Dave Chinner [Mon, 9 Oct 2017 18:37:22 +0000 (11:37 -0700)]
xfs: move more RT specific code under CONFIG_XFS_RT
Various utility functions and interfaces that iterate internal
devices try to reference the realtime device even when RT support is
not compiled into the kernel.
Make sure this code is excluded from the CONFIG_XFS_RT=n build,
and where appropriate stub functions to return fatal errors if
they ever get called when RT support is not present.
Dave Chinner [Mon, 9 Oct 2017 18:37:22 +0000 (11:37 -0700)]
xfs: Don't log uninitialised fields in inode structures
Prevent kmemcheck from throwing warnings about reading uninitialised
memory when formatting inodes into the incore log buffer. There are
several issues here - we don't always log all the fields in the
inode log format item, and we never log the inode the
di_next_unlinked field.
In the case of the inode log format item, this is exacerbated
by the old xfs_inode_log_format structure padding issue. Hence make
the padded, 64 bit aligned version of the structure the one we always
use for formatting the log and get rid of the 64 bit variant. This
means we'll always log the 64-bit version and so recovery only needs
to convert from the unpadded 32 bit version from older 32 bit
kernels.
Linus Torvalds [Wed, 11 Oct 2017 16:16:38 +0000 (09:16 -0700)]
Merge tag 'gpio-v4.14-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio
Pull GPIO fixes from Linus Walleij:
"Here are some smallish GPIO fixes for v4.14. Like with pin control:
some build/Kconfig noise and one serious bug in a specific driver.
- Three Kconfig/build warning fixes
- A fix for lost edge IRQs in the OMAP driver"
* tag 'gpio-v4.14-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio:
gpio: omap: Fix lost edge interrupts
gpio: omap: omap_gpio_show_rev is not __init
gpio: acpi: work around false-positive -Wstring-overflow warning
gpio: thunderx: select IRQ_DOMAIN_HIERARCHY instead of depends on
Linus Torvalds [Wed, 11 Oct 2017 16:00:22 +0000 (09:00 -0700)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Pull vfs fixes from Al Viro:
"Fairly old DIO bug caught by Andreas (3.10+) and several slightly
younger blk_rq_map_user_iov() bugs, both on map and copy codepaths
(Vitaly and me)"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
bio_copy_user_iov(): don't ignore ->iov_offset
more bio_map_user_iov() leak fixes
fix unbalanced page refcounting in bio_map_user_iov
direct-io: Prevent NULL pointer access in submit_page_section
Tom Lendacky [Tue, 10 Oct 2017 19:45:04 +0000 (14:45 -0500)]
x86/mm: Disable various instrumentations of mm/mem_encrypt.c and mm/tlb.c
Some routines in mem_encrypt.c are called very early in the boot process,
e.g. sme_enable(). When CONFIG_KCOV=y is defined the resulting code added
to sme_enable() (and others) for KCOV instrumentation results in a kernel
crash. Disable the KCOV instrumentation for mem_encrypt.c by adding
KCOV_INSTRUMENT_mem_encrypt.o := n to arch/x86/mm/Makefile.
In order to avoid other possible early boot issues, model mem_encrypt.c
after head64.c in regards to tools. In addition to disabling KCOV as
stated above and a previous patch that disables branch profiling, also
remove the "-pg" CFLAG if CONFIG_FUNCTION_TRACER is enabled and set
KASAN_SANITIZE to "n", each of which are done on a file basis.
Takashi Iwai [Wed, 11 Oct 2017 14:39:02 +0000 (16:39 +0200)]
ALSA: caiaq: Fix stray URB at probe error path
caiaq driver doesn't kill the URB properly at its error path during
the probe, which may lead to a use-after-free error later. This patch
addresses it.
HID: hid-elecom: extend to fix descriptor for HUGE trackball
In addition to DEFT, Elecom introduced a larger trackball called HUGE, in
both wired (M-HT1URBK) and wireless (M-HT1DRBK) versions. It has the same
buttons and behavior as the DEFT. This patch adds the two relevant USB IDs
to enable operation of the three Fn buttons on the top of the device.
Jaejoong Kim [Thu, 28 Sep 2017 10:16:30 +0000 (19:16 +0900)]
HID: usbhid: fix out-of-bounds bug
The hid descriptor identifies the length and type of subordinate
descriptors for a device. If the received hid descriptor is smaller than
the size of the struct hid_descriptor, it is possible to cause
out-of-bounds.
In addition, if bNumDescriptors of the hid descriptor have an incorrect
value, this can also cause out-of-bounds while approaching hdesc->desc[n].
So check the size of hid descriptor and bNumDescriptors.
BUG: KASAN: slab-out-of-bounds in usbhid_parse+0x9b1/0xa20
Read of size 1 at addr ffff88006c5f8edf by task kworker/1:2/1261
Joe Lawrence [Mon, 2 Oct 2017 15:56:48 +0000 (11:56 -0400)]
livepatch: unpatch all klp_objects if klp_module_coming fails
When an incoming module is considered for livepatching by
klp_module_coming(), it iterates over multiple patches and multiple
kernel objects in this order:
which means that if one of the kernel objects fails to patch,
klp_module_coming()'s error path needs to unpatch and cleanup any kernel
objects that were already patched by a previous patch.
Alan Stern [Fri, 29 Sep 2017 14:54:24 +0000 (10:54 -0400)]
usb: usbtest: fix NULL pointer dereference
If the usbtest driver encounters a device with an IN bulk endpoint but
no OUT bulk endpoint, it will try to dereference a NULL pointer
(out->desc.bEndpointAddress). The problem can be solved by adding a
missing test.
Andrew Gabbasov [Sat, 30 Sep 2017 15:54:52 +0000 (08:54 -0700)]
usb: gadget: configfs: Fix memory leak of interface directory data
Kmemleak checking configuration reports a memory leak in
usb_os_desc_prepare_interf_dir function when rndis function
instance is freed and then allocated again. For example, this
happens with FunctionFS driver with RNDIS function enabled
when "ffs-test" test application is run several times in a row.
The data for intermediate "os_desc" group for interface directories
is allocated as a single VLA chunk and (after a change of default
groups handling) is not ever freed and actually not stored anywhere
besides inside a list of default groups of a parent group.
The fix is to make usb_os_desc_prepare_interf_dir function return
a pointer to allocated data (as a pointer to the first VLA item)
instead of (an unused) integer and to make the caller component
(currently the only one is RNDIS function) responsible for storing
the pointer and freeing the memory when appropriate.
Andrew Gabbasov [Sat, 30 Sep 2017 15:55:55 +0000 (08:55 -0700)]
usb: gadget: composite: Fix use-after-free in usb_composite_overwrite_options
KASAN enabled configuration reports an error
BUG: KASAN: use-after-free in usb_composite_overwrite_options+...
[libcomposite] at addr ...
Read of size 1 by task ...
when some driver is un-bound and then bound again.
For example, this happens with FunctionFS driver when "ffs-test"
test application is run several times in a row.
If the driver has empty manufacturer ID string in initial static data,
it is then replaced with generated string. After driver unbinding
the generated string is freed, but the driver data still keep that
pointer. And if the driver is then bound again, that pointer
is re-used for string emptiness check.
The fix is to clean up the driver string data upon its unbinding
to drop the pointer to freed memory.
Fixes: cc2683c318a5 ("usb: gadget: Provide a default implementation of default manufacturer string") Cc: [email protected] Signed-off-by: Andrew Gabbasov <[email protected]> Signed-off-by: Felipe Balbi <[email protected]>
Dan Carpenter [Sat, 30 Sep 2017 08:15:29 +0000 (11:15 +0300)]
usb: misc: usbtest: Fix overflow in usbtest_do_ioctl()
There used to be a test against "if (param->sglen > MAX_SGLEN)" but it
was removed during a refactor. It leads to an integer overflow and a
stack overflow in test_queue() if we try to create a too large urbs[]
array on the stack.
There is a second integer overflow in test_queue() as well if
"param->iterations" is too high. I don't immediately see that it's
harmful but I've added a check to prevent it and silence the static
checker warning.
usb: renesas_usbhs: Fix DMAC sequence for receiving zero-length packet
The DREQE bit of the DnFIFOSEL should be set to 1 after the DE bit of
USB-DMAC on R-Car SoCs is set to 1 after the USB-DMAC received a
zero-length packet. Otherwise, a transfer completion interruption
of USB-DMAC doesn't happen. Even if the driver changes the sequence,
normal operations (transmit/receive without zero-length packet) will
not cause any side-effects. So, this patch fixes the sequence anyway.
Alan Stern [Fri, 6 Oct 2017 14:27:44 +0000 (10:27 -0400)]
USB: dummy-hcd: Fix deadlock caused by disconnect detection
The dummy-hcd driver calls the gadget driver's disconnect callback
under the wrong conditions. It should invoke the callback when Vbus
power is turned off, but instead it does so when the D+ pullup is
turned off.
This can cause a deadlock in the composite core when a gadget driver
is unregistered:
Jon Hunter [Mon, 2 Oct 2017 11:22:53 +0000 (12:22 +0100)]
usb: phy: tegra: Fix phy suspend for UDC
Commit dfebb5f43a78 ("usb: chipidea: Add support for Tegra20/30/114/124")
added UDC support for Tegra but with UDC support enabled, is was found
that Tegra30, Tegra114 and Tegra124 would hang on entry to suspend.
The hang occurred during the suspend of the USB PHY when the Tegra PHY
driver attempted to disable the PHY clock. The problem is that before
the Tegra PHY driver is suspended, the chipidea driver already disabled
the PHY clock and when the Tegra PHY driver suspended, it could not read
DEVLC register and caused the device to hang.
The Tegra USB PHY driver is used by both the Tegra EHCI driver and now
the chipidea UDC driver and so simply removing the disabling of the PHY
clock from the USB PHY driver would not work for the Tegra EHCI driver.
Fortunately, the status of the USB PHY clock can be read from the
USB_SUSP_CTRL register and therefore, to workaround this issue, simply
poll the register prior to disabling the clock in USB PHY driver to see
if clock gating has already been initiated. Please note that it can take
a few uS for the clock to disable and so simply reading this status
register once on entry is not sufficient.
Similarly when turning on the PHY clock, it is possible that the clock
is already enabled or in the process of being enabled, and so check for
this when enabling the PHY.
Please note that no issues are seen with Tegra20 because it has a slightly
different PHY to Tegra30/114/124.
Lucas Stach [Mon, 18 Sep 2017 15:45:07 +0000 (17:45 +0200)]
gpu: ipu-v3: pre: implement workaround for ERR009624
The PRE has a bug where a software write to the CTRL register can block
the setting of the ENABLE bit by the hardware in auto repeat mode. When
this happens the PRE will fail to handle new jobs. To work around this
software must not write to CTRL register when the PRE store engine is
inside the unsafe window, where a hardware update to the ENABLE bit
may happen.
Lucas Stach [Fri, 24 Mar 2017 17:01:53 +0000 (18:01 +0100)]
gpu: ipu-v3: prg: wait for double buffers to be filled on channel startup
Wait for both double buffer to be filled when first starting a channel.
This makes channel startup a lot more reliable, probably because it allows
the internal state machine to settle before the requests from the IPU are
coming in.
Philipp Zabel [Tue, 10 Oct 2017 13:13:55 +0000 (15:13 +0200)]
gpu: ipu-v3: Allow channel burst locking on i.MX6 only
The IDMAC_LOCK_EN registers on i.MX51 have a different layout, and on
i.MX53 enabling the lock feature causes bursts to get lost. Restrict
enabling the burst lock feature to i.MX6.
Reported-by: Patrick Brünn <[email protected]> Fixes: 790cb4c7c954 ("drm/imx: lock scanout transfers for consecutive bursts") Tested-by: Patrick Brünn <[email protected]> Signed-off-by: Philipp Zabel <[email protected]>
Takashi Iwai [Mon, 9 Oct 2017 09:09:20 +0000 (11:09 +0200)]
ALSA: seq: Fix use-after-free at creating a port
There is a potential race window opened at creating and deleting a
port via ioctl, as spotted by fuzzing. snd_seq_create_port() creates
a port object and returns its pointer, but it doesn't take the
refcount, thus it can be deleted immediately by another thread.
Meanwhile, snd_seq_ioctl_create_port() still calls the function
snd_seq_system_client_ev_port_start() with the created port object
that is being deleted, and this triggers use-after-free like:
We may fix this in a few different ways, and in this patch, it's fixed
simply by taking the refcount properly at snd_seq_create_port() and
letting the caller unref the object after use. Also, there is another
potential use-after-free by sprintf() call in snd_seq_create_port(),
and this is moved inside the lock.
Al Viro [Sun, 24 Sep 2017 14:21:15 +0000 (10:21 -0400)]
bio_copy_user_iov(): don't ignore ->iov_offset
Since "block: support large requests in blk_rq_map_user_iov" we
started to call it with partially drained iter; that works fine
on the write side, but reads create a copy of iter for completion
time. And that needs to take the possibility of ->iov_iter != 0
into account...
Al Viro [Sat, 23 Sep 2017 19:51:23 +0000 (15:51 -0400)]
more bio_map_user_iov() leak fixes
we need to take care of failure exit as well - pages already
in bio should be dropped by analogue of bio_unmap_pages(),
since their refcounts had been bumped only once per reference
in bio.
fix unbalanced page refcounting in bio_map_user_iov
bio_map_user_iov and bio_unmap_user do unbalanced pages refcounting if
IO vector has small consecutive buffers belonging to the same page.
bio_add_pc_page merges them into one, but the page reference is never
dropped.
direct-io: Prevent NULL pointer access in submit_page_section
In the code added to function submit_page_section by commit b1058b981,
sdio->bio can currently be NULL when calling dio_bio_submit. This then
leads to a NULL pointer access in dio_bio_submit, so check for a NULL
bio in submit_page_section before trying to submit it instead.
PCI: aardvark: Move to struct pci_host_bridge IRQ mapping functions
struct pci_host_bridge gained hooks to map/swizzle IRQs, so that the IRQ
mapping can be done automatically by PCI core code through the
pci_assign_irq() function instead of resorting to arch-specific
implementation callbacks to carry out the same task which force PCI host
bridge drivers implementation to implement per-arch kludges to carry out a
task that is inherently architecture agnostic.
Commit 769b461fc0c0 ("arm64: PCI: Drop DT IRQ allocation from
pcibios_alloc_irq()") was assuming all PCI host controller drivers had been
converted to use ->map_irq(), but that wasn't the case: pci-aardvark had
not been converted. Due to this, it broke the support for legacy PCI
interrupts when using the pci-aardvark driver (used on Marvell Armada 3720
platforms).
In order to fix this, we make sure the ->map_irq and ->swizzle_irq fields
of pci_host_bridge are properly filled in.
Fixes: 769b461fc0c0 ("arm64: PCI: Drop DT IRQ allocation from pcibios_alloc_irq()") Signed-off-by: Thomas Petazzoni <[email protected]> Signed-off-by: Bjorn Helgaas <[email protected]> Cc: [email protected] # v4.13+
It turns out that Tegra20 has a bug in the implementation of the MSI
target address register (which is worked around by the existence of the
struct tegra_pcie_soc.msi_base_shift parameter) that restricts the MSI
target memory to the lower 32 bits of physical memory on that particular
generation. The offending patch causes a regression on TrimSlice, which
is a Tegra20-based device and has a PCI network interface card.
An initial, simpler fix was to change the MSI target address for Tegra20
only, but it was pointed out that the offending commit also prevents the
use of 32-bit only MSI capable devices, even on later chips. Technically
this was never guaranteed to work with the prior code in the first place
because the allocated page could have resided beyond the 4 GiB boundary,
but it is still possible that this could've introduced a regression.
The proper fix that was settled on is to select a fixed address within
the lowest 32 bits of physical address space that is otherwise unused,
but testing of that patch has provided mixed results that are not fully
understood yet.
Given all of the above and the relative urgency to get this fixed in
v4.13, revert the offending commit until a universal fix is found.
Linus Torvalds [Tue, 10 Oct 2017 20:08:59 +0000 (13:08 -0700)]
Merge tag 'seccomp-v4.14-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux
Pull seccomp fixlet from Kees Cook:
"Minor seccomp fix for v4.14-rc5. I debated sending this at all for
v4.14, but since it fixes a minor issue in the prior fix, which also
went to -stable, it seemed better to just get all of it cleaned up
right now.
- fix missed "static" to avoid Sparse warning (Colin King)"
* tag 'seccomp-v4.14-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
seccomp: make function __get_seccomp_filter static
Linus Torvalds [Tue, 10 Oct 2017 20:01:51 +0000 (13:01 -0700)]
Merge tag 'nfsd-4.14-1' of git://linux-nfs.org/~bfields/linux
Pull nfsd fix from Bruce Fields:
"One fix for a 4.14 regression, and one minor fix to the MAINTAINERs
file. (I was weirdly flattered by the idea that lots of random people
suddenly seemed to think Jeff and I were VFS experts. Turns out it was
just a typo)"
* tag 'nfsd-4.14-1' of git://linux-nfs.org/~bfields/linux:
nfsd4: define nfsd4_secinfo_no_name_release()
MAINTAINERS: associate linux/fs.h with VFS instead of file locking
Colin Ian King [Fri, 29 Sep 2017 13:26:48 +0000 (14:26 +0100)]
seccomp: make function __get_seccomp_filter static
The function __get_seccomp_filter is local to the source and does
not need to be in global scope, so make it static.
Cleans up sparse warning:
symbol '__get_seccomp_filter' was not declared. Should it be static?
Signed-off-by: Colin Ian King <[email protected]> Fixes: 66a733ea6b61 ("seccomp: fix the usage of get/put_seccomp_filter() in seccomp_get_filter()") Cc: [email protected] Signed-off-by: Kees Cook <[email protected]>
When RPMSG_QCOM_GLINK_SMEM=m and one driver causes the qcom_common.c file
to be compiled as built-in, we get a link error:
drivers/remoteproc/qcom_common.o: In function `glink_subdev_remove':
qcom_common.c:(.text+0x130): undefined reference to `qcom_glink_smem_unregister'
qcom_common.c:(.text+0x130): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `qcom_glink_smem_unregister'
drivers/remoteproc/qcom_common.o: In function `glink_subdev_probe':
qcom_common.c:(.text+0x160): undefined reference to `qcom_glink_smem_register'
qcom_common.c:(.text+0x160): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `qcom_glink_smem_register'
Out of the three PIL driver instances, QCOM_ADSP_PIL already has a
Kconfig dependency to prevent this from happening, but the other two
do not. This adds the same dependency there.
Linus Torvalds [Tue, 10 Oct 2017 17:57:46 +0000 (10:57 -0700)]
Merge tag 'linux-kselftest-4.14-rc5-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest
Pull kselftest fixes from Shuah Khan:
- fix for x86: sysret_ss_attrs test build failure preventing the x86
tests from running
- fix mqueue: fix regression in silencing test run output
* tag 'linux-kselftest-4.14-rc5-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest:
selftests: mqueue: fix regression in silencing output from RUN_TESTS
selftests: x86: sysret_ss_attrs doesn't build on a PIE build
Tom Lendacky [Fri, 6 Oct 2017 21:35:40 +0000 (16:35 -0500)]
iommu/amd: Do not disable SWIOTLB if SME is active
When SME memory encryption is active it will rely on SWIOTLB to handle
DMA for devices that cannot support the addressing requirements of
having the encryption mask set in the physical address. The IOMMU
currently disables SWIOTLB if it is not running in passthrough mode.
This is not desired as non-PCI devices attempting DMA may fail. Update
the code to check if SME is active and not disable SWIOTLB.
Fixes: 2543a786aa25 ("iommu/amd: Allow the AMD IOMMU to work with memory encryption") Signed-off-by: Tom Lendacky <[email protected]> Signed-off-by: Joerg Roedel <[email protected]>
The shash ahash digest adaptor function may crash if given a
zero-length input together with a null SG list. This is because
it tries to read the SG list before looking at the length.
Jan Kara [Tue, 10 Oct 2017 12:40:42 +0000 (14:40 +0200)]
quota: Generate warnings for DQUOT_SPACE_NOFAIL allocations
Eryu has reported that since commit 7b9ca4c61bc2 "quota: Reduce
contention on dq_data_lock" test generic/233 occasionally fails. This is
caused by the fact that since that commit we don't generate warning and
set grace time for quota allocations that have DQUOT_SPACE_NOFAIL set
(these are for example some metadata allocations in ext4). We need these
allocations to behave regularly wrt warning generation and grace time
setting so fix the code to return to the original behavior.
Ladi Prosek [Thu, 5 Oct 2017 09:10:23 +0000 (11:10 +0200)]
KVM: MMU: always terminate page walks at level 1
is_last_gpte() is not equivalent to the pseudo-code given in commit 6bb69c9b69c31 ("KVM: MMU: simplify last_pte_bitmap") because an incorrect
value of last_nonleaf_level may override the result even if level == 1.
It is critical for is_last_gpte() to return true on level == 1 to
terminate page walks. Otherwise memory corruption may occur as level
is used as an index to various data structures throughout the page
walking code. Even though the actual bug would be wherever the MMU is
initialized (as in the previous patch), be defensive and ensure here
that is_last_gpte() returns the correct value.
This patch is also enough to fix CVE-2017-12188.
Fixes: 6bb69c9b69c315200ddc2bc79aee14c0184cf5b2 Cc: [email protected] Cc: Andy Honig <[email protected]> Signed-off-by: Ladi Prosek <[email protected]>
[Panic if walk_addr_generic gets an incorrect level; this is a serious
bug and it's not worth a WARN_ON where the recovery path might hide
further exploitable issues; suggested by Andrew Honig. - Paolo] Signed-off-by: Paolo Bonzini <[email protected]>
Ladi Prosek [Thu, 5 Oct 2017 09:10:22 +0000 (11:10 +0200)]
KVM: nVMX: update last_nonleaf_level when initializing nested EPT
The function updates context->root_level but didn't call
update_last_nonleaf_level so the previous and potentially wrong value
was used for page walks. For example, a zero value of last_nonleaf_level
would allow a potential out-of-bounds access in arch/x86/mmu/paging_tmpl.h's
walk_addr_generic function (CVE-2017-12188).
Takashi Iwai [Tue, 10 Oct 2017 12:10:32 +0000 (14:10 +0200)]
ALSA: usb-audio: Kill stray URB at exiting
USB-audio driver may leave a stray URB for the mixer interrupt when it
exits by some error during probe. This leads to a use-after-free
error as spotted by syzkaller like:
==================================================================
BUG: KASAN: use-after-free in snd_usb_mixer_interrupt+0x604/0x6f0
Call Trace:
<IRQ>
__dump_stack lib/dump_stack.c:16
dump_stack+0x292/0x395 lib/dump_stack.c:52
print_address_description+0x78/0x280 mm/kasan/report.c:252
kasan_report_error mm/kasan/report.c:351
kasan_report+0x23d/0x350 mm/kasan/report.c:409
__asan_report_load8_noabort+0x19/0x20 mm/kasan/report.c:430
snd_usb_mixer_interrupt+0x604/0x6f0 sound/usb/mixer.c:2490
__usb_hcd_giveback_urb+0x2e0/0x650 drivers/usb/core/hcd.c:1779
....
Actually such a URB is killed properly at disconnection when the
device gets probed successfully, and what we need is to apply it for
the error-path, too.
In this patch, we apply snd_usb_mixer_disconnect() at releasing.
Also introduce a new flag, disconnected, to struct usb_mixer_interface
for not performing the disconnection procedure twice.
x86/hyperv: Fix hypercalls with extended CPU ranges for TLB flushing
Do not consider the fixed size of hv_vp_set when passing the variable
header size to hv_do_rep_hypercall().
The Hyper-V hypervisor specification states that for a hypercall with a
variable header only the size of the variable portion should be supplied
via the input control.
For HVCALL_FLUSH_VIRTUAL_ADDRESS_SPACE_EX/LIST_EX calls that means the
fixed portion of hv_vp_set should not be considered.
That fixes random failures of some applications that are unexpectedly
killed with SIGBUS or SIGSEGV.
x86/hyperv: Don't use percpu areas for pcpu_flush/pcpu_flush_ex structures
hv_do_hypercall() does virt_to_phys() translation and with some configs
(CONFIG_SLAB) this doesn't work for percpu areas, we pass wrong memory to
hypervisor and get #GP. We could use working slow_virt_to_phys() instead
but doing so kills the performance.
Move pcpu_flush/pcpu_flush_ex structures out of percpu areas and
allocate memory on first call. The additional level of indirection gives
us a small performance penalty, in future we may consider introducing
hypercall functions which avoid virt_to_phys() conversion and cache
physical addresses of pcpu_flush/pcpu_flush_ex structures somewhere.
x86/hyperv: Clear vCPU banks between calls to avoid flushing unneeded vCPUs
hv_flush_pcpu_ex structures are not cleared between calls for performance
reasons (they're variable size up to PAGE_SIZE each) but we must clear
hv_vp_set.bank_contents part of it to avoid flushing unneeded vCPUs. The
rest of the structure is formed correctly.
To do the clearing in an efficient way stash the maximum possible vCPU
number (this may differ from Linux CPU id).
Colin Ian King [Mon, 9 Oct 2017 17:26:55 +0000 (18:26 +0100)]
perf/x86/intel/uncore: Fix memory leaks on allocation failures
Currently if an allocation fails then the error return paths
don't free up any currently allocated pmus[].boxes and pmus causing
a memory leak. Add an error clean up exit path that frees these
objects.
Detected by CoverityScan, CID#711632 ("Resource Leak")
Josh Poimboeuf [Tue, 10 Oct 2017 01:20:04 +0000 (20:20 -0500)]
x86/unwind: Align stack pointer in unwinder dump
When printing the unwinder dump, the stack pointer could be unaligned,
for one of two reasons:
- stack corruption; or
- GCC created an unaligned stack.
There's no way for the unwinder to tell the difference between the two,
so we have to assume one or the other. GCC unaligned stacks are very
rare, and have only been spotted before GCC 5. Presumably, if we're
doing an unwinder stack dump, stack corruption is more likely than a
GCC unaligned stack. So always align the stack before starting the
dump.
Josh Poimboeuf [Tue, 10 Oct 2017 01:20:03 +0000 (20:20 -0500)]
x86/unwind: Use MSB for frame pointer encoding on 32-bit
On x86-32, Tetsuo Handa and Fengguang Wu reported unwinder warnings
like:
WARNING: kernel stack regs at f60bb9c8 in swapper:1 has bad 'bp' value 0ba00000
And also there were some stack dumps with a bunch of unreliable '?'
symbols after an apic_timer_interrupt symbol, meaning the unwinder got
confused when it tried to read the regs.
The cause of those issues is that, with GCC 4.8 (and possibly older),
there are cases where GCC misaligns the stack pointer in a leaf function
for no apparent reason:
If an interrupt occurs in such a function, the regs on the stack will be
unaligned, which breaks the frame pointer encoding assumption. So on
32-bit, use the MSB instead of the LSB to encode the regs.
This isn't an issue on 64-bit, because interrupts align the stack before
writing to it.
On x86-32, after decoding a frame pointer to get a regs address,
regs_size() dereferences the regs pointer when it checks regs->cs to see
if the regs are user mode. This is dangerous because it's possible that
what looks like a decoded frame pointer is actually a corrupt value, and
we don't want the unwinder to make things worse.
Instead of calling regs_size() on an unsafe pointer, just assume they're
kernel regs to start with. Later, once it's safe to access the regs, we
can do the user mode check and corresponding safety check for the
remaining two regs.
powerpc: Don't call lockdep_assert_cpus_held() from arch_update_cpu_topology()
It turns out that not all paths calling arch_update_cpu_topology() hold
cpu_hotplug_lock, but that's OK because those paths can't race with
any concurrent hotplug events.
Which is safe because it's called as part of CPU hotplug, so although
we don't hold the CPU hotplug lock, there is another thread driving
the CPU hotplug operation which does hold the lock, and there is no
race.
Thanks to tglx for deciphering it for us.
Fixes: 3e401f7a2e51 ("powerpc: Only obtain cpu_hotplug_lock if called by rtasd") Signed-off-by: Thiago Jung Bauermann <[email protected]> Signed-off-by: Michael Ellerman <[email protected]>
Sandipan Das [Tue, 10 Oct 2017 06:45:30 +0000 (12:15 +0530)]
powerpc/lib/sstep: Fix count leading zeros instructions
According to the GCC documentation, the behaviour of __builtin_clz()
and __builtin_clzl() is undefined if the value of the input argument
is zero. Without handling this special case, these builtins have been
used for emulating the following instructions:
* Count Leading Zeros Word (cntlzw[.])
* Count Leading Zeros Doubleword (cntlzd[.])
This fixes the emulated behaviour of these instructions by adding an
additional check for this special case.
Fixes: 3cdfcbfd32b9d ("powerpc: Change analyse_instr so it doesn't modify *regs") Signed-off-by: Sandipan Das <[email protected]> Reviewed-by: Naveen N. Rao <[email protected]> Signed-off-by: Michael Ellerman <[email protected]>
Peter Zijlstra [Mon, 9 Oct 2017 08:36:53 +0000 (10:36 +0200)]
sched/core: Ensure load_balance() respects the active_mask
While load_balance() masks the source CPUs against active_mask, it had
a hole against the destination CPU. Ensure the destination CPU is also
part of the 'domain-mask & active-mask' set.
Peter Zijlstra [Fri, 6 Oct 2017 07:23:24 +0000 (09:23 +0200)]
sched/core: Address more wake_affine() regressions
The trivial wake_affine_idle() implementation is very good for a
number of workloads, but it comes apart at the moment there are no
idle CPUs left, IOW. the overloaded case.
sysbench-mysql-2 : 2135.17 per sec. 2142.51 per sec.
sysbench-mysql-5 : 4809.68 per sec. 4800.19 per sec.
sysbench-mysql-10 : 9158.59 per sec. 9157.05 per sec.
sysbench-mysql-20 : 14570.70 per sec. 14543.55 per sec.
sysbench-mysql-40 : 22130.56 per sec. 22184.82 per sec.
sysbench-mysql-80 : 20995.56 per sec. 21904.18 per sec.
sysbench-psql-2 : 1679.58 per sec. 1705.06 per sec.
sysbench-psql-5 : 3797.69 per sec. 3879.93 per sec.
sysbench-psql-10 : 7253.22 per sec. 7258.06 per sec.
sysbench-psql-20 : 11166.75 per sec. 11220.00 per sec.
sysbench-psql-40 : 17277.28 per sec. 17359.78 per sec.
sysbench-psql-80 : 17112.44 per sec. 17221.16 per sec.
(increase on the top end)
tbench:
NO_WA_WEIGHT
Throughput 685.211 MB/sec 2 clients 2 procs max_latency=0.123 ms
Throughput 1596.64 MB/sec 5 clients 5 procs max_latency=0.119 ms
Throughput 2985.47 MB/sec 10 clients 10 procs max_latency=0.262 ms
Throughput 4521.15 MB/sec 20 clients 20 procs max_latency=0.506 ms
Throughput 9438.1 MB/sec 40 clients 40 procs max_latency=2.052 ms
Throughput 8210.5 MB/sec 80 clients 80 procs max_latency=8.310 ms
WA_WEIGHT
Throughput 697.292 MB/sec 2 clients 2 procs max_latency=0.127 ms
Throughput 1596.48 MB/sec 5 clients 5 procs max_latency=0.080 ms
Throughput 2975.22 MB/sec 10 clients 10 procs max_latency=0.254 ms
Throughput 4575.14 MB/sec 20 clients 20 procs max_latency=0.502 ms
Throughput 9468.65 MB/sec 40 clients 40 procs max_latency=2.069 ms
Throughput 8631.73 MB/sec 80 clients 80 procs max_latency=8.605 ms
Eric reported a sysbench regression against commit:
3fed382b46ba ("sched/numa: Implement NUMA node level wake_affine()")
Similarly, Rik was looking at the NAS-lu.C benchmark, which regressed
against his v3.10 enterprise kernel.
PRE (current tip/master):
ivb-ep sysbench:
2: [30 secs] transactions: 64110 (2136.94 per sec.)
5: [30 secs] transactions: 143644 (4787.99 per sec.)
10: [30 secs] transactions: 274298 (9142.93 per sec.)
20: [30 secs] transactions: 418683 (13955.45 per sec.)
40: [30 secs] transactions: 320731 (10690.15 per sec.)
80: [30 secs] transactions: 355096 (11834.28 per sec.)
hsw-ex NAS:
OMP_PROC_BIND/lu.C.x_threads_144_run_1.log: Time in seconds = 18.01
OMP_PROC_BIND/lu.C.x_threads_144_run_2.log: Time in seconds = 17.89
OMP_PROC_BIND/lu.C.x_threads_144_run_3.log: Time in seconds = 17.93
lu.C.x_threads_144_run_1.log: Time in seconds = 434.68
lu.C.x_threads_144_run_2.log: Time in seconds = 405.36
lu.C.x_threads_144_run_3.log: Time in seconds = 433.83
POST (+patch):
ivb-ep sysbench:
2: [30 secs] transactions: 64494 (2149.75 per sec.)
5: [30 secs] transactions: 145114 (4836.99 per sec.)
10: [30 secs] transactions: 278311 (9276.69 per sec.)
20: [30 secs] transactions: 437169 (14571.60 per sec.)
40: [30 secs] transactions: 669837 (22326.73 per sec.)
80: [30 secs] transactions: 631739 (21055.88 per sec.)
hsw-ex NAS:
lu.C.x_threads_144_run_1.log: Time in seconds = 23.36
lu.C.x_threads_144_run_2.log: Time in seconds = 22.96
lu.C.x_threads_144_run_3.log: Time in seconds = 22.52
This patch takes out all the shiny wake_affine() stuff and goes back to
utter basics. Between the two CPUs involved with the wakeup (the CPU
doing the wakeup and the CPU we ran on previously) pick the CPU we can
run on _now_.
This restores much of the regressions against the older kernels,
but leaves some ground in the overloaded case. The default-enabled
WA_WEIGHT (which will be introduced in the next patch) is an attempt
to address the overloaded situation.
... when a PMU is unregistered then its associated ->pmu_cpu_context is
unconditionally freed. Whilst this is fine for dynamically allocated
context types (i.e. those registered using perf_invalid_context), this
causes a problem for sharing of static contexts such as
perf_{sw,hw}_context, which are used by multiple built-in PMUs and
effectively have a global lifetime.
Whilst testing the ARM SPE driver, which must use perf_sw_context to
support per-task AUX tracing, unregistering the driver as a result of a
module unload resulted in:
Unable to handle kernel NULL pointer dereference at virtual address 00000038
Internal error: Oops: 96000004 [#1] PREEMPT SMP
Modules linked in: [last unloaded: arm_spe_pmu]
PC is at ctx_resched+0x38/0xe8
LR is at perf_event_exec+0x20c/0x278
[...]
ctx_resched+0x38/0xe8
perf_event_exec+0x20c/0x278
setup_new_exec+0x88/0x118
load_elf_binary+0x26c/0x109c
search_binary_handler+0x90/0x298
do_execveat_common.isra.14+0x540/0x618
SyS_execve+0x38/0x48
since the software context has been freed and the ctx.pmu->pmu_disable_count
field has been set to NULL.
This patch fixes the problem by avoiding the freeing of static PMU contexts
altogether. Whilst the sharing of dynamic contexts is questionable, this
actually requires the caller to share their context pointer explicitly
and so the burden is on them to manage the object lifetime.
Peter Zijlstra [Wed, 4 Oct 2017 08:50:25 +0000 (10:50 +0200)]
locking/selftest: Avoid false BUG report
The work-around for the expected failure is providing another failure :/
Only when CONFIG_PROVE_LOCKING=y do we increment unexpected_testcase_failures,
so only then do we need to decrement, otherwise we'll end up with a negative
number and that will again trigger a BUG (printout, not crash).
Peter Zijlstra [Wed, 4 Oct 2017 09:13:37 +0000 (11:13 +0200)]
locking/lockdep: Fix stacktrace mess
There is some complication between check_prevs_add() and
check_prev_add() wrt. saving stack traces. The problem is that we want
to be frugal with saving stack traces, since it consumes static
resources.
We'll only know in check_prev_add() if we need the trace, but we can
call into it multiple times. So we want to do on-demand and re-use.
A further complication is that check_prev_add() can drop graph_lock
and mess with our static resources.
In any case, the current state; after commit:
ce07a9415f26 ("locking/lockdep: Make check_prev_add() able to handle external stack_trace")
is that we'll assume the trace contains valid data once
check_prev_add() returns '2'. However, as noted by Josh, this is
false, check_prev_add() can return '2' before having saved a trace,
this then result in the possibility of using uninitialized data.
Testing, as reported by Wu, shows a NULL deref.
So simplify.
Since the graph_lock() thing is a debug path that hasn't
really been used in a long while, take it out back and avoid the
head-ache.
Further initialize the stack_trace to a known 'empty' state; as long
as nr_entries == 0, nothing should deref entries. We can then use the
'entries == NULL' test for a valid trace / on-demand saving.
lis r2, STACK_END_MAGIC@h
ori r2, r2, STACK_END_MAGIC@l
ld r12, -8(r1)
Fix the corruption by using r11 register for livepatch stack
manipulation, instead of shuffling task stack and livepatch stack into
r1 register. Using r11 register also avoids disabling/enabling irq's
while setting up the livepatch stack.
Jarkko Nikula [Mon, 9 Oct 2017 13:28:37 +0000 (16:28 +0300)]
device property: Track owner device of device property
Deletion of subdevice will remove device properties associated to parent
when they share the same firmware node after commit 478573c93abd (driver
core: Don't leak secondary fwnode on device removal). This was observed
with a driver adding subdevice that driver wasn't able to read device
properties after rmmod/modprobe cycle.
Parent device will have its primary firmware node pointing to an ACPI
node and secondary firmware node point to device properties.
ACPI_COMPANION_SET() call in parent probe will set the subdevice's
firmware node to point to the same 'struct fwnode_handle' and the
associated secondary firmware node, i.e. the device properties as the
parent.
When subdevice is deleted in parent remove that will remove those
device properties and attempt to read device properties in next
parent probe call will fail.
Fix this by tracking the owner device of device properties and delete
them only when owner device is being deleted.
Fixes: 478573c93abd (driver core: Don't leak secondary fwnode on device removal) Cc: 4.9+ <[email protected]> # 4.9+ Signed-off-by: Jarkko Nikula <[email protected]> Signed-off-by: Rafael J. Wysocki <[email protected]>
Linus Torvalds [Tue, 10 Oct 2017 02:08:32 +0000 (19:08 -0700)]
Merge branch 'ppc-bundle' (bundle from Michael Ellerman)
Merge powerpc transactional memory fixes from Michael Ellerman:
"I figured I'd still send you the commits using a bundle to make sure
it works in case I need to do it again in future"
This fixes transactional memory state restore for powerpc.
* bundle'd patches from Michael Ellerman:
powerpc/tm: Fix illegal TM state in signal handler
powerpc/64s: Use emergency stack for kernel TM Bad Thing program checks
1) Fix object leak on IPSEC offload failure, from Steffen Klassert.
2) Fix range checks in ipset address range addition operations, from
Jozsef Kadlecsik.
3) Fix pernet ops unregistration order in ipset, from Florian Westphal.
4) Add missing netlink attribute policy for nl80211 packet pattern
attrs, from Peng Xu.
5) Fix PPP device destruction race, from Guillaume Nault.
6) Write marks get lost when BPF verifier processes R1=R2 register
assignments, causing incorrect liveness information and less state
pruning. Fix from Alexei Starovoitov.
7) Fix blockhole routes so that they are marked dead and therefore not
cached in sockets, otherwise IPSEC stops working. From Steffen
Klassert.
8) Fix broadcast handling of UDP socket early demux, from Paolo Abeni.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (37 commits)
cdc_ether: flag the u-blox TOBY-L2 and SARA-U2 as wwan
net: thunderx: mark expected switch fall-throughs in nicvf_main()
udp: fix bcast packet reception
netlink: do not set cb_running if dump's start() errs
ipv4: Fix traffic triggered IPsec connections.
ipv6: Fix traffic triggered IPsec connections.
ixgbe: incorrect XDP ring accounting in ethtool tx_frame param
net: ixgbe: Use new PCI_DEV_FLAGS_NO_RELAXED_ORDERING flag
Revert commit 1a8b6d76dc5b ("net:add one common config...")
ixgbe: fix masking of bits read from IXGBE_VXLANCTRL register
ixgbe: Return error when getting PHY address if PHY access is not supported
netfilter: xt_bpf: Fix XT_BPF_MODE_FD_PINNED mode of 'xt_bpf_info_v1'
netfilter: SYNPROXY: skip non-tcp packet in {ipv4, ipv6}_synproxy_hook
tipc: Unclone message at secondary destination lookup
tipc: correct initialization of skb list
gso: fix payload length when gso_size is zero
mlxsw: spectrum_router: Avoid expensive lookup during route removal
bpf: fix liveness marking
doc: Fix typo "8023.ad" in bonding documentation
ipv6: fix net.ipv6.conf.all.accept_dad behaviour for real
...
cdc_ether: flag the u-blox TOBY-L2 and SARA-U2 as wwan
The u-blox TOBY-L2 is a LTE Cat 4 module with HSPA+ and 2G fallback.
This module allows switching to different USB profiles with the
'AT+UUSBCONF' command, and provides a ECM network interface when the
'AT+UUSBCONF=2' profile is selected.
The u-blox SARA-U2 is a HSPA module with 2G fallback. The default USB
configuration includes a ECM network interface.
Both these modules are controlled via AT commands through one of the
TTYs exposed. Connecting these modules may be done just by activating
the desired PDP context with 'AT+CGACT=1,<cid>' and then running DHCP
on the ECM interface.
Mark Rutland [Fri, 6 Oct 2017 18:38:22 +0000 (19:38 +0100)]
perf pmu: Unbreak perf record for arm/arm64 with events with explicit PMU
Currently, perf record is broken on arm/arm64 systems when the PMU is
specified explicitly as part of the event, e.g.
$ ./perf record -e armv8_cortex_a53/cpu_cycles/u true
In such cases, perf record fails to open events unless
perf_event_paranoid is set to -1, even if the PMU in question supports
mode exclusion. Further, even when perf_event_paranoid is toggled, no
samples are recorded.
This is an unintended side effect of commit:
e3ba76deef23064f ("perf tools: Force uncore events to system wide monitoring)
... which assumes that if a PMU has an associated cpu_map, it is an
uncore PMU, and forces events for such PMUs to be system-wide.
This is not true for arm/arm64 systems, which can have heterogeneous
CPUs. To account for this, multiple CPU PMUs are exposed, each with a
"cpus" field under sysfs, which the perf tool parses into a cpu_map. ARM
PMUs do not have a "cpumask" file, and only have a "cpus" file. For the
gory details as to why, see commit:
7e3fcffe95544010 ("perf pmu: Support alternative sysfs cpumask")
Given all of this, we can instead identify uncore PMUs by explicitly
checking for a "cpumask" file, and restore arm/arm64 PMU support back to
a working state. This patch does so, adding a new perf_pmu::is_uncore
field, and splitting the existing cpumask parsing so that it can be
reused.
bugfixes:
- NFSv4/pnfs: Fix an infinite layoutget loop
- nfs: RPC_MAX_AUTH_SIZE is in bytes"
* tag 'nfs-for-4.14-3' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
NFSv4/pnfs: Fix an infinite layoutget loop
nfs/filelayout: fix oops when freeing filelayout segment
sunrpc: remove redundant initialization of sock
NFS: Fix uninitialized rpc_wait_queue
NFS: Cleanup error handling in nfs_idmap_request_key()
nfs: RPC_MAX_AUTH_SIZE is in bytes
David S. Miller [Mon, 9 Oct 2017 17:36:25 +0000 (10:36 -0700)]
Merge branch '10GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net-queue
Jeff Kirsher says:
====================
Intel Wired LAN Driver Updates 2017-10-09
This series contains updates to ixgbe and arch/Kconfig.
Mark fixes a case where PHY register access is not supported and we were
returning a PHY address, when we should have been returning -EOPNOTSUPP.
Sabrina Dubroca fixes the use of a logical "and" when it should have been
the bitwise "and" operator.
Ding Tianhong reverts the commit that added the Kconfig bool option
ARCH_WANT_RELAX_ORDER, since there is now a new flag
PCI_DEV_FLAGS_NO_RELAXED_ORDERING that has been added to indicate that
Relaxed Ordering Attributes should not be used for Transaction Layer
Packets. Then follows up with making the needed changes to ixgbe to
use the new PCI_DEV_FLAGS_NO_RELAXED_ORDERING flag.
John Fastabend fixes an issue in the ring accounting when the transmit
ring parameters are changed via ethtool when an XDP program is attached.
====================
Ville Syrjälä [Fri, 1 Apr 2016 15:37:25 +0000 (18:37 +0300)]
drm/i915: Read timings from the correct transcoder in intel_crtc_mode_get()
intel_crtc->config->cpu_transcoder isn't yet filled out when
intel_crtc_mode_get() gets called during output probing, so we should
not use it there. Instead intel_crtc_mode_get() figures out the correct
transcoder on its own, and that's what we should use.
If the BIOS boots LVDS on pipe B, intel_crtc_mode_get() would actually
end up reading the timings from pipe A instead (since PIPE_A==0),
which clearly isn't what we want.
It looks to me like this may have been broken by
commit eccb140bca67 ("drm/i915: hw state readout&check support for cpu_transcoder")
as that one removed the early initialization of cpu_transcoder from
intel_crtc_init().
Chris Wilson [Fri, 6 Oct 2017 11:56:17 +0000 (12:56 +0100)]
drm/i915: Order two completing nop_submit_request
If two nop's (requests in-flight following a wedged device) complete at
the same time, the global_seqno value written to the HWSP is undefined
as the two threads are not serialized.
v2: Use irqsafe spinlock. We expect the callback may be called from
inside another irq spinlock, so we can't unconditionally restore irqs.
drm/i915: Use crtc_state_is_legacy_gamma in intel_color_check
crtc_state_is_legacy_gamma also checks for CTM, which was missing from
intel_color_check. By using the same condition for commit and check
we reduce the chance of mismatches.
This was spotted by KASAN while trying to rework kms_color igt test.
Manasi Navare [Tue, 3 Oct 2017 23:37:25 +0000 (16:37 -0700)]
drm/i915/edp: Increase the T12 delay quirk to 1300ms
For this specific PCI device, the eDP panel requires a higher panel
power cycle delay of 1300ms where the minimum spec requirement of panel
power cycle delay is 500ms. This fix in combination with correct
timestamp at which we get the panel power off time fixes the dP AUX CH
timeouts seen on various IGT tests.
Manasi Navare [Wed, 4 Oct 2017 16:48:26 +0000 (09:48 -0700)]
drm/i915/edp: Get the Panel Power Off timestamp after panel is off
Kernel stores the time in jiffies at which the eDP panel is turned
off. This should be obtained after the panel is off (after the
wait_panel_off). When we next attempt to turn the panel on, we use the
difference between the timestamp at which we want to turn the panel on
and timestamp at which panel was turned off to ensure that this is equal
to panel power cycle delay and if not we wait for the remaining
time. Not waiting for the panel power cycle delay can cause the panel to
not turn on giving rise to AUX timeouts for the attempted AUX
transactions.
v2:
* Separate lines for bugzilla (Jani Nikula)
* Suggested by tag (Daniel Vetter)
Paolo Abeni [Mon, 9 Oct 2017 12:52:10 +0000 (14:52 +0200)]
udp: fix bcast packet reception
The commit bc044e8db796 ("udp: perform source validation for
mcast early demux") does not take into account that broadcast packets
lands in the same code path and they need different checks for the
source address - notably, zero source address are valid for bcast
and invalid for mcast.
As a result, 2nd and later broadcast packets with 0 source address
landing to the same socket are dropped. This breaks dhcp servers.
Since we don't have stringent performance requirements for ingress
broadcast traffic, fix it by disabling UDP early demux such traffic.
Reported-by: Hannes Frederic Sowa <[email protected]> Fixes: bc044e8db796 ("udp: perform source validation for mcast early demux") Signed-off-by: Paolo Abeni <[email protected]> Signed-off-by: David S. Miller <[email protected]>
netlink: do not set cb_running if dump's start() errs
It turns out that multiple places can call netlink_dump(), which means
it's still possible to dereference partially initialized values in
dump() that were the result of a faulty returned start().
This fixes the issue by calling start() _before_ setting cb_running to
true, so that there's no chance at all of hitting the dump() function
through any indirect paths.
It also moves the call to start() to be when the mutex is held. This has
the nice side effect of serializing invocations to start(), which is
likely desirable anyway. It also prevents any possible other races that
might come out of this logic.
In testing this with several different pieces of tricky code to trigger
these issues, this commit fixes all avenues that I'm aware of.
The QCA folks found another netlink problem - we were missing validation
of some attributes. It's not super problematic since one can only read a
few bytes beyond the message (and that memory must exist), but here's the
fix for it.
I thought perhaps we can make nla_parse_nested() require a policy, but
given the two-stage validation/parsing in regular netlink that won't work.
Please pull and let me know if there's any problem.
====================
A recent patch removed the dst_free() on the allocated
dst_entry in ipv4_blackhole_route(). The dst_free() marked the
dst_entry as dead and added it to the gc list. I.e. it was setup
for a one time usage. As a result we may now have a blackhole
route cached at a socket on some IPsec scenarios. This makes the
connection unusable.
Fix this by marking the dst_entry directly at allocation time
as 'dead', so it is used only once.
Fixes: b838d5e1c5b6 ("ipv4: mark DST_NOGC and remove the operation of dst_free()") Reported-by: Tobias Brunner <[email protected]> Signed-off-by: Steffen Klassert <[email protected]> Signed-off-by: David S. Miller <[email protected]>
A recent patch removed the dst_free() on the allocated
dst_entry in ipv6_blackhole_route(). The dst_free() marked
the dst_entry as dead and added it to the gc list. I.e. it
was setup for a one time usage. As a result we may now have
a blackhole route cached at a socket on some IPsec scenarios.
This makes the connection unusable.
Fix this by marking the dst_entry directly at allocation time
as 'dead', so it is used only once.
Fixes: 587fea741134 ("ipv6: mark DST_NOGC and remove the operation of dst_free()") Reported-by: Tobias Brunner <[email protected]> Signed-off-by: Steffen Klassert <[email protected]> Signed-off-by: David S. Miller <[email protected]>
sync_file: Return consistent status in SYNC_IOC_FILE_INFO
sync_file_ioctl_fence_info has a race between filling the status
of the underlying fences and the overall status of the sync_file.
If fence transitions in the time frame between its sync_fill_fence_info
and the later dma_fence_is_signaled for the sync_file, the returned
information is inconsistent showing non-signaled underlying fences but
an overall signaled state.
This patch changes sync_file_ioctl_fence_info to track what has been
encoded and using that as the overall sync_file status.
Ding Tianhong [Fri, 18 Aug 2017 06:21:05 +0000 (14:21 +0800)]
net: ixgbe: Use new PCI_DEV_FLAGS_NO_RELAXED_ORDERING flag
The ixgbe driver use the compile check to determine if it can
send TLPs to Root Port with the Relaxed Ordering Attribute set,
this is too inconvenient, now the new flag PCI_DEV_FLAGS_NO_RELAXED_ORDERING
has been added to the kernel and we could check the bit4 in the PCIe
Device Control register to determine whether we should use the Relaxed
Ordering Attributes or not, so use this new way in the ixgbe driver.
Ding Tianhong [Fri, 18 Aug 2017 06:21:04 +0000 (14:21 +0800)]
Revert commit 1a8b6d76dc5b ("net:add one common config...")
The new flag PCI_DEV_FLAGS_NO_RELAXED_ORDERING has been added
to indicate that Relaxed Ordering Attributes (RO) should not
be used for Transaction Layer Packets (TLP) targeted toward
these affected Root Port, it will clear the bit4 in the PCIe
Device Control register, so the PCIe device drivers could
query PCIe configuration space to determine if it can send
TLPs to Root Port with the Relaxed Ordering Attributes set.
With this new flag we don't need the config ARCH_WANT_RELAX_ORDER
to control the Relaxed Ordering Attributes for the ixgbe drivers
just like the commit 1a8b6d76dc5b ("net:add one common config...") did,
so revert this commit.
ixgbe: fix masking of bits read from IXGBE_VXLANCTRL register
In ixgbe_clear_udp_tunnel_port(), we read the IXGBE_VXLANCTRL register
and then try to mask some bits out of the value, using the logical
instead of bitwise and operator.
Fixes: a21d0822ff69 ("ixgbe: add support for geneve Rx offload") Signed-off-by: Sabrina Dubroca <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
Mark D Rustad [Wed, 31 Aug 2016 17:34:28 +0000 (10:34 -0700)]
ixgbe: Return error when getting PHY address if PHY access is not supported
In cases where PHY register access is not supported, don't mislead
a caller into thinking that it is supported by returning a PHY
address. Instead, return -EOPNOTSUPP when PHY access is not
supported.
Paul Burton [Fri, 8 Sep 2017 22:12:21 +0000 (15:12 -0700)]
MIPS: math-emu: Remove pr_err() calls from fpu_emu()
The FPU emulator includes 2 calls to pr_err() which are triggered by
invalid instruction encodings for MIPSr6 cmp.cond.fmt instructions.
These cases are not kernel errors, merely invalid instructions which are
already handled by delivering a SIGILL which will provide notification
that something failed in cases where that makes sense.
In cases where that SIGILL is somewhat expected & being handled, for
example when crashme happens to generate one of the affected bad
encodings, the message is printed with no useful context about what
triggered it & spams the kernel log for no good reason.
Remove the pr_err() calls to make crashme run silently & treat the bad
encodings the same way we do others, with a SIGILL & no further kernel
log output.
Signed-off-by: Paul Burton <[email protected]> Fixes: f8c3c6717a71 ("MIPS: math-emu: Add support for the CMP.condn.fmt R6 instruction") Cc: [email protected] Cc: stable <[email protected]> # v4.3+
Patchwork: https://patchwork.linux-mips.org/patch/17253/ Signed-off-by: Ralf Baechle <[email protected]>
Paul Burton [Sun, 3 Sep 2017 17:24:58 +0000 (10:24 -0700)]
MIPS: Fix generic-board-config.sh for builds using O=
When configuring the kernel using one of the generic MIPS defconfig
targets, the generic-board-config.sh script is used to check
requirements listed in board config fragments against a reference config
in order to determine which board config fragments to merge into the
final config.
When specifying O= to configure in a directory other than the kernel
source directory, this generic-board-config.sh script is invoked in the
directory that we are configuring in (ie. the directory that O equals),
and the path to the reference config is relative to the current
directory. The script then changes the current directory to the source
tree, which unfortunately breaks later access to the reference file
since its path is relative to a directory that is no longer the current
working directory. This results in configuration failing with errors
such as:
$ make ARCH=mips O=tmp 32r2_defconfig
make[1]: Entering directory '/home/pburton/src/linux/tmp'
Using ../arch/mips/configs/generic_defconfig as base
Merging ../arch/mips/configs/generic/32r2.config
Merging ../arch/mips/configs/generic/eb.config
grep: ./.config.32r2_defconfig: No such file or directory
grep: ./.config.32r2_defconfig: No such file or directory
The base file '.config' does not exist. Exit.
make[1]: *** [arch/mips/Makefile:505: 32r2_defconfig] Error 1
make[1]: Leaving directory '/home/pburton/src/linux-ingenic/tmp'
make: *** [Makefile:145: sub-make] Error 2
Fix this by avoiding changing the working directory in
generic-board-config.sh, instead using full paths to files under
$(srctree)/ where necessary.
Paul Burton [Fri, 1 Sep 2017 21:46:50 +0000 (14:46 -0700)]
MIPS: Fix cmpxchg on 32b signed ints for 64b kernel with !kernel_uses_llsc
Commit 8263db4d7768 ("MIPS: cmpxchg: Implement __cmpxchg() as a
function") refactored our implementation of __cmpxchg() to be a function
rather than a macro, with the aim of making it easier to read & modify.
Unfortunately the commit breaks use of cmpxchg() for signed 32 bit
values when we have a 64 bit kernel with kernel_uses_llsc == false,
because:
- In cmpxchg_local() we cast the old value to the type the pointer
points to, and then to an unsigned long. If the pointer points to a
signed type smaller than 64 bits then the old value will be sign
extended to 64 bits. That is, bits beyond the size of the pointed to
type will be set to 1 if the old value is negative. In the case of a
signed 32 bit integer with a negative value, bits 63:32 will all be
set.
- In __cmpxchg_asm() we load the value from memory, ie. dereference the
pointer, and store the value as an unsigned integer (__ret) whose
size matches the pointer. For a 32 bit cmpxchg() this means we store
the value in a u32, because the pointer provided to __cmpxchg_asm()
by __cmpxchg() is of type volatile u32 *.
- __cmpxchg_asm() then checks whether the value in memory (__ret)
matches the provided old value, by comparing the two values. This
results in the u32 being promoted to a 64 bit unsigned long to match
the old argument - however because both types are unsigned the value
is zero extended, which does not match the sign extension performed
on the old value in cmpxchg_local() earlier.
This mismatch means that unfortunate cmpxchg() calls can incorrectly
fail for 64 bit kernels with kernel_uses_llsc == false. This is the case
on at least non-SMP Cavium Octeon kernels, which hardcode
kernel_uses_llsc in their cpu-feature-overrides.h header. Using a
v4.13-rc7 kernel configured using cavium_octeon_defconfig with SMP
manually disabled, this presents itself as oddity when we reach
userland - for example:
can't run '/bin/mount': Text file busy
can't run '/bin/mkdir': Text file busy
can't run '/bin/mkdir': Text file busy
can't run '/bin/mount': Text file busy
can't run '/bin/hostname': Text file busy
can't run '/etc/init.d/rcS': Text file busy
can't run '/sbin/getty': Text file busy
can't run '/sbin/getty': Text file busy
It appears that some part of the init process, which is in this case
buildroot's busybox init, is running successfully. It never manages to
reach the login prompt though, and complains about /sbin/getty being
busy repeatedly and indefinitely.
Fix this by casting the old value provided to __cmpxchg_asm() to an
appropriately sized unsigned integer, such that we consistently
zero-extend avoiding the mismatch. The __cmpxchg_small() case for 8 & 16
bit values is unaffected because __cmpxchg_small() already masks
provided values appropriately.
Takashi Iwai [Mon, 9 Oct 2017 12:51:23 +0000 (14:51 +0200)]
ALSA: line6: Fix leftover URB at error-path during probe
While line6_probe() may kick off URB for a control MIDI endpoint, the
function doesn't clean up it properly at its error path. This results
in a leftover URB action that is eventually triggered later and causes
an Oops like:
general protection fault: 0000 [#1] PREEMPT SMP KASAN
CPU: 1 PID: 0 Comm: swapper/1 Not tainted
RIP: 0010:usb_fill_bulk_urb ./include/linux/usb.h:1619
RIP: 0010:line6_start_listen+0x3fe/0x9e0 sound/usb/line6/driver.c:76
Call Trace:
<IRQ>
line6_data_received+0x1f7/0x470 sound/usb/line6/driver.c:326
__usb_hcd_giveback_urb+0x2e0/0x650 drivers/usb/core/hcd.c:1779
usb_hcd_giveback_urb+0x337/0x420 drivers/usb/core/hcd.c:1845
dummy_timer+0xba9/0x39f0 drivers/usb/gadget/udc/dummy_hcd.c:1965
call_timer_fn+0x2a2/0x940 kernel/time/timer.c:1281
....
Since the whole clean-up procedure is done in line6_disconnect()
callback, we can simply call it in the error path instead of
open-coding the whole again. It'll fix such an issue automagically.
Takashi Iwai [Mon, 9 Oct 2017 12:26:27 +0000 (14:26 +0200)]
ALSA: line6: Fix NULL dereference at podhd_disconnect()
When podhd_init() failed with the acquiring a ctrl i/f, the line6
helper still calls the disconnect callback that eventually calls again
usb_driver_release_interface() with the NULL intf.
Put the proper NULL check before calling it for avoiding an Oops.
Takashi Iwai [Mon, 9 Oct 2017 12:32:15 +0000 (14:32 +0200)]
ALSA: line6: Fix missing initialization before error path
The error path in podhd_init() tries to clear the pending timer, while
the timer object is initialized at the end of init sequence, thus it
may hit the uninitialized object, as spotted by syzkaller:
INFO: trying to register non-static key.
the code is fine but needs lockdep annotation.
turning off the locking correctness validator.
CPU: 1 PID: 1845 Comm: kworker/1:2 Not tainted 4.14.0-rc2-42613-g1488251d1a98 #238
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
Workqueue: usb_hub_wq hub_event
Call Trace:
__dump_stack lib/dump_stack.c:16
dump_stack+0x292/0x395 lib/dump_stack.c:52
register_lock_class+0x6c4/0x1a00 kernel/locking/lockdep.c:769
__lock_acquire+0x27e/0x4550 kernel/locking/lockdep.c:3385
lock_acquire+0x259/0x620 kernel/locking/lockdep.c:4002
del_timer_sync+0x12c/0x280 kernel/time/timer.c:1237
podhd_disconnect+0x8c/0x160 sound/usb/line6/podhd.c:299
line6_probe+0x844/0x1310 sound/usb/line6/driver.c:783
podhd_probe+0x64/0x70 sound/usb/line6/podhd.c:474
....
For addressing it, assure the initializations of timer and work by
moving them to the beginning of podhd_init().
Shmulik Ladkani [Mon, 9 Oct 2017 12:27:15 +0000 (15:27 +0300)]
netfilter: xt_bpf: Fix XT_BPF_MODE_FD_PINNED mode of 'xt_bpf_info_v1'
Commit 2c16d6033264 ("netfilter: xt_bpf: support ebpf") introduced
support for attaching an eBPF object by an fd, with the
'bpf_mt_check_v1' ABI expecting the '.fd' to be specified upon each
IPT_SO_SET_REPLACE call.
However this breaks subsequent iptables calls:
# iptables -A INPUT -m bpf --object-pinned /sys/fs/bpf/xxx -j ACCEPT
# iptables -A INPUT -s 5.6.7.8 -j ACCEPT
iptables: Invalid argument. Run `dmesg' for more information.
That's because iptables works by loading existing rules using
IPT_SO_GET_ENTRIES to userspace, then issuing IPT_SO_SET_REPLACE with
the replacement set.
However, the loaded 'xt_bpf_info_v1' has an arbitrary '.fd' number
(from the initial "iptables -m bpf" invocation) - so when 2nd invocation
occurs, userspace passes a bogus fd number, which leads to
'bpf_mt_check_v1' to fail.
One suggested solution [1] was to hack iptables userspace, to perform a
"entries fixup" immediatley after IPT_SO_GET_ENTRIES, by opening a new,
process-local fd per every 'xt_bpf_info_v1' entry seen.
However, in [2] both Pablo Neira Ayuso and Willem de Bruijn suggested to
depricate the xt_bpf_info_v1 ABI dealing with pinned ebpf objects.
This fix changes the XT_BPF_MODE_FD_PINNED behavior to ignore the given
'.fd' and instead perform an in-kernel lookup for the bpf object given
the provided '.path'.
It also defines an alias for the XT_BPF_MODE_FD_PINNED mode, named
XT_BPF_MODE_PATH_PINNED, to better reflect the fact that the user is
expected to provide the path of the pinned object.
Existing XT_BPF_MODE_FD_ELF behavior (non-pinned fd mode) is preserved.
projects / linux.git / log