endpoints.
For PCI hotplugged controllers this will prevent 5 second command timeouts
For static xhci controllers the configure_endpoint command is not needed
in the removal case as everything will be returned, freed, and the
controller is reset.
For now the flag is only set for PCI connected host controllers.
On some xHCI controllers (e.g. R-Car SoCs), the AC64 bit (bit 0) of
HCCPARAMS1 is set to 1. However, the xHCs don't support 64-bit
address memory pointers actually. So, in this case, this driver should
call dma_set_coherent_mask(dev, DMA_BIT_MASK(32)) in xhci_gen_setup().
Otherwise, the xHCI controller will be died after a usb device is
connected if it runs on above 4GB physical memory environment.
So, this patch adds a new quirk XHCI_NO_64BIT_SUPPORT to resolve
such an issue.
Give USB3 devices a better chance to enumerate at USB 3 speeds if
they are connected to a suspended host.
Solves an issue with NEC uPD720200 host hanging when partially
enumerating a USB3 device as USB2 after host controller runtime resume.
Changes since V1: fixed the description and added KASan warning.
In assoc_array_insert_into_terminal_node(), we call the
compare_object() method on all non-empty slots, even when they're
not leaves, passing a pointer to an unexpected structure to
compare_object(). Currently it causes an out-of-bound read access
in keyring_compare_object detected by KASan (see below). The issue
is easily reproduced with keyutils testsuite.
Only call compare_object() when the slot is a leave.
KASan warning:
==================================================================
BUG: KASAN: slab-out-of-bounds in keyring_compare_object+0x213/0x240 at addr ffff880060a6f838
Read of size 8 by task keyctl/1655
=============================================================================
BUG kmalloc-192 (Not tainted): kasan: bad access detected
-----------------------------------------------------------------------------
Commit 127500ccb766f ("ARM: OMAP2+: Only write the sysconfig on idle
when necessary") talks about verification of sysconfig cache value before
updating it, only during idle path. But the patch is adding the
verification in the enable path. So, adding the check in a proper place
as per the commit description.
Not keeping this check during enable path as there is a chance of losing
context and it is safe to do on idle as the context of the register will
never be lost while the device is active.
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com> Acked-by: Tero Kristo <t-kristo@ti.com> Cc: Jon Hunter <jonathanh@nvidia.com> Fixes: commit 127500ccb766 "ARM: OMAP2+: Only write the sysconfig on idle when necessary"
[paul@pwsan.com: appears to have been caused by my own mismerge of the
originally posted patch] Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The following commits:
commit 3fa609755c11 ("ARM: omap2: restore OMAP4 barrier behaviour")
commit f746929ffdc8 ("Revert "ARM: OMAP4: remove dead kconfig option OMAP4_ERRATA_I688"")
and
commit ea827ad5ffbb ("ARM: DRA7: Provide proper IO map table")
came in around the same time, unfortunately this seem to have missed
initializing the barrier for DRA7 platforms - omap5_map_io was reused
for dra7 till it was split out by the last patch. barrier_init
needs to be hence carried forward as it is valid for DRA7 family of
processors as they are for OMAP5.
Fixes: ea827ad5ffbb7 ("ARM: DRA7: Provide proper IO map table") Reported-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Reported-by: Tomi Valkeinen <tomi.valkeinen@ti.com> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Signed-off-by: Nishanth Menon <nm@ti.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The USB2 port for Armada 38x is defined to be at 58000, not at
50000.
Fixes: 2d0a7addbd10 ("ARM: Kirkwood: Add support for many Synology NAS devices") Signed-off-by: Patrick Uiterwijk <patrick@puiterwijk.org> Acked-by: Imre Kaloz <kaloz@openwrt.org> Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
EDMA was allocating DMA channels 32 and 33 for memcpy usage, out of which
channel 33 is actually used by DES crypto engine. This bad allocation of
the channel causes a crash in the DES crypto engine, as the channel
gets configured for memcpy usage instead of hardware <-> memory DMA.
Fixed by allocating DMA channels 58 and 59 for memcpy usage (I2C0 RX/TX),
which are not used by anybody.
Fixes: cce1ee000187 ("ARM: DTS: am437x: Use the new DT bindings for
the eDMA3") Signed-off-by: Tero Kristo <t-kristo@ti.com> Suggested-by: Peter Ujfalusi <peter.ujfalusi@ti.com> Signed-off-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 55ee7017ee31 ("arm: omap2: board-generic: use omap4_local_timer_init
for AM437x") makes synctimer32k as the clocksource on AM43xx. By default
the synctimer32k is clocked by 32K RTC OSC on AM43xx. But this 32K RTC OSC
is not available on epos boards which makes it fail to boot.
Synctimer32k can also be clocked by a peripheral PLL, so making this as
clock parent for synctimer3k on epos boards.
Fixes: 55ee7017ee31 ("arm: omap2: board-generic: use omap4_local_timer_init for AM437x") Reported-by: Nishanth Menon <nm@ti.com> Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com> Signed-off-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On a host that runs NTP, corrections can have a direct impact on
the background timer that we program on the behalf of a vcpu.
In particular, NTP performing a forward correction will result in
a timer expiring sooner than expected from a guest point of view.
Not a big deal, we kick the vcpu anyway.
But on wake-up, the vcpu thread is going to perform a check to
find out whether or not it should block. And at that point, the
timer check is going to say "timer has not expired yet, go back
to sleep". This results in the timer event being lost forever.
There are multiple ways to handle this. One would be record that
the timer has expired and let kvm_cpu_has_pending_timer return
true in that case, but that would be fairly invasive. Another is
to check for the "short sleep" condition in the hrtimer callback,
and restart the timer for the remaining time when the condition
is detected.
This patch implements the latter, with a bit of refactoring in
order to avoid too much code duplication.
Reported-by: Alexander Graf <agraf@suse.de> Reviewed-by: Alexander Graf <agraf@suse.de> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
An interrupt handler that uses the fpu can kill a KVM VM, if it runs
under the following conditions:
- the guest's xcr0 register is loaded on the cpu
- the guest's fpu context is not loaded
- the host is using eagerfpu
Note that the guest's xcr0 register and fpu context are not loaded as
part of the atomic world switch into "guest mode". They are loaded by
KVM while the cpu is still in "host mode".
