Building an allmodconfig kernel fails horribly because of
endian mismatch. It turns out that the -mlittle-endian
switch was not honored at all as we were using the wrong
Kconfig symbol and failing to apply CPUFLAGS to the CFLAGS.
Finally, the linker flags did not get set right.
This addresses all three of those issues, which now lets
me build both big-endian and little-endian kernels for
testing.
Fixes: 428dbf156cc5 ("arch: change default endian for microblaze") Fixes: 206d3642d8ee ("arch/microblaze: add choice for endianness and update Makefile") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Michal Simek <michal.simek@xilinx.com> Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When a request is preempted, it is unsubmitted from the HW queue and
removed from the active list of breadcrumbs. In the process, this
however triggers the signaler and it may see the clear rbtree with the
old, and still valid, seqno, or it may match the cleared seqno with the
now zero rq->global_seqno. This confuses the signaler into action and
signaling the fence.
__show_regs pretty prints PC and LR by attempting to map them to kernel
function names to improve the utility of crash reports. Unfortunately,
this mapping is applied even when the pt_regs corresponds to user mode,
resulting in a KASLR oracle.
Avoid this issue by only looking up the function symbols when the register
state indicates that we're actually running at EL1.
Add quirks for handling PX/HG systems. In this case, add
a quirk for a weston dGPU that only seems to properly power
down using ATPX power control rather than HG (_PR3).
v2: append a new weston XT
Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com> (v2) Reviewed-and-Tested-by: Junwei Zhang <Jerry.Zhang@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Cc: stable@vger.kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Similar to the CPU address space the VA on Vega10 has a hole in it.
v2: use dev_dbg instead of dev_err
v3: add some more comments to explain how the hw works
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> CC: stable@vger.kernel.org Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
During eviction, the driver may free more than one hole in the drm_mm
due to the side-effects in evicting the scanned nodes. However,
drm_mm_scan_color_evict() expects that the scan result is the first
available hole (in the mru freed hole_stack list):
During a non-blocking commit, it is possible to return before the
commit_tail work is queued (-ERESTARTSYS, for example).
Since a reference on the crtc commit object is obtained for the pending
vblank event when preparing the commit, the above situation will leave
us with an extra reference.
Therefore, if the commit_tail worker has not consumed the event at the
end of a commit, release it's reference.
Changes since v1:
- Also check for state->event->base.completion being set, to
handle the case where stall_checks() fails in setup_crtc_commit().
Changes since v2:
- Add a flag to drm_crtc_commit, to prevent dereferencing a freed event.
i915 may unreference the state in a worker.
Fixes: 24835e442f28 ("drm: reference count event->completion") Cc: <stable@vger.kernel.org> # v4.11+ Signed-off-by: Leo (Sunpeng) Li <sunpeng.li@amd.com> Acked-by: Harry Wentland <harry.wentland@amd.com> #v1 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180117115108.29608-1-maarten.lankhorst@linux.intel.com Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This fixes an issue that a gadget driver (usb_f_fs) is possible to
stop rx transactions after the usb-dmac is used because the following
functions missed to set/check the "running" flag.
- usbhsf_dma_prepare_pop_with_usb_dmac()
- usbhsf_dma_pop_done_with_usb_dmac()
So, if next transaction uses pio, the usbhsf_prepare_pop() can not
start the transaction because the "running" flag is 0.
Fixes: 8355b2b3082d ("usb: renesas_usbhs: fix the behavior of some usbhs_pkt_handle") Cc: <stable@vger.kernel.org> # v3.19+ Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In commit 2bfa0719ac2a ("usb: gadget: function: f_fs: pass
companion descriptor along") there is a pointer arithmetic
bug where the comp_desc is obtained as follows:
Since ds is a pointer to usb_endpoint_descriptor, adding
7 to it ends up going out of bounds (7 * sizeof(struct
usb_endpoint_descriptor), which is actually 7*9 bytes) past
the SS descriptor. As a result the maxburst value will be
read incorrectly, and the UDC driver will also get a garbage
comp_desc (assuming it uses it).
Since Felipe wrote, "Eventually, f_fs.c should be converted
to use config_ep_by_speed() like all other functions, though",
let's finally do it. This allows the other usb_ep fields to
be properly populated, such as maxpacket and mult. It also
eliminates the awkward speed-based descriptor lookup since
config_ep_by_speed() does that already using the ones found
in struct usb_function.
During _ffs_func_bind(), the received descriptors are evaluated
to prepare for binding with the gadget in order to allocate
endpoints and optionally set up OS descriptors. However, the
high- and super-speed descriptors are only parsed based on
whether the gadget_is_dualspeed() and gadget_is_superspeed()
calls are true, respectively.
This is a problem in case a userspace program always provides
all of the {full,high,super,OS} descriptors when configuring a
function. Then, for example if a gadget device is not capable
of SuperSpeed, the call to ffs_do_descs() for the SS descriptors
is skipped, resulting in an incorrect offset calculation for
the vla_ptr when moving on to the OS descriptors that follow.
This causes ffs_do_os_descs() to fail as it is now looking at
the SS descriptors' offset within the raw_descs buffer instead.
_ffs_func_bind() should evaluate the descriptors unconditionally,
so remove the checks for gadget speed.
Fixes: f0175ab51993 ("usb: gadget: f_fs: OS descriptors support") Cc: stable@vger.kernel.org Co-Developed-by: Mayank Rana <mrana@codeaurora.org> Signed-off-by: Mayank Rana <mrana@codeaurora.org> Signed-off-by: Jack Pham <jackp@codeaurora.org> Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit dbac5d07d13e ("usb: musb: host: don't start next rx urb if current one failed")
along with commit b5801212229f ("usb: musb: host: clear rxcsr error bit if set")
try to solve the issue described in [1], but the latter alone is
sufficient, and the former causes the issue as in [2], so now revert it.
Commit e93650994a95 ("usb: phy: mxs: add usb charger type detection")
causes the following kernel hang on i.MX28:
[ 2.207973] usbcore: registered new interface driver usb-storage
[ 2.235659] Unable to handle kernel NULL pointer dereference at virtual address 00000188
[ 2.244195] pgd = (ptrval)
[ 2.246994] [00000188] *pgd=00000000
[ 2.250676] Internal error: Oops: 5 [#1] ARM
[ 2.254979] Modules linked in:
[ 2.258089] CPU: 0 PID: 1 Comm: swapper Not tainted 4.15.0-rc8-next-20180117-00002-g75d5f21 #7
[ 2.266724] Hardware name: Freescale MXS (Device Tree)
[ 2.271921] PC is at regmap_read+0x0/0x5c
[ 2.275977] LR is at mxs_phy_charger_detect+0x34/0x1dc
mxs_phy_charger_detect() makes accesses to the anatop registers via regmap,
however i.MX23/28 do not have such registers, which causes a NULL pointer
dereference.
Fix the issue by doing a NULL check on the 'regmap' pointer.
Fixes: e93650994a95 ("usb: phy: mxs: add usb charger type detection") Cc: <stable@vger.kernel.org> # v4.15 Reviewed-by: Li Jun <jun.li@nxp.com> Acked-by: Peter Chen <peter.chen@nxp.com> Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com> Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
DWC3 tracks TRB counter for each ep0 direction separately. In control
read transfer completion handler, the driver needs to reset the TRB
enqueue counter for ep0 IN direction. Currently the driver only resets
the TRB counter for control OUT endpoint. Check for the data direction
and properly reset the TRB counter from correct control endpoint.
