linux.git/arch, branch v5.0.9

ARM: 8839/1: kprobe: make patch_lock a raw_spinlock_t

2019-04-20T07:17:04+00:00

[ Upstream commit 143c2a89e0e5fda6c6fd08d7bc1126438c19ae90 ]

When running kprobe on -rt kernel, the below bug is caught:

|BUG: sleeping function called from invalid context at kernel/locking/rtmutex.c:931
|in_atomic(): 1, irqs_disabled(): 128, pid: 14, name: migration/0
|Preemption disabled at:[<802f2b98>] cpu_stopper_thread+0xc0/0x140
|CPU: 0 PID: 14 Comm: migration/0 Tainted: G O 4.8.3-rt2 #1
|Hardware name: Freescale LS1021A
|[<8025a43c>] (___might_sleep)
|[<80b5b324>] (rt_spin_lock)
|[<80b5c31c>] (__patch_text_real)
|[<80b5c3ac>] (patch_text_stop_machine)
|[<802f2920>] (multi_cpu_stop)

Since patch_text_stop_machine() is called in stop_machine() which
disables IRQ, sleepable lock should be not used in this atomic context,
 so replace patch_lock to raw lock.

Signed-off-by: Yang Shi 
Signed-off-by: Sebastian Andrzej Siewior 
Reviewed-by: Arnd Bergmann 
Signed-off-by: Russell King 
Signed-off-by: Sasha Levin

x86/gart: Exclude GART aperture from kcore

2019-04-20T07:17:02+00:00

[ Upstream commit ffc8599aa9763f39f6736a79da4d1575e7006f9a ]

On machines where the GART aperture is mapped over physical RAM,
/proc/kcore contains the GART aperture range. Accessing the GART range via
/proc/kcore results in a kernel crash.

vmcore used to have the same issue, until it was fixed with commit
2a3e83c6f96c ("x86/gart: Exclude GART aperture from vmcore")', leveraging
existing hook infrastructure in vmcore to let /proc/vmcore return zeroes
when attempting to read the aperture region, and so it won't read from the
actual memory.

Apply the same workaround for kcore. First implement the same hook
infrastructure for kcore, then reuse the hook functions introduced in the
previous vmcore fix. Just with some minor adjustment, rename some functions
for more general usage, and simplify the hook infrastructure a bit as there
is no module usage yet.

Suggested-by: Baoquan He 
Signed-off-by: Kairui Song 
Signed-off-by: Thomas Gleixner 
Reviewed-by: Jiri Bohac 
Acked-by: Baoquan He 
Cc: Borislav Petkov 
Cc: "H. Peter Anvin" 
Cc: Alexey Dobriyan 
Cc: Andrew Morton 
Cc: Omar Sandoval 
Cc: Dave Young 
Link: https://lkml.kernel.org/r/20190308030508.13548-1-kasong@redhat.com
Signed-off-by: Sasha Levin

x86/hw_breakpoints: Make default case in hw_breakpoint_arch_parse() return an error

2019-04-20T07:17:02+00:00

[ Upstream commit e898e69d6b9475bf123f99b3c5d1a67bb7cb2361 ]

When building with -Wsometimes-uninitialized, Clang warns:

arch/x86/kernel/hw_breakpoint.c:355:2: warning: variable 'align' is used
uninitialized whenever switch default is taken
[-Wsometimes-uninitialized]

The default cannot be reached because arch_build_bp_info() initializes
hw->len to one of the specified cases. Nevertheless the warning is valid
and returning -EINVAL makes sure that this cannot be broken by future
modifications.

Suggested-by: Nick Desaulniers 
Signed-off-by: Nathan Chancellor 
Signed-off-by: Thomas Gleixner 
Reviewed-by: Nick Desaulniers 
Cc: Borislav Petkov 
Cc: "H. Peter Anvin" 
Cc: clang-built-linux@googlegroups.com
Link: https://github.com/ClangBuiltLinux/linux/issues/392
Link: https://lkml.kernel.org/r/20190307212756.4648-1-natechancellor@gmail.com
Signed-off-by: Sasha Levin

x86/cpu/cyrix: Use correct macros for Cyrix calls on Geode processors

2019-04-20T07:17:01+00:00

[ Upstream commit 18fb053f9b827bd98cfc64f2a35df8ab19745a1d ]

There are comments in processor-cyrix.h advising you to _not_ make calls
using the deprecated macros in this style:

  setCx86_old(CX86_CCR4, getCx86_old(CX86_CCR4) | 0x80);

This is because it expands the macro into a non-functioning calling
sequence. The calling order must be:

  outb(CX86_CCR2, 0x22);
  inb(0x23);

From the comments:

 * When using the old macros a line like
 *   setCx86(CX86_CCR2, getCx86(CX86_CCR2) | 0x88);
 * gets expanded to:
 *  do {
 *    outb((CX86_CCR2), 0x22);
 *    outb((({
 *        outb((CX86_CCR2), 0x22);
 *        inb(0x23);
 *    }) | 0x88), 0x23);
 *  } while (0);

The new macros fix this problem, so use them instead. Tested on an
actual Geode processor.