Usage of the fpu in interrupt context is gated by irq_fpu_usable(). The
interrupt handler will look something like this:
if (irq_fpu_usable()) {
kernel_fpu_begin();
[... code that uses the fpu ...]
kernel_fpu_end();
}
As long as the guest's fpu is not loaded and the host is using eager
fpu, irq_fpu_usable() returns true (interrupted_kernel_fpu_idle()
returns true). The interrupt handler proceeds to use the fpu with
the guest's xcr0 live.
kernel_fpu_begin() saves the current fpu context. If this uses
XSAVE[OPT], it may leave the xsave area in an undesirable state.
According to the SDM, during XSAVE bit i of XSTATE_BV is not modified
if bit i is 0 in xcr0. So it's possible that XSTATE_BV[i] == 1 and
xcr0[i] == 0 following an XSAVE.
kernel_fpu_end() restores the fpu context. Now if any bit i in
XSTATE_BV == 1 while xcr0[i] == 0, XRSTOR generates a #GP. The
fault is trapped and SIGSEGV is delivered to the current process.
Only pre-4.2 kernels appear to be vulnerable to this sequence of
events. Commit 653f52c ("kvm,x86: load guest FPU context more eagerly")
from 4.2 forces the guest's fpu to always be loaded on eagerfpu hosts.
This patch fixes the bug by keeping the host's xcr0 loaded outside
of the interrupts-disabled region where KVM switches into guest mode.
Suggested-by: Andy Lutomirski <luto@amacapital.net> Signed-off-by: David Matlack <dmatlack@google.com>
[Move load after goto cancel_injection. - Paolo] Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When we loop over all queued machine check error records to pass them
to the registered notifiers we use llist_for_each_entry(). But the loop
calls gen_pool_free() for the entry in the body of the loop - and then
the iterator looks at node->next after the free.
Use llist_for_each_entry_safe() instead.
Signed-off-by: Tony Luck <tony.luck@intel.com> Signed-off-by: Borislav Petkov <bp@suse.de> Cc: Gong Chen <gong.chen@linux.intel.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: linux-edac <linux-edac@vger.kernel.org> Link: http://lkml.kernel.org/r/0205920@agluck-desk.sc.intel.com Link: http://lkml.kernel.org/r/1459929916-12852-4-git-send-email-bp@alien8.de Signed-off-by: Ingo Molnar <mingo@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 4dffbfc48d65 ("arm64/efi: mark UEFI reserved regions as
MEMBLOCK_NOMAP") updated the mapping logic of both the RuntimeServices
regions as well as the kernel's copy of the UEFI memory map to set the
MEMBLOCK_NOMAP flag, which causes these regions to be omitted from the
kernel direct mapping, and from being covered by a struct page.
For the RuntimeServices regions, this is an obvious win, since the contents
of these regions have significance to the firmware executable code itself,
and are mapped in the EFI page tables using attributes that are described in
the UEFI memory map, and which may differ from the attributes we use for
mapping system RAM. It also prevents the contents from being modified
inadvertently, since the EFI page tables are only live during runtime
service invocations.
None of these concerns apply to the allocation that covers the UEFI memory
map, since it is entirely owned by the kernel. Setting the MEMBLOCK_NOMAP on
the region did allow us to use ioremap_cache() to map it both on arm64 and
on ARM, since the latter does not allow ioremap_cache() to be used on
regions that are covered by a struct page.
The ioremap_cache() on ARM restriction will be lifted in the v4.7 timeframe,
but in the mean time, it has been reported that commit 4dffbfc48d65 causes
a regression on 64k granule kernels. This is due to the fact that, given
the 64 KB page size, the region that we end up removing from the kernel
direct mapping is rounded up to 64 KB, and this 64 KB page frame may be
shared with the initrd when booting via GRUB (which does not align its
EFI_LOADER_DATA allocations to 64 KB like the stub does). This will crash
the kernel as soon as it tries to access the initrd.
Since the issue is specific to arm64, revert back to memblock_reserve()'ing
the UEFI memory map when running on arm64. This is a temporary fix for v4.5
and v4.6, and will be superseded in the v4.7 timeframe when we will be able
to move back to memblock_reserve() unconditionally.
Fixes: 4dffbfc48d65 ("arm64/efi: mark UEFI reserved regions as MEMBLOCK_NOMAP") Reported-by: Mark Salter <msalter@redhat.com> Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Acked-by: Will Deacon <will.deacon@arm.com> Cc: Leif Lindholm <leif.lindholm@linaro.org> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Jeremy Linton <jeremy.linton@arm.com> Cc: Mark Langsdorf <mlangsdo@redhat.com> Signed-off-by: Matt Fleming <matt@codeblueprint.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Starting from commit e36f620428(block: split bios to max possible length),
block core starts to split bio in the middle of bvec.
Unfortunately loop dio/aio doesn't consider this situation, and
always treat 'iter.iov_offset' as zero. Then filesystem corruption
is observed.
This patch figures out the offset of the base bvevc via
'bio->bi_iter.bi_bvec_done' and fixes the issue by passing the offset
to iov iterator.
Fixes: e36f6204288088f (block: split bios to max possible length) Cc: Keith Busch <keith.busch@intel.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Ming Lei <ming.lei@canonical.com> Signed-off-by: Jens Axboe <axboe@fb.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The initialization of partition's percpu_ref should have been done before
sending out KOBJ_ADD uevent, which may cause userspace to read partition
table. So the uninitialized percpu_ref may be accessed in data path.
Commit 520bd7a8b415 ("mmc: core: Optimize boot time by detecting cards
simultaneously") causes regressions for some platforms.
These platforms relies on fixed mmcblk device indexes, instead of
deploying the defacto standard with UUID/PARTUUID. In other words their
rootfs needs to be available at hardcoded paths, like /dev/mmcblk0p2.
Such guarantees have never been made by the kernel, but clearly the above
commit changes the behaviour. More precisely, because of that the order
changes of how cards becomes detected, so do their corresponding mmcblk
device indexes.
As the above commit significantly improves boot time for some platforms
(magnitude of seconds), let's avoid reverting this change but instead
restore the behaviour of how mmcblk device indexes becomes picked.
By using the same index for the mmcblk device as for the corresponding mmc
host device, the probe order of mmc host devices decides the index we get
for the mmcblk device.
For those platforms that suffers from a regression, one could expect that
this updated behaviour should be sufficient to meet their expectations of
"fixed" mmcblk device indexes.
Another side effect from this change, is that the same index is used for
the mmc host device, the mmcblk device and the mmc block queue. That
should clarify their relationship.
