firefly-linux-kernel-4.4.55.git
7 years agofs/proc/kcore.c: Make bounce buffer global for read
Jiri Olsa [Thu, 8 Sep 2016 07:57:07 +0000 (09:57 +0200)]
fs/proc/kcore.c: Make bounce buffer global for read

Next patch adds bounce buffer for ktext area, so it's
convenient to have single bounce buffer for both
vmalloc/module and ktext cases.

Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit f5beeb1851ea6f8cfcf2657f26cb24c0582b4945)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agofs/proc/kcore.c: Add bounce buffer for ktext data
Jiri Olsa [Thu, 8 Sep 2016 07:57:08 +0000 (09:57 +0200)]
fs/proc/kcore.c: Add bounce buffer for ktext data

We hit hardened usercopy feature check for kernel text access by reading
kcore file:

  usercopy: kernel memory exposure attempt detected from ffffffff8179a01f (<kernel text>) (4065 bytes)
  kernel BUG at mm/usercopy.c:75!

Bypassing this check for kcore by adding bounce buffer for ktext data.

Reported-by: Steve Best <sbest@redhat.com>
Fixes: f5509cc18daa ("mm: Hardened usercopy")
Suggested-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit df04abfd181acc276ba6762c8206891ae10ae00d)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agoMerge branch 'v4.4/topic/mm-kaslr' into v4.4/topic/mm-kaslr-pax_usercopy
Alex Shi [Thu, 23 Mar 2017 07:39:18 +0000 (15:39 +0800)]
Merge branch 'v4.4/topic/mm-kaslr' into v4.4/topic/mm-kaslr-pax_usercopy

7 years agoarm64: enable CONFIG_DEBUG_RODATA by default
Ard Biesheuvel [Thu, 3 Mar 2016 14:10:59 +0000 (15:10 +0100)]
arm64: enable CONFIG_DEBUG_RODATA by default

In spite of its name, CONFIG_DEBUG_RODATA is an important hardening feature
for production kernels, and distros all enable it by default in their
kernel configs. However, since enabling it used to result in more granular,
and thus less efficient kernel mappings, it is not enabled by default for
performance reasons.

However, since commit 2f39b5f91eb4 ("arm64: mm: Mark .rodata as RO"), the
various kernel segments (.text, .rodata, .init and .data) are already
mapped individually, and the only effect of setting CONFIG_DEBUG_RODATA is
that the existing .text and .rodata mappings are updated late in the boot
sequence to have their read-only attributes set, which means that any
performance concerns related to enabling CONFIG_DEBUG_RODATA are no longer
valid.

So from now on, make CONFIG_DEBUG_RODATA default to 'y'

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
(cherry picked from commit 57efac2f7108e3255d0dfe512290c9896f4ed55f)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agoarm64: fix alignment when RANDOMIZE_TEXT_OFFSET is enabled
Mark Rutland [Tue, 31 May 2016 14:58:00 +0000 (15:58 +0100)]
arm64: fix alignment when RANDOMIZE_TEXT_OFFSET is enabled

With ARM64_64K_PAGES and RANDOMIZE_TEXT_OFFSET enabled, we hit the
following issue on the boot:

kernel BUG at arch/arm64/mm/mmu.c:480!
Internal error: Oops - BUG: 0 [#1] PREEMPT SMP
Modules linked in:
CPU: 0 PID: 0 Comm: swapper Not tainted 4.6.0 #310
Hardware name: ARM Juno development board (r2) (DT)
task: ffff000008d58a80 ti: ffff000008d30000 task.ti: ffff000008d30000
PC is at map_kernel_segment+0x44/0xb0
LR is at paging_init+0x84/0x5b0
pc : [<ffff000008c450b4>] lr : [<ffff000008c451a4>] pstate: 600002c5

Call trace:
[<ffff000008c450b4>] map_kernel_segment+0x44/0xb0
[<ffff000008c451a4>] paging_init+0x84/0x5b0
[<ffff000008c42728>] setup_arch+0x198/0x534
[<ffff000008c40848>] start_kernel+0x70/0x388
[<ffff000008c401bc>] __primary_switched+0x30/0x74

Commit 7eb90f2ff7e3 ("arm64: cover the .head.text section in the .text
segment mapping") removed the alignment between the .head.text and .text
sections, and used the _text rather than the _stext interval for mapping
the .text segment.

Prior to this commit _stext was always section aligned and didn't cause
any issue even when RANDOMIZE_TEXT_OFFSET was enabled. Since that
alignment has been removed and _text is used to map the .text segment,
we need ensure _text is always page aligned when RANDOMIZE_TEXT_OFFSET
is enabled.

This patch adds logic to TEXT_OFFSET fuzzing to ensure that the offset
is always aligned to the kernel page size. To ensure this, we rely on
the PAGE_SHIFT being available via Kconfig.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reported-by: Sudeep Holla <sudeep.holla@arm.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will.deacon@arm.com>
Fixes: 7eb90f2ff7e3 ("arm64: cover the .head.text section in the .text segment mapping")
Signed-off-by: Will Deacon <will.deacon@arm.com>
(cherry picked from commit aed7eb8367939244ba19445292ffdfc398e0d66a)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agocoresight: Remove erroneous dma_free_coherent in tmc_probe
Suzuki K Poulose [Thu, 25 Aug 2016 21:18:54 +0000 (15:18 -0600)]
coresight: Remove erroneous dma_free_coherent in tmc_probe

commit de5461970b3e9e194 ("coresight: tmc: allocating memory when needed")
removed the static allocation of buffer for the trace data in ETR mode in
tmc_probe. However it failed to remove the "devm_free_coherent" in
tmc_probe when the probe fails due to other reasons. This patch gets
rid of the incorrect dma_free_coherent() call.

Fixes: commit de5461970b3e9e194 ("coresight: tmc: allocating memory when needed")
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
(cherry picked from commit 481e46fe7a88557b66330cbb047b25cc13eff4b9)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agoarm64: avoid TLB conflict with CONFIG_RANDOMIZE_BASE
Mark Rutland [Wed, 24 Aug 2016 17:02:08 +0000 (18:02 +0100)]
arm64: avoid TLB conflict with CONFIG_RANDOMIZE_BASE

When CONFIG_RANDOMIZE_BASE is selected, we modify the page tables to remap the
kernel at a newly-chosen VA range. We do this with the MMU disabled, but do not
invalidate TLBs prior to re-enabling the MMU with the new tables. Thus the old
mappings entries may still live in TLBs, and we risk violating
Break-Before-Make requirements, leading to TLB conflicts and/or other issues.

We invalidate TLBs when we uninsall the idmap in early setup code, but prior to
this we are subject to issues relating to the Break-Before-Make violation.

Avoid these issues by invalidating the TLBs before the new mappings can be
used by the hardware.

Fixes: f80fb3a3d508 ("arm64: add support for kernel ASLR")
Cc: <stable@vger.kernel.org> # 4.6+
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Acked-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
(cherry picked from commit fd363bd417ddb6103564c69cfcbd92d9a7877431)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agoperf/core: Use this_cpu_ptr() when stopping AUX events
Will Deacon [Wed, 24 Aug 2016 09:07:14 +0000 (10:07 +0100)]
perf/core: Use this_cpu_ptr() when stopping AUX events

When tearing down an AUX buf for an event via perf_mmap_close(),
__perf_event_output_stop() is called on the event's CPU to ensure that
trace generation is halted before the process of unmapping and
freeing the buffer pages begins.

The callback is performed via cpu_function_call(), which ensures that it
runs with interrupts disabled and is therefore not preemptible.
Unfortunately, the current code grabs the per-cpu context pointer using
get_cpu_ptr(), which unnecessarily disables preemption and doesn't pair
the call with put_cpu_ptr(), leading to a preempt_count() imbalance and
a BUG when freeing the AUX buffer later on:

  WARNING: CPU: 1 PID: 2249 at kernel/events/ring_buffer.c:539 __rb_free_aux+0x10c/0x120
  Modules linked in:
  [...]
  Call Trace:
   [<ffffffff813379dd>] dump_stack+0x4f/0x72
   [<ffffffff81059ff6>] __warn+0xc6/0xe0
   [<ffffffff8105a0c8>] warn_slowpath_null+0x18/0x20
   [<ffffffff8112761c>] __rb_free_aux+0x10c/0x120
   [<ffffffff81128163>] rb_free_aux+0x13/0x20
   [<ffffffff8112515e>] perf_mmap_close+0x29e/0x2f0
   [<ffffffff8111da30>] ? perf_iterate_ctx+0xe0/0xe0
   [<ffffffff8115f685>] remove_vma+0x25/0x60
   [<ffffffff81161796>] exit_mmap+0x106/0x140
   [<ffffffff8105725c>] mmput+0x1c/0xd0
   [<ffffffff8105cac3>] do_exit+0x253/0xbf0
   [<ffffffff8105e32e>] do_group_exit+0x3e/0xb0
   [<ffffffff81068d49>] get_signal+0x249/0x640
   [<ffffffff8101c273>] do_signal+0x23/0x640
   [<ffffffff81905f42>] ? _raw_write_unlock_irq+0x12/0x30
   [<ffffffff81905f69>] ? _raw_spin_unlock_irq+0x9/0x10
   [<ffffffff81901896>] ? __schedule+0x2c6/0x710
   [<ffffffff810022a4>] exit_to_usermode_loop+0x74/0x90
   [<ffffffff81002a56>] prepare_exit_to_usermode+0x26/0x30
   [<ffffffff81906d1b>] retint_user+0x8/0x10

This patch uses this_cpu_ptr() instead of get_cpu_ptr(), since preemption is
already disabled by the caller.

Signed-off-by: Will Deacon <will.deacon@arm.com>
Reviewed-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Fixes: 95ff4ca26c49 ("perf/core: Free AUX pages in unmap path")
Link: http://lkml.kernel.org/r/20160824091905.GA16944@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit 8b6a3fe8fab97716990a3abde1a01fb5a34552a3)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agoarm64: Only select ARM64_MODULE_PLTS if MODULES=y
Catalin Marinas [Tue, 26 Jul 2016 17:16:55 +0000 (10:16 -0700)]
arm64: Only select ARM64_MODULE_PLTS if MODULES=y

Selecting CONFIG_RANDOMIZE_BASE=y and CONFIG_MODULES=n fails to build
the module PLTs support:

  CC      arch/arm64/kernel/module-plts.o
/work/Linux/linux-2.6-aarch64/arch/arm64/kernel/module-plts.c: In function ‘module_emit_plt_entry’:
/work/Linux/linux-2.6-aarch64/arch/arm64/kernel/module-plts.c:32:49: error: dereferencing pointer to incomplete type ‘struct module’

This patch selects ARM64_MODULE_PLTS conditionally only if MODULES is
enabled.

Fixes: f80fb3a3d508 ("arm64: add support for kernel ASLR")
Cc: <stable@vger.kernel.org> # 4.6+
Reported-by: Jeff Vander Stoep <jeffv@google.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
(cherry picked from commit b9c220b589daaf140f5b8ebe502c98745b94e65c)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agoarm64: spinlock: fix spin_unlock_wait for LSE atomics
Will Deacon [Wed, 8 Jun 2016 14:10:57 +0000 (15:10 +0100)]
arm64: spinlock: fix spin_unlock_wait for LSE atomics

Commit d86b8da04dfa ("arm64: spinlock: serialise spin_unlock_wait against
concurrent lockers") fixed spin_unlock_wait for LL/SC-based atomics under
the premise that the LSE atomics (in particular, the LDADDA instruction)
are indivisible.

Unfortunately, these instructions are only indivisible when used with the
-AL (full ordering) suffix and, consequently, the same issue can
theoretically be observed with LSE atomics, where a later (in program
order) load can be speculated before the write portion of the atomic
operation.

This patch fixes the issue by performing a CAS of the lock once we've
established that it's unlocked, in much the same way as the LL/SC code.

