platform/kernel/linux-rpi.git
6 years agolocking/lockdep: Fix debug_locks off performance problem
Waiman Long [Fri, 19 Oct 2018 01:45:17 +0000 (21:45 -0400)]
locking/lockdep: Fix debug_locks off performance problem

It was found that when debug_locks was turned off because of a problem
found by the lockdep code, the system performance could drop quite
significantly when the lock_stat code was also configured into the
kernel. For instance, parallel kernel build time on a 4-socket x86-64
server nearly doubled.

Further analysis into the cause of the slowdown traced back to the
frequent call to debug_locks_off() from the __lock_acquired() function
probably due to some inconsistent lockdep states with debug_locks
off. The debug_locks_off() function did an unconditional atomic xchg
to write a 0 value into debug_locks which had already been set to 0.
This led to severe cacheline contention in the cacheline that held
debug_locks.  As debug_locks is being referenced in quite a few different
places in the kernel, this greatly slow down the system performance.

To prevent that trashing of debug_locks cacheline, lock_acquired()
and lock_contended() now checks the state of debug_locks before
proceeding. The debug_locks_off() function is also modified to check
debug_locks before calling __debug_locks_off().

Signed-off-by: Waiman Long <longman@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Link: http://lkml.kernel.org/r/1539913518-15598-1-git-send-email-longman@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agolocking/pvqspinlock: Extend node size when pvqspinlock is configured
Waiman Long [Tue, 16 Oct 2018 13:45:07 +0000 (09:45 -0400)]
locking/pvqspinlock: Extend node size when pvqspinlock is configured

The qspinlock code supports up to 4 levels of slowpath nesting using
four per-CPU mcs_spinlock structures. For 64-bit architectures, they
fit nicely in one 64-byte cacheline.

For para-virtualized (PV) qspinlocks it needs to store more information
in the per-CPU node structure than there is space for. It uses a trick
to use a second cacheline to hold the extra information that it needs.
So PV qspinlock needs to access two extra cachelines for its information
whereas the native qspinlock code only needs one extra cacheline.

Freshly added counter profiling of the qspinlock code, however, revealed
that it was very rare to use more than two levels of slowpath nesting.
So it doesn't make sense to penalize PV qspinlock code in order to have
four mcs_spinlock structures in the same cacheline to optimize for a case
in the native qspinlock code that rarely happens.

Extend the per-CPU node structure to have two more long words when PV
qspinlock locks are configured to hold the extra data that it needs.

As a result, the PV qspinlock code will enjoy the same benefit of using
just one extra cacheline like the native counterpart, for most cases.

[ mingo: Minor changelog edits. ]

Signed-off-by: Waiman Long <longman@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Link: http://lkml.kernel.org/r/1539697507-28084-2-git-send-email-longman@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agolocking/qspinlock_stat: Count instances of nested lock slowpaths
Waiman Long [Tue, 16 Oct 2018 13:45:06 +0000 (09:45 -0400)]
locking/qspinlock_stat: Count instances of nested lock slowpaths

Queued spinlock supports up to 4 levels of lock slowpath nesting -
user context, soft IRQ, hard IRQ and NMI. However, we are not sure how
often the nesting happens.

So add 3 more per-CPU stat counters to track the number of instances where
nesting index goes to 1, 2 and 3 respectively.

On a dual-socket 64-core 128-thread Zen server, the following were the
new stat counter values under different circumstances:

         State                         slowpath   index1   index2   index3
         -----                         --------   ------   ------   -------
  After bootup                         1,012,150    82       0        0
  After parallel build + perf-top    125,195,009    82       0        0

So the chance of having more than 2 levels of nesting is extremely low.

[ mingo: Minor changelog edits. ]

Signed-off-by: Waiman Long <longman@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Link: http://lkml.kernel.org/r/1539697507-28084-1-git-send-email-longman@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agolocking/qspinlock, x86: Provide liveness guarantee
Peter Zijlstra [Wed, 26 Sep 2018 11:01:20 +0000 (13:01 +0200)]
locking/qspinlock, x86: Provide liveness guarantee

On x86 we cannot do fetch_or() with a single instruction and thus end up
using a cmpxchg loop, this reduces determinism. Replace the fetch_or()
with a composite operation: tas-pending + load.

Using two instructions of course opens a window we previously did not
have. Consider the scenario:

CPU0 CPU1 CPU2

 1) lock
  trylock -> (0,0,1)

 2) lock
  trylock /* fail */

 3) unlock -> (0,0,0)

 4) lock
  trylock -> (0,0,1)

 5)   tas-pending -> (0,1,1)
  load-val <- (0,1,0) from 3

 6)   clear-pending-set-locked -> (0,0,1)

  FAIL: _2_ owners

where 5) is our new composite operation. When we consider each part of
the qspinlock state as a separate variable (as we can when
_Q_PENDING_BITS == 8) then the above is entirely possible, because
tas-pending will only RmW the pending byte, so the later load is able
to observe prior tail and lock state (but not earlier than its own
trylock, which operates on the whole word, due to coherence).