Reported-by: Peter Hurley <peter@hurleysoftware.com> Reported-by: Laszlo Fiat <laszlo.fiat@gmail.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Fixes: 520bd7a8b415 ("mmc: core: Optimize boot time by detecting cards
simultaneously") Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The usbhid driver has inconsistently duplicated code in its post-reset,
resume, and reset-resume pathways.
reset-resume doesn't check HID_STARTED before trying to
restart the I/O queues.
resume fails to clear the HID_SUSPENDED flag if HID_STARTED
isn't set.
resume calls usbhid_restart_queues() with usbhid->lock held
and the others call it without holding the lock.
The first item in particular causes a problem following a reset-resume
if the driver hasn't started up its I/O. URB submission fails because
usbhid->urbin is NULL, and this triggers an unending reset-retry loop.
This patch fixes the problem by creating a new subroutine,
hid_restart_io(), to carry out all the common activities. It also
adds some checks that were missing in the original code:
After a reset, there's no need to clear any halted endpoints.
After a resume, if a reset is pending there's no need to
restart any I/O until the reset is finished.
After a resume, if the interrupt-IN endpoint is halted there's
no need to submit the input URB until the halt has been
cleared.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Reported-by: Daniel Fraga <fragabr@gmail.com> Tested-by: Daniel Fraga <fragabr@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Looks like recent changes in the Wacom driver made the Bamboo ONE crashes.
The tablet behaves as if it was a regular Bamboo device with pen, touch
and pad, but there is no physical pad connected to it.
The weird part is that the pad is still sending events and given that
there is no input node connected to it, we get anull pointer exception.
This is Dell usb dock audio workaround.
It was fixed the master volume keep lower.
[Some background: the patch essentially skips the controls of a couple
of FU volumes. Although the firmware exposes the dB and the value
information via the usb descriptor, changing the values (we set the
min volume as default) screws up the device. Although this has been
fixed in the newer firmware, the devices are shipped with the old
firmware, thus we need the workaround in the driver side. -- tiwai]
Plantronics BT300 does not support reading the sample rate which leads
to many lines of "cannot get freq at ep 0x1". This patch adds the USB
ID of the BT300 to quirks.c and avoids those error messages.
Phoenix Audio TMX320 gives the similar error when the sample rate is
asked:
usb 2-1.3: 2:1: cannot get freq at ep 0x85
usb 2-1.3: 1:1: cannot get freq at ep 0x2
....
Add the corresponding USB-device ID (1de7:0014) to
snd_usb_get_sample_rate_quirk() list.
The commit [bd48128539ab: ALSA: hda - Fix forgotten HDMI
monitor_present update] covered the missing update of monitor_present
flag, but this caused a regression for devices without the i915 eld
notifier. Since the old code supposed that pin_eld->monitor_present
was updated by the caller side, the hdmi_present_sense_via_verbs()
doesn't update the temporary eld->monitor_present but only
pin_eld->monitor_present, which is now overridden in update_eld().
The fix is to update pin_eld->monitor_present as well before calling
update_eld().
Note that this may still leave monitor_present flag in an inconsistent
state when the driver repolls, but this is at least the old behavior.
More proper fix will follow in the later patch.
Add support and PCI IDs for more Broxton host controllers
Other BXT IDs were added in v4.4 so cc'ing stable. This patch
is dependent on commit 163cbe31e516 ("mmc: sdhci-pci: Fix card
detect race for Intel BXT/APL") but that is already in stable
since v4.4.4.
Several commits relating to setting power have been introducing
problems by putting driver-specific rules into generic SDHCI code.
Krzysztof Kozlowski reported that after commit 918f4cbd4340 ("mmc:
sdhci: restore behavior when setting VDD via external regulator")
on Trats2 board there are warnings for invalid VDD value (2.8V):
[ 3.119656] ------------[ cut here ]------------
[ 3.119666] WARNING: CPU: 3 PID: 90 at
../drivers/mmc/host/sdhci.c:1234 sdhci_do_set_ios+0x4cc/0x5e0
[ 3.119669] mmc0: Invalid vdd 0x10
[ 3.119673] Modules linked in:
[ 3.119679] CPU: 3 PID: 90 Comm: kworker/3:1 Tainted: G W
4.5.0-next-20160324 #23
[ 3.119681] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[ 3.119690] Workqueue: events_freezable mmc_rescan
[ 3.119708] [<c010e0ac>] (unwind_backtrace) from [<c010ae10>]
(show_stack+0x10/0x14)
[ 3.119719] [<c010ae10>] (show_stack) from [<c0323260>]
(dump_stack+0x88/0x9c)
[ 3.119728] [<c0323260>] (dump_stack) from [<c011b754>] (__warn+0xe8/0x100)
[ 3.119734] [<c011b754>] (__warn) from [<c011b7a4>]
(warn_slowpath_fmt+0x38/0x48)
[ 3.119740] [<c011b7a4>] (warn_slowpath_fmt) from [<c0527d28>]
(sdhci_do_set_ios+0x4cc/0x5e0)
[ 3.119748] [<c0527d28>] (sdhci_do_set_ios) from [<c0528018>]
(sdhci_runtime_resume_host+0x60/0x114)
[ 3.119758] [<c0528018>] (sdhci_runtime_resume_host) from
[<c0402570>] (__rpm_callback+0x2c/0x60)
[ 3.119767] [<c0402570>] (__rpm_callback) from [<c04025c4>]
(rpm_callback+0x20/0x80)
[ 3.119773] [<c04025c4>] (rpm_callback) from [<c04034b8>]
(rpm_resume+0x36c/0x558)
[ 3.119780] [<c04034b8>] (rpm_resume) from [<c04036f0>]
(__pm_runtime_resume+0x4c/0x64)
[ 3.119788] [<c04036f0>] (__pm_runtime_resume) from [<c0512728>]
(__mmc_claim_host+0x170/0x1b0)
[ 3.119795] [<c0512728>] (__mmc_claim_host) from [<c0514e2c>]
(mmc_rescan+0x54/0x348)
[ 3.119807] [<c0514e2c>] (mmc_rescan) from [<c0130dac>]
(process_one_work+0x120/0x3f4)
[ 3.119815] [<c0130dac>] (process_one_work) from [<c01310b8>]
(worker_thread+0x38/0x554)
[ 3.119823] [<c01310b8>] (worker_thread) from [<c01365a4>]
(kthread+0xdc/0xf4)
[ 3.119831] [<c01365a4>] (kthread) from [<c0107878>]
(ret_from_fork+0x14/0x3c)
[ 3.119834] ---[ end trace a22d652aa3276886 ]---
Fix by adding a 'set_power' callback and restoring the default
behaviour prior to commit 918f4cbd4340 ("mmc: sdhci: restore
behavior when setting VDD via external regulator"). The desired
behaviour of that commit is gotten by having sdhci-pxav3 provide
its own set_power callback.