Fixes: d86b8da04dfa ("arm64: spinlock: serialise spin_unlock_wait against concurrent lockers")
Signed-off-by: Will Deacon <will.deacon@arm.com>
(cherry picked from commit 3a5facd09da848193f5bcb0dea098a298bc1a29d)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agoarm64: spinlock: order spin_{is_locked,unlock_wait} against local locks
Will Deacon [Thu, 2 Jun 2016 14:27:04 +0000 (15:27 +0100)]
arm64: spinlock: order spin_{is_locked,unlock_wait} against local locks

spin_is_locked has grown two very different use-cases:

(1) [The sane case] API functions may require a certain lock to be held
    by the caller and can therefore use spin_is_locked as part of an
    assert statement in order to verify that the lock is indeed held.
    For example, usage of assert_spin_locked.

(2) [The insane case] There are two locks, where a CPU takes one of the
    locks and then checks whether or not the other one is held before
    accessing some shared state. For example, the "optimized locking" in
    ipc/sem.c.

In the latter case, the sequence looks like:

  spin_lock(&sem->lock);
  if (!spin_is_locked(&sma->sem_perm.lock))
    /* Access shared state */

and requires that the spin_is_locked check is ordered after taking the
sem->lock. Unfortunately, since our spinlocks are implemented using a
LDAXR/STXR sequence, the read of &sma->sem_perm.lock can be speculated
before the STXR and consequently return a stale value.

Whilst this hasn't been seen to cause issues in practice, PowerPC fixed
the same issue in 51d7d5205d33 ("powerpc: Add smp_mb() to
arch_spin_is_locked()") and, although we did something similar for
spin_unlock_wait in d86b8da04dfa ("arm64: spinlock: serialise
spin_unlock_wait against concurrent lockers") that doesn't actually take
care of ordering against local acquisition of a different lock.

This patch adds an smp_mb() to the start of our arch_spin_is_locked and
arch_spin_unlock_wait routines to ensure that the lock value is always
loaded after any other locks have been taken by the current CPU.

Reported-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>
(cherry picked from commit 38b850a73034f075c4088e7511b36ebbef9dce00)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agoarm64: fix dump_instr when PAN and UAO are in use
Mark Rutland [Mon, 13 Jun 2016 10:15:14 +0000 (11:15 +0100)]
arm64: fix dump_instr when PAN and UAO are in use

If the kernel is set to show unhandled signals, and a user task does not
handle a SIGILL as a result of an instruction abort, we will attempt to
log the offending instruction with dump_instr before killing the task.

We use dump_instr to log the encoding of the offending userspace
instruction. However, dump_instr is also used to dump instructions from
kernel space, and internally always switches to KERNEL_DS before dumping
the instruction with get_user. When both PAN and UAO are in use, reading
a user instruction via get_user while in KERNEL_DS will result in a
permission fault, which leads to an Oops.

As we have regs corresponding to the context of the original instruction
abort, we can inspect this and only flip to KERNEL_DS if the original
abort was taken from the kernel, avoiding this issue. At the same time,
remove the redundant (and incorrect) comments regarding the order
dump_mem and dump_instr are called in.

Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Robin Murphy <robin.murphy@arm.com>
Cc: <stable@vger.kernel.org> #4.6+
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reported-by: Vladimir Murzin <vladimir.murzin@arm.com>
Tested-by: Vladimir Murzin <vladimir.murzin@arm.com>
Fixes: 57f4959bad0a154a ("arm64: kernel: Add support for User Access Override")
Signed-off-by: Will Deacon <will.deacon@arm.com>
(cherry picked from commit c5cea06be060f38e5400d796e61cfc8c36e52924)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agoMerge branch 'v4.4/topic/mm-kaslr' into v4.4/topic/mm-kaslr-pax_usercopy
Alex Shi [Wed, 22 Mar 2017 10:48:45 +0000 (18:48 +0800)]
Merge branch 'v4.4/topic/mm-kaslr' into v4.4/topic/mm-kaslr-pax_usercopy

7 years agoarm64: hugetlb: fix the wrong return value for huge_ptep_set_access_flags
Huang Shijie [Wed, 11 Jan 2017 06:02:00 +0000 (14:02 +0800)]
arm64: hugetlb: fix the wrong return value for huge_ptep_set_access_flags

In current code, the @changed always returns the last one's status for
the huge page with the contiguous bit set. This is really not what we
want. Even one of the PTEs is changed, we should tell it to the caller.

This patch fixes this issue.

Fixes: 66b3923a1a0f ("arm64: hugetlb: add support for PTE contiguous bit")
Cc: <stable@vger.kernel.org> # 4.5.x-
Signed-off-by: Huang Shijie <shijie.huang@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
(cherry picked from commit 69d012345a1a32d3f03957f14d972efccc106a98)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agoarm64: hugetlb: fix the wrong address for several functions
Huang Shijie [Tue, 8 Nov 2016 05:44:39 +0000 (13:44 +0800)]
arm64: hugetlb: fix the wrong address for several functions

The libhugetlbfs meets several failures since the following functions
do not use the correct address:
   huge_ptep_get_and_clear()
   huge_ptep_set_access_flags()
   huge_ptep_set_wrprotect()
   huge_ptep_clear_flush()

This patch fixes the wrong address for them.

Signed-off-by: Huang Shijie <shijie.huang@arm.com>
Reviewed-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
(cherry picked from commit 0c2f0afe3582c58efeef93bc57bc07d502132618)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agoarm64: hugetlb: remove the wrong pmd check in find_num_contig()
Huang Shijie [Tue, 8 Nov 2016 05:44:38 +0000 (13:44 +0800)]
arm64: hugetlb: remove the wrong pmd check in find_num_contig()

The find_num_contig() will return 1 when the pmd is not present.
It will cause a kernel dead loop in the following scenaro:

   1.) pmd entry is not present.

   2.) the page fault occurs:
       ... hugetlb_fault() --> hugetlb_no_page() --> set_huge_pte_at()

   3.) set_huge_pte_at() will only set the first PMD entry, since the
       find_num_contig just return 1 in this case. So the PMD entries
       are all empty except the first one.

   4.) when kernel accesses the address mapped by the second PMD entry,
       a new page fault occurs:
       ... hugetlb_fault() --> huge_ptep_set_access_flags()

       The second PMD entry is still empty now.

   5.) When the kernel returns, the access will cause a page fault again.
       The kernel will run like the "4)" above.
       We will see a dead loop since here.

The dead loop is caught in the 32M hugetlb page (2M PMD + Contiguous bit).

This patch removes wrong pmd check, and fixes this dead loop.

This patch also removes the redundant checks for PGD/PUD in
the find_num_contig().

Acked-by: Steve Capper <steve.capper@arm.com>
Signed-off-by: Huang Shijie <shijie.huang@arm.com>
Reviewed-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
(cherry picked from commit 20156ce2365d61beaa6f5a78a7a789044e0e7acc)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agoarm64: kaslr: fix breakage with CONFIG_MODVERSIONS=y
Ard Biesheuvel [Thu, 13 Oct 2016 16:42:09 +0000 (17:42 +0100)]
arm64: kaslr: fix breakage with CONFIG_MODVERSIONS=y

As it turns out, the KASLR code breaks CONFIG_MODVERSIONS, since the
kcrctab has an absolute address field that is relocated at runtime
when the kernel offset is randomized.

This has been fixed already for PowerPC in the past, so simply wire up
the existing code dealing with this issue.

Cc: <stable@vger.kernel.org>
Fixes: f80fb3a3d508 ("arm64: add support for kernel ASLR")
Tested-by: Timur Tabi <timur@codeaurora.org>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>
(cherry picked from commit 9c0e83c371cf4696926c95f9c8c77cd6ea803426)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agoarm64: fix dump_backtrace/unwind_frame with NULL tsk
Mark Rutland [Fri, 23 Sep 2016 16:55:05 +0000 (17:55 +0100)]
arm64: fix dump_backtrace/unwind_frame with NULL tsk

In some places, dump_backtrace() is called with a NULL tsk parameter,
e.g. in bug_handler() in arch/arm64, or indirectly via show_stack() in
core code. The expectation is that this is treated as if current were
passed instead of NULL. Similar is true of unwind_frame().

Commit a80a0eb70c358f8c ("arm64: make irq_stack_ptr more robust") didn't
take this into account. In dump_backtrace() it compares tsk against
current *before* we check if tsk is NULL, and in unwind_frame() we never
set tsk if it is NULL.

Due to this, we won't initialise irq_stack_ptr in either function. In
dump_backtrace() this results in calling dump_mem() for memory
immediately above the IRQ stack range, rather than for the relevant
range on the task stack. In unwind_frame we'll reject unwinding frames
on the IRQ stack.

In either case this results in incomplete or misleading backtrace
information, but is not otherwise problematic. The initial percpu areas
(including the IRQ stacks) are allocated in the linear map, and dump_mem
uses __get_user(), so we shouldn't access anything with side-effects,
and will handle holes safely.

This patch fixes the issue by having both functions handle the NULL tsk
case before doing anything else with tsk.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Fixes: a80a0eb70c358f8c ("arm64: make irq_stack_ptr more robust")
Acked-by: James Morse <james.morse@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Yang Shi <yang.shi@linaro.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>
(cherry picked from commit b5e7307d9d5a340d2c9fabbe1cee137d4c682c71)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
7 years agousercopy: fold builtin_const check into inline function
Kees Cook [Wed, 31 Aug 2016 23:04:21 +0000 (16:04 -0700)]
usercopy: fold builtin_const check into inline function

Instead of having each caller of check_object_size() need to remember to
check for a const size parameter, move the check into check_object_size()
itself. This actually matches the original implementation in PaX, though
this commit cleans up the now-redundant builtin_const() calls in the
various architectures.

Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit 81409e9e28058811c9ea865345e1753f8f677e44)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agounsafe_[get|put]_user: change interface to use a error target label
Linus Torvalds [Mon, 8 Aug 2016 20:02:01 +0000 (13:02 -0700)]
unsafe_[get|put]_user: change interface to use a error target label

When I initially added the unsafe_[get|put]_user() helpers in commit
5b24a7a2aa20 ("Add 'unsafe' user access functions for batched
accesses"), I made the mistake of modeling the interface on our
traditional __[get|put]_user() functions, which return zero on success,
or -EFAULT on failure.

That interface is fairly easy to use, but it's actually fairly nasty for
good code generation, since it essentially forces the caller to check
the error value for each access.

In particular, since the error handling is already internally
implemented with an exception handler, and we already use "asm goto" for
various other things, we could fairly easily make the error cases just
jump directly to an error label instead, and avoid the need for explicit
checking after each operation.

So switch the interface to pass in an error label, rather than checking
the error value in the caller.  Best do it now before we start growing
more users (the signal handling code in particular would be a good place
to use the new interface).

So rather than

if (unsafe_get_user(x, ptr))
... handle error ..

the interface is now

unsafe_get_user(x, ptr, label);

where an error during the user mode fetch will now just cause a jump to
'label' in the caller.

Right now the actual _implementation_ of this all still ends up being a
"if (err) goto label", and does not take advantage of any exception
label tricks, but for "unsafe_put_user()" in particular it should be
fairly straightforward to convert to using the exception table model.

Note that "unsafe_get_user()" is much harder to convert to a clever
exception table model, because current versions of gcc do not allow the
use of "asm goto" (for the exception) with output values (for the actual
value to be fetched).  But that is hopefully not a limitation in the
long term.

[ Also note that it might be a good idea to switch unsafe_get_user() to
  actually _return_ the value it fetches from user space, but this
  commit only changes the error handling semantics ]

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit 1bd4403d86a1c06cb6cc9ac87664a0c9d3413d51)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agousercopy: remove page-spanning test for now
Kees Cook [Wed, 7 Sep 2016 16:54:34 +0000 (09:54 -0700)]
usercopy: remove page-spanning test for now

A custom allocator without __GFP_COMP that copies to userspace has been
found in vmw_execbuf_process[1], so this disables the page-span checker
by placing it behind a CONFIG for future work where such things can be
tracked down later.