Signed-off-by: Matthew Whitehead 
Signed-off-by: Thomas Gleixner 
Cc: luto@kernel.org
Link: https://lkml.kernel.org/r/1552596361-8967-2-git-send-email-tedheadster@gmail.com
Signed-off-by: Sasha Levin

x86/hyperv: Prevent potential NULL pointer dereference

2019-04-20T07:17:01+00:00

[ Upstream commit 534c89c22e26b183d838294f0937ee092c82ad3a ]

The page allocation in hv_cpu_init() can fail, but the code does not
have a check for that.

Add a check and return -ENOMEM when the allocation fails.

[ tglx: Massaged changelog ]

Signed-off-by: Kangjie Lu 
Signed-off-by: Thomas Gleixner 
Reviewed-by: Mukesh Ojha 
Acked-by: "K. Y. Srinivasan" 
Cc: pakki001@umn.edu
Cc: Haiyang Zhang 
Cc: Stephen Hemminger 
Cc: Sasha Levin 
Cc: Borislav Petkov 
Cc: "H. Peter Anvin" 
Cc: linux-hyperv@vger.kernel.org
Link: https://lkml.kernel.org/r/20190314054651.1315-1-kjlu@umn.edu
Signed-off-by: Sasha Levin

x86/hpet: Prevent potential NULL pointer dereference

2019-04-20T07:17:01+00:00

[ Upstream commit 2e84f116afca3719c9d0a1a78b47b48f75fd5724 ]

hpet_virt_address may be NULL when ioremap_nocache fail, but the code lacks
a check.

Add a check to prevent NULL pointer dereference.

Signed-off-by: Aditya Pakki 
Signed-off-by: Thomas Gleixner 
Cc: kjlu@umn.edu
Cc: Borislav Petkov 
Cc: "H. Peter Anvin" 
Cc: Kees Cook 
Cc: Joe Perches 
Cc: Nicolai Stange 
Cc: Roland Dreier 
Link: https://lkml.kernel.org/r/20190319021958.17275-1-pakki001@umn.edu
Signed-off-by: Sasha Levin

x86/mm: Don't leak kernel addresses

2019-04-20T07:16:50+00:00

[ Upstream commit a3151724437f54076cc10bc02b1c4f0003ae36cd ]

Since commit:

  ad67b74d2469d9b8 ("printk: hash addresses printed with %p")

at boot "____ptrval____" is printed instead of actual addresses:

    found SMP MP-table at [mem 0x000f5cc0-0x000f5ccf] mapped at [(____ptrval____)]

Instead of changing the print to "%px", and leaking a kernel addresses,
just remove the print completely, like in:

  071929dbdd865f77 ("arm64: Stop printing the virtual memory layout").

Signed-off-by: Matteo Croce 
Cc: Borislav Petkov 
Cc: Linus Torvalds 
Cc: Peter Zijlstra 
Cc: Thomas Gleixner 
Signed-off-by: Ingo Molnar 
Signed-off-by: Sasha Levin

mips: bcm47xx: Enable USB power on Netgear WNDR3400v2

2019-04-20T07:16:45+00:00

[ Upstream commit cdb8faa00e3fcdd0ad10add743516d616dc7d38e ]

Eric has reported on OpenWrt's bug tracking system[1], that he's not
able to use USB devices on his WNDR3400v2 device after the boot, until
he turns on GPIO #21 manually through sysfs.

1. https://bugs.openwrt.org/index.php?do=details&task_id=2170

Cc: Rafał Miłecki 
Cc: Hauke Mehrtens 
Reported-by: Eric Bohlman 
Tested-by: Eric Bohlman 
Signed-off-by: Petr Štetiar 
Signed-off-by: Paul Burton 
Signed-off-by: Sasha Levin

arc: hsdk_defconfig: Enable CONFIG_BLK_DEV_RAM

2019-04-20T07:16:45+00:00

[ Upstream commit 0728aeb7ead99a9b0dac2f3c92b3752b4e02ff97 ]

We have now a HSDK device in our kernelci lab, but kernel builded via
the hsdk_defconfig lacks ramfs supports, so it cannot boot kernelci jobs
yet.

So this patch enable CONFIG_BLK_DEV_RAM in hsdk_defconfig.

Signed-off-by: Corentin Labbe 
Acked-by: Alexey Brodkin 
Signed-off-by: Vineet Gupta 
Signed-off-by: Sasha Levin

ARC: u-boot args: check that magic number is correct

2019-04-20T07:16:45+00:00

[ Upstream commit edb64bca50cd736c6894cc6081d5263c007ce005 ]

In case of devboards we really often disable bootloader and load
Linux image in memory via JTAG. Even if kernel tries to verify
uboot_tag and uboot_arg there is sill a chance that we treat some
garbage in registers as valid u-boot arguments in JTAG case.
E.g. it is enough to have '1' in r0 to treat any value in r2 as
a boot command line.

So check that magic number passed from u-boot is correct and drop
u-boot arguments otherwise. That helps to reduce the possibility
of using garbage as u-boot arguments in JTAG case.

We can safely check U-boot magic value (0x0) in linux passed via
r1 register as U-boot pass it from the beginning. So there is no
backward-compatibility issues.

Signed-off-by: Eugeniy Paltsev 
Signed-off-by: Vineet Gupta 
Signed-off-by: Sasha Levin