BUG: sleeping function called from invalid context at kernel/locking/rtmutex.c:917
in_atomic(): 1, irqs_disabled(): 128, pid: 383, name: sh
Preemption disabled at:[<ffff800000124c18>] kgdb_cpu_enter+0x158/0x6b8
If we rename an inode A (be it a file or a directory), create a new
inode B with the old name of inode A and under the same parent directory,
fsync inode B and then power fail, at log tree replay time we end up
removing inode A completely. If inode A is a directory then all its files
are gone too.
Example scenarios where this happens:
This is reproducible with the following steps, taken from a couple of
test cases written for fstests which are going to be submitted upstream
soon:
The next time the fs is mounted, log tree replay happens and
the directory "y" does not exist nor do the files "foo" and
"bar" exist anywhere (neither in "y" nor in "x", nor the root
nor anywhere).
The next time the fs is mounted, log tree replay happens and the
file "bar" does not exists anymore. A file with the name "foo"
exists and it matches the second file we created.
Another related problem that does not involve file/data loss is when a
new inode is created with the name of a deleted snapshot and we fsync it:
The next time the fs is mounted the log replay procedure fails because
it attempts to delete the snapshot entry (which has dir item key type
of BTRFS_ROOT_ITEM_KEY) as if it were a regular (non-root) entry,
resulting in the following error that causes mount to fail:
Fix this by forcing a transaction commit when such cases happen.
This means we check in the commit root of the subvolume tree if there
was any other inode with the same reference when the inode we are
fsync'ing is a new inode (created in the current transaction).
Test cases for fstests, covering all the scenarios given above, were
submitted upstream for fstests:
* fstests: generic test for fsync after renaming directory
https://patchwork.kernel.org/patch/8694281/
* fstests: generic test for fsync after renaming file
https://patchwork.kernel.org/patch/8694301/
* fstests: add btrfs test for fsync after snapshot deletion
https://patchwork.kernel.org/patch/8670671/
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>
IOMMU drivers that do not support default domains, but make
use of the the group->domain pointer can get that pointer
overwritten with NULL on device add/remove.
Make sure this can't happen by only overwriting the domain
pointer when it is NULL.
Previously, ext4 would fail the mount if the file system had the quota
feature enabled and quota mount options (used for the older quota
setups) were present. This broke xfstests, since xfs silently ignores
the usrquote and grpquota mount options if they are specified. This
commit changes things so that we are consistent with xfs; having the
mount options specified is harmless, so no sense break users by
forbidding them.
With the internal Quota feature, mke2fs creates empty quota inodes and
quota usage tracking is enabled as soon as the file system is mounted.
Since quotacheck is no longer preallocating all of the blocks in the
quota inode that are likely needed to be written to, we are now seeing
a lockdep false positive caused by needing to allocate a quota block
from inside ext4_map_blocks(), while holding i_data_sem for a data
inode. This results in this complaint:
If the lower or upper directory of an overlayfs mount belong to a btrfs
file system and we fsync the file through the overlayfs' merged directory
we ended up accessing an inode that didn't belong to btrfs as if it were
a btrfs inode at btrfs_sync_file() resulting in a crash like the following:
This started happening since commit 4bacc9c9234 (overlayfs: Make f_path
always point to the overlay and f_inode to the underlay) and even though
after this change we could still access the btrfs inode through
struct file->f_mapping->host or struct file->f_inode, we would end up
resulting in more similar issues later on at check_parent_dirs_for_sync()
because the dentry we got (from struct file->f_path.dentry) was from
overlayfs and not from btrfs, that is, we had no way of getting the dentry
that belonged to btrfs (we always got the dentry that belonged to
overlayfs).
The new patch from Miklos Szeredi, titled "vfs: add file_dentry()" and
recently submitted to linux-fsdevel, adds a file_dentry() API that allows
us to get the btrfs dentry from the input file and therefore being able
to fsync when the upper and lower directories belong to btrfs filesystems.
This issue has been reported several times by users in the mailing list
and bugzilla. A test case for xfstests is being submitted as well.
Fixes: 4bacc9c9234c ("overlayfs: Make f_path always point to the overlay and f_inode to the underlay")
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=101951
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=109791 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>
EXT4 may be used as lower layer of overlayfs and accessing f_path.dentry
can lead to a crash.
Fix by replacing direct access of file->f_path.dentry with the
file_dentry() accessor, which will always return a native object.
Reported-by: Daniel Axtens <dja@axtens.net> Fixes: 4bacc9c9234c ("overlayfs: Make f_path always point to the overlay and f_inode to the underlay") Fixes: ff978b09f973 ("ext4 crypto: move context consistency check to ext4_file_open()") Signed-off-by: Miklos Szeredi <mszeredi@redhat.com> Signed-off-by: Theodore Ts'o <tytso@mit.edu> Cc: David Howells <dhowells@redhat.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This avoids potential problems caused by a race where the inode gets
renamed out from its parent directory and the parent directory is
deleted while ext4_d_revalidate() is running.
NFS may be used as lower layer of overlayfs and accessing f_path.dentry can
lead to a crash.
Fix by replacing direct access of file->f_path.dentry with the
file_dentry() accessor, which will always return a native object.
Fixes: 4bacc9c9234c ("overlayfs: Make f_path always point to the overlay and f_inode to the underlay") Signed-off-by: Miklos Szeredi <mszeredi@redhat.com> Tested-by: Goldwyn Rodrigues <rgoldwyn@suse.com> Acked-by: Trond Myklebust <trond.myklebust@primarydata.com> Signed-off-by: Theodore Ts'o <tytso@mit.edu> Cc: David Howells <dhowells@redhat.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This series fixes bugs in nfs and ext4 due to 4bacc9c9234c ("overlayfs:
Make f_path always point to the overlay and f_inode to the underlay").
Regular files opened on overlayfs will result in the file being opened on
the underlying filesystem, while f_path points to the overlayfs
mount/dentry.
This confuses filesystems which get the dentry from struct file and assume
it's theirs.
Add a new helper, file_dentry() [*], to get the filesystem's own dentry
from the file. This checks file->f_path.dentry->d_flags against
DCACHE_OP_REAL, and returns file->f_path.dentry if DCACHE_OP_REAL is not
set (this is the common, non-overlayfs case).
In the uncommon case it will call into overlayfs's ->d_real() to get the
underlying dentry, matching file_inode(file).
The reason we need to check against the inode is that if the file is copied
up while being open, d_real() would return the upper dentry, while the open
file comes from the lower dentry.
[*] If possible, it's better simply to use file_inode() instead.
During revalidate we check whether device capacity has changed before we
decide whether to output disk information or not.
The check for old capacity failed to take into account that we scaled
sdkp->capacity based on the reported logical block size. And therefore
the capacity test would always fail for devices with sectors bigger than
512 bytes and we would print several copies of the same discovery
information.
Avoid scaling sdkp->capacity and instead adjust the value on the fly
when setting the block device capacity and generating fake C/H/S
geometry.
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Reported-by: Hannes Reinecke <hare@suse.de> Reviewed-by: Hannes Reinicke <hare@suse.de> Reviewed-by: Ewan Milne <emilne@redhat.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For big endian platforms, reading the axes will return
invalid values.
The device stores each axis value in a 16 bit little
endian register. The driver uses regmap_read_bulk to get
the axis value, resulting in a 16 bit little endian value.
This needs to be converted to cpu endianness to work
on big endian platforms.
Fix endianness for big endian platforms by converting
the values for the axes read from little endian to
cpu.
This is also partially fixed in commit 82d8e5da1a33 ("iio:
accel: bmg160: optimize transfers in trigger handler").
When reading gyroscope axes using iio buffers, the values
returned are always 0. In the interrupt handler, the return
value of the read operation is returned to the user instead
of the value read. Return the value read to the user.
This is also fixed in commit 82d8e5da1a33 ("iio:
accel: bmg160: optimize transfers in trigger handler").
For big endian platforms, reading the axes will return
invalid values.
The device stores each axis value in a 16 bit little
endian register. The driver uses regmap_read_bulk to get
the axis value, resulting in a 16 bit little endian value.
This needs to be converted to cpu endianness to work
on big endian platforms.
Fix endianness for big endian platforms by converting
the values for the axes read from little endian to
cpu.
This is also partially fixed in commit b6fb9b6d6552 ("iio:
accel: bmc150: optimize transfers in trigger handler").
When CONFIG_IIO_TRIGGER is enabled but CONFIG_IIO_BUFFER is
not, we get a build error in the st_magn driver:
drivers/iio/magnetometer/st_magn_core.c:573:23: error: 'ST_MAGN_TRIGGER_SET_STATE' undeclared here (not in a function)
.set_trigger_state = ST_MAGN_TRIGGER_SET_STATE,
^~~~~~~~~~~~~~~~~~~~~~~~~
Apparently, this ST_MAGN_TRIGGER_SET_STATE macro was meant to
be set to NULL when the definition is not available because
st_magn_buffer.c is not compiled, but the alternative definition
was not included in the original patch. This adds it.
config structure is set to 0 when updating the buffers, so by
default config->watermark will be 0. When computing the minimum
between config->watermark and the buffer->watermark or
insert_buffer-watermark, this will always be 0 regardless of the
value set by the user for the buffer.
Set as initial value for config->watermark the maximum allowed
value so that the minimum value will always be set from one of the
buffers.
Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Fixes: f0566c0c405d ("iio: Set device watermark based on watermark of all
attached buffers") Signed-off-by: Jonathan Cameron <jic23@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It was reported that a kernel with CONFIG_ARM_PATCH_IDIV=y stopped
booting when compiled with the upcoming gcc 6. Turns out that turning
a function address into a writable array is undefined and gcc 6 decided
it was OK to omit the store to the first word of the function while
still preserving the store to the second word.
Even though gcc 6 is now fixed to behave more coherently, it is a
mystery that gcc 4 and gcc 5 actually produce wanted code in the kernel.
And in fact the reduced test case to illustrate the issue does indeed
break with gcc < 6 as well.
In any case, let's guard the kernel against undefined compiler behavior
by hiding the nature of the array location as suggested by gcc
developers.
Signed-off-by: Nicolas Pitre <nico@linaro.org> Reported-by: Marcin Juszkiewicz <mjuszkiewicz@redhat.com> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The patch "scsi: rescan VPD attributes" introduced a regression in which
devices that don't support VPD were being scanned for VPD attributes
anyway. This could cause issues for some devices and should be avoided
so the check for scsi_level has been moved out of scsi_add_lun and into
scsi_attach_vpd so that all callers will not scan VPD for devices that
don't support it.
[mkp: Merge fix]
Fixes: 09e2b0b14690 ("scsi: rescan VPD attributes") Suggested-by: Alexander Duyck <aduyck@mirantis.com> Signed-off-by: Hannes Reinecke <hare@suse.com> Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch fixes an issue that usbhsg_queue_done() may cause kernel
panic when dma callback is running and usb_ep_disable() is called
by interrupt handler. (Especially, we can reproduce this issue using
g_audio with usb-dmac driver.)
For example of a flow:
usbhsf_dma_complete (on tasklet)
--> usbhsf_pkt_handler (on tasklet)
--> usbhsg_queue_done (on tasklet)
*** interrupt happened and usb_ep_disable() is called ***
--> usbhsg_queue_pop (on tasklet)
Then, oops happened.
This patch adds a code to surely disable TX IRQ of the pipe before
starting TX DMAC transfer. Otherwise, a lot of unnecessary TX IRQs
may happen in rare cases when DMAC is used.
When unexpected situation happened (e.g. tx/rx irq happened while
DMAC is used), the usbhsf_pkt_handler() was possible to cause NULL
pointer dereference like the followings:
The current implementation only uses the first byte in val,
the second byte is always 0. Change it to use cpu_to_le16
to write the two bytes into the register
Signed-off-by: Yong Li <sdliyong@gmail.com> Reviewed-by: Phil Reid <preid@electromag.com.au> Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The driver can access the queue simultanously
while mac80211 tears down the interface. Without
spinlock protection this could lead to corrupting
sk_buff_head and subsequently to an invalid
pointer dereference.
Fixes: ba8c3d6f16a1 ("mac80211: add an intermediate software queue implementation") Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The ieee80211_queue_stopped() expects hw queue
number but it was given raw WMM AC number instead.
This could cause frame drops and problems with
traffic in some cases - most notably if driver
doesn't map AC numbers to queue numbers 1:1 and
uses ieee80211_stop_queues() and
ieee80211_wake_queue() only without ever calling
ieee80211_wake_queues().
On ath10k it was possible to hit this problem in
the following case:
1. wlan0 uses queue 0
(ath10k maps queues per vif)
2. offchannel uses queue 15
3. queues 1-14 are unused
4. ieee80211_stop_queues()
5. ieee80211_wake_queue(q=0)
6. ieee80211_wake_queue(q=15)
(other queues are not woken up because both
driver and mac80211 know other queues are
unused)
7. ieee80211_rx_h_mesh_fwding()
8. ieee80211_select_queue_80211() returns 2
9. ieee80211_queue_stopped(q=2) returns true
10. frame is dropped (oops!)
When joining IBSS a full scan should be initiated in order to search
for existing cell, unless the fixed_channel parameter was set.
A default channel to create the IBSS on if no cell was found is
provided as well.
However - a scan is initiated only on the default channel provided
regardless of whether ifibss->fixed_channel is set or not, with the
obvious result of the cell not joining existing IBSS cell that is
on another channel.
When CONFIG_OPTIMIZE_INLINING is set, the sta_info_insert_finish
function consumes more stack than normally, exceeding the
1024 byte limit on ARM:
net/mac80211/sta_info.c: In function 'sta_info_insert_finish':
net/mac80211/sta_info.c:561:1: error: the frame size of 1080 bytes is larger than 1024 bytes [-Werror=frame-larger-than=]
It turns out that there are two functions that put a 'struct station_info'
on the stack: __sta_info_destroy_part2 and sta_info_insert_finish, and
this structure alone requires up to 792 bytes.
Hoping that both are called rarely enough, this replaces the
on-stack structure with a dynamic allocation, which unfortunately
requires some suboptimal error handling for out-of-memory.
The __sta_info_destroy_part2 function is actually affected by the
stack usage twice because it calls cfg80211_del_sta_sinfo(), which
has another instance of struct station_info on its stack.
Signed-off-by: Arnd Bergmann <arnd@arndb.de> Fixes: 98b6218388e3 ("mac80211/cfg80211: add station events") Fixes: 6f7a8d26e266 ("mac80211: send statistics with delete station event") Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The original hand-implemented hash-table in mac80211 couldn't result
in insertion errors, and while converting to rhashtable I evidently
forgot to check the errors.
This surfaced now only because Ben is adding many identical keys and
that resulted in hidden insertion errors.
Fixes: 7bedd0cfad4e1 ("mac80211: use rhashtable for station table") Reported-by: Ben Greear <greearb@candelatech.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The spec says: after writing 0 to device_status, the driver MUST wait
for a read of device_status to return 0 before reinitializing the
device.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Acked-by: Jason Wang <jasowang@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
As of 5a60e87603c4c533492c515b7f62578189b03c9c, RBD object request
allocations are made via rbd_obj_request_create() with GFP_NOIO.
However, subsequent OSD request allocations in rbd_osd_req_create*()
use GFP_ATOMIC.
With heavy page cache usage (e.g. OSDs running on same host as krbd
client), rbd_osd_req_create() order-1 GFP_ATOMIC allocations have been
observed to fail, where direct reclaim would have allowed GFP_NOIO
allocations to succeed.
Suggested-by: Vlastimil Babka <vbabka@suse.cz> Suggested-by: Neil Brown <neilb@suse.com> Signed-off-by: David Disseldorp <ddiss@suse.de> Signed-off-by: Ilya Dryomov <idryomov@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
remove the usage of removed irq_to_gpio() function. On pre-DB1200
boards, pass the actual carddetect GPIO number instead of the IRQ,
because we need the gpio to actually test card status (inserted or
not) and can get the irq number with gpio_to_irq() instead.
Tested on DB1300 and DB1500, this patch fixes PCMCIA on the DB1500,
which used irq_to_gpio().
Fixes: 832f5dacfa0b ("MIPS: Remove all the uses of custom gpio.h") Signed-off-by: Manuel Lauss <manuel.lauss@gmail.com> Acked-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Linus Walleij <linus.walleij@linaro.org> Cc: linux-pcmcia@lists.infradead.org Cc: Linux-MIPS <linux-mips@linux-mips.org>
Patchwork: https://patchwork.linux-mips.org/patch/12747/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 58f896d859ce ("[media] v4l: vsp1: sru: Make the intensity
controllable during streaming") refactored the stream start code and
removed the SRU CTRL0 register write by mistake. Add it back.
Fixes: 58f896d859ce ("[media] v4l: vsp1: sru: Make the intensity controllable during streaming") Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The au0828 dev_state is actually a bit mask. It should not be
checking with "==" but, instead, with a logic and. There are some
places where it was doing it wrong.
Fix that by replacing the dev_state set/clear/test with the
bitops.
As reviewed by Shuah:
"Looks good. Tested running bind/unbind au0828 loop for 1000 times.
Didn't see any problems and the v4l2_querycap() problem has been
fixed with this patch.
After the above test, ran bind/unbind snd_usb_audio 1000 times.
Didn't see any problems. Generated media graph and the graph
looks good."
au0828_v4l2_close() check for dev_state == DEV_DISCONNECTED will fail to
detect the device disconnected state correctly, if au0828_v4l2_open() runs
to set the DEV_INITIALIZED bit. A loop test of bind/unbind found this bug
by increasing the likelihood of au0828_v4l2_open() occurring while unbind
is in progress. When au0828_v4l2_close() fails to detect that the device
is in disconnect state, it attempts to power down the device and fails with
the following general protection fault:
In legacy pxa builds, ie. non device-tree and platform-data only builds,
pinctrl is not yet available. As a consequence, the pinctrl gpio
direction change function is a stub, returning always success.
In the current state, the gpio driver direction function believes the
pinctrl direction change was successful, and exits without actually
changing the gpio direction.
This patch changes the logic :
- if the pinctrl direction function fails, gpio direction will report
that failure
- if the pinctrl direction function succeeds, gpio direction is changed
by the gpio driver anyway.
This is sub optimal in the pinctrl aware case, as the gpio direction
will be changed twice: once by pinctrl function and another time by
the gpio direction function.
Yet it should be acceptable in this form, as this is functional for all
pxa platforms (device-tree and platform-data), and moreover changing a
gpio direction is very very seldom, usually in machine initialization,
seldom in drivers probe, and an exception for ac97 reset bug.
Fixes: a770d946371e ("gpio: pxa: add pin control gpio direction and request") Reported-by: Guenter Roeck <guenter@roeck-us.net> Tested-by: Guenter Roeck <guenter@roeck-us.net> Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr> Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
pinctrl-sun8i-a33.c (and the dts) declare only 2 interrupt banks,
where as the closely related a23 has 3 banks. This matches with the
datasheet for the A33 where only interrupt banks B and G are specified
where as the A23 has banks A, B and G.
However the A33 being the A23 derative it is means that the interrupt
configure/status io-addresses for the 2 banks it has are not changed
from the A23, iow they have the same address as if bank A was still
present. Where as the sunxi pinctrl currently tries to use the A23 bank
A addresses for bank B, since the pinctrl code does not know about the
removed bank A.
Add a irq_bank_base parameter and use this where appropriate to take
the missing bank A into account.
This fixes external interrupts not working on the A33 (tested with
an i2c touchscreen controller which uses an external interrupt).
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If pinctrl_provide_dummies() is used unconditionally, then the dummy
state will be used even on DT platforms when the "init" state was
intentionally left out. Instead of "default", the dummy "init" state
will then be used during probe. Thus, when probing an I2C controller on
cold boot, communication triggered by bus notifiers broke because the
pins were not initialized.
Do it like OMAP2: use the dummy state only for non-DT platforms.
Fixes: ef0eebc05130 ("drivers/pinctrl: Add the concept of an "init" state") Reported-by: Geert Uytterhoeven <geert+renesas@glider.be> Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Acked-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> Tested-by: Geert Uytterhoeven <geert+renesas@glider.be> Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Copying the content of an MSA vector from user memory may involve TLB
faults & mapping in pages. This will fail when preemption is disabled
due to an inability to acquire mmap_sem from do_page_fault, which meant
such vector loads to unmapped pages would always fail to be emulated.
Fix this by disabling preemption later only around the updating of
vector register state.
This change does however introduce a race between performing the load
into thread context & the thread being preempted, saving its current
live context & clobbering the loaded value. This should be a rare
occureence, so optimise for the fast path by simply repeating the load if
we are preempted.
Additionally if the copy failed then the failure path was taken with
preemption left disabled, leading to the kernel typically encountering
further issues around sleeping whilst atomic. The change to where
preemption is disabled avoids this issue.
Fixes: e4aa1f153add "MIPS: MSA unaligned memory access support" Reported-by: James Hogan <james.hogan@imgtec.com> Signed-off-by: Paul Burton <paul.burton@imgtec.com> Reviewed-by: James Hogan <james.hogan@imgtec.com> Cc: Leonid Yegoshin <Leonid.Yegoshin@imgtec.com> Cc: Maciej W. Rozycki <macro@linux-mips.org> Cc: James Cowgill <James.Cowgill@imgtec.com> Cc: Markos Chandras <markos.chandras@imgtec.com> Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/12345/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Windows lets applications choose the frequency of the timer tick,
and in Windows 10 the maximum rate was changed from 1024 Hz to
2048 Hz. Unfortunately, because of the way the Windows API
works, most applications who need a higher rate than the default
64 Hz will just do
and pick the maximum rate. This causes very high CPU usage when
playing media or games on Windows 10, even if the guest does not
actually use the CPU very much, because the frequent timer tick
causes halt_poll_ns to kick in.
There is no really good solution, especially because Microsoft
could sooner or later bump the limit to 4096 Hz, but for now
the best we can do is lower a bit the upper limit for
halt_poll_ns. :-(
Reported-by: Jon Panozzo <jonp@lime-technology.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Non maskable interrupts (NMI) are preferred to interrupts in current
implementation. If a NMI is pending and NMI is blocked by the result
of nmi_allowed(), pending interrupt is not injected and
enable_irq_window() is not executed, even if interrupts injection is
allowed.
In old kernel (e.g. 2.6.32), schedule() is often called in NMI context.
In this case, interrupts are needed to execute iret that intends end
of NMI. The flag of blocking new NMI is not cleared until the guest
execute the iret, and interrupts are blocked by pending NMI. Due to
this, iret can't be invoked in the guest, and the guest is starved
until block is cleared by some events (e.g. canceling injection).
This patch injects pending interrupts, when it's allowed, even if NMI
is blocked. And, If an interrupts is pending after executing
inject_pending_event(), enable_irq_window() is executed regardless of
NMI pending counter.
Signed-off-by: Yuki Shibuya <shibuya.yk@ncos.nec.co.jp> Suggested-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add a new NO_REPORT_LUNS quirk and set it for Seagate drives with
an usb-id of: 0bc2:331a, as these will fail to respond to a
REPORT_LUNS command.
Reported-and-tested-by: David Webb <djw@noc.ac.uk> Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 64d513ac31bd ("scsi: use host wide tags by default") causes
the SCSI core to queue more commands then we can handle on devices with
multiple LUNs, limit the queue depth at the scsi-host level instead of
per slave to fix this.
find_outdev calls inet{,6}_fib_lookup_dev() or dev_get_by_index() to
find the output device. In case of an error, inet{,6}_fib_lookup_dev()
returns error pointer and dev_get_by_index() returns NULL. But the function
only checks for NULL and thus can end up calling dev_put on an ERR_PTR.
This patch adds an additional check for err ptr after the NULL check.
After patch:
$ip -f mpls route add 100 as 200 via inet 10.1.1.8
RTNETLINK answers: Network is unreachable
Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reported-by: David Miller <davem@davemloft.net> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When sending a UDPv6 message longer than MTU, account for the length
of fragmentable IPv6 extension headers in skb->network_header offset.
Same as we do in alloc_new_skb path in __ip6_append_data().
This ensures that later on __ip6_make_skb() will make space in
headroom for fragmentable extension headers:
/* move skb->data to ip header from ext header */
if (skb->data < skb_network_header(skb))
__skb_pull(skb, skb_network_offset(skb));
Reported-by: Ji Jianwen <jiji@redhat.com> Signed-off-by: Jakub Sitnicki <jkbs@redhat.com> Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When creating an ip6tnl tunnel with ip tunnel, rtnl_link_ops is not set
before ip6_tnl_create2 is called. When register_netdevice is called, there
is no linkinfo attribute in the NEWLINK message because of that.
Setting rtnl_link_ops before calling register_netdevice fixes that.
Fixes: 0b112457229d ("ip6tnl: add support of link creation via rtnl") Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@redhat.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
pskb_may_pull() can change skb->data, so we have to load ptr/optr at the
right place.
Signed-off-by: Haishuang Yan <yanhaishuang@cmss.chinamobile.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
pskb_may_pull() can change skb->data, so we have to load ptr/optr at the
right place.
Signed-off-by: Haishuang Yan <yanhaishuang@cmss.chinamobile.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
After commit f84bb1eac027 ("net: fix IFF_NO_QUEUE for drivers using
alloc_netdev"), default qdisc was changed to noqueue because
tuntap does not set tx_queue_len during .setup(). This patch restores
default qdisc by setting tx_queue_len in tun_setup().
Fixes: f84bb1eac027 ("net: fix IFF_NO_QUEUE for drivers using alloc_netdev") Cc: Phil Sutter <phil@nwl.cc> Signed-off-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Acked-by: Phil Sutter <phil@nwl.cc> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Same can be triggered with PROVE_RCU (+ PROVE_RCU_REPEATEDLY) enabled
from tun_attach_filter() when user space calls ioctl(tun_fd, TUN{ATTACH,
DETACH}FILTER, ...) for adding/removing a BPF filter on tap devices.
Since the fix in f91ff5b9ff52 ("net: sk_{detach|attach}_filter() rcu
fixes") sk_attach_filter()/sk_detach_filter() now dereferences the
filter with rcu_dereference_protected(), checking whether socket lock
is held in control path.
Since its introduction in 994051625981 ("tun: socket filter support"),
tap filters are managed under RTNL lock from __tun_chr_ioctl(). Thus the
sock_owned_by_user(sk) doesn't apply in this specific case and therefore
triggers the false positive.
Extend the BPF API with __sk_attach_filter()/__sk_detach_filter() pair
that is used by tap filters and pass in lockdep_rtnl_is_held() for the
rcu_dereference_protected() checks instead.
Reported-by: Sasha Levin <sasha.levin@oracle.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Size of the attribute IFLA_PHYS_PORT_NAME was missing.
Fixes: db24a9044ee1 ("net: add support for phys_port_name") CC: David Ahern <dsahern@gmail.com> Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Acked-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When NET_SWITCHDEV=n, switchdev_port_attr_set will return -EOPNOTSUPP,
we should ignore this error code and continue to set the ageing time.
Fixes: c62987bbd8a1 ("bridge: push bridge setting ageing_time down to switchdev") Signed-off-by: Haishuang Yan <yanhaishuang@cmss.chinamobile.com> Acked-by: Ido Schimmel <idosch@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
IPv6 counters updates use a different macro than IPv4.
Fixes: 36cbb2452cbaf ("udp: Increment UDP_MIB_IGNOREDMULTI for arriving unmatched multicasts") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Rick Jones <rick.jones2@hp.com> Cc: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reported-by: Thomas Schäfer <tschaefer@t-online.de> Signed-off-by: Bjørn Mork <bjorn@mork.no> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A crash is observed when a decrypted packet is processed in receive
path. get_rps_cpus() tries to dereference the skb->dev fields but it
appears that the device is freed from the poison pattern.
-013|get_rps_cpu(
| dev = 0xFFFFFFC08B688000,
| skb = 0xFFFFFFC0C76AAC00 -> (
| dev = 0xFFFFFFC08B688000 -> (
| name =
"......................................................
| name_hlist = (next = 0xAAAAAAAAAAAAAAAA, pprev =
0xAAAAAAAAAAA
Following are the sequence of events observed -
- Encrypted packet in receive path from netdevice is queued
- Encrypted packet queued for decryption (asynchronous)
- Netdevice brought down and freed
- Packet is decrypted and returned through callback in esp_input_done
- Packet is queued again for process in network stack using netif_rx
Since the device appears to have been freed, the dereference of
skb->dev in get_rps_cpus() leads to an unhandled page fault
exception.
Fix this by holding on to device reference when queueing packets
asynchronously and releasing the reference on call back return.
v2: Make the change generic to xfrm as mentioned by Steffen and
update the title to xfrm
Suggested-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Jerome Stanislaus <jeromes@codeaurora.org> Signed-off-by: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Let channels hold a reference on their network namespace.
Some channel types, like ppp_async and ppp_synctty, can have their
userspace controller running in a different namespace. Therefore they
can't rely on them to preclude their netns from being removed from
under them.
==================================================================
BUG: KASAN: use-after-free in ppp_unregister_channel+0x372/0x3a0 at
addr ffff880064e217e0
Read of size 8 by task syz-executor/11581
=============================================================================
BUG net_namespace (Not tainted): kasan: bad access detected
-----------------------------------------------------------------------------
Field fl4.flowi4_flags is not initialized in fib_compute_spec_dst()
before calling fib_lookup(), which means fib_table_lookup() is
using non-deterministic data at this line:
if (!(flp->flowi4_flags & FLOWI_FLAG_SKIP_NH_OIF)) {
Fix by initializing the entire fl4 structure, which will prevent
similar issues as fields are added in the future by ensuring that
all fields are initialized to zero unless explicitly initialized
to another value.
Fixes: 58189ca7b2741 ("net: Fix vti use case with oif in dst lookups") Suggested-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: Lance Richardson <lrichard@redhat.com> Acked-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, ingress ipv4 broadcast datagrams are dropped since,
in udp_v4_early_demux(), ip_check_mc_rcu() is invoked even on
bcast packets.
This patch addresses the issue, invoking ip_check_mc_rcu()
only for mcast packets.
Fixes: 6e5403093261 ("ipv4/udp: Verify multicast group is ours in upd_v4_early_demux()") Signed-off-by: Paolo Abeni <pabeni@redhat.com> Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
bond_get_stats() can be called from rtnetlink (with RTNL held)
or from /proc/net/dev seq handler (with RCU held)
The logic added in commit 5f0c5f73e5ef ("bonding: make global bonding
stats more reliable") kind of assumed only one cpu could run there.
If multiple threads are reading /proc/net/dev, stats can be really
messed up after a while.
A second problem is that some fields are 32bit, so we need to properly
handle the wrap around problem.
Given that RTNL is not always held, we need to use
bond_for_each_slave_rcu().
Fixes: 5f0c5f73e5ef ("bonding: make global bonding stats more reliable") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Andy Gospodarek <gospo@cumulusnetworks.com> Cc: Jay Vosburgh <j.vosburgh@gmail.com> Cc: Veaceslav Falico <vfalico@gmail.com> Reviewed-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 un-mapping skb->data in __bcmgenet_tx_reclaim(),
we must use the length that was used in original dma_map_single(),
instead of skb->len that might be bigger (includes the frags)
We simply can store skb_len into tx_cb_ptr->dma_len and use it
at unmap time.
Fixes: 1c1008c793fa ("net: bcmgenet: add main driver file") Signed-off-by: Eric Dumazet <edumazet@google.com> Acked-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
projects / thirdparty / kernel / stable.git / log