[1] https://bugzilla.redhat.com/show_bug.cgi?id=1373326

Reported-by: Vinson Lee <vlee@freedesktop.org>
Fixes: f5509cc18daa ("mm: Hardened usercopy")
Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit 8e1f74ea02cf4562404c48c6882214821552c13f)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agousercopy: fix overlap check for kernel text
Josh Poimboeuf [Mon, 22 Aug 2016 16:53:59 +0000 (11:53 -0500)]
usercopy: fix overlap check for kernel text

When running with a local patch which moves the '_stext' symbol to the
very beginning of the kernel text area, I got the following panic with
CONFIG_HARDENED_USERCOPY:

  usercopy: kernel memory exposure attempt detected from ffff88103dfff000 (<linear kernel text>) (4096 bytes)
  ------------[ cut here ]------------
  kernel BUG at mm/usercopy.c:79!
  invalid opcode: 0000 [#1] SMP
  ...
  CPU: 0 PID: 4800 Comm: cp Not tainted 4.8.0-rc3.after+ #1
  Hardware name: Dell Inc. PowerEdge R720/0X3D66, BIOS 2.5.4 01/22/2016
  task: ffff880817444140 task.stack: ffff880816274000
  RIP: 0010:[<ffffffff8121c796>] __check_object_size+0x76/0x413
  RSP: 0018:ffff880816277c40 EFLAGS: 00010246
  RAX: 000000000000006b RBX: ffff88103dfff000 RCX: 0000000000000000
  RDX: 0000000000000000 RSI: ffff88081f80dfa8 RDI: ffff88081f80dfa8
  RBP: ffff880816277c90 R08: 000000000000054c R09: 0000000000000000
  R10: 0000000000000005 R11: 0000000000000006 R12: 0000000000001000
  R13: ffff88103e000000 R14: ffff88103dffffff R15: 0000000000000001
  FS:  00007fb9d1750800(0000) GS:ffff88081f800000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 00000000021d2000 CR3: 000000081a08f000 CR4: 00000000001406f0
  Stack:
   ffff880816277cc8 0000000000010000 000000043de07000 0000000000000000
   0000000000001000 ffff880816277e60 0000000000001000 ffff880816277e28
   000000000000c000 0000000000001000 ffff880816277ce8 ffffffff8136c3a6
  Call Trace:
   [<ffffffff8136c3a6>] copy_page_to_iter_iovec+0xa6/0x1c0
   [<ffffffff8136e766>] copy_page_to_iter+0x16/0x90
   [<ffffffff811970e3>] generic_file_read_iter+0x3e3/0x7c0
   [<ffffffffa06a738d>] ? xfs_file_buffered_aio_write+0xad/0x260 [xfs]
   [<ffffffff816e6262>] ? down_read+0x12/0x40
   [<ffffffffa06a61b1>] xfs_file_buffered_aio_read+0x51/0xc0 [xfs]
   [<ffffffffa06a6692>] xfs_file_read_iter+0x62/0xb0 [xfs]
   [<ffffffff812224cf>] __vfs_read+0xdf/0x130
   [<ffffffff81222c9e>] vfs_read+0x8e/0x140
   [<ffffffff81224195>] SyS_read+0x55/0xc0
   [<ffffffff81003a47>] do_syscall_64+0x67/0x160
   [<ffffffff816e8421>] entry_SYSCALL64_slow_path+0x25/0x25
  RIP: 0033:[<00007fb9d0c33c00>] 0x7fb9d0c33c00
  RSP: 002b:00007ffc9c262f28 EFLAGS: 00000246 ORIG_RAX: 0000000000000000
  RAX: ffffffffffffffda RBX: fffffffffff8ffff RCX: 00007fb9d0c33c00
  RDX: 0000000000010000 RSI: 00000000021c3000 RDI: 0000000000000004
  RBP: 00000000021c3000 R08: 0000000000000000 R09: 00007ffc9c264d6c
  R10: 00007ffc9c262c50 R11: 0000000000000246 R12: 0000000000010000
  R13: 00007ffc9c2630b0 R14: 0000000000000004 R15: 0000000000010000
  Code: 81 48 0f 44 d0 48 c7 c6 90 4d a3 81 48 c7 c0 bb b3 a2 81 48 0f 44 f0 4d 89 e1 48 89 d9 48 c7 c7 68 16 a3 81 31 c0 e8 f4 57 f7 ff <0f> 0b 48 8d 90 00 40 00 00 48 39 d3 0f 83 22 01 00 00 48 39 c3
  RIP  [<ffffffff8121c796>] __check_object_size+0x76/0x413
   RSP <ffff880816277c40>

The checked object's range [ffff88103dfff000ffff88103e000000) is
valid, so there shouldn't have been a BUG.  The hardened usercopy code
got confused because the range's ending address is the same as the
kernel's text starting address at 0xffff88103e000000.  The overlap check
is slightly off.

Fixes: f5509cc18daa ("mm: Hardened usercopy")
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit 94cd97af690dd9537818dc9841d0ec68bb1dd877)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agomm/slub: support left redzone
Joonsoo Kim [Tue, 15 Mar 2016 21:55:12 +0000 (14:55 -0700)]
mm/slub: support left redzone

SLUB already has a redzone debugging feature.  But it is only positioned
at the end of object (aka right redzone) so it cannot catch left oob.
Although current object's right redzone acts as left redzone of next
object, first object in a slab cannot take advantage of this effect.
This patch explicitly adds a left red zone to each object to detect left
oob more precisely.

Background:

Someone complained to me that left OOB doesn't catch even if KASAN is
enabled which does page allocation debugging.  That page is out of our
control so it would be allocated when left OOB happens and, in this
case, we can't find OOB.  Moreover, SLUB debugging feature can be
enabled without page allocator debugging and, in this case, we will miss
that OOB.

Before trying to implement, I expected that changes would be too
complex, but, it doesn't look that complex to me now.  Almost changes
are applied to debug specific functions so I feel okay.

Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit d86bd1bece6fc41d59253002db5441fe960a37f6)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agomm: SLUB hardened usercopy support
Kees Cook [Thu, 23 Jun 2016 22:24:05 +0000 (15:24 -0700)]
mm: SLUB hardened usercopy support

Under CONFIG_HARDENED_USERCOPY, this adds object size checking to the
SLUB allocator to catch any copies that may span objects. Includes a
redzone handling fix discovered by Michael Ellerman.

Based on code from PaX and grsecurity.

Signed-off-by: Kees Cook <keescook@chromium.org>
Tested-by: Michael Ellerman <mpe@ellerman.id.au>
Reviwed-by: Laura Abbott <labbott@redhat.com>
(cherry picked from commit ed18adc1cdd00a5c55a20fbdaed4804660772281)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agomm: SLAB hardened usercopy support
Kees Cook [Thu, 23 Jun 2016 22:20:59 +0000 (15:20 -0700)]
mm: SLAB hardened usercopy support

Under CONFIG_HARDENED_USERCOPY, this adds object size checking to the
SLAB allocator to catch any copies that may span objects.

Based on code from PaX and grsecurity.

Signed-off-by: Kees Cook <keescook@chromium.org>
Tested-by: Valdis Kletnieks <valdis.kletnieks@vt.edu>
(cherry picked from commit 04385fc5e8fffed84425d909a783c0f0c587d847)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agos390/uaccess: Enable hardened usercopy
Kees Cook [Thu, 7 Jul 2016 18:38:39 +0000 (11:38 -0700)]
s390/uaccess: Enable hardened usercopy

Enables CONFIG_HARDENED_USERCOPY checks on s390.

Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit 97433ea4fda62349bfa42089455593cbcb57e06c)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agosparc/uaccess: Enable hardened usercopy
Kees Cook [Thu, 23 Jun 2016 22:10:13 +0000 (15:10 -0700)]
sparc/uaccess: Enable hardened usercopy

Enables CONFIG_HARDENED_USERCOPY checks on sparc.

Based on code from PaX and grsecurity.

Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit 9d9208a15800f9f06f102f9aac1e8b323c3b8575)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agopowerpc/uaccess: Enable hardened usercopy
Kees Cook [Thu, 23 Jun 2016 22:10:01 +0000 (15:10 -0700)]
powerpc/uaccess: Enable hardened usercopy

Enables CONFIG_HARDENED_USERCOPY checks on powerpc.

Based on code from PaX and grsecurity.

Signed-off-by: Kees Cook <keescook@chromium.org>
Tested-by: Michael Ellerman <mpe@ellerman.id.au>
(cherry picked from commit 1d3c1324746fed0e34a5b94d3ed303e7521ed603)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agoia64/uaccess: Enable hardened usercopy
Kees Cook [Thu, 23 Jun 2016 22:09:50 +0000 (15:09 -0700)]
ia64/uaccess: Enable hardened usercopy

Enables CONFIG_HARDENED_USERCOPY checks on ia64.

Based on code from PaX and grsecurity.

Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit 73d35887e24da77e8d1321b2e92bd9b9128e2fc2)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agoarm64/uaccess: Enable hardened usercopy
Kees Cook [Thu, 23 Jun 2016 22:59:42 +0000 (15:59 -0700)]
arm64/uaccess: Enable hardened usercopy

Enables CONFIG_HARDENED_USERCOPY checks on arm64. As done by KASAN in -next,
renames the low-level functions to __arch_copy_*_user() so a static inline
can do additional work before the copy.

Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit faf5b63e294151d6ac24ca6906d6f221bd3496cd)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agoARM: uaccess: Enable hardened usercopy
Kees Cook [Thu, 23 Jun 2016 22:06:53 +0000 (15:06 -0700)]
ARM: uaccess: Enable hardened usercopy

Enables CONFIG_HARDENED_USERCOPY checks on arm.

Based on code from PaX and grsecurity.

Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit dfd45b6103c973bfcea2341d89e36faf947dbc33)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agox86/uaccess: Enable hardened usercopy
Kees Cook [Thu, 23 Jun 2016 22:04:01 +0000 (15:04 -0700)]
x86/uaccess: Enable hardened usercopy

Enables CONFIG_HARDENED_USERCOPY checks on x86. This is done both in
copy_*_user() and __copy_*_user() because copy_*_user() actually calls
down to _copy_*_user() and not __copy_*_user().

Based on code from PaX and grsecurity.

Signed-off-by: Kees Cook <keescook@chromium.org>
Tested-by: Valdis Kletnieks <valdis.kletnieks@vt.edu>
(cherry picked from commit 5b710f34e194c6b7710f69fdb5d798fdf35b98c1)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agox86: remove more uaccess_32.h complexity
Linus Torvalds [Mon, 23 May 2016 00:21:27 +0000 (17:21 -0700)]
x86: remove more uaccess_32.h complexity

I'm looking at trying to possibly merge the 32-bit and 64-bit versions
of the x86 uaccess.h implementation, but first this needs to be cleaned
up.

For example, the 32-bit version of "__copy_from_user_inatomic()" is
mostly the special cases for the constant size, and it's actually almost
never relevant.  Most users aren't actually using a constant size
anyway, and the few cases that do small constant copies are better off
just using __get_user() instead.

So get rid of the unnecessary complexity.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit bd28b14591b98f696bc9f94c5ba2e598ca487dfd)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agox86: remove pointless uaccess_32.h complexity
Linus Torvalds [Sun, 22 May 2016 21:19:37 +0000 (14:19 -0700)]
x86: remove pointless uaccess_32.h complexity

I'm looking at trying to possibly merge the 32-bit and 64-bit versions
of the x86 uaccess.h implementation, but first this needs to be cleaned
up.

For example, the 32-bit version of "__copy_to_user_inatomic()" is mostly
the special cases for the constant size, and it's actually never
relevant.  Every user except for one aren't actually using a constant
size anyway, and the one user that uses it is better off just using
__put_user() instead.

So get rid of the unnecessary complexity.

[ The same cleanup should likely happen to __copy_from_user_inatomic()
  as well, but that one has a lot more users that I need to take a look
  at first ]

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit 5b09c3edecd37ec1a52fbd5ae97a19734edc7a77)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agox86: fix SMAP in 32-bit environments
Linus Torvalds [Tue, 23 Feb 2016 22:58:52 +0000 (14:58 -0800)]
x86: fix SMAP in 32-bit environments

In commit 11f1a4b9755f ("x86: reorganize SMAP handling in user space
accesses") I changed how the stac/clac instructions were generated
around the user space accesses, which then made it possible to do
batched accesses efficiently for user string copies etc.

However, in doing so, I completely spaced out, and didn't even think
about the 32-bit case.  And nobody really even seemed to notice, because
SMAP doesn't even exist until modern Skylake processors, and you'd have
to be crazy to run 32-bit kernels on a modern CPU.

Which brings us to Andy Lutomirski.

He actually tested the 32-bit kernel on new hardware, and noticed that
it doesn't work.  My bad.  The trivial fix is to add the required
uaccess begin/end markers around the raw accesses in <asm/uaccess_32.h>.

I feel a bit bad about this patch, just because that header file really
should be cleaned up to avoid all the duplicated code in it, and this
commit just expands on the problem.  But this just fixes the bug without
any bigger cleanup surgery.

Reported-and-tested-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit de9e478b9d49f3a0214310d921450cf5bb4a21e6)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agoUse the new batched user accesses in generic user string handling
Linus Torvalds [Thu, 17 Dec 2015 18:05:19 +0000 (10:05 -0800)]
Use the new batched user accesses in generic user string handling

This converts the generic user string functions to use the batched user
access functions.

It makes a big difference on Skylake, which is the first x86
microarchitecture to implement SMAP.  The STAC/CLAC instructions are not
very fast, and doing them for each access inside the loop that copies
strings from user space (which is what the pathname handling does for
every pathname the kernel uses, for example) is very inefficient.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit 9fd4470ff4974c41b1db43c3b355b9085af9c12a)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agoAdd 'unsafe' user access functions for batched accesses
Linus Torvalds [Thu, 17 Dec 2015 17:57:27 +0000 (09:57 -0800)]
Add 'unsafe' user access functions for batched accesses

The naming is meant to discourage random use: the helper functions are
not really any more "unsafe" than the traditional double-underscore
functions (which need the address range checking), but they do need even
more infrastructure around them, and should not be used willy-nilly.

In addition to checking the access range, these user access functions
require that you wrap the user access with a "user_acess_{begin,end}()"
around it.

That allows architectures that implement kernel user access control
(x86: SMAP, arm64: PAN) to do the user access control in the wrapping
user_access_begin/end part, and then batch up the actual user space
accesses using the new interfaces.

The main (and hopefully only) use for these are for core generic access
helpers, initially just the generic user string functions
(strnlen_user() and strncpy_from_user()).

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit 5b24a7a2aa2040c8c50c3b71122901d01661ff78)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agox86: reorganize SMAP handling in user space accesses
Linus Torvalds [Thu, 17 Dec 2015 17:45:09 +0000 (09:45 -0800)]
x86: reorganize SMAP handling in user space accesses

This reorganizes how we do the stac/clac instructions in the user access
code.  Instead of adding the instructions directly to the same inline
asm that does the actual user level access and exception handling, add
them at a higher level.

This is mainly preparation for the next step, where we will expose an
interface to allow users to mark several accesses together as being user
space accesses, but it does already clean up some code:

 - the inlined trivial cases of copy_in_user() now do stac/clac just
   once over the accesses: they used to do one pair around the user
   space read, and another pair around the write-back.

 - the {get,put}_user_ex() macros that are used with the catch/try
   handling don't do any stac/clac at all, because that happens in the
   try/catch surrounding them.

Other than those two cleanups that happened naturally from the
re-organization, this should not make any difference. Yet.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit 11f1a4b9755f5dbc3e822a96502ebe9b044b14d8)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agomm: Hardened usercopy
Kees Cook [Tue, 7 Jun 2016 18:05:33 +0000 (11:05 -0700)]
mm: Hardened usercopy

This is the start of porting PAX_USERCOPY into the mainline kernel. This
is the first set of features, controlled by CONFIG_HARDENED_USERCOPY. The
work is based on code by PaX Team and Brad Spengler, and an earlier port
from Casey Schaufler. Additional non-slab page tests are from Rik van Riel.

This patch contains the logic for validating several conditions when
performing copy_to_user() and copy_from_user() on the kernel object
being copied to/from:
- address range doesn't wrap around
- address range isn't NULL or zero-allocated (with a non-zero copy size)
- if on the slab allocator:
  - object size must be less than or equal to copy size (when check is
    implemented in the allocator, which appear in subsequent patches)
- otherwise, object must not span page allocations (excepting Reserved
  and CMA ranges)
- if on the stack
  - object must not extend before/after the current process stack
  - object must be contained by a valid stack frame (when there is
    arch/build support for identifying stack frames)
- object must not overlap with kernel text

Signed-off-by: Kees Cook <keescook@chromium.org>
Tested-by: Valdis Kletnieks <valdis.kletnieks@vt.edu>
Tested-by: Michael Ellerman <mpe@ellerman.id.au>
(cherry picked from commit f5509cc18daa7f82bcc553be70df2117c8eedc16)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
Conflicts:
skip debug_page_ref and KCOV_INSTRUMENT in mm/Makefile

8 years agomm: Implement stack frame object validation
Kees Cook [Tue, 12 Jul 2016 23:19:48 +0000 (16:19 -0700)]
mm: Implement stack frame object validation

This creates per-architecture function arch_within_stack_frames() that
should validate if a given object is contained by a kernel stack frame.
Initial implementation is on x86.

This is based on code from PaX.

Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit 0f60a8efe4005ab5e65ce000724b04d4ca04a199)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
Conflicts:
skip EBPF_JIT in arch/x86/Kconfig

8 years agomm: Add is_migrate_cma_page
Laura Abbott [Tue, 19 Jul 2016 22:00:04 +0000 (15:00 -0700)]
mm: Add is_migrate_cma_page

Code such as hardened user copy[1] needs a way to tell if a
page is CMA or not. Add is_migrate_cma_page in a similar way
to is_migrate_isolate_page.

[1]http://article.gmane.org/gmane.linux.kernel.mm/155238

Signed-off-by: Laura Abbott <labbott@redhat.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit 7c15d9bb8231f998ae7dc0b72415f5215459f7fb)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agoMerge tag 'lsk-v4.4-arm64-v4.8-kaslr-updates' of git://git.linaro.org/people/ard...
Alex Shi [Mon, 15 Aug 2016 09:42:55 +0000 (17:42 +0800)]
Merge tag 'lsk-v4.4-arm64-v4.8-kaslr-updates' of git://git.linaro.org/people/ard.biesheuvel/linux-arm into v4.4/topic/mm-kaslr

8 years agoarm64: relocatable: suppress R_AARCH64_ABS64 relocations in vmlinux
Ard Biesheuvel [Sun, 24 Jul 2016 12:00:13 +0000 (14:00 +0200)]
arm64: relocatable: suppress R_AARCH64_ABS64 relocations in vmlinux

The linker routines that we rely on to produce a relocatable PIE binary
treat it as a shared ELF object in some ways, i.e., it emits symbol based
R_AARCH64_ABS64 relocations into the final binary since doing so would be
appropriate when linking a shared library that is subject to symbol
preemption. (This means that an executable can override certain symbols
that are exported by a shared library it is linked with, and that the
shared library *must* update all its internal references as well, and point
them to the version provided by the executable.)

Symbol preemption does not occur for OS hosted PIE executables, let alone
for vmlinux, and so we would prefer to get rid of these symbol based
relocations. This would allow us to simplify the relocation routines, and
to strip the .dynsym, .dynstr and .hash sections from the binary. (Note
that these are tiny, and are placed in the .init segment, but they clutter
up the vmlinux binary.)

Note that these R_AARCH64_ABS64 relocations are only emitted for absolute
references to symbols defined in the linker script, all other relocatable
quantities are covered by anonymous R_AARCH64_RELATIVE relocations that
simply list the offsets to all 64-bit values in the binary that need to be
fixed up based on the offset between the link time and run time addresses.

Fortunately, GNU ld has a -Bsymbolic option, which is intended for shared
libraries to allow them to ignore symbol preemption, and unconditionally
bind all internal symbol references to its own definitions. So set it for
our PIE binary as well, and get rid of the asoociated sections and the
relocation code that processes them.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
[will: fixed conflict with __dynsym_offset linker script entry]
Signed-off-by: Will Deacon <will.deacon@arm.com>
(cherry picked from commit 08cc55b2afd97a654f71b3bebf8bb0ec89fdc498)
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
8 years agoarm64: vmlinux.lds: make __rela_offset and __dynsym_offset ABSOLUTE
Ard Biesheuvel [Thu, 28 Jul 2016 14:15:14 +0000 (16:15 +0200)]
arm64: vmlinux.lds: make __rela_offset and __dynsym_offset ABSOLUTE

Due to the untyped KIMAGE_VADDR constant, the linker may not notice
that the __rela_offset and __dynsym_offset expressions are absolute
values (i.e., are not subject to relocation). This does not matter for
KASLR, but it does confuse kallsyms in relative mode, since it uses
the lowest non-absolute symbol address as the anchor point, and expects
all other symbol addresses to be within 4 GB of it.

Fix this by qualifying these expressions as ABSOLUTE() explicitly.

Fixes: 0cd3defe0af4 ("arm64: kernel: perform relocation processing from ID map")
Cc: <stable@vger.kernel.org>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>
(cherry picked from commit d6732fc402c2665f61e72faf206a0268e65236e9)
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
8 years ago Merge tag 'v4.4.17' into linux-linaro-lsk-v4.4
Alex Shi [Thu, 11 Aug 2016 04:15:51 +0000 (12:15 +0800)]
 Merge tag 'v4.4.17' into linux-linaro-lsk-v4.4

 This is the 4.4.17 stable release

8 years agoLinux 4.4.17
Greg Kroah-Hartman [Wed, 10 Aug 2016 09:49:43 +0000 (11:49 +0200)]
Linux 4.4.17

8 years agovfs: fix deadlock in file_remove_privs() on overlayfs
Miklos Szeredi [Wed, 3 Aug 2016 11:44:27 +0000 (13:44 +0200)]
vfs: fix deadlock in file_remove_privs() on overlayfs

commit c1892c37769cf89c7e7ba57528ae2ccb5d153c9b upstream.

file_remove_privs() is called with inode lock on file_inode(), which
proceeds to calling notify_change() on file->f_path.dentry.  Which triggers
the WARN_ON_ONCE(!inode_is_locked(inode)) in addition to deadlocking later
when ovl_setattr tries to lock the underlying inode again.

Fix this mess by not mixing the layers, but doing everything on underlying
dentry/inode.

Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Fixes: 07a2daab49c5 ("ovl: Copy up underlying inode's ->i_mode to overlay inode")
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agointel_th: Fix a deadlock in modprobing
Alexander Shishkin [Thu, 30 Jun 2016 08:51:44 +0000 (11:51 +0300)]
intel_th: Fix a deadlock in modprobing

commit a36aa80f3cb2540fb1dbad6240852de4365a2e82 upstream.

Driver initialization tries to request a hub (GTH) driver module from
its probe callback, resulting in a deadlock.

This patch solves the problem by adding a deferred work for requesting
the hub module.

Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agointel_th: pci: Add Kaby Lake PCH-H support
Alexander Shishkin [Tue, 28 Jun 2016 15:55:23 +0000 (18:55 +0300)]
intel_th: pci: Add Kaby Lake PCH-H support

commit 7a1a47ce35821b40f5b2ce46379ba14393bc3873 upstream.

This adds Intel(R) Trace Hub PCI ID for Kaby Lake PCH-H.

Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agonet: mvneta: set real interrupt per packet for tx_done
Dmitri Epshtein [Wed, 6 Jul 2016 02:18:58 +0000 (04:18 +0200)]
net: mvneta: set real interrupt per packet for tx_done

commit 06708f81528725148473c0869d6af5f809c6824b upstream.

Commit aebea2ba0f74 ("net: mvneta: fix Tx interrupt delay") intended to
set coalescing threshold to a value guaranteeing interrupt generation
per each sent packet, so that buffers can be released with no delay.

In fact setting threshold to '1' was wrong, because it causes interrupt
every two packets. According to the documentation a reason behind it is
following - interrupt occurs once sent buffers counter reaches a value,
which is higher than one specified in MVNETA_TXQ_SIZE_REG(q). This
behavior was confirmed during tests. Also when testing the SoC working
as a NAS device, better performance was observed with int-per-packet,
as it strongly depends on the fact that all transmitted packets are
released immediately.