Cc: stable@vger.kernel.org Fixes: c2da2ff00606 ("usb: dwc3: ep0: don't use ep0in for transfers") Signed-off-by: Thinh Nguyen <thinhn@synopsys.com> Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There are 2 control endpoint structures for DWC3. However, the driver
only updates the OUT direction control endpoint structure during
ConnectDone event. DWC3 driver needs to update the endpoint max packet
size for control IN endpoint as well. If the max packet size is not
properly set, then the driver will incorrectly calculate the data
transfer size and fail to send ZLP for HS/FS 3-stage control read
transfer.
The fix is simply to update the max packet size for the ep0 IN direction
during ConnectDone event.
commit a8c06e407ef9 ("usb: separate out sysdev pointer from usb_bus")
converted to use hcd->self.sysdev for DMA operations instead of
hcd->self.controller, but forgot to do it for hcd test mode. Replace
the correct one in this commit.
Fixes: a8c06e407ef9 ("usb: separate out sysdev pointer from usb_bus") Signed-off-by: Peter Chen <peter.chen@nxp.com> Acked-by: Alan Stern <stern@rowland.harvard.edu> Cc: stable <stable@vger.kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Similar to commit e10aec652f31 ("drm/edid: Add 6 bpc quirk for display
AEO model 0."), the EDID reports "DFP 1.x compliant TMDS" but it support
6bpc instead of 8 bpc.
Following on from this patch: https://lkml.org/lkml/2017/11/3/516,
Corsair K70 RGB keyboards also require the DELAY_INIT quirk to
start correctly at boot.
Device ids found here:
usb 3-3: New USB device found, idVendor=1b1c, idProduct=1b13
usb 3-3: New USB device strings: Mfr=1, Product=2, SerialNumber=3
usb 3-3: Product: Corsair K70 RGB Gaming Keyboard
Stop printing a (ratelimited) kernel message for each instance of an
unimplemented syscall being called. Userland making an unimplemented
syscall is not necessarily misbehaviour and to be expected with a
current userland running on an older kernel. Also, the current message
looks scary to users but does not actually indicate a real problem nor
help them narrow down the cause. Just rely on sys_ni_syscall() to return
-ENOSYS.
Cc: <stable@vger.kernel.org> Acked-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Michael Weiser <michael.weiser@gmx.de> Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is a race condition between finish_unlinks->finish_urb() function
and usb_kill_urb() in ohci controller case. The finish_urb calls
spin_unlock(&ohci->lock) before usb_hcd_giveback_urb() function call,
then if during this time, usb_kill_urb is called for another endpoint,
then new ed will be added to ed_rm_list at beginning for unlink, and
ed_rm_list will point to newly added.
When finish_urb() is completed in finish_unlinks() and ed->td_list
becomes empty as in below code (in finish_unlinks() function):
The *last = ed->ed_next will make ed_rm_list to point to ed->ed_next
and previously added ed by usb_kill_urb will be left unreferenced by
ed_rm_list. This causes usb_kill_urb() hang forever waiting for
finish_unlink to remove added ed from ed_rm_list.
The main reason for hang in this race condtion is addition and removal
of ed from ed_rm_list in the beginning during usb_kill_urb and later
last* is modified in finish_unlinks().
As suggested by Alan Stern, the solution for proper handling of
ohci->ed_rm_list is to remove ed from the ed_rm_list before finishing
any URBs. Then at the end, we can add ed back to the list if necessary.
This properly handle the updated ohci->ed_rm_list in usb_kill_urb().
Fixes: 977dcfdc6031 ("USB: OHCI: don't lose track of EDs when a controller dies") Acked-by: Alan Stern <stern@rowland.harvard.edu> CC: <stable@vger.kernel.org> Signed-off-by: Aman Deep <aman.deep@samsung.com> Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Running io_watchdog_func() while ohci_urb_enqueue() is running can
cause a race condition where ohci->prev_frame_no is corrupted and the
watchdog can mis-detect following error:
ohci-platform 664a0800.usb: frame counter not updating; disabled
ohci-platform 664a0800.usb: HC died; cleaning up
Specifically, following scenario causes a race condition:
1. ohci_urb_enqueue() calls spin_lock_irqsave(&ohci->lock, flags)
and enters the critical section
2. ohci_urb_enqueue() calls timer_pending(&ohci->io_watchdog) and it
returns false
3. ohci_urb_enqueue() sets ohci->prev_frame_no to a frame number
read by ohci_frame_no(ohci)
4. ohci_urb_enqueue() schedules io_watchdog_func() with mod_timer()
5. ohci_urb_enqueue() calls spin_unlock_irqrestore(&ohci->lock,
flags) and exits the critical section
6. Later, ohci_urb_enqueue() is called
7. ohci_urb_enqueue() calls spin_lock_irqsave(&ohci->lock, flags)
and enters the critical section
8. The timer scheduled on step 4 expires and io_watchdog_func() runs
9. io_watchdog_func() calls spin_lock_irqsave(&ohci->lock, flags)
and waits on it because ohci_urb_enqueue() is already in the
critical section on step 7
10. ohci_urb_enqueue() calls timer_pending(&ohci->io_watchdog) and it
returns false
11. ohci_urb_enqueue() sets ohci->prev_frame_no to new frame number
read by ohci_frame_no(ohci) because the frame number proceeded
between step 3 and 6
12. ohci_urb_enqueue() schedules io_watchdog_func() with mod_timer()
13. ohci_urb_enqueue() calls spin_unlock_irqrestore(&ohci->lock,
flags) and exits the critical section, then wake up
io_watchdog_func() which is waiting on step 9
14. io_watchdog_func() enters the critical section
15. io_watchdog_func() calls ohci_frame_no(ohci) and set frame_no
variable to the frame number
16. io_watchdog_func() compares frame_no and ohci->prev_frame_no
On step 16, because this calling of io_watchdog_func() is scheduled on
step 4, the frame number set in ohci->prev_frame_no is expected to the
number set on step 3. However, ohci->prev_frame_no is overwritten on
step 11. Because step 16 is executed soon after step 11, the frame
number might not proceed, so ohci->prev_frame_no must equals to
frame_no.
To address above scenario, this patch introduces a special sentinel
value IO_WATCHDOG_OFF and set this value to ohci->prev_frame_no when
the watchdog is not pending or running. When ohci_urb_enqueue()
schedules the watchdog (step 4 and 12 above), it compares
ohci->prev_frame_no to IO_WATCHDOG_OFF so that ohci->prev_frame_no is
not overwritten while io_watchdog_func() is running.
Signed-off-by: Shigeru Yoshida <Shigeru.Yoshida@windriver.com> Signed-off-by: Haiqing Bai <Haiqing.Bai@windriver.com> Acked-by: Alan Stern <stern@rowland.harvard.edu> Cc: stable <stable@vger.kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The control channel calls registered callbacks when control messages
such as XDomain protocol messages are received. The control channel
handling is done in a worker running on system workqueue which means the
networking driver can't run tear down flow which includes sending
disconnect request and waiting for a reply in the same worker. Otherwise
reply is never received (as the work is already running) and the
operation times out.
To fix this run disconnect ThunderboltIP flow asynchronously once
ThunderboltIP logout message is received.
Fixes: e69b6c02b4c3 ("net: Add support for networking over Thunderbolt cable") Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com> Cc: stable@vger.kernel.org Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When suspending to mem or disk the Thunderbolt controller typically goes
down as well tearing down the connection automatically. However, when
suspend to idle is used this does not happen so we need to make sure the
connection is properly disconnected before it can be re-established
during resume.
Fixes: e69b6c02b4c3 ("net: Add support for networking over Thunderbolt cable") Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com> Cc: stable@vger.kernel.org Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We've run into a problem where our device is attached
to a Virtual Machine and the use of the new pci_set_vpd_size()
API doesn't help. The VM kernel has been informed that
the accesses are okay, but all of the actual VPD Capability
Accesses are trapped down into the KVM Hypervisor where it
goes ahead and imposes the silent denials.
The right idea is to follow the kernel.org
commit 1c7de2b4ff88 ("PCI: Enable access to non-standard VPD for
Chelsio devices (cxgb3)") which Alexey Kardashevskiy authored
to establish a PCI Quirk for our T3-based adapters. This commit
extends that PCI Quirk to cover Chelsio T4 devices and later.
The advantage of this approach is that the VPD Size gets set early
in the Base OS/Hypervisor Boot and doesn't require that the cxgb4
driver even be available in the Base OS/Hypervisor. Thus PF4 can
be exported to a Virtual Machine and everything should work.
Fixes: 67e658794ca1 ("cxgb4: Set VPD size so we can read both VPD structures") Cc: <stable@vger.kernel.org> # v4.9+ Signed-off-by: Casey Leedom <leedom@chelsio.com> Signed-off-by: Arjun Vynipadath <arjun@chelsio.com> Signed-off-by: Ganesh Goudar <ganeshgr@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 7778c4b27cbe ("irqchip: mips-gic: Use pcpu_masks to avoid reading
GIC_SH_MASK*") removed the read of the hardware mask register when
handling shared interrupts, instead using the driver's shadow pcpu_masks
entry as the effective mask. Unfortunately this did not take account of
the write to pcpu_masks during gic_shared_irq_domain_map, which
effectively unmasks the interrupt early. If an interrupt is asserted,
gic_handle_shared_int decodes and processes the interrupt even though it
has not yet been unmasked via gic_unmask_irq, which also sets the
appropriate bit in pcpu_masks.
On the MIPS Boston board, when a console command line of
"console=ttyS0,115200n8r" is passed, the modem status IRQ is enabled in
the UART, which is immediately raised to the GIC. The interrupt has been
mapped, but no handler has yet been registered, nor is it expected to be
unmasked. However, the write to pcpu_masks in gic_shared_irq_domain_map
has effectively unmasked it, resulting in endless reports of:
To fix this, just remove the write to pcpu_masks in
gic_shared_irq_domain_map. The existing write in gic_unmask_irq is the
correct place for what is now the effective unmasking.
Cc: stable@vger.kernel.org Fixes: 7778c4b27cbe ("irqchip: mips-gic: Use pcpu_masks to avoid reading GIC_SH_MASK*") Signed-off-by: Matt Redfearn <matt.redfearn@mips.com> Reviewed-by: Paul Burton <paul.burton@mips.com> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A DMB instruction can be used to ensure the relative order of only
memory accesses before and after the barrier. Since writes to system
registers are not memory operations, barrier DMB is not sufficient
for observability of memory accesses that occur before ICC_SGI1R_EL1
writes.
A DSB instruction ensures that no instructions that appear in program
order after the DSB instruction, can execute until the DSB instruction
has completed.
Cc: stable@vger.kernel.org Acked-by: Will Deacon <will.deacon@arm.com>, Signed-off-by: Shanker Donthineni <shankerd@codeaurora.org> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This fixes a compile problem of some user space applications by not
including linux/libc-compat.h in uapi/if_ether.h.
linux/libc-compat.h checks which "features" the header files, included
from the libc, provide to make the Linux kernel uapi header files only
provide no conflicting structures and enums. If a user application mixes
kernel headers and libc headers it could happen that linux/libc-compat.h
gets included too early where not all other libc headers are included
yet. Then the linux/libc-compat.h would not prevent all the
redefinitions and we run into compile problems.
This patch removes the include of linux/libc-compat.h from
uapi/if_ether.h to fix the recently introduced case, but not all as this
is more or less impossible.
It is no problem to do the check directly in the if_ether.h file and not
in libc-compat.h as this does not need any fancy glibc header detection
as glibc never provided struct ethhdr and should define
__UAPI_DEF_ETHHDR by them self when they will provide this.
The following test program did not compile correctly any more:
Commit f7f99100d8d9 ("mm: stop zeroing memory during allocation in
vmemmap") broke Xen pv domains in some configurations, as the "Pinned"
information in struct page of early page tables could get lost.
This will lead to the kernel trying to write directly into the page
tables instead of asking the hypervisor to do so. The result is a crash
like the following:
Avoid this problem by not deferring struct page initialization when
running as Xen pv guest.
Pavel said:
: This is unique for Xen, so this particular issue won't effect other
: configurations. I am going to investigate if there is a way to
: re-enable deferred page initialization on xen guests.
[akpm@linux-foundation.org: explicitly include xen.h] Link: http://lkml.kernel.org/r/20180216154101.22865-1-jgross@suse.com Fixes: f7f99100d8d95d ("mm: stop zeroing memory during allocation in vmemmap") Signed-off-by: Juergen Gross <jgross@suse.com> Reviewed-by: Pavel Tatashin <pasha.tatashin@oracle.com> Cc: Steven Sistare <steven.sistare@oracle.com> Cc: Daniel Jordan <daniel.m.jordan@oracle.com> Cc: Bob Picco <bob.picco@oracle.com> Cc: <stable@vger.kernel.org> [4.15.x] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It was reported by Sergey Senozhatsky that if THP (Transparent Huge
Page) and frontswap (via zswap) are both enabled, when memory goes low
so that swap is triggered, segfault and memory corruption will occur in
random user space applications as follow,
After bisection, it was found the first bad commit is bd4c82c22c36 ("mm,
THP, swap: delay splitting THP after swapped out").
The root cause is as follows:
When the pages are written to swap device during swapping out in
swap_writepage(), zswap (fontswap) is tried to compress the pages to
improve performance. But zswap (frontswap) will treat THP as a normal
page, so only the head page is saved. After swapping in, tail pages
will not be restored to their original contents, causing memory
corruption in the applications.
This is fixed by refusing to save page in the frontswap store functions
if the page is a THP. So that the THP will be swapped out to swap
device.
Another choice is to split THP if frontswap is enabled. But it is found
that the frontswap enabling isn't flexible. For example, if
CONFIG_ZSWAP=y (cannot be module), frontswap will be enabled even if
zswap itself isn't enabled.
Frontswap has multiple backends, to make it easy for one backend to
enable THP support, the THP checking is put in backend frontswap store
functions instead of the general interfaces.
Link: http://lkml.kernel.org/r/20180209084947.22749-1-ying.huang@intel.com Fixes: bd4c82c22c367e068 ("mm, THP, swap: delay splitting THP after swapped out") Signed-off-by: "Huang, Ying" <ying.huang@intel.com> Reported-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com> Tested-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com> Suggested-by: Minchan Kim <minchan@kernel.org> [put THP checking in backend] Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Dan Streetman <ddstreet@ieee.org> Cc: Seth Jennings <sjenning@redhat.com> Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp> Cc: Shaohua Li <shli@kernel.org> Cc: Michal Hocko <mhocko@suse.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Mel Gorman <mgorman@techsingularity.net> Cc: Shakeel Butt <shakeelb@google.com> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com> Cc: Juergen Gross <jgross@suse.com> Cc: <stable@vger.kernel.org> [4.14] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
GCC-8 shows a warning for the x86 oprofile code that copies per-CPU
data from CPU 0 to all other CPUs, which when building a non-SMP
kernel turns into a memcpy() with identical source and destination
pointers:
arch/x86/oprofile/nmi_int.c: In function 'mux_clone':
arch/x86/oprofile/nmi_int.c:285:2: error: 'memcpy' source argument is the same as destination [-Werror=restrict]
memcpy(per_cpu(cpu_msrs, cpu).multiplex,
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
per_cpu(cpu_msrs, 0).multiplex,
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
sizeof(struct op_msr) * model->num_virt_counters);
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
arch/x86/oprofile/nmi_int.c: In function 'nmi_setup':
arch/x86/oprofile/nmi_int.c:466:3: error: 'memcpy' source argument is the same as destination [-Werror=restrict]
arch/x86/oprofile/nmi_int.c:470:3: error: 'memcpy' source argument is the same as destination [-Werror=restrict]
I have analyzed a number of such warnings now: some are valid and the
GCC warning is welcome. Others turned out to be false-positives, and
GCC was changed to not warn about those any more. This is a corner case
that is a false-positive but the GCC developers feel it's better to keep
warning about it.
In this case, it seems best to work around it by telling GCC
a little more clearly that this code path is never hit with
an IS_ENABLED() configuration check.
Cc:stable as we also want old kernels to build cleanly with GCC-8.
When a irq vector is replaced, then the previous vector is normally
released when the first interrupt happens on the new vector. If the target
CPU of the previous vector is already offline when the new vector is
installed, then the previous vector is silently discarded, which leads to
accounting issues causing suspend failures and other problems.
Adjust the logic so that the previous vector is freed in the underlying
matrix allocator to ensure that the accounting stays correct.
Fixes: 69cde0004a4b ("x86/vector: Use matrix allocator for vector assignment") Reported-by: Yuriy Vostrikov <delamonpansie@gmail.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Yuriy Vostrikov <delamonpansie@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Randy Dunlap <rdunlap@infradead.org> Cc: stable@vger.kernel.org Link: https://lkml.kernel.org/r/20180222112316.930791749@linutronix.de Signed-off-by: Ingo Molnar <mingo@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Build testing with LTO found a couple of files that get compiled
differently depending on whether asm/byteorder.h gets included early
enough or not. In particular, include/asm-generic/qrwlock_types.h is
affected by this, but there are probably others as well.
The symptom is a series of LTO link time warnings, including these:
net/netlabel/netlabel_unlabeled.h:223: error: type of 'netlbl_unlhsh_add' does not match original declaration [-Werror=lto-type-mismatch]
int netlbl_unlhsh_add(struct net *net,
net/netlabel/netlabel_unlabeled.c:377: note: 'netlbl_unlhsh_add' was previously declared here
include/net/ipv6.h:360: error: type of 'ipv6_renew_options_kern' does not match original declaration [-Werror=lto-type-mismatch]
ipv6_renew_options_kern(struct sock *sk,
net/ipv6/exthdrs.c:1162: note: 'ipv6_renew_options_kern' was previously declared here
net/core/dev.c:761: note: 'dev_get_by_name_rcu' was previously declared here
struct net_device *dev_get_by_name_rcu(struct net *net, const char *name)
net/core/dev.c:761: note: code may be misoptimized unless -fno-strict-aliasing is used
drivers/gpu/drm/i915/i915_drv.h:3377: error: type of 'i915_gem_object_set_to_wc_domain' does not match original declaration [-Werror=lto-type-mismatch]
i915_gem_object_set_to_wc_domain(struct drm_i915_gem_object *obj, bool write);
drivers/gpu/drm/i915/i915_gem.c:3639: note: 'i915_gem_object_set_to_wc_domain' was previously declared here
include/linux/debugfs.h:92:9: error: type of 'debugfs_attr_read' does not match original declaration [-Werror=lto-type-mismatch]
ssize_t debugfs_attr_read(struct file *file, char __user *buf,
fs/debugfs/file.c:318: note: 'debugfs_attr_read' was previously declared here
include/linux/rwlock_api_smp.h:30: error: type of '_raw_read_unlock' does not match original declaration [-Werror=lto-type-mismatch]
void __lockfunc _raw_read_unlock(rwlock_t *lock) __releases(lock);
kernel/locking/spinlock.c:246:26: note: '_raw_read_unlock' was previously declared here
include/linux/fs.h:3308:5: error: type of 'simple_attr_open' does not match original declaration [-Werror=lto-type-mismatch]
int simple_attr_open(struct inode *inode, struct file *file,
fs/libfs.c:795: note: 'simple_attr_open' was previously declared here
All of the above are caused by include/asm-generic/qrwlock_types.h
failing to include asm/byteorder.h after commit e0d02285f16e
("locking/qrwlock: Use 'struct qrwlock' instead of 'struct __qrwlock'")
in linux-4.15.
Similar bugs may or may not exist in older kernels as well, but there is
no easy way to test those with link-time optimizations, and kernels
before 4.14 are harder to fix because they don't have Babu's patch
series
We had similar issues with CONFIG_ symbols in the past and ended up
always including the configuration headers though linux/kconfig.h. This
works around the issue through that same file, defining either
__BIG_ENDIAN or __LITTLE_ENDIAN depending on CONFIG_CPU_BIG_ENDIAN,
which is now always set on all architectures since commit 4c97a0c8fee3
("arch: define CPU_BIG_ENDIAN for all fixed big endian archs").
The adis_probe_trigger() creates a new IIO trigger and requests an
interrupt associated with the trigger. The interrupt uses the generic
iio_trigger_generic_data_rdy_poll() function as its interrupt handler.
Currently the driver initializes some fields of the trigger structure after
the interrupt has been requested. But an interrupt can fire as soon as it
has been requested. This opens up a race condition.
iio_trigger_generic_data_rdy_poll() will access the trigger data structure
and dereference the ops field. If the ops field is not yet initialized this
will result in a NULL pointer deref.
It is not expected that the device generates an interrupt at this point, so
typically this issue did not surface unless e.g. due to a hardware
misconfiguration (wrong interrupt number, wrong polarity, etc.).
But some newer devices from the ADIS family start to generate periodic
interrupts in their power-on reset configuration and unfortunately the
interrupt can not be masked in the device. This makes the race condition
much more visible and the following crash has been observed occasionally
when booting a system using the ADIS16460.
Unable to handle kernel NULL pointer dereference at virtual address 00000008
pgd = c0004000
[00000008] *pgd=00000000
Internal error: Oops: 5 [#1] PREEMPT SMP ARM
Modules linked in:
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.9.0-04126-gf9739f0-dirty #257
Hardware name: Xilinx Zynq Platform
task: ef04f640 task.stack: ef050000
PC is at iio_trigger_notify_done+0x30/0x68
LR is at iio_trigger_generic_data_rdy_poll+0x18/0x20
pc : [<c042d868>] lr : [<c042d924>] psr: 60000193
sp : ef051bb8 ip : 00000000 fp : ef106400
r10: c081d80a r9 : ef3bfa00 r8 : 00000087
r7 : ef051bec r6 : 00000000 r5 : ef3bfa00 r4 : ee92ab00
r3 : 00000000 r2 : 00000000 r1 : 00000000 r0 : ee97e400
Flags: nZCv IRQs off FIQs on Mode SVC_32 ISA ARM Segment none
Control: 18c5387d Table: 0000404a DAC: 00000051
Process swapper/0 (pid: 1, stack limit = 0xef050210)
[<c042d868>] (iio_trigger_notify_done) from [<c0065b10>] (__handle_irq_event_percpu+0x88/0x118)
[<c0065b10>] (__handle_irq_event_percpu) from [<c0065bbc>] (handle_irq_event_percpu+0x1c/0x58)
[<c0065bbc>] (handle_irq_event_percpu) from [<c0065c30>] (handle_irq_event+0x38/0x5c)
[<c0065c30>] (handle_irq_event) from [<c0068e28>] (handle_level_irq+0xa4/0x130)
[<c0068e28>] (handle_level_irq) from [<c0064e74>] (generic_handle_irq+0x24/0x34)
[<c0064e74>] (generic_handle_irq) from [<c021ab7c>] (zynq_gpio_irqhandler+0xb8/0x13c)
[<c021ab7c>] (zynq_gpio_irqhandler) from [<c0064e74>] (generic_handle_irq+0x24/0x34)
[<c0064e74>] (generic_handle_irq) from [<c0065370>] (__handle_domain_irq+0x5c/0xb4)
[<c0065370>] (__handle_domain_irq) from [<c000940c>] (gic_handle_irq+0x48/0x8c)
[<c000940c>] (gic_handle_irq) from [<c0013e8c>] (__irq_svc+0x6c/0xa8)
To fix this make sure that the trigger is fully initialized before
requesting the interrupt.
Fixes: ccd2b52f4ac6 ("staging:iio: Add common ADIS library") Reported-by: Robin Getz <Robin.Getz@analog.com> Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Cc: <Stable@vger.kernel.org> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If no iio buffer has been set up and poll is called return 0.
Without this check there will be a null pointer dereference when
calling poll on a iio driver without an iio buffer.
Cc: stable@vger.kernel.org Signed-off-by: Stefan Windfeldt-Prytz <stefan.windfeldt@axis.com> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Functions for triggered buffer support are needed by this module.
If they are not defined accidentally by another driver, there's an error
thrown out while linking.
Add a select of IIO_BUFFER and IIO_TRIGGERED_BUFFER in the Kconfig file.
Error handling in stm32h7_adc_enable routine doesn't unwind enable
sequence correctly. ADEN can only be cleared by hardware (e.g. by
writing one to ADDIS).
It's also better to clear ADRDY just after it's been set by hardware.
Fixes: 95e339b6e85d ("iio: adc: stm32: add support for STM32H7") Signed-off-by: Fabrice Gasnier <fabrice.gasnier@st.com> Cc: <Stable@vger.kernel.org> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
==================================================================
BUG: KASAN: use-after-free in copy_ah_attr_from_uverbs+0x6f2/0x8c0
Read of size 4 at addr ffff88006476a198 by task syzkaller697701/265
There is no matching lock for this mutex. Git history suggests this is
just a missed remnant from an earlier version of the function before
this locking was moved into uverbs_free_xrcd.
Originally this lock was protecting the xrcd_table_delete()
=====================================
WARNING: bad unlock balance detected!
4.15.0+ #87 Not tainted
-------------------------------------
syzkaller223405/269 is trying to release lock (&uverbs_dev->xrcd_tree_mutex) at:
[<00000000b8703372>] ib_uverbs_close_xrcd+0x195/0x1f0
but there are no more locks to release!
other info that might help us debug this:
1 lock held by syzkaller223405/269:
#0: (&uverbs_dev->disassociate_srcu){....}, at: [<000000005af3b960>] ib_uverbs_write+0x265/0xef0
The command number is not bounds checked against the command mask before it
is shifted, resulting in an ubsan hit. This does not cause malfunction since
the command number is eventually bounds checked, but we can make this ubsan
clean by moving the bounds check to before the mask check.
The race is between lookup_get_idr_uobject and
uverbs_idr_remove_uobj -> uverbs_uobject_put.
We deliberately do not call sychronize_rcu after the idr_remove in
uverbs_idr_remove_uobj for performance reasons, instead we call
kfree_rcu() during uverbs_uobject_put.
However, this means we can obtain pointers to uobj's that have
already been released and must protect against krefing them
using kref_get_unless_zero.
==================================================================
BUG: KASAN: use-after-free in copy_ah_attr_from_uverbs.isra.2+0x860/0xa00
Read of size 4 at addr ffff88005fda1ac8 by task syz-executor2/441
Memory state around the buggy address: ffff88005fda1980: fc fc fb fb fb fb fc fc fb fb fb fb fc fc fb fb ffff88005fda1a00: fb fb fc fc fb fb fb fb fc fc 00 00 00 00 fc fc ffff88005fda1a80: fb fb fb fb fc fc fb fb fb fb fc fc fb fb fb fb ffff88005fda1b00: fc fc 00 00 00 00 fc fc fb fb fb fb fc fc fb fb ffff88005fda1b80: fb fb fc fc fb fb fb fb fc fc fb fb fb fb fc fc
==================================================================@
Cc: syzkaller <syzkaller@googlegroups.com> Cc: <stable@vger.kernel.org> # 4.11 Fixes: 3832125624b7 ("IB/core: Add support for idr types") Reported-by: Noa Osherovich <noaos@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
At CPU hotunplug the corresponding per cpu matrix allocator is shut down and
the allocated interrupt bits are discarded under the assumption that all
allocated bits have been either migrated away or shut down through the
managed interrupts mechanism.
This is not true because interrupts which are not started up might have a
vector allocated on the outgoing CPU. When the interrupt is started up
later or completely shutdown and freed then the allocated vector is handed
back, triggering warnings or causing accounting issues which result in
suspend failures and other issues.
Change the CPU hotplug mechanism of the matrix allocator so that the
remaining allocations at unplug time are preserved and global accounting at
hotplug is correctly readjusted to take the dormant vectors into account.
Some other drivers may be waiting for our extcon to show-up, exiting their
probe methods with -EPROBE_DEFER until we show up.
These drivers will typically get the cable state directly after getting
the extcon, this commit changes the int3496 code to wait for the initial
processing of the id-pin to complete before exiting probe() with 0, which
will cause devices waiting on the defered probe to get reprobed.
This fixes a race where the initial work might still be running while other
drivers were already calling extcon_get_state().
If there is a blacklisted certificate in a SignerInfo's certificate
chain, then pkcs7_verify_sig_chain() sets sinfo->blacklisted and returns
0. But, pkcs7_verify() fails to handle this case appropriately, as it
actually continues on to the line 'actual_ret = 0;', indicating that the
SignerInfo has passed verification. Consequently, PKCS#7 signature
verification ignores the certificate blacklist.
Fix this by not considering blacklisted SignerInfos to have passed
verification.
Also fix the function comment with regards to when 0 is returned.
When pkcs7_verify_sig_chain() is building the certificate chain for a
SignerInfo using the certificates in the PKCS#7 message, it is passing
the wrong arguments to public_key_verify_signature(). Consequently,
when the next certificate is supposed to be used to verify the previous
certificate, the next certificate is actually used to verify itself.
An attacker can use this bug to create a bogus certificate chain that
has no cryptographic relationship between the beginning and end.
Fortunately I couldn't quite find a way to use this to bypass the
overall signature verification, though it comes very close. Here's the
reasoning: due to the bug, every certificate in the chain beyond the
first actually has to be self-signed (where "self-signed" here refers to
the actual key and signature; an attacker might still manipulate the
certificate fields such that the self_signed flag doesn't actually get
set, and thus the chain doesn't end immediately). But to pass trust
validation (pkcs7_validate_trust()), either the SignerInfo or one of the
certificates has to actually be signed by a trusted key. Since only
self-signed certificates can be added to the chain, the only way for an
attacker to introduce a trusted signature is to include a self-signed
trusted certificate.