To avoid this we need 2 things:

 - the load must come after the tas-pending (obviously, otherwise it
   can trivially observe prior state).

 - the tas-pending must be a full word RmW instruction, it cannot be an XCHGB for
   example, such that we cannot observe other state prior to setting
   pending.

On x86 we can realize this by using "LOCK BTS m32, r32" for
tas-pending followed by a regular load.

Note that observing later state is not a problem:

 - if we fail to observe a later unlock, we'll simply spin-wait for
   that store to become visible.

 - if we observe a later xchg_tail(), there is no difference from that
   xchg_tail() having taken place before the tas-pending.

Suggested-by: Will Deacon <will.deacon@arm.com>
Reported-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Will Deacon <will.deacon@arm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: andrea.parri@amarulasolutions.com
Cc: longman@redhat.com
Fixes: 59fb586b4a07 ("locking/qspinlock: Remove unbounded cmpxchg() loop from locking slowpath")
Link: https://lkml.kernel.org/r/20181003130957.183726335@infradead.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agox86/asm: 'Simplify' GEN_*_RMWcc() macros
Peter Zijlstra [Wed, 3 Oct 2018 10:34:10 +0000 (12:34 +0200)]
x86/asm: 'Simplify' GEN_*_RMWcc() macros

Currently the GEN_*_RMWcc() macros include a return statement, which
pretty much mandates we directly wrap them in a (inline) function.

Macros with return statements are tricky and, as per the above, limit
use, so remove the return statement and make them
statement-expressions. This allows them to be used more widely.

Also, shuffle the arguments a bit. Place the @cc argument as 3rd, this
makes it consistent between UNARY and BINARY, but more importantly, it
makes the @arg0 argument last.

Since the @arg0 argument is now last, we can do CPP trickery and make
it an optional argument, simplifying the users; 17 out of 18
occurences do not need this argument.

Finally, change to asm symbolic names, instead of the numeric ordering
of operands, which allows us to get rid of __BINARY_RMWcc_ARG and get
cleaner code overall.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: JBeulich@suse.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: bp@alien8.de
Cc: hpa@linux.intel.com
Link: https://lkml.kernel.org/r/20181003130957.108960094@infradead.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agolocking/qspinlock: Rework some comments
Peter Zijlstra [Wed, 26 Sep 2018 11:01:19 +0000 (13:01 +0200)]
locking/qspinlock: Rework some comments

While working my way through the code again; I felt the comments could
use help.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Will Deacon <will.deacon@arm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: andrea.parri@amarulasolutions.com
Cc: longman@redhat.com
Link: https://lkml.kernel.org/r/20181003130257.156322446@infradead.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agolocking/qspinlock: Re-order code
Peter Zijlstra [Wed, 26 Sep 2018 11:01:18 +0000 (13:01 +0200)]
locking/qspinlock: Re-order code

Flip the branch condition after atomic_fetch_or_acquire(_Q_PENDING_VAL)
such that we loose the indent. This also result in a more natural code
flow IMO.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Will Deacon <will.deacon@arm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: andrea.parri@amarulasolutions.com
Cc: longman@redhat.com
Link: https://lkml.kernel.org/r/20181003130257.156322446@infradead.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agoMerge branch 'x86/build' into locking/core, to pick up dependent patches and unify...
Ingo Molnar [Tue, 16 Oct 2018 15:30:11 +0000 (17:30 +0200)]
Merge branch 'x86/build' into locking/core, to pick up dependent patches and unify jump-label work

Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agolocking/lockdep: Remove duplicated 'lock_class_ops' percpu array
Waiman Long [Fri, 12 Oct 2018 21:42:27 +0000 (17:42 -0400)]
locking/lockdep: Remove duplicated 'lock_class_ops' percpu array

Remove the duplicated 'lock_class_ops' percpu array that is not used
anywhere.

Signed-off-by: Waiman Long <longman@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Fixes: 8ca2b56cd7da ("locking/lockdep: Make class->ops a percpu counter and move it under CONFIG_DEBUG_LOCKDEP=y")
Link: http://lkml.kernel.org/r/1539380547-16726-1-git-send-email-longman@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agox86/defconfig: Enable CONFIG_USB_XHCI_HCD=y
Adam Borowski [Tue, 9 Oct 2018 06:28:03 +0000 (08:28 +0200)]
x86/defconfig: Enable CONFIG_USB_XHCI_HCD=y

A spanking new machine I just got has all but one USB ports wired as 3.0.
Booting defconfig resulted in no keyboard or mouse, which was pretty
uncool.  Let's enable that -- USB3 is ubiquitous rather than an oddity.
As 'y' not 'm' -- recovering from initrd problems needs a keyboard.

Also add it to the 32-bit defconfig.