This commit enables NETA controller work in interrupt per sent packet mode
by setting coalescing threshold to 0.

Signed-off-by: Dmitri Epshtein <dima@marvell.com>
Signed-off-by: Marcin Wojtas <mw@semihalf.com>
Fixes aebea2ba0f74 ("net: mvneta: fix Tx interrupt delay")
Acked-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agolibceph: apply new_state before new_up_client on incrementals
Ilya Dryomov [Tue, 19 Jul 2016 01:50:28 +0000 (03:50 +0200)]
libceph: apply new_state before new_up_client on incrementals

commit 930c532869774ebf8af9efe9484c597f896a7d46 upstream.

Currently, osd_weight and osd_state fields are updated in the encoding
order.  This is wrong, because an incremental map may look like e.g.

    new_up_client: { osd=6, addr=... } # set osd_state and addr
    new_state: { osd=6, xorstate=EXISTS } # clear osd_state

Suppose osd6's current osd_state is EXISTS (i.e. osd6 is down).  After
applying new_up_client, osd_state is changed to EXISTS | UP.  Carrying
on with the new_state update, we flip EXISTS and leave osd6 in a weird
"!EXISTS but UP" state.  A non-existent OSD is considered down by the
mapping code

2087    for (i = 0; i < pg->pg_temp.len; i++) {
2088            if (ceph_osd_is_down(osdmap, pg->pg_temp.osds[i])) {
2089                    if (ceph_can_shift_osds(pi))
2090                            continue;
2091
2092                    temp->osds[temp->size++] = CRUSH_ITEM_NONE;

and so requests get directed to the second OSD in the set instead of
the first, resulting in OSD-side errors like:

[WRN] : client.4239 192.168.122.21:0/2444980242 misdirected client.4239.1:2827 pg 2.5df899f2 to osd.4 not [1,4,6] in e680/680

and hung rbds on the client:

[  493.566367] rbd: rbd0: write 400000 at 11cc00000 (0)
[  493.566805] rbd: rbd0:   result -6 xferred 400000
[  493.567011] blk_update_request: I/O error, dev rbd0, sector 9330688

The fix is to decouple application from the decoding and:
- apply new_weight first
- apply new_state before new_up_client
- twiddle osd_state flags if marking in
- clear out some of the state if osd is destroyed

Fixes: http://tracker.ceph.com/issues/14901
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agolibata: LITE-ON CX1-JB256-HP needs lower max_sectors
Tejun Heo [Mon, 18 Jul 2016 22:40:00 +0000 (18:40 -0400)]
libata: LITE-ON CX1-JB256-HP needs lower max_sectors

commit 1488a1e3828d60d74c9b802a05e24c0487babe4e upstream.

Since 34b48db66e08 ("block: remove artifical max_hw_sectors cap"),
max_sectors is no longer limited to BLK_DEF_MAX_SECTORS and LITE-ON
CX1-JB256-HP keeps timing out with higher max_sectors.  Revert it to
the previous value.

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: dgerasimov@gmail.com
Link: https://bugzilla.kernel.org/show_bug.cgi?id=121671
Fixes: 34b48db66e08 ("block: remove artifical max_hw_sectors cap")
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoi2c: mux: reg: wrong condition checked for of_address_to_resource return value
Lukasz Gemborowski [Mon, 27 Jun 2016 10:57:47 +0000 (12:57 +0200)]
i2c: mux: reg: wrong condition checked for of_address_to_resource return value

commit 22ebf00eb56fe77922de8138aa9af9996582c2b3 upstream.

of_address_to_resource return 0 on successful call but
devm_ioremap_resource is called only if it returns non-zero value

Signed-off-by: Lukasz Gemborowski <lukasz.gemborowski@nokia.com>
Reviewed-by: Alexander Sverdlin <alexander.sverdlin@nokia.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoposix_cpu_timer: Exit early when process has been reaped
Alexey Dobriyan [Thu, 7 Jul 2016 22:39:11 +0000 (01:39 +0300)]
posix_cpu_timer: Exit early when process has been reaped

commit 2c13ce8f6b2f6fd9ba2f9261b1939fc0f62d1307 upstream.

Variable "now" seems to be genuinely used unintialized
if branch

if (CPUCLOCK_PERTHREAD(timer->it_clock)) {

is not taken and branch

if (unlikely(sighand == NULL)) {

is taken. In this case the process has been reaped and the timer is marked as
disarmed anyway. So none of the postprocessing of the sample is
required. Return right away.

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Link: http://lkml.kernel.org/r/20160707223911.GA26483@p183.telecom.by
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agomedia: fix airspy usb probe error path
James Patrick-Evans [Fri, 15 Jul 2016 15:40:45 +0000 (16:40 +0100)]
media: fix airspy usb probe error path

commit aa93d1fee85c890a34f2510a310e55ee76a27848 upstream.

Fix a memory leak on probe error of the airspy usb device driver.

The problem is triggered when more than 64 usb devices register with
v4l2 of type VFL_TYPE_SDR or VFL_TYPE_SUBDEV.

The memory leak is caused by the probe function of the airspy driver
mishandeling errors and not freeing the corresponding control structures
when an error occours registering the device to v4l2 core.

A badusb device can emulate 64 of these devices, and then through
continual emulated connect/disconnect of the 65th device, cause the
kernel to run out of RAM and crash the kernel, thus causing a local DOS
vulnerability.

Fixes CVE-2016-5400

Signed-off-by: James Patrick-Evans <james@jmp-e.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoipr: Clear interrupt on croc/crocodile when running with LSI
Brian King [Mon, 27 Jun 2016 14:09:40 +0000 (09:09 -0500)]
ipr: Clear interrupt on croc/crocodile when running with LSI

commit 54e430bbd490e18ab116afa4cd90dcc45787b3df upstream.

If we fall back to using LSI on the Croc or Crocodile chip we need to
clear the interrupt so we don't hang the system.

Tested-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoSCSI: fix new bug in scsi_dev_info_list string matching
Alan Stern [Thu, 23 Jun 2016 19:05:26 +0000 (15:05 -0400)]
SCSI: fix new bug in scsi_dev_info_list string matching

commit 5e7ff2ca7f2da55fe777167849d0c93403bd0dc8 upstream.

Commit b704f70ce200 ("SCSI: fix bug in scsi_dev_info_list matching")
changed the way vendor- and model-string matching was carried out in the
routine that looks up entries in a SCSI devinfo list.  The new matching
code failed to take into account the case of a maximum-length string; in
such cases it could end up testing for a terminating '\0' byte beyond
the end of the memory allocated to the string.  This out-of-bounds bug
was detected by UBSAN.

I don't know if anybody has actually encountered this bug.  The symptom
would be that a device entry in the blacklist might not be matched
properly if it contained an 8-character vendor name or a 16-character
model name.  Such entries certainly exist in scsi_static_device_list.

This patch fixes the problem by adding a check for a maximum-length
string before the '\0' test.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Fixes: b704f70ce200 ("SCSI: fix bug in scsi_dev_info_list matching")
Tested-by: Wilfried Klaebe <linux-kernel@lebenslange-mailadresse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoRDS: fix rds_tcp_init() error path
Vegard Nossum [Sun, 3 Jul 2016 08:54:54 +0000 (10:54 +0200)]
RDS: fix rds_tcp_init() error path

commit 3dad5424adfb346c871847d467f97dcdca64ea97 upstream.

If register_pernet_subsys() fails, we shouldn't try to call
unregister_pernet_subsys().

Fixes: 467fa15356 ("RDS-TCP: Support multiple RDS-TCP listen endpoints, one per netns.")
Cc: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Vegard Nossum <vegard.nossum@oracle.com>
Acked-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>
8 years agocan: fix oops caused by wrong rtnl dellink usage
Oliver Hartkopp [Tue, 21 Jun 2016 13:45:47 +0000 (15:45 +0200)]
can: fix oops caused by wrong rtnl dellink usage

commit 25e1ed6e64f52a692ba3191c4fde650aab3ecc07 upstream.

For 'real' hardware CAN devices the netlink interface is used to set CAN
specific communication parameters. Real CAN hardware can not be created nor
removed with the ip tool ...

This patch adds a private dellink function for the CAN device driver interface
that does just nothing.

It's a follow up to commit 993e6f2fd ("can: fix oops caused by wrong rtnl
newlink usage") but for dellink.

Reported-by: ajneu <ajneu1@gmail.com>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agocan: fix handling of unmodifiable configuration options fix
Oliver Hartkopp [Tue, 21 Jun 2016 10:14:07 +0000 (12:14 +0200)]
can: fix handling of unmodifiable configuration options fix

commit bce271f255dae8335dc4d2ee2c4531e09cc67f5a upstream.

With upstream commit bb208f144cf3f59 (can: fix handling of unmodifiable
configuration options) a new can_validate() function was introduced.

When invoking 'ip link set can0 type can' without any configuration data
can_validate() tries to validate the content without taking into account that
there's totally no content. This patch adds a check for missing content.

Reported-by: ajneu <ajneu1@gmail.com>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agocan: c_can: Update D_CAN TX and RX functions to 32 bit - fix Altera Cyclone access
Thor Thayer [Thu, 16 Jun 2016 16:10:19 +0000 (11:10 -0500)]
can: c_can: Update D_CAN TX and RX functions to 32 bit - fix Altera Cyclone access

commit 427460c83cdf55069eee49799a0caef7dde8df69 upstream.

When testing CAN write floods on Altera's CycloneV, the first 2 bytes
are sometimes 0x00, 0x00 or corrupted instead of the values sent. Also
observed bytes 4 & 5 were corrupted in some cases.

The D_CAN Data registers are 32 bits and changing from 16 bit writes to
32 bit writes fixes the problem.

Testing performed on Altera CycloneV (D_CAN).  Requesting tests on other
C_CAN & D_CAN platforms.

Reported-by: Richard Andrysek <richard.andrysek@gomtec.de>
Signed-off-by: Thor Thayer <tthayer@opensource.altera.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agocan: at91_can: RX queue could get stuck at high bus load
Wolfgang Grandegger [Mon, 13 Jun 2016 13:44:19 +0000 (15:44 +0200)]
can: at91_can: RX queue could get stuck at high bus load

commit 43200a4480cbbe660309621817f54cbb93907108 upstream.

At high bus load it could happen that "at91_poll()" enters with all RX
message boxes filled up. If then at the end the "quota" is exceeded as
well, "rx_next" will not be reset to the first RX mailbox and hence the
interrupts remain disabled.

Signed-off-by: Wolfgang Grandegger <wg@grandegger.com>
Tested-by: Amr Bekhit <amrbekhit@gmail.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoperf/x86: fix PEBS issues on Intel Atom/Core2
Stephane Eranian [Thu, 3 Dec 2015 22:33:18 +0000 (23:33 +0100)]
perf/x86: fix PEBS issues on Intel Atom/Core2

commit 1424a09a9e1839285e948d4ea9fdfca26c9a2086 upstream.

This patch fixes broken PEBS support on Intel Atom and Core2
due to wrong pointer arithmetic in intel_pmu_drain_pebs_core().

The get_next_pebs_record_by_bit() was called on PEBS format fmt0
which does not use the pebs_record_nhm layout.

Signed-off-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: kan.liang@intel.com
Fixes: 21509084f999 ("perf/x86/intel: Handle multiple records in the PEBS buffer")
Link: http://lkml.kernel.org/r/1449182000-31524-3-git-send-email-eranian@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoovl: handle ATTR_KILL*
Miklos Szeredi [Mon, 4 Jul 2016 14:49:48 +0000 (16:49 +0200)]
ovl: handle ATTR_KILL*

commit b99c2d913810e56682a538c9f2394d76fca808f8 upstream.

Before 4bacc9c9234c ("overlayfs: Make f_path...") file->f_path pointed to
the underlying file, hence suid/sgid removal on write worked fine.

After that patch file->f_path pointed to the overlay file, and the file
mode bits weren't copied to overlay_inode->i_mode.  So the suid/sgid
removal simply stopped working.

The fix is to copy the mode bits, but then ovl_setattr() needs to clear
ATTR_MODE to avoid the BUG() in notify_change().  So do this first, then in
the next patch copy the mode.

Reported-by: Eryu Guan <eguan@redhat.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Fixes: 4bacc9c9234c ("overlayfs: Make f_path always point to the overlay and f_inode to the underlay")
Cc: Eric Schultz <eric@startuperic.com>
Cc: Eric Hameleers <alien@slackware.com>
[backported by Eric Hameleers as seen in https://bugzilla.kernel.org/show_bug.cgi?id=150711]
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agosched/fair: Fix effective_load() to consistently use smoothed load
Peter Zijlstra [Fri, 24 Jun 2016 13:53:54 +0000 (15:53 +0200)]
sched/fair: Fix effective_load() to consistently use smoothed load

commit 7dd4912594daf769a46744848b05bd5bc6d62469 upstream.

Starting with the following commit:

  fde7d22e01aa ("sched/fair: Fix overly small weight for interactive group entities")

calc_tg_weight() doesn't compute the right value as expected by effective_load().

The difference is in the 'correction' term. In order to ensure \Sum
rw_j >= rw_i we cannot use tg->load_avg directly, since that might be
lagging a correction on the current cfs_rq->avg.load_avg value.
Therefore we use tg->load_avg - cfs_rq->tg_load_avg_contrib +
cfs_rq->avg.load_avg.

Now, per the referenced commit, calc_tg_weight() doesn't use
cfs_rq->avg.load_avg, as is later used in @w, but uses
cfs_rq->load.weight instead.

So stop using calc_tg_weight() and do it explicitly.

The effects of this bug are wake_affine() making randomly
poor choices in cgroup-intense workloads.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: fde7d22e01aa ("sched/fair: Fix overly small weight for interactive group entities")
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agommc: block: fix packed command header endianness
Taras Kondratiuk [Wed, 13 Jul 2016 22:05:38 +0000 (22:05 +0000)]
mmc: block: fix packed command header endianness

commit f68381a70bb2b26c31b13fdaf67c778f92fd32b4 upstream.

The code that fills packed command header assumes that CPU runs in
little-endian mode. Hence the header is malformed in big-endian mode
and causes MMC data transfer errors:

[  563.200828] mmcblk0: error -110 transferring data, sector 2048, nr 8, cmd response 0x900, card status 0xc40
[  563.219647] mmcblk0: packed cmd failed, nr 2, sectors 16, failure index: -1

Convert header data to LE.

Signed-off-by: Taras Kondratiuk <takondra@cisco.com>
Fixes: ce39f9d17c14 ("mmc: support packed write command for eMMC4.5 devices")
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoblock: fix use-after-free in sys_ioprio_get()
Omar Sandoval [Fri, 1 Jul 2016 07:39:35 +0000 (00:39 -0700)]
block: fix use-after-free in sys_ioprio_get()

commit 8ba8682107ee2ca3347354e018865d8e1967c5f4 upstream.

get_task_ioprio() accesses the task->io_context without holding the task
lock and thus can race with exit_io_context(), leading to a
use-after-free. The reproducer below hits this within a few seconds on
my 4-core QEMU VM:

#define _GNU_SOURCE
#include <assert.h>
#include <unistd.h>
#include <sys/syscall.h>
#include <sys/wait.h>

int main(int argc, char **argv)
{
pid_t pid, child;
long nproc, i;

/* ioprio_set(IOPRIO_WHO_PROCESS, 0, IOPRIO_PRIO_VALUE(IOPRIO_CLASS_IDLE, 0)); */
syscall(SYS_ioprio_set, 1, 0, 0x6000);

nproc = sysconf(_SC_NPROCESSORS_ONLN);

for (i = 0; i < nproc; i++) {
pid = fork();
assert(pid != -1);
if (pid == 0) {
for (;;) {
pid = fork();
assert(pid != -1);
if (pid == 0) {
_exit(0);
} else {
child = wait(NULL);
assert(child == pid);
}
}
}

pid = fork();
assert(pid != -1);
if (pid == 0) {
for (;;) {
/* ioprio_get(IOPRIO_WHO_PGRP, 0); */
syscall(SYS_ioprio_get, 2, 0);
}
}
}

for (;;) {
/* ioprio_get(IOPRIO_WHO_PGRP, 0); */
syscall(SYS_ioprio_get, 2, 0);
}

return 0;
}

This gets us KASAN dumps like this:

[   35.526914] ==================================================================
[   35.530009] BUG: KASAN: out-of-bounds in get_task_ioprio+0x7b/0x90 at addr ffff880066f34e6c
[   35.530009] Read of size 2 by task ioprio-gpf/363
[   35.530009] =============================================================================
[   35.530009] BUG blkdev_ioc (Not tainted): kasan: bad access detected
[   35.530009] -----------------------------------------------------------------------------

[   35.530009] Disabling lock debugging due to kernel taint
[   35.530009] INFO: Allocated in create_task_io_context+0x2b/0x370 age=0 cpu=0 pid=360
[   35.530009]  ___slab_alloc+0x55d/0x5a0
[   35.530009]  __slab_alloc.isra.20+0x2b/0x40
[   35.530009]  kmem_cache_alloc_node+0x84/0x200
[   35.530009]  create_task_io_context+0x2b/0x370
[   35.530009]  get_task_io_context+0x92/0xb0
[   35.530009]  copy_process.part.8+0x5029/0x5660
[   35.530009]  _do_fork+0x155/0x7e0
[   35.530009]  SyS_clone+0x19/0x20
[   35.530009]  do_syscall_64+0x195/0x3a0
[   35.530009]  return_from_SYSCALL_64+0x0/0x6a
[   35.530009] INFO: Freed in put_io_context+0xe7/0x120 age=0 cpu=0 pid=1060
[   35.530009]  __slab_free+0x27b/0x3d0
[   35.530009]  kmem_cache_free+0x1fb/0x220
[   35.530009]  put_io_context+0xe7/0x120
[   35.530009]  put_io_context_active+0x238/0x380
[   35.530009]  exit_io_context+0x66/0x80
[   35.530009]  do_exit+0x158e/0x2b90
[   35.530009]  do_group_exit+0xe5/0x2b0
[   35.530009]  SyS_exit_group+0x1d/0x20
[   35.530009]  entry_SYSCALL_64_fastpath+0x1a/0xa4
[   35.530009] INFO: Slab 0xffffea00019bcd00 objects=20 used=4 fp=0xffff880066f34ff0 flags=0x1fffe0000004080
[   35.530009] INFO: Object 0xffff880066f34e58 @offset=3672 fp=0x0000000000000001
[   35.530009] ==================================================================

Fix it by grabbing the task lock while we poke at the io_context.

Reported-by: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoqeth: delete napi struct when removing a qeth device
Ursula Braun [Mon, 4 Jul 2016 12:07:16 +0000 (14:07 +0200)]
qeth: delete napi struct when removing a qeth device

commit 7831b4ff0d926e0deeaabef9db8800ed069a2757 upstream.

A qeth_card contains a napi_struct linked to the net_device during
device probing. This struct must be deleted when removing the qeth
device, otherwise Panic on oops can occur when qeth devices are
repeatedly removed and added.

Fixes: a1c3ed4c9ca ("qeth: NAPI support for l2 and l3 discipline")
Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
Tested-by: Alexander Klein <ALKL@de.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoplatform/chrome: cros_ec_dev - double fetch bug in ioctl
Dan Carpenter [Tue, 21 Jun 2016 13:58:46 +0000 (16:58 +0300)]
platform/chrome: cros_ec_dev - double fetch bug in ioctl

commit 096cdc6f52225835ff503f987a0d68ef770bb78e upstream.

We verify "u_cmd.outsize" and "u_cmd.insize" but we need to make sure
that those values have not changed between the two copy_from_user()
calls.  Otherwise it could lead to a buffer overflow.

Additionally, cros_ec_cmd_xfer() can set s_cmd->insize to a lower value.
We should use the new smaller value so we don't copy too much data to
the user.

Reported-by: Pengfei Wang <wpengfeinudt@gmail.com>
Fixes: a841178445bb ('mfd: cros_ec: Use a zero-length array for command data')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Tested-by: Gwendal Grignou <gwendal@chromium.org>
Signed-off-by: Olof Johansson <olof@lixom.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoclk: rockchip: initialize flags of clk_init_data in mmc-phase clock
Heiko Stuebner [Tue, 17 May 2016 18:57:50 +0000 (20:57 +0200)]
clk: rockchip: initialize flags of clk_init_data in mmc-phase clock

commit 595144c1141c951a3c6bb9004ae6a2bc29aad66f upstream.

The flags element of clk_init_data was never initialized for mmc-
phase-clocks resulting in the element containing a random value
and thus possibly enabling unwanted clock flags.

Fixes: 89bf26cbc1a0 ("clk: rockchip: Add support for the mmc clock phases using the framework")
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agospi: sun4i: fix FIFO limit
Michal Suchanek [Mon, 13 Jun 2016 17:46:49 +0000 (17:46 +0000)]
spi: sun4i: fix FIFO limit

commit 6d9fe44bd73d567d04d3a68a2d2fa521ab9532f2 upstream.

When testing SPI without DMA I noticed that filling the FIFO on the
spi controller causes timeout.

Always leave room for one byte in the FIFO.

Signed-off-by: Michal Suchanek <hramrach@gmail.com>
Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agospi: sunxi: fix transfer timeout
Michal Suchanek [Mon, 13 Jun 2016 17:46:49 +0000 (17:46 +0000)]
spi: sunxi: fix transfer timeout

commit 719bd6542044efd9b338a53dba1bef45f40ca169 upstream.

The trasfer timeout is fixed at 1000 ms. Reading a 4Mbyte flash over
1MHz SPI bus takes way longer than that. Calculate the timeout from the
actual time the transfer is supposed to take and multiply by 2 for good
measure.

Signed-off-by: Michal Suchanek <hramrach@gmail.com>
Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agonamespace: update event counter when umounting a deleted dentry
Andrey Ulanov [Fri, 15 Apr 2016 21:24:41 +0000 (14:24 -0700)]
namespace: update event counter when umounting a deleted dentry

commit e06b933e6ded42384164d28a2060b7f89243b895 upstream.

- m_start() in fs/namespace.c expects that ns->event is incremented each
  time a mount added or removed from ns->list.
- umount_tree() removes items from the list but does not increment event
  counter, expecting that it's done before the function is called.
- There are some codepaths that call umount_tree() without updating
  "event" counter. e.g. from __detach_mounts().
- When this happens m_start may reuse a cached mount structure that no
  longer belongs to ns->list (i.e. use after free which usually leads
  to infinite loop).

This change fixes the above problem by incrementing global event counter
before invoking umount_tree().

Change-Id: I622c8e84dcb9fb63542372c5dbf0178ee86bb589
Signed-off-by: Andrey Ulanov <andreyu@google.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years ago9p: use file_dentry()
Miklos Szeredi [Wed, 29 Jun 2016 08:54:23 +0000 (10:54 +0200)]
9p: use file_dentry()

commit b403f0e37a11f84f7ceaf40b0075499e5bcfd220 upstream.

v9fs may be used as lower layer of overlayfs and accessing f_path.dentry
can lead to a crash.  In this case it's a NULL pointer dereference in
p9_fid_create().

Fix by replacing direct access of file->f_path.dentry with the
file_dentry() accessor, which will always return a native object.

Reported-by: Alessio Igor Bogani <alessioigorbogani@gmail.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Tested-by: Alessio Igor Bogani <alessioigorbogani@gmail.com>
Fixes: 4bacc9c9234c ("overlayfs: Make f_path always point to the overlay and f_inode to the underlay")
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoext4: verify extent header depth
Vegard Nossum [Fri, 15 Jul 2016 04:22:07 +0000 (00:22 -0400)]
ext4: verify extent header depth

commit 7bc9491645118c9461bd21099c31755ff6783593 upstream.

Although the extent tree depth of 5 should enough be for the worst
case of 2*32 extents of length 1, the extent tree code does not
currently to merge nodes which are less than half-full with a sibling
node, or to shrink the tree depth if possible.  So it's possible, at
least in theory, for the tree depth to be greater than 5.  However,
even in the worst case, a tree depth of 32 is highly unlikely, and if
the file system is maliciously corrupted, an insanely large eh_depth
can cause memory allocation failures that will trigger kernel warnings
(here, eh_depth = 65280):

    JBD2: ext4.exe wants too many credits credits:195849 rsv_credits:0 max:256
    ------------[ cut here ]------------
    WARNING: CPU: 0 PID: 50 at fs/jbd2/transaction.c:293 start_this_handle+0x569/0x580
    CPU: 0 PID: 50 Comm: ext4.exe Not tainted 4.7.0-rc5+ #508
    Stack:
     604a8947 625badd8 0002fd09 00000000
     60078643 00000000 62623910 601bf9bc
     62623970 6002fc84 626239b0 900000125
    Call Trace:
     [<6001c2dc>] show_stack+0xdc/0x1a0
     [<601bf9bc>] dump_stack+0x2a/0x2e
     [<6002fc84>] __warn+0x114/0x140
     [<6002fdff>] warn_slowpath_null+0x1f/0x30
     [<60165829>] start_this_handle+0x569/0x580
     [<60165d4e>] jbd2__journal_start+0x11e/0x220
     [<60146690>] __ext4_journal_start_sb+0x60/0xa0
     [<60120a81>] ext4_truncate+0x131/0x3a0
     [<60123677>] ext4_setattr+0x757/0x840
     [<600d5d0f>] notify_change+0x16f/0x2a0
     [<600b2b16>] do_truncate+0x76/0xc0
     [<600c3e56>] path_openat+0x806/0x1300
     [<600c55c9>] do_filp_open+0x89/0xf0
     [<600b4074>] do_sys_open+0x134/0x1e0
     [<600b4140>] SyS_open+0x20/0x30
     [<6001ea68>] handle_syscall+0x88/0x90
     [<600295fd>] userspace+0x3fd/0x500
     [<6001ac55>] fork_handler+0x85/0x90

    ---[ end trace 08b0b88b6387a244 ]---

[ Commit message modified and the extent tree depath check changed
from 5 to 32 -- tytso ]

Cc: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Vegard Nossum <vegard.nossum@oracle.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoecryptfs: don't allow mmap when the lower fs doesn't support it
Jeff Mahoney [Tue, 5 Jul 2016 21:32:30 +0000 (17:32 -0400)]
ecryptfs: don't allow mmap when the lower fs doesn't support it

commit f0fe970df3838c202ef6c07a4c2b36838ef0a88b upstream.

There are legitimate reasons to disallow mmap on certain files, notably
in sysfs or procfs.  We shouldn't emulate mmap support on file systems
that don't offer support natively.

CVE-2016-1583

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
[tyhicks: clean up f_op check by using ecryptfs_file_to_lower()]
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoRevert "ecryptfs: forbid opening files without mmap handler"
Jeff Mahoney [Tue, 5 Jul 2016 21:32:29 +0000 (17:32 -0400)]
Revert "ecryptfs: forbid opening files without mmap handler"

commit 78c4e172412de5d0456dc00d2b34050aa0b683b5 upstream.

This reverts commit 2f36db71009304b3f0b95afacd8eba1f9f046b87.

It fixed a local root exploit but also introduced a dependency on
the lower file system implementing an mmap operation just to open a file,
which is a bit of a heavy hammer.  The right fix is to have mmap depend
on the existence of the mmap handler instead.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agolocks: use file_inode()
Miklos Szeredi [Fri, 1 Jul 2016 12:56:07 +0000 (14:56 +0200)]
locks: use file_inode()

commit 6343a2120862f7023006c8091ad95c1f16a32077 upstream.

(Another one for the f_path debacle.)

ltp fcntl33 testcase caused an Oops in selinux_file_send_sigiotask.

The reason is that generic_add_lease() used filp->f_path.dentry->inode
while all the others use file_inode().  This makes a difference for files
opened on overlayfs since the former will point to the overlay inode the
latter to the underlying inode.

So generic_add_lease() added the lease to the overlay inode and
generic_delete_lease() removed it from the underlying inode.  When the file
was released the lease remained on the overlay inode's lock list, resulting
in use after free.

Reported-by: Eryu Guan <eguan@redhat.com>
Fixes: 4bacc9c9234c ("overlayfs: Make f_path always point to the overlay and f_inode to the underlay")
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agopower_supply: power_supply_read_temp only if use_cnt > 0
Rhyland Klein [Thu, 9 Jun 2016 21:28:39 +0000 (17:28 -0400)]
power_supply: power_supply_read_temp only if use_cnt > 0

commit 5bc28b93a36e3cb3acc2870fb75cb6ffb182fece upstream.

Change power_supply_read_temp() to use power_supply_get_property()
so that it will check the use_cnt and ensure it is > 0. The use_cnt
will be incremented at the end of __power_supply_register, so this
will block to case where get_property can be called before the supply
is fully registered. This fixes the issue show in the stack below:

[    1.452598] power_supply_read_temp+0x78/0x80
[    1.458680] thermal_zone_get_temp+0x5c/0x11c
[    1.464765] thermal_zone_device_update+0x34/0xb4
[    1.471195] thermal_zone_device_register+0x87c/0x8cc
[    1.477974] __power_supply_register+0x364/0x424
[    1.484317] power_supply_register_no_ws+0x10/0x18
[    1.490833] bq27xxx_battery_setup+0x10c/0x164
[    1.497003] bq27xxx_battery_i2c_probe+0xd0/0x1b0
[    1.503435] i2c_device_probe+0x174/0x240
[    1.509172] driver_probe_device+0x1fc/0x29c
[    1.515167] __driver_attach+0xa4/0xa8
[    1.520643] bus_for_each_dev+0x58/0x98
[    1.526204] driver_attach+0x20/0x28
[    1.531505] bus_add_driver+0x1c8/0x22c
[    1.537067] driver_register+0x68/0x108
[    1.542630] i2c_register_driver+0x38/0x7c
[    1.548457] bq27xxx_battery_i2c_driver_init+0x18/0x20
[    1.555321] do_one_initcall+0x38/0x12c
[    1.560886] kernel_init_freeable+0x148/0x1ec
[    1.566972] kernel_init+0x10/0xfc
[    1.572101] ret_from_fork+0x10/0x40

Also make the same change to ps_get_max_charge_cntl_limit() and
ps_get_cur_chrage_cntl_limit() to be safe. Lastly, change the return
value of power_supply_get_property() to -EAGAIN from -ENODEV if
use_cnt <= 0.

Fixes: 297d716f6260 ("power_supply: Change ownership from driver to core")
Signed-off-by: Rhyland Klein <rklein@nvidia.com>
Reviewed-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agocgroup: set css->id to -1 during init
Tejun Heo [Thu, 26 May 2016 19:42:13 +0000 (15:42 -0400)]
cgroup: set css->id to -1 during init

commit 8fa3b8d689a54d6d04ff7803c724fb7aca6ce98e upstream.

If percpu_ref initialization fails during css_create(), the free path
can end up trying to free css->id of zero.  As ID 0 is unused, it
doesn't cause a critical breakage but it does trigger a warning
message.  Fix it by setting css->id to -1 from init_and_link_css().

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Wenwei Tao <ww.tao0320@gmail.com>
Fixes: 01e586598b22 ("cgroup: release css->id after css_free")
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agopinctrl: imx: Do not treat a PIN without MUX register as an error
Alexander Shiyan [Wed, 1 Jun 2016 19:21:53 +0000 (22:21 +0300)]
pinctrl: imx: Do not treat a PIN without MUX register as an error

commit ba562d5e54fd3136bfea0457add3675850247774 upstream.

Some PINs do not have a MUX register, it is not an error.
It is necessary to allow the continuation of the PINs configuration,
otherwise the whole PIN-group will be configured incorrectly.

Signed-off-by: Alexander Shiyan <shc_work@mail.ru>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agopinctrl: single: Fix missing flush of posted write for a wakeirq
Tony Lindgren [Tue, 31 May 2016 21:17:06 +0000 (14:17 -0700)]
pinctrl: single: Fix missing flush of posted write for a wakeirq

commit 0ac3c0a4025f41748a083bdd4970cb3ede802b15 upstream.

With many repeated suspend resume cycles, the pin specific wakeirq
may not always work on omaps. This is because the write to enable the
pin interrupt may not have reached the device over the interconnect
before suspend happens.

Let's fix the issue with a flush of posted write with a readback.

Reported-by: Nishanth Menon <nm@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agopvclock: Add CPU barriers to get correct version value
Minfei Huang [Fri, 27 May 2016 06:17:10 +0000 (14:17 +0800)]
pvclock: Add CPU barriers to get correct version value

commit 749d088b8e7f4b9826ede02b9a043e417fa84aa1 upstream.

Protocol for the "version" fields is: hypervisor raises it (making it
uneven) before it starts updating the fields and raises it again (making
it even) when it is done.  Thus the guest can make sure the time values
it got are consistent by checking the version before and after reading
them.

Add CPU barries after getting version value just like what function
vread_pvclock does, because all of callees in this function is inline.

Fixes: 502dfeff239e8313bfbe906ca0a1a6827ac8481b
Signed-off-by: Minfei Huang <mnghuan@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoInput: tsc200x - report proper input_dev name
Michael Welling [Wed, 20 Jul 2016 17:02:07 +0000 (10:02 -0700)]
Input: tsc200x - report proper input_dev name

commit e9003c9cfaa17d26991688268b04244adb67ee2b upstream.

Passes input_id struct to the common probe function for the tsc200x drivers
instead of just the bustype.

This allows for the use of the product variable to set the input_dev->name
variable according to the type of touchscreen used. Note that when we
introduced support for TSC2004 we started calling everything TSC200X, so
let's keep this quirk.

Signed-off-by: Michael Welling <mwelling@ieee.org>
Acked-by: Pavel Machek <pavel@ucw.cz>
Acked-by: Pali Rohár <pali.rohar@gmail.com>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoInput: xpad - validate USB endpoint count during probe
Cameron Gutman [Wed, 29 Jun 2016 16:51:35 +0000 (09:51 -0700)]
Input: xpad - validate USB endpoint count during probe

commit caca925fca4fb30c67be88cacbe908eec6721e43 upstream.

This prevents a malicious USB device from causing an oops.

Signed-off-by: Cameron Gutman <aicommander@gmail.com>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoInput: wacom_w8001 - w8001_MAX_LENGTH should be 13
Ping Cheng [Thu, 23 Jun 2016 17:54:17 +0000 (10:54 -0700)]
Input: wacom_w8001 - w8001_MAX_LENGTH should be 13

commit 12afb34400eb2b301f06b2aa3535497d14faee59 upstream.

Somehow the patch that added two-finger touch support forgot to update
W8001_MAX_LENGTH from 11 to 13.

Signed-off-by: Ping Cheng <pingc@wacom.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoInput: xpad - fix oops when attaching an unknown Xbox One gamepad
Cameron Gutman [Thu, 23 Jun 2016 17:24:42 +0000 (10:24 -0700)]
Input: xpad - fix oops when attaching an unknown Xbox One gamepad

commit c7f1429389ec1aa25e042bb13451385fbb596f8c upstream.

Xbox One controllers have multiple interfaces which all have the
same class, subclass, and protocol. One of the these interfaces
has only a single endpoint. When Xpad attempts to bind to this
interface, it causes an oops when trying initialize the output URB
by trying to access the second endpoint's descriptor.

This situation was avoided for known Xbox One devices by checking
the XTYPE constant associated with the VID and PID tuple. However,
this breaks when new or previously unknown Xbox One controllers
are attached to the system.

This change addresses the problem by deriving the XTYPE for Xbox
One controllers based on the interface protocol before checking
the interface number.

Fixes: 1a48ff81b391 ("Input: xpad - add support for Xbox One controllers")
Signed-off-by: Cameron Gutman <aicommander@gmail.com>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoInput: elantech - add more IC body types to the list
Dmitry Torokhov [Tue, 21 Jun 2016 23:09:00 +0000 (16:09 -0700)]
Input: elantech - add more IC body types to the list

commit 226ba707744a51acb4244724e09caacb1d96aed9 upstream.

The touchpad in HP Pavilion 14-ab057ca reports it's version as 12 and
according to Elan both 11 and 12 are valid IC types and should be
identified as hw_version 4.

Reported-by: Patrick Lessard <Patrick.Lessard@cogeco.com>
Tested-by: Patrick Lessard <Patrick.Lessard@cogeco.com>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoInput: vmmouse - remove port reservation
Sinclair Yeh [Fri, 24 Jun 2016 00:37:34 +0000 (17:37 -0700)]
Input: vmmouse - remove port reservation

commit 60842ef8128e7bf58c024814cd0dc14319232b6c upstream.

The VMWare EFI BIOS will expose port 0x5658 as an ACPI resource.  This
causes the port to be reserved by the APCI module as the system comes up,
making it unavailable to be reserved again by other drivers, thus
preserving this VMWare port for special use in a VMWare guest.

This port is designed to be shared among multiple VMWare services, such as
the VMMOUSE.  Because of this, VMMOUSE should not try to reserve this port
on its own.

The VMWare non-EFI BIOS does not do this to preserve compatibility with
existing/legacy VMs.  It is known that there is small chance a VM may be
configured such that these ports get reserved by other non-VMWare devices,
and if this ever happens, the result is undefined.

Signed-off-by: Sinclair Yeh <syeh@vmware.com>
Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoALSA: timer: Fix leak in events via snd_timer_user_tinterrupt
Kangjie Lu [Tue, 3 May 2016 20:44:32 +0000 (16:44 -0400)]
ALSA: timer: Fix leak in events via snd_timer_user_tinterrupt

commit e4ec8cc8039a7063e24204299b462bd1383184a5 upstream.

The stack object “r1” has a total size of 32 bytes. Its field
“event” and “val” both contain 4 bytes padding. These 8 bytes
padding bytes are sent to user without being initialized.

Signed-off-by: Kangjie Lu <kjlu@gatech.edu>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoALSA: timer: Fix leak in events via snd_timer_user_ccallback
Kangjie Lu [Tue, 3 May 2016 20:44:20 +0000 (16:44 -0400)]
ALSA: timer: Fix leak in events via snd_timer_user_ccallback

commit 9a47e9cff994f37f7f0dbd9ae23740d0f64f9fe6 upstream.

The stack object “r1” has a total size of 32 bytes. Its field
“event” and “val” both contain 4 bytes padding. These 8 bytes
padding bytes are sent to user without being initialized.

Signed-off-by: Kangjie Lu <kjlu@gatech.edu>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoALSA: timer: Fix leak in SNDRV_TIMER_IOCTL_PARAMS
Kangjie Lu [Tue, 3 May 2016 20:44:07 +0000 (16:44 -0400)]
ALSA: timer: Fix leak in SNDRV_TIMER_IOCTL_PARAMS

commit cec8f96e49d9be372fdb0c3836dcf31ec71e457e upstream.

The stack object “tread” has a total size of 32 bytes. Its field
“event” and “val” both contain 4 bytes padding. These 8 bytes
padding bytes are sent to user without being initialized.

Signed-off-by: Kangjie Lu <kjlu@gatech.edu>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoxenbus: don't bail early from xenbus_dev_request_and_reply()
Jan Beulich [Thu, 7 Jul 2016 07:32:04 +0000 (01:32 -0600)]
xenbus: don't bail early from xenbus_dev_request_and_reply()

commit 7469be95a487319514adce2304ad2af3553d2fc9 upstream.

xenbus_dev_request_and_reply() needs to track whether a transaction is
open.  For XS_TRANSACTION_START messages it calls transaction_start()
and for XS_TRANSACTION_END messages it calls transaction_end().

If sending an XS_TRANSACTION_START message fails or responds with an
an error, the transaction is not open and transaction_end() must be
called.

If sending an XS_TRANSACTION_END message fails, the transaction is
still open, but if an error response is returned the transaction is
closed.

Commit 027bd7e89906 ("xen/xenbus: Avoid synchronous wait on XenBus
stalling shutdown/restart") introduced a regression where failed
XS_TRANSACTION_START messages were leaving the transaction open.  This
can cause problems with suspend (and migration) as all transactions
must be closed before suspending.

It appears that the problematic change was added accidentally, so just
remove it.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoxenbus: don't BUG() on user mode induced condition
Jan Beulich [Thu, 7 Jul 2016 07:23:57 +0000 (01:23 -0600)]
xenbus: don't BUG() on user mode induced condition

commit 0beef634b86a1350c31da5fcc2992f0d7c8a622b upstream.

Inability to locate a user mode specified transaction ID should not
lead to a kernel crash. For other than XS_TRANSACTION_START also
don't issue anything to xenbus if the specified ID doesn't match that
of any active transaction.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoxen/pciback: Fix conf_space read/write overlap check.
Andrey Grodzovsky [Tue, 21 Jun 2016 18:26:36 +0000 (14:26 -0400)]
xen/pciback: Fix conf_space read/write overlap check.

commit 02ef871ecac290919ea0c783d05da7eedeffc10e upstream.

Current overlap check is evaluating to false a case where a filter
field is fully contained (proper subset) of a r/w request.  This
change applies classical overlap check instead to include all the
scenarios.

More specifically, for (Hilscher GmbH CIFX 50E-DP(M/S)) device driver
the logic is such that the entire confspace is read and written in 4
byte chunks. In this case as an example, CACHE_LINE_SIZE,
LATENCY_TIMER and PCI_BIST are arriving together in one call to
xen_pcibk_config_write() with offset == 0xc and size == 4.  With the
exsisting overlap check the LATENCY_TIMER field (offset == 0xd, length
== 1) is fully contained in the write request and hence is excluded
from write, which is incorrect.

Signed-off-by: Andrey Grodzovsky <andrey2805@gmail.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Reviewed-by: Jan Beulich <JBeulich@suse.com>
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoARC: unwind: ensure that .debug_frame is generated (vs. .eh_frame)
Vineet Gupta [Tue, 28 Jun 2016 04:12:25 +0000 (09:42 +0530)]
ARC: unwind: ensure that .debug_frame is generated (vs. .eh_frame)

commit f52e126cc7476196f44f3c313b7d9f0699a881fc upstream.

With recent binutils update to support dwarf CFI pseudo-ops in gas, we
now get .eh_frame vs. .debug_frame. Although the call frame info is
exactly the same in both, the CIE differs, which the current kernel
unwinder can't cope with.

This broke both the kernel unwinder as well as loadable modules (latter
because of a new unhandled relo R_ARC_32_PCREL from .rela.eh_frame in
the module loader)

The ideal solution would be to switch unwinder to .eh_frame.
For now however we can make do by just ensureing .debug_frame is
generated by removing -fasynchronous-unwind-tables

 .eh_frame    generated with -gdwarf-2 -fasynchronous-unwind-tables
 .debug_frame generated with -gdwarf-2

Fixes STAR 9001058196

Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoarc: unwind: warn only once if DW2_UNWIND is disabled
Alexey Brodkin [Thu, 23 Jun 2016 08:00:39 +0000 (11:00 +0300)]
arc: unwind: warn only once if DW2_UNWIND is disabled

commit 9bd54517ee86cb164c734f72ea95aeba4804f10b upstream.

If CONFIG_ARC_DW2_UNWIND is disabled every time arc_unwind_core()
gets called following message gets printed in debug console:
----------------->8---------------
CONFIG_ARC_DW2_UNWIND needs to be enabled
----------------->8---------------

That message makes sense if user indeed wants to see a backtrace or
get nice function call-graphs in perf but what if user disabled
unwinder for the purpose? Why pollute his debug console?

So instead we'll warn user about possibly missing feature once and
let him decide if that was what he or she really wanted.

Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agokernel/sysrq, watchdog, sched/core: Reset watchdog on all CPUs while processing sysrq-w
Andrey Ryabinin [Thu, 9 Jun 2016 12:20:05 +0000 (15:20 +0300)]
kernel/sysrq, watchdog, sched/core: Reset watchdog on all CPUs while processing sysrq-w

commit 57675cb976eff977aefb428e68e4e0236d48a9ff upstream.

Lengthy output of sysrq-w may take a lot of time on slow serial console.

Currently we reset NMI-watchdog on the current CPU to avoid spurious
lockup messages. Sometimes this doesn't work since softlockup watchdog
might trigger on another CPU which is waiting for an IPI to proceed.
We reset softlockup watchdogs on all CPUs, but we do this only after
listing all tasks, and this may be too late on a busy system.

So, reset watchdogs CPUs earlier, in for_each_process_thread() loop.

Signed-off-by: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1465474805-14641-1-git-send-email-aryabinin@virtuozzo.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agopps: do not crash when failed to register
Jiri Slaby [Wed, 20 Jul 2016 22:45:08 +0000 (15:45 -0700)]
pps: do not crash when failed to register

commit 368301f2fe4b07e5fb71dba3cc566bc59eb6705f upstream.

With this command sequence:

  modprobe plip
  modprobe pps_parport
  rmmod pps_parport

the partport_pps modules causes this crash:

  BUG: unable to handle kernel NULL pointer dereference at (null)
  IP: parport_detach+0x1d/0x60 [pps_parport]
  Oops: 0000 [#1] SMP
  ...
  Call Trace:
    parport_unregister_driver+0x65/0xc0 [parport]
    SyS_delete_module+0x187/0x210

The sequence that builds up to this is:

 1) plip is loaded and takes the parport device for exclusive use:

    plip0: Parallel port at 0x378, using IRQ 7.

 2) pps_parport then fails to grab the device:

    pps_parport: parallel port PPS client
    parport0: cannot grant exclusive access for device pps_parport
    pps_parport: couldn't register with parport0

 3) rmmod of pps_parport is then killed because it tries to access
    pardev->name, but pardev (taken from port->cad) is NULL.

So add a check for NULL in the test there too.

Link: http://lkml.kernel.org/r/20160714115245.12651-1-jslaby@suse.cz
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Acked-by: Rodolfo Giometti <giometti@enneenne.com>
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>
8 years agovmlinux.lds: account for destructor sections
Dmitry Vyukov [Thu, 14 Jul 2016 19:07:29 +0000 (12:07 -0700)]
vmlinux.lds: account for destructor sections

commit e41f501d391265ff568f3e49d6128cc30856a36f upstream.

If CONFIG_KASAN is enabled and gcc is configured with
--disable-initfini-array and/or gold linker is used, gcc emits
.ctors/.dtors and .text.startup/.text.exit sections instead of
.init_array/.fini_array.  .dtors section is not explicitly accounted in
the linker script and messes vvar/percpu layout.

We want:
  ffffffff822bfd80 D _edata
  ffffffff822c0000 D __vvar_beginning_hack
  ffffffff822c0000 A __vvar_page
  ffffffff822c0080 0000000000000098 D vsyscall_gtod_data
  ffffffff822c1000 A __init_begin
  ffffffff822c1000 D init_per_cpu__irq_stack_union
  ffffffff822c1000 A __per_cpu_load
  ffffffff822d3000 D init_per_cpu__gdt_page

We got:
  ffffffff8279a600 D _edata
  ffffffff8279b000 A __vvar_page
  ffffffff8279c000 A __init_begin
  ffffffff8279c000 D init_per_cpu__irq_stack_union
  ffffffff8279c000 A __per_cpu_load
  ffffffff8279e000 D __vvar_beginning_hack
  ffffffff8279e080 0000000000000098 D vsyscall_gtod_data
  ffffffff827ae000 D init_per_cpu__gdt_page

This happens because __vvar_page and .vvar get different addresses in
arch/x86/kernel/vmlinux.lds.S:

. = ALIGN(PAGE_SIZE);
__vvar_page = .;

.vvar : AT(ADDR(.vvar) - LOAD_OFFSET) {
/* work around gold bug 13023 */
__vvar_beginning_hack = .;

Discard .dtors/.fini_array/.text.exit, since we don't call dtors.
Merge .text.startup into init text.

Link: http://lkml.kernel.org/r/1467386363-120030-1-git-send-email-dvyukov@google.com
Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Reviewed-by: Andrey Ryabinin <aryabinin@virtuozzo.com>
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>