But, when pkcs7_validate_trust_one() reaches that certificate, instead
of trying to verify the signature on that certificate, it will actually
look up the corresponding trusted key, which will succeed, and then try
to verify the *previous* certificate, which will fail. Thus, disaster
is narrowly averted (as far as I could tell).
Fixes: 6c2dc5ae4ab7 ("X.509: Extract signature digest and make self-signed cert checks earlier") Cc: <stable@vger.kernel.org> # v4.7+ Signed-off-by: Eric Biggers <ebiggers@google.com> Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The asymmetric key type allows an X.509 certificate to be added even if
its signature's hash algorithm is not available in the crypto API. In
that case 'payload.data[asym_auth]' will be NULL. But the key
restriction code failed to check for this case before trying to use the
signature, resulting in a NULL pointer dereference in
key_or_keyring_common() or in restrict_link_by_signature().
Fix this by returning -ENOPKG when the signature is unsupported.
Reproducer when all the CONFIG_CRYPTO_SHA512* options are disabled and
keyctl has support for the 'restrict_keyring' command:
Fixes: a511e1af8b12 ("KEYS: Move the point of trust determination to __key_link()") Cc: <stable@vger.kernel.org> # v4.7+ Signed-off-by: Eric Biggers <ebiggers@google.com> Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The X.509 parser mishandles the case where the certificate's signature's
hash algorithm is not available in the crypto API. In this case,
x509_get_sig_params() doesn't allocate the cert->sig->digest buffer;
this part seems to be intentional. However,
public_key_verify_signature() is still called via
x509_check_for_self_signed(), which triggers the 'BUG_ON(!sig->digest)'.
Fix this by making public_key_verify_signature() return -ENOPKG if the
hash buffer has not been allocated.
Reproducer when all the CONFIG_CRYPTO_SHA512* options are disabled:
We were leaving them in the power on state (or the state the firmware
had set up for some client, if we were taking over from them). The
boot state was 30 core clocks, when we actually want to sample some
time after (to make sure that the new input bit has actually arrived).
Signed-off-by: Eric Anholt <eric@anholt.net> Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de> Cc: stable@kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
One I2C bus on my Atom E3845 board has been broken since 4.9.
It has two devices, both declared by ACPI and with built-in drivers.
There are two back-to-back transactions originating from the kernel, one
targeting each device. The first transaction works, the second one locks
up the I2C controller. The controller never recovers.
These kernel logs show up whenever an I2C transaction is attempted after
this failure.
i2c-designware-pci 0000:00:18.3: timeout in disabling adapter
i2c-designware-pci 0000:00:18.3: timeout waiting for bus ready
Waiting for the I2C controller status to indicate that it is enabled
before programming it fixes the issue.
I have tested this patch on 4.14 and 4.15.
Fixes: commit 2702ea7dbec5 ("i2c: designware: wait for disable/enable only if necessary") Cc: linux-stable <stable@vger.kernel.org> #4.13+ Signed-off-by: Ben Gardner <gardner.ben@gmail.com> Acked-by: Jarkko Nikula <jarkko.nikula@linux.intel.com> Reviewed-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
net/mac80211/cfg.c: In function 'cfg80211_beacon_dup':
net/mac80211/cfg.c:2896:3: error: 'memcpy' source argument is the same as destination [-Werror=restrict]
From the context, I conclude that we want to copy from beacon into
new_beacon, as we do in the rest of the function.
MIPS' struct compat_flock doesn't match the 32-bit struct flock, as it
has an extra short __unused before pad[4], which combined with alignment
increases the size to 40 bytes compared with struct flock's 36 bytes.
Since commit 8c6657cb50cb ("Switch flock copyin/copyout primitives to
copy_{from,to}_user()"), put_compat_flock() writes the full compat_flock
struct to userland, which results in corruption of the userland word
after the struct flock when running 32-bit userlands on 64-bit kernels.
This was observed to cause a bus error exception when starting Firefox
on Debian 8 (Jessie).
Reported-by: Peter Mamonov <pmamonov@gmail.com> Signed-off-by: James Hogan <jhogan@kernel.org> Tested-by: Peter Mamonov <pmamonov@gmail.com> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: linux-mips@linux-mips.org Cc: <stable@vger.kernel.org> # 4.13+
Patchwork: https://patchwork.linux-mips.org/patch/18646/ Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The fcp_rsp_info structure as defined in the FC spec has an initial 3
bytes reserved field. The ibmvfc driver mistakenly defined this field as
4 bytes resulting in the rsp_code field being defined in what should be
the start of the second reserved field and thus always being reported as
zero by the driver.
Ideally, we should wire ibmvfc up with libfc for the sake of code
deduplication, and ease of maintaining standardized structures in a
single place. However, for now simply fixup the definition in ibmvfc for
backporting to distros on older kernels. Wiring up with libfc will be
done in a followup patch.
Cc: <stable@vger.kernel.org> Reported-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Tyrel Datwyler <tyreld@linux.vnet.ibm.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Xtensa memory initialization code frees high memory pages without
checking whether they are in the reserved memory regions or not. That
results in invalid value of totalram_pages and duplicate page usage by
CMA and highmem. It produces a bunch of BUGs at startup looking like
this:
The MIPS %.its.S compiler command did not define __ASSEMBLY__, which meant
when compiler_types.h was added to kconfig.h, unexpected things appeared
(e.g. struct declarations) which should not have been present. As done in
the general %.S compiler command, __ASSEMBLY__ is now included here too.
The failure was:
Error: arch/mips/boot/vmlinux.gz.its:201.1-2 syntax error
FATAL ERROR: Unable to parse input tree
/usr/bin/mkimage: Can't read arch/mips/boot/vmlinux.gz.itb.tmp: Invalid argument
/usr/bin/mkimage Can't add hashes to FIT blob
Reported-by: kbuild test robot <lkp@intel.com> Fixes: 28128c61e08e ("kconfig.h: Include compiler types to avoid missed struct attributes") Signed-off-by: Kees Cook <keescook@chromium.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The header files for some structures could get included in such a way
that struct attributes (specifically __randomize_layout from path.h) would
be parsed as variable names instead of attributes. This could lead to
some instances of a structure being unrandomized, causing nasty GPFs, etc.
This patch makes sure the compiler_types.h header is included in
kconfig.h so that we've always got types and struct attributes defined,
since kconfig.h is included from the compiler command line.
Reported-by: Patrick McLean <chutzpah@gentoo.org> Root-caused-by: Maciej S. Szmigiero <mail@maciej.szmigiero.name> Suggested-by: Linus Torvalds <torvalds@linux-foundation.org> Tested-by: Maciej S. Szmigiero <mail@maciej.szmigiero.name> Fixes: 3859a271a003 ("randstruct: Mark various structs for randomization") Signed-off-by: Kees Cook <keescook@chromium.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Ard Biesheuvel [Fri, 23 Feb 2018 18:29:02 +0000 (18:29 +0000)]
arm64: mm: don't write garbage into TTBR1_EL1 register
Stable backport commit 173358a49173 ("arm64: kpti: Add ->enable callback
to remap swapper using nG mappings") of upstream commit f992b4dfd58b did
not survive the backporting process unscathed, and ends up writing garbage
into the TTBR1_EL1 register, rather than pointing it to the zero page to
disable translations. Fix that.
Cc: <stable@vger.kernel.org> #v4.14 Reported-by: Nicolas Dechesne <nicolas.dechesne@linaro.org> Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Acked-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The Syzbot reported a possible deadlock in the netfilter area caused by
rtnl lock, xt lock and socket lock being acquired with a different order
on different code paths, leading to the following backtrace: Reviewed-by: Xin Long <lucien.xin@gmail.com>
======================================================
WARNING: possible circular locking dependency detected
4.15.0+ #301 Not tainted
------------------------------------------------------
syzkaller233489/4179 is trying to acquire lock:
(rtnl_mutex){+.+.}, at: [<0000000048e996fd>] rtnl_lock+0x17/0x20
net/core/rtnetlink.c:74
but task is already holding lock:
(&xt[i].mutex){+.+.}, at: [<00000000328553a2>]
xt_find_table_lock+0x3e/0x3e0 net/netfilter/x_tables.c:1041
which lock already depends on the new lock.
===
Since commit 3f34cfae1230 ("netfilter: on sockopt() acquire sock lock
only in the required scope"), we already acquire the socket lock in
the innermost scope, where needed. In such commit I forgot to remove
the outer-most socket lock from the getsockopt() path, this commit
addresses the issues dropping it now.
v1 -> v2: fix bad subj, added relavant 'fixes' tag
Fixes: 22265a5c3c10 ("netfilter: xt_TEE: resolve oif using netdevice notifiers") Fixes: 202f59afd441 ("netfilter: ipt_CLUSTERIP: do not hold dev") Fixes: 3f34cfae1230 ("netfilter: on sockopt() acquire sock lock only in the required scope") Reported-by: syzbot+ddde1c7b7ff7442d7f2d@syzkaller.appspotmail.com Suggested-by: Florian Westphal <fw@strlen.de> Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Tested-by: Krzysztof Piotr Oledzki <ole@ans.pl> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Kai Heng Feng has noticed that BUG_ON(PageHighMem(pg)) triggers in
drivers/media/common/saa7146/saa7146_core.c since 19809c2da28a ("mm,
vmalloc: use __GFP_HIGHMEM implicitly").
saa7146_vmalloc_build_pgtable uses vmalloc_32 and it is reasonable to
expect that the resulting page is not in highmem. The above commit
aimed to add __GFP_HIGHMEM only for those requests which do not specify
any zone modifier gfp flag. vmalloc_32 relies on GFP_VMALLOC32 which
should do the right thing. Except it has been missed that GFP_VMALLOC32
is an alias for GFP_KERNEL on 32b architectures. Thanks to Matthew to
notice this.
Fix the problem by unconditionally setting GFP_DMA32 in GFP_VMALLOC32
for !64b arches (as a bailout). This should do the right thing and use
ZONE_NORMAL which should be always below 4G on 32b systems.
Debugged by Matthew Wilcox.
[akpm@linux-foundation.org: coding-style fixes] Link: http://lkml.kernel.org/r/20180212095019.GX21609@dhcp22.suse.cz Fixes: 19809c2da28a ("mm, vmalloc: use __GFP_HIGHMEM implicitly”) Signed-off-by: Michal Hocko <mhocko@suse.com> Reported-by: Kai Heng Feng <kai.heng.feng@canonical.com> Cc: Matthew Wilcox <willy@infradead.org> Cc: Laura Abbott <labbott@redhat.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In function xhci_stop, xhci_debugfs_exit called before xhci_mem_cleanup.
xhci_debugfs_exit removed the xhci debugfs root nodes, xhci_mem_cleanup
called function xhci_free_virt_devices_depth_first which in turn called
function xhci_debugfs_remove_slot.
Function xhci_debugfs_remove_slot removed the nodes for devices, the nodes
folders are sub folder of xhci debugfs.
It is unreasonable to remove xhci debugfs root folder before
xhci debugfs sub folder. Function xhci_mem_cleanup should be called
before function xhci_debugfs_exit.
There is a bug after plugged out USB device, the device and its ep00
nodes are still kept, we need to remove the nodes in xhci_free_dev when
USB device is plugged out.
During system resume from hibernation, xhci host is reset, all the
nodes in devices folder are removed in xhci_mem_cleanup function.
Later nodes in /sys/kernel/debug/usb/xhci/* are created again in
function xhci_run, but the nodes already exist, so the nodes still
keep the old ones, finally device nodes in xhci debugfs folder
/sys/kernel/debug/usb/xhci/*/devices/* are disappeared.
This fix removed xhci debugfs nodes before the nodes are re-created,
so all the nodes in xhci debugfs can be re-created successfully.
Commit dde634057da7 ("xhci: Fix use-after-free in xhci debugfs") causes a
null pointer dereference while fixing xhci-debugfs usage of ring pointers
that were freed during hibernate.
The fix passed addresses to ring pointers instead, but forgot to do this
change for the xhci_ring_trb_show function.
The address of the ring pointer passed to xhci-debugfs was of a temporary
ring pointer "new_ring" instead of the actual ring "ring" pointer. The
temporary new_ring pointer will be set to NULL later causing the NULL
pointer dereference.
This issue was seen when reading xhci related files in debugfs:
Since commit 152a6a884ae1 ("staging:iio:accel:sca3000 move
to hybrid hard / soft buffer design.")
the buffer mechanism has changed and the
INDIO_BUFFER_HARDWARE flag has been unused.
Since commit 2d6ca60f3284 ("iio: Add a DMAengine framework
based buffer")
the INDIO_BUFFER_HARDWARE flag has been re-purposed for
DMA buffers.
This driver has lagged behind these changes, and
in order for buffers to work, the INDIO_BUFFER_SOFTWARE
needs to be used.
Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Fixes: 2d6ca60f3284 ("iio: Add a DMAengine framework based buffer") Cc: <Stable@vger.kernel.org> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Selecting GENERIC_MSI_IRQ_DOMAIN on x86 causes a compile-time error in
some configurations:
drivers/base/platform-msi.c:37:19: error: field 'arg' has incomplete type
On the other architectures, we are fine, but here we should have an additional
dependency on X86_LOCAL_APIC so we can get the PCI_MSI_IRQ_DOMAIN symbol.
binder_send_failed_reply() is called when a synchronous
transaction fails. It reports an error to the thread that
is waiting for the completion. Given that the transaction
is synchronous, there should never be more than 1 error
response to that thread -- this was being asserted with
a WARN().
However, when exercising the driver with syzbot tests, cases
were observed where multiple "synchronous" requests were
sent without waiting for responses, so it is possible that
multiple errors would be reported to the thread. This testing
was conducted with panic_on_warn set which forced the crash.
This is easily reproduced by sending back-to-back
"synchronous" transactions without checking for any
response (eg, set read_size to 0):
The first transaction is sent to the servicemanager and the reply
fails because no VMA is set up by this client. After
binder_send_failed_reply() is called, the BINDER_WORK_RETURN_ERROR
is sitting on the thread's todo list since the read_size was 0 and
the client is not waiting for a response.
The 2nd transaction is sent and the BINDER_WORK_RETURN_ERROR has not
been consumed, so the thread's reply_error.cmd is still set (normally
cleared when the BINDER_WORK_RETURN_ERROR is handled). Therefore
when the servicemanager attempts to reply to the 2nd failed
transaction, the error is already set and it triggers this warning.
This is a user error since it is not waiting for the synchronous
transaction to complete. If it ever does check, it will see an
error.
After commit 3f34cfae1238 ("netfilter: on sockopt() acquire sock lock
only in the required scope"), the caller of nf_{get/set}sockopt() must
not hold any lock, but, in such changeset, I forgot to cope with DECnet.
This commit addresses the issue moving the nf call outside the lock,
in the dn_{get,set}sockopt() with the same schema currently used by
ipv4 and ipv6. Also moves the unhandled sockopts of the end of the main
switch statements, to improve code readability.
Reported-by: Petr Vandrovec <petr@vandrovec.name> BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=198791#c2 Fixes: 3f34cfae1238 ("netfilter: on sockopt() acquire sock lock only in the required scope") Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
dtc complains about the lack of #coolin-cells properties for the
CPU nodes that are referred to as "cooling-device":
arch/arm64/boot/dts/mediatek/mt8173-evb.dtb: Warning (cooling_device_property): Missing property '#cooling-cells' in node /cpus/cpu@0 or bad phandle (referred from /thermal-zones/cpu_thermal/cooling-maps/map@0:cooling-device[0])
arch/arm64/boot/dts/mediatek/mt8173-evb.dtb: Warning (cooling_device_property): Missing property '#cooling-cells' in node /cpus/cpu@100 or bad phandle (referred from /thermal-zones/cpu_thermal/cooling-maps/map@1:cooling-device[0])
Apparently this property must be '<2>' to match the binding.
syzbot reported a lockdep splat in gen_new_estimator() /
est_fetch_counters() when attempting to lock est->stats_lock.
Since est_fetch_counters() is called from BH context from timer
interrupt, we need to block BH as well when calling it from process
context.
Most qdiscs use per cpu counters and are immune to the problem,
but net/sched/act_api.c and net/netfilter/xt_RATEEST.c are using
a spinlock to protect their data. They both call gen_new_estimator()
while object is created and not yet alive, so this bug could
not trigger a deadlock, only a lockdep splat.
Fixes: 1c0d32fde5bd ("net_sched: gen_estimator: complete rewrite of rate estimators") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Acked-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
skb_warn_bad_offload warns when packets enter the GSO stack that
require skb_checksum_help or vice versa. Do not warn on arbitrary
bad packets. Packet sockets can craft many. Syzkaller was able to
demonstrate another one with eth_type games.
In particular, suppress the warning when segmentation returns an
error, which is for reasons other than checksum offload.
See also commit 36c92474498a ("net: WARN if skb_checksum_help() is
called on skb requiring segmentation") for context on this warning.
Signed-off-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The rds_tcp_kill_sock() function parses the rds_tcp_conn_list
to find the rds_connection entries marked for deletion as part
of the netns deletion under the protection of the rds_tcp_conn_lock.
Since the rds_tcp_conn_list tracks rds_tcp_connections (which
have a 1:1 mapping with rds_conn_path), multiple tc entries in
the rds_tcp_conn_list will map to a single rds_connection, and will
be deleted as part of the rds_conn_destroy() operation that is
done outside the rds_tcp_conn_lock.
The rds_tcp_conn_list traversal done under the protection of
rds_tcp_conn_lock should not leave any doomed tc entries in
the list after the rds_tcp_conn_lock is released, else another
concurrently executiong netns delete (for a differnt netns) thread
may trip on these entries.
Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com> Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 8edc3affc077 ("rds: tcp: Take explicit refcounts on struct net")
introduces a regression in rds-tcp netns cleanup. The cleanup_net(),
(and thus rds_tcp_dev_event notification) is only called from put_net()
when all netns refcounts go to 0, but this cannot happen if the
rds_connection itself is holding a c_net ref that it expects to
release in rds_tcp_kill_sock.
Instead, the rds_tcp_kill_sock callback should make sure to
tear down state carefully, ensuring that the socket teardown
is only done after all data-structures and workqs that depend
on it are quiesced.
The original motivation for commit 8edc3affc077 ("rds: tcp: Take explicit
refcounts on struct net") was to resolve a race condition reported by
syzkaller where workqs for tx/rx/connect were triggered after the
namespace was deleted. Those worker threads should have been
cancelled/flushed before socket tear-down and indeed,
rds_conn_path_destroy() does try to sequence this by doing
/* cancel cp_send_w */
/* cancel cp_recv_w */
/* flush cp_down_w */
/* free data structures */
Here the "flush cp_down_w" will trigger rds_conn_shutdown and thus
invoke rds_tcp_conn_path_shutdown() to close the tcp socket, so that
we ought to have satisfied the requirement that "socket-close is
done after all other dependent state is quiesced". However,
rds_conn_shutdown has a bug in that it *always* triggers the reconnect
workq (and if connection is successful, we always restart tx/rx
workqs so with the right timing, we risk the race conditions reported
by syzkaller).
Netns deletion is like module teardown- no need to restart a
reconnect in this case. We can use the c_destroy_in_prog bit
to avoid restarting the reconnect.
Fixes: 8edc3affc077 ("rds: tcp: Take explicit refcounts on struct net") Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com> Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
rateest_hash is supposed to be protected by xt_rateest_mutex,
and, as suggested by Eric, lookup and insert should be atomic,
so we should acquire the xt_rateest_mutex once for both.
So introduce a non-locking helper for internal use and keep the
locking one for external.
Reported-by: <syzbot+5cb189720978275e4c75@syzkaller.appspotmail.com> Fixes: 5859034d7eb8 ("[NETFILTER]: x_tables: add RATEEST target") Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Reviewed-by: Florian Westphal <fw@strlen.de> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Syzbot reported several deadlocks in the netfilter area caused by
rtnl lock and socket lock being acquired with a different order on
different code paths, leading to backtraces like the following one:
======================================================
WARNING: possible circular locking dependency detected
4.15.0-rc9+ #212 Not tainted
------------------------------------------------------
syzkaller041579/3682 is trying to acquire lock:
(sk_lock-AF_INET6){+.+.}, at: [<000000008775e4dd>] lock_sock
include/net/sock.h:1463 [inline]
(sk_lock-AF_INET6){+.+.}, at: [<000000008775e4dd>]
do_ipv6_setsockopt.isra.8+0x3c5/0x39d0 net/ipv6/ipv6_sockglue.c:167
but task is already holding lock:
(rtnl_mutex){+.+.}, at: [<000000004342eaa9>] rtnl_lock+0x17/0x20
net/core/rtnetlink.c:74
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
1 lock held by syzkaller041579/3682:
#0: (rtnl_mutex){+.+.}, at: [<000000004342eaa9>] rtnl_lock+0x17/0x20
net/core/rtnetlink.c:74
The problem, as Florian noted, is that nf_setsockopt() is always
called with the socket held, even if the lock itself is required only
for very tight scopes and only for some operation.
This patch addresses the issues moving the lock_sock() call only
where really needed, namely in ipv*_getorigdst(), so that nf_setsockopt()
does not need anymore to acquire both locks.
Fixes: 22265a5c3c10 ("netfilter: xt_TEE: resolve oif using netdevice notifiers") Reported-by: syzbot+a4c2dc980ac1af699b36@syzkaller.appspotmail.com Suggested-by: Florian Westphal <fw@strlen.de> Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 136e92bbec0a switched local_nodes from an array to a bitmask
but did not add proper bounds checks. As the result
clusterip_config_init_nodelist() can both over-read
ipt_clusterip_tgt_info.local_nodes and over-write
clusterip_config.local_nodes.
Add bounds checks for both.
Fixes: 136e92bbec0a ("[NETFILTER] CLUSTERIP: use a bitmap to store node responsibility data") Signed-off-by: Dmitry Vyukov <dvyukov@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
syzkaller triggered OOM kills by passing ipt_replace.size = -1
to IPT_SO_SET_REPLACE. The root cause is that SMP_ALIGN() in
xt_alloc_table_info() causes int overflow and the size check passes
when it should not. SMP_ALIGN() is no longer needed leftover.