Signed-off-by: Adam Borowski <kilobyte@angband.pl>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-usb@vger.kernel.org
Link: http://lkml.kernel.org/r/20181009062803.4332-1-kilobyte@angband.pl
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agofutex: Replace spin_is_locked() with lockdep
Lance Roy [Wed, 3 Oct 2018 05:38:57 +0000 (22:38 -0700)]
futex: Replace spin_is_locked() with lockdep

lockdep_assert_held() is better suited for checking locking requirements,
since it won't get confused when the lock is held by some other task. This
is also a step towards possibly removing spin_is_locked().

Signed-off-by: Lance Roy <ldr709@gmail.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: "Paul E. McKenney" <paulmck@linux.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Darren Hart <dvhart@infradead.org>
Link: https://lkml.kernel.org/r/20181003053902.6910-12-ldr709@gmail.com
6 years agolocking/lockdep: Make class->ops a percpu counter and move it under CONFIG_DEBUG_LOCK...
Waiman Long [Wed, 3 Oct 2018 17:07:18 +0000 (13:07 -0400)]
locking/lockdep: Make class->ops a percpu counter and move it under CONFIG_DEBUG_LOCKDEP=y

A sizable portion of the CPU cycles spent on the __lock_acquire() is used
up by the atomic increment of the class->ops stat counter. By taking it out
from the lock_class structure and changing it to a per-cpu per-lock-class
counter, we can reduce the amount of cacheline contention on the class
structure when multiple CPUs are trying to acquire locks of the same
class simultaneously.

To limit the increase in memory consumption because of the percpu nature
of that counter, it is now put back under the CONFIG_DEBUG_LOCKDEP
config option. So the memory consumption increase will only occur if
CONFIG_DEBUG_LOCKDEP is defined. The lock_class structure, however,
is reduced in size by 16 bytes on 64-bit archs after ops removal and
a minor restructuring of the fields.

This patch also fixes a bug in the increment code as the counter is of
the 'unsigned long' type, but atomic_inc() was used to increment it.

Signed-off-by: Waiman Long <longman@redhat.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Link: http://lkml.kernel.org/r/d66681f3-8781-9793-1dcf-2436a284550b@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agox86/jump-labels: Macrofy inline assembly code to work around GCC inlining bugs
Nadav Amit [Fri, 5 Oct 2018 20:27:18 +0000 (13:27 -0700)]
x86/jump-labels: Macrofy inline assembly code to work around GCC inlining bugs

As described in:

  77b0bf55bc67: ("kbuild/Makefile: Prepare for using macros in inline assembly code to work around asm() related GCC inlining bugs")

GCC's inlining heuristics are broken with common asm() patterns used in
kernel code, resulting in the effective disabling of inlining.

The workaround is to set an assembly macro and call it from the inline
assembly block - which is also a minor cleanup for the jump-label code.

As a result the code size is slightly increased, but inlining decisions
are better:

      text     data     bss      dec     hex  filename
  18163528 10226300 2957312 31347140 1de51c4  ./vmlinux before
  18163608 10227348 2957312 31348268 1de562c  ./vmlinux after (+1128)

And functions such as intel_pstate_adjust_policy_max(),
kvm_cpu_accept_dm_intr(), kvm_register_readl() are inlined.

Tested-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Nadav Amit <namit@vmware.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Kate Stewart <kstewart@linuxfoundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Philippe Ombredanne <pombredanne@nexb.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20181005202718.229565-4-namit@vmware.com
Link: https://lore.kernel.org/lkml/20181003213100.189959-11-namit@vmware.com/T/#u
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agox86/cpufeature: Macrofy inline assembly code to work around GCC inlining bugs
Nadav Amit [Fri, 5 Oct 2018 20:27:17 +0000 (13:27 -0700)]
x86/cpufeature: Macrofy inline assembly code to work around GCC inlining bugs

As described in:

  77b0bf55bc67: ("kbuild/Makefile: Prepare for using macros in inline assembly code to work around asm() related GCC inlining bugs")

GCC's inlining heuristics are broken with common asm() patterns used in
kernel code, resulting in the effective disabling of inlining.

The workaround is to set an assembly macro and call it from the inline
assembly block - which is pretty pointless indirection in the static_cpu_has()
case, but is worth it to improve overall inlining quality.

The patch slightly increases the kernel size:

      text     data     bss      dec     hex  filename
  18162879 10226256 2957312 31346447 1de4f0f  ./vmlinux before
  18163528 10226300 2957312 31347140 1de51c4  ./vmlinux after (+693)

And enables the inlining of function such as free_ldt_pgtables().

Tested-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Nadav Amit <namit@vmware.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.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/20181005202718.229565-3-namit@vmware.com
Link: https://lore.kernel.org/lkml/20181003213100.189959-10-namit@vmware.com/T/#u
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agox86/extable: Macrofy inline assembly code to work around GCC inlining bugs
Nadav Amit [Fri, 5 Oct 2018 20:27:16 +0000 (13:27 -0700)]
x86/extable: Macrofy inline assembly code to work around GCC inlining bugs

As described in:

  77b0bf55bc67: ("kbuild/Makefile: Prepare for using macros in inline assembly code to work around asm() related GCC inlining bugs")

GCC's inlining heuristics are broken with common asm() patterns used in
kernel code, resulting in the effective disabling of inlining.

The workaround is to set an assembly macro and call it from the inline
assembly block - which is also a minor cleanup for the exception table
code.

Text size goes up a bit:

      text     data     bss      dec     hex  filename
  18162555 10226288 2957312 31346155 1de4deb  ./vmlinux before
  18162879 10226256 2957312 31346447 1de4f0f  ./vmlinux after (+292)

But this allows the inlining of functions such as nested_vmx_exit_reflected(),
set_segment_reg(), __copy_xstate_to_user() which is a net benefit.

Tested-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Nadav Amit <namit@vmware.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Josh Poimboeuf <jpoimboe@redhat.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/20181005202718.229565-2-namit@vmware.com
Link: https://lore.kernel.org/lkml/20181003213100.189959-9-namit@vmware.com/T/#u
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agoMerge branch 'core/core' into x86/build, to prevent conflicts
Ingo Molnar [Sat, 6 Oct 2018 13:51:56 +0000 (15:51 +0200)]
Merge branch 'core/core' into x86/build, to prevent conflicts

Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agoMerge branch 'x86/core' into x86/build, to avoid conflicts
Ingo Molnar [Fri, 5 Oct 2018 09:27:23 +0000 (11:27 +0200)]
Merge branch 'x86/core' into x86/build, to avoid conflicts

Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agox86/paravirt: Work around GCC inlining bugs when compiling paravirt ops
Nadav Amit [Wed, 3 Oct 2018 21:30:57 +0000 (14:30 -0700)]
x86/paravirt: Work around GCC inlining bugs when compiling paravirt ops

As described in:

  77b0bf55bc67: ("kbuild/Makefile: Prepare for using macros in inline assembly code to work around asm() related GCC inlining bugs")

GCC's inlining heuristics are broken with common asm() patterns used in
kernel code, resulting in the effective disabling of inlining.

The workaround is to set an assembly macro and call it from the inline
assembly block. As a result GCC considers the inline assembly block as
a single instruction. (Which it isn't, but that's the best we can get.)

In this patch we wrap the paravirt call section tricks in a macro,
to hide it from GCC.

The effect of the patch is a more aggressive inlining, which also
causes a size increase of kernel.

      text     data     bss      dec     hex  filename
  18147336 10226688 2957312 31331336 1de1408  ./vmlinux before
  18162555 10226288 2957312 31346155 1de4deb  ./vmlinux after (+14819)

The number of static text symbols (non-inlined functions) goes down:

  Before: 40053
  After:  39942 (-111)

[ mingo: Rewrote the changelog. ]

Tested-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Nadav Amit <namit@vmware.com>
Reviewed-by: Juergen Gross <jgross@suse.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alok Kataria <akataria@vmware.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: virtualization@lists.linux-foundation.org
Link: http://lkml.kernel.org/r/20181003213100.189959-8-namit@vmware.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agox86/bug: Macrofy the BUG table section handling, to work around GCC inlining bugs
Nadav Amit [Wed, 3 Oct 2018 21:30:56 +0000 (14:30 -0700)]
x86/bug: Macrofy the BUG table section handling, to work around GCC inlining bugs

As described in:

  77b0bf55bc67: ("kbuild/Makefile: Prepare for using macros in inline assembly code to work around asm() related GCC inlining bugs")

GCC's inlining heuristics are broken with common asm() patterns used in
kernel code, resulting in the effective disabling of inlining.

The workaround is to set an assembly macro and call it from the inline
assembly block. As a result GCC considers the inline assembly block as
a single instruction. (Which it isn't, but that's the best we can get.)

This patch increases the kernel size:

      text     data     bss      dec     hex  filename
  18146889 10225380 2957312 31329581 1de0d2d  ./vmlinux before
  18147336 10226688 2957312 31331336 1de1408  ./vmlinux after (+1755)

But enables more aggressive inlining (and probably better branch decisions).

The number of static text symbols in vmlinux is much lower:

 Before: 40218
 After:  40053 (-165)

The assembly code gets harder to read due to the extra macro layer.

[ mingo: Rewrote the changelog. ]

Tested-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Nadav Amit <namit@vmware.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Josh Poimboeuf <jpoimboe@redhat.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/20181003213100.189959-7-namit@vmware.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agox86/alternatives: Macrofy lock prefixes to work around GCC inlining bugs
Nadav Amit [Wed, 3 Oct 2018 21:30:55 +0000 (14:30 -0700)]
x86/alternatives: Macrofy lock prefixes to work around GCC inlining bugs

As described in:

  77b0bf55bc67: ("kbuild/Makefile: Prepare for using macros in inline assembly code to work around asm() related GCC inlining bugs")

GCC's inlining heuristics are broken with common asm() patterns used in
kernel code, resulting in the effective disabling of inlining.

The workaround is to set an assembly macro and call it from the inline
assembly block - i.e. to macrify the affected block.

As a result GCC considers the inline assembly block as a single instruction.

This patch handles the LOCK prefix, allowing more aggresive inlining:

      text     data     bss      dec     hex  filename
  18140140 10225284 2957312 31322736 1ddf270  ./vmlinux before
  18146889 10225380 2957312 31329581 1de0d2d  ./vmlinux after (+6845)

This is the reduction in non-inlined functions:

  Before: 40286
  After:  40218 (-68)

Tested-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Nadav Amit <namit@vmware.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Josh Poimboeuf <jpoimboe@redhat.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/20181003213100.189959-6-namit@vmware.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agox86/refcount: Work around GCC inlining bug
Nadav Amit [Wed, 3 Oct 2018 21:30:54 +0000 (14:30 -0700)]
x86/refcount: Work around GCC inlining bug

As described in:

  77b0bf55bc67: ("kbuild/Makefile: Prepare for using macros in inline assembly code to work around asm() related GCC inlining bugs")

GCC's inlining heuristics are broken with common asm() patterns used in
kernel code, resulting in the effective disabling of inlining.

The workaround is to set an assembly macro and call it from the inline
assembly block. As a result GCC considers the inline assembly block as
a single instruction. (Which it isn't, but that's the best we can get.)

This patch allows GCC to inline simple functions such as __get_seccomp_filter().

To no-one's surprise the result is that GCC performs more aggressive (read: correct)
inlining decisions in these senarios, which reduces the kernel size and presumably
also speeds it up:

      text     data     bss      dec     hex  filename
  18140970 10225412 2957312 31323694 1ddf62e  ./vmlinux before
  18140140 10225284 2957312 31322736 1ddf270  ./vmlinux after (-958)

16 fewer static text symbols:

   Before: 40302
    After: 40286 (-16)

these got inlined instead.

Functions such as kref_get(), free_user(), fuse_file_get() now get inlined. Hurray!

[ mingo: Rewrote the changelog. ]

Tested-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Nadav Amit <namit@vmware.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Jan Beulich <JBeulich@suse.com>
Cc: Josh Poimboeuf <jpoimboe@redhat.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/20181003213100.189959-5-namit@vmware.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agox86/objtool: Use asm macros to work around GCC inlining bugs
Nadav Amit [Wed, 3 Oct 2018 21:30:53 +0000 (14:30 -0700)]
x86/objtool: Use asm macros to work around GCC inlining bugs

As described in:

  77b0bf55bc67: ("kbuild/Makefile: Prepare for using macros in inline assembly code to work around asm() related GCC inlining bugs")

GCC's inlining heuristics are broken with common asm() patterns used in
kernel code, resulting in the effective disabling of inlining.

In the case of objtool the resulting borkage can be significant, since all the
annotations of objtool are discarded during linkage and never inlined,
yet GCC bogusly considers most functions affected by objtool annotations
as 'too large'.

The workaround is to set an assembly macro and call it from the inline
assembly block. As a result GCC considers the inline assembly block as
a single instruction. (Which it isn't, but that's the best we can get.)

This increases the kernel size slightly:

      text     data     bss      dec     hex filename
  18140829 10224724 2957312 31322865 1ddf2f1 ./vmlinux before
  18140970 10225412 2957312 31323694 1ddf62e ./vmlinux after (+829)

The number of static text symbols (i.e. non-inlined functions) is reduced:

  Before:  40321
  After:   40302 (-19)

[ mingo: Rewrote the changelog. ]

Tested-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Nadav Amit <namit@vmware.com>
Reviewed-by: Josh Poimboeuf <jpoimboe@redhat.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Christopher Li <sparse@chrisli.org>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-sparse@vger.kernel.org
Link: http://lkml.kernel.org/r/20181003213100.189959-4-namit@vmware.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agokbuild/Makefile: Prepare for using macros in inline assembly code to work around...
Nadav Amit [Wed, 3 Oct 2018 21:30:52 +0000 (14:30 -0700)]
kbuild/Makefile: Prepare for using macros in inline assembly code to work around asm() related GCC inlining bugs

Using macros in inline assembly allows us to work around bugs
in GCC's inlining decisions.

Compile macros.S and use it to assemble all C files.
Currently only x86 will use it.

Background:

The inlining pass of GCC doesn't include an assembler, so it's not aware
of basic properties of the generated code, such as its size in bytes,
or that there are such things as discontiuous blocks of code and data
due to the newfangled linker feature called 'sections' ...

Instead GCC uses a lazy and fragile heuristic: it does a linear count of
certain syntactic and whitespace elements in inlined assembly block source
code, such as a count of new-lines and semicolons (!), as a poor substitute
for "code size and complexity".

Unsurprisingly this heuristic falls over and breaks its neck whith certain
common types of kernel code that use inline assembly, such as the frequent
practice of putting useful information into alternative sections.

As a result of this fresh, 20+ years old GCC bug, GCC's inlining decisions
are effectively disabled for inlined functions that make use of such asm()
blocks, because GCC thinks those sections of code are "large" - when in
reality they are often result in just a very low number of machine
instructions.

This absolute lack of inlining provess when GCC comes across such asm()
blocks both increases generated kernel code size and causes performance
overhead, which is particularly noticeable on paravirt kernels, which make
frequent use of these inlining facilities in attempt to stay out of the
way when running on baremetal hardware.

Instead of fixing the compiler we use a workaround: we set an assembly macro
and call it from the inlined assembly block. As a result GCC considers the
inline assembly block as a single instruction. (Which it often isn't but I digress.)

This uglifies and bloats the source code - for example just the refcount
related changes have this impact:

 Makefile                 |    9 +++++++--
 arch/x86/Makefile        |    7 +++++++
 arch/x86/kernel/macros.S |    7 +++++++
 scripts/Kbuild.include   |    4 +++-
 scripts/mod/Makefile     |    2 ++
 5 files changed, 26 insertions(+), 3 deletions(-)

Yay readability and maintainability, it's not like assembly code is hard to read
and maintain ...

We also hope that GCC will eventually get fixed, but we are not holding
our breath for that. Yet we are optimistic, it might still happen, any decade now.

[ mingo: Wrote new changelog describing the background. ]

Tested-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Nadav Amit <namit@vmware.com>
Acked-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Michal Marek <michal.lkml@markovi.net>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kbuild@vger.kernel.org
Link: http://lkml.kernel.org/r/20181003213100.189959-3-namit@vmware.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agokbuild/arch/xtensa: Define LINKER_SCRIPT for the linker script
Nadav Amit [Wed, 3 Oct 2018 21:30:51 +0000 (14:30 -0700)]
kbuild/arch/xtensa: Define LINKER_SCRIPT for the linker script

Define the LINKER_SCRIPT when building the linker script as being done
in other architectures. This is required, because upcoming Makefile changes
would otherwise break things.

Signed-off-by: Nadav Amit <namit@vmware.com>
Acked-by: Max Filippov <jcmvbkbc@gmail.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Chris Zankel <chris@zankel.net>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Masahiro Yamada <yamada.masahiro@socionext.com>
Cc: Michal Marek <michal.lkml@markovi.net>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-xtensa@linux-xtensa.org
Link: http://lkml.kernel.org/r/20181003213100.189959-2-namit@vmware.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agoMerge branch 'linus' into x86/core, to pick up fixes
Ingo Molnar [Thu, 4 Oct 2018 06:23:03 +0000 (08:23 +0200)]
Merge branch 'linus' into x86/core, to pick up fixes

Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agolocking/lockdep: Add a faster path in __lock_release()
Waiman Long [Tue, 2 Oct 2018 20:19:18 +0000 (16:19 -0400)]
locking/lockdep: Add a faster path in __lock_release()

When __lock_release() is called, the most likely unlock scenario is
on the innermost lock in the chain.  In this case, we can skip some of
the checks and provide a faster path to completion.

Signed-off-by: Waiman Long <longman@redhat.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Link: http://lkml.kernel.org/r/1538511560-10090-4-git-send-email-longman@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agolocking/lockdep: Eliminate redundant IRQs check in __lock_acquire()
Waiman Long [Tue, 2 Oct 2018 20:19:17 +0000 (16:19 -0400)]
locking/lockdep: Eliminate redundant IRQs check in __lock_acquire()

The static __lock_acquire() function has only two callers:

 1) lock_acquire()
 2) reacquire_held_locks()

In lock_acquire(), raw_local_irq_save() is called beforehand. So
IRQs must have been disabled. So the check:

DEBUG_LOCKS_WARN_ON(!irqs_disabled())

is kind of redundant in this case. So move the above check
to reacquire_held_locks() to eliminate redundant code in the
lock_acquire() path.

Signed-off-by: Waiman Long <longman@redhat.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Link: http://lkml.kernel.org/r/1538511560-10090-3-git-send-email-longman@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agolocking/lockdep: Remove add_chain_cache_classes()
Waiman Long [Tue, 2 Oct 2018 20:19:16 +0000 (16:19 -0400)]
locking/lockdep: Remove add_chain_cache_classes()

The inline function add_chain_cache_classes() is defined, but has no
caller. Just remove it.

Signed-off-by: Waiman Long <longman@redhat.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Link: http://lkml.kernel.org/r/1538511560-10090-2-git-send-email-longman@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agoMerge tag 'fbdev-v4.19-rc7' of https://github.com/bzolnier/linux
Greg Kroah-Hartman [Tue, 2 Oct 2018 12:19:43 +0000 (05:19 -0700)]
Merge tag 'fbdev-v4.19-rc7' of https://github.com/bzolnier/linux

Bartlomiej writes:
  "fbdev fixes for v4.19-rc7:

   - fix OMAPFB_MEMORY_READ ioctl to not leak kernel memory in omapfb driver
     (Tomi Valkeinen)

   - add missing prepare/unprepare clock operations in pxa168fb driver
     (Lubomir Rintel)

   - add nobgrt option in efifb driver to disable ACPI BGRT logo restore
     (Hans de Goede)

   - fix spelling mistake in fall-through annotation in stifb driver
     (Gustavo A. R. Silva)

   - fix URL for uvesafb repository in the documentation (Adam Jackson)"

* tag 'fbdev-v4.19-rc7' of https://github.com/bzolnier/linux:
  video/fbdev/stifb: Fix spelling mistake in fall-through annotation
  uvesafb: Fix URLs in the documentation
  efifb: BGRT: Add nobgrt option
  fbdev/omapfb: fix omapfb_memory_read infoleak
  pxa168fb: prepare the clock

6 years agoMerge tag 'mmc-v4.19-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc
Greg Kroah-Hartman [Tue, 2 Oct 2018 12:19:04 +0000 (05:19 -0700)]
Merge tag 'mmc-v4.19-rc4' of git://git./linux/kernel/git/ulfh/mmc

Ulf writes:
  "MMC core:
    - Fixup conversion of debounce time to/from ms/us

   MMC host:
    - sdhi: Fixup whitelisting for Gen3 types"

* tag 'mmc-v4.19-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:
  mmc: slot-gpio: Fix debounce time to use miliseconds again
  mmc: core: Fix debounce time to use microseconds
  mmc: sdhi: sys_dmac: check for all Gen3 types when whitelisting

6 years agoDocumentation/lockstat: Fix trivial typo
Andrew Murray [Mon, 1 Oct 2018 11:01:03 +0000 (12:01 +0100)]
Documentation/lockstat: Fix trivial typo

Fix incorrect line number in example output

Signed-off-by: Andrew Murray <andrew.murray@arm.com>
Cc: Jiri Kosina <trivial@kernel.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Cc: linux-doc@vger.kernel.org
Link: http://lkml.kernel.org/r/1538391663-54524-1-git-send-email-andrew.murray@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agolocking/memory-barriers: Replace smp_cond_acquire() with smp_cond_load_acquire()
Andrea Parri [Wed, 26 Sep 2018 18:29:20 +0000 (11:29 -0700)]
locking/memory-barriers: Replace smp_cond_acquire() with smp_cond_load_acquire()

Amend the changes in commit:

  1f03e8d2919270 ("locking/barriers: Replace smp_cond_acquire() with smp_cond_load_acquire()")

... by updating the documentation accordingly.

Also remove some obsolete information related to the implementation.

Signed-off-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Will Deacon <will.deacon@arm.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Cc: Akira Yokosawa <akiyks@gmail.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Daniel Lustig <dlustig@nvidia.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Jade Alglave <j.alglave@ucl.ac.uk>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Luc Maranget <luc.maranget@inria.fr>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: linux-arch@vger.kernel.org
Cc: parri.andrea@gmail.com
Link: http://lkml.kernel.org/r/20180926182920.27644-5-paulmck@linux.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agotools/memory-model: Add more LKMM limitations
Paul E. McKenney [Wed, 26 Sep 2018 18:29:19 +0000 (11:29 -0700)]
tools/memory-model: Add more LKMM limitations

This commit adds more detail about compiler optimizations and
not-yet-modeled Linux-kernel APIs.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
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: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: akiyks@gmail.com
Cc: boqun.feng@gmail.com
Cc: dhowells@redhat.com
Cc: j.alglave@ucl.ac.uk
Cc: linux-arch@vger.kernel.org
Cc: luc.maranget@inria.fr
Cc: npiggin@gmail.com
Cc: parri.andrea@gmail.com
Cc: stern@rowland.harvard.edu
Cc: will.deacon@arm.com
Link: http://lkml.kernel.org/r/20180926182920.27644-4-paulmck@linux.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agotools/memory-model: Fix a README typo
SeongJae Park [Wed, 26 Sep 2018 18:29:18 +0000 (11:29 -0700)]
tools/memory-model: Fix a README typo

This commit fixes a duplicate-"the" typo in README.

Signed-off-by: SeongJae Park <sj38.park@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
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: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: akiyks@gmail.com
Cc: boqun.feng@gmail.com
Cc: dhowells@redhat.com
Cc: j.alglave@ucl.ac.uk
Cc: linux-arch@vger.kernel.org
Cc: luc.maranget@inria.fr
Cc: npiggin@gmail.com
Cc: parri.andrea@gmail.com
Cc: will.deacon@arm.com
Link: http://lkml.kernel.org/r/20180926182920.27644-3-paulmck@linux.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agotools/memory-model: Add extra ordering for locks and remove it for ordinary release...
Alan Stern [Wed, 26 Sep 2018 18:29:17 +0000 (11:29 -0700)]
tools/memory-model: Add extra ordering for locks and remove it for ordinary release/acquire

More than one kernel developer has expressed the opinion that the LKMM
should enforce ordering of writes by locking.  In other words, given
the following code:

WRITE_ONCE(x, 1);
spin_unlock(&s):
spin_lock(&s);
WRITE_ONCE(y, 1);

the stores to x and y should be propagated in order to all other CPUs,
even though those other CPUs might not access the lock s.  In terms of
the memory model, this means expanding the cumul-fence relation.

Locks should also provide read-read (and read-write) ordering in a
similar way.  Given:

READ_ONCE(x);
spin_unlock(&s);
spin_lock(&s);
READ_ONCE(y); // or WRITE_ONCE(y, 1);

the load of x should be executed before the load of (or store to) y.
The LKMM already provides this ordering, but it provides it even in
the case where the two accesses are separated by a release/acquire
pair of fences rather than unlock/lock.  This would prevent
architectures from using weakly ordered implementations of release and
acquire, which seems like an unnecessary restriction.  The patch
therefore removes the ordering requirement from the LKMM for that
case.

There are several arguments both for and against this change.  Let us
refer to these enhanced ordering properties by saying that the LKMM
would require locks to be RCtso (a bit of a misnomer, but analogous to
RCpc and RCsc) and it would require ordinary acquire/release only to
be RCpc.  (Note: In the following, the phrase "all supported
architectures" is meant not to include RISC-V.  Although RISC-V is
indeed supported by the kernel, the implementation is still somewhat
in a state of flux and therefore statements about it would be
premature.)

Pros:

The kernel already provides RCtso ordering for locks on all
supported architectures, even though this is not stated
explicitly anywhere.  Therefore the LKMM should formalize it.

In theory, guaranteeing RCtso ordering would reduce the need
for additional barrier-like constructs meant to increase the
ordering strength of locks.

Will Deacon and Peter Zijlstra are strongly in favor of
formalizing the RCtso requirement.  Linus Torvalds and Will
would like to go even further, requiring locks to have RCsc
behavior (ordering preceding writes against later reads), but
they recognize that this would incur a noticeable performance
degradation on the POWER architecture.  Linus also points out
that people have made the mistake, in the past, of assuming
that locking has stronger ordering properties than is
currently guaranteed, and this change would reduce the
likelihood of such mistakes.

Not requiring ordinary acquire/release to be any stronger than
RCpc may prove advantageous for future architectures, allowing
them to implement smp_load_acquire() and smp_store_release()
with more efficient machine instructions than would be
possible if the operations had to be RCtso.  Will and Linus
approve this rationale, hypothetical though it is at the
moment (it may end up affecting the RISC-V implementation).
The same argument may or may not apply to RMW-acquire/release;
see also the second Con entry below.

Linus feels that locks should be easy for people to use
without worrying about memory consistency issues, since they
are so pervasive in the kernel, whereas acquire/release is
much more of an "experts only" tool.  Requiring locks to be
RCtso is a step in this direction.

Cons:

Andrea Parri and Luc Maranget think that locks should have the
same ordering properties as ordinary acquire/release (indeed,
Luc points out that the names "acquire" and "release" derive
from the usage of locks).  Andrea points out that having
different ordering properties for different forms of acquires
and releases is not only unnecessary, it would also be
confusing and unmaintainable.

Locks are constructed from lower-level primitives, typically
RMW-acquire (for locking) and ordinary release (for unlock).
It is illogical to require stronger ordering properties from
the high-level operations than from the low-level operations
they comprise.  Thus, this change would make

while (cmpxchg_acquire(&s, 0, 1) != 0)
cpu_relax();

an incorrect implementation of spin_lock(&s) as far as the
LKMM is concerned.  In theory this weakness can be ameliorated
by changing the LKMM even further, requiring
RMW-acquire/release also to be RCtso (which it already is on
all supported architectures).

As far as I know, nobody has singled out any examples of code
in the kernel that actually relies on locks being RCtso.
(People mumble about RCU and the scheduler, but nobody has
pointed to any actual code.  If there are any real cases,
their number is likely quite small.)  If RCtso ordering is not
needed, why require it?

A handful of locking constructs (qspinlocks, qrwlocks, and
mcs_spinlocks) are built on top of smp_cond_load_acquire()
instead of an RMW-acquire instruction.  It currently provides
only the ordinary acquire semantics, not the stronger ordering
this patch would require of locks.  In theory this could be
ameliorated by requiring smp_cond_load_acquire() in
combination with ordinary release also to be RCtso (which is
currently true on all supported architectures).

On future weakly ordered architectures, people may be able to
implement locks in a non-RCtso fashion with significant
performance improvement.  Meeting the RCtso requirement would
necessarily add run-time overhead.

Overall, the technical aspects of these arguments seem relatively
minor, and it appears mostly to boil down to a matter of opinion.
Since the opinions of senior kernel maintainers such as Linus,
Peter, and Will carry more weight than those of Luc and Andrea, this
patch changes the model in accordance with the maintainers' wishes.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Will Deacon <will.deacon@arm.com>
Reviewed-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
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: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: akiyks@gmail.com
Cc: boqun.feng@gmail.com
Cc: dhowells@redhat.com
Cc: j.alglave@ucl.ac.uk
Cc: linux-arch@vger.kernel.org
Cc: luc.maranget@inria.fr
Cc: npiggin@gmail.com
Cc: parri.andrea@gmail.com
Link: http://lkml.kernel.org/r/20180926182920.27644-2-paulmck@linux.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agotools/memory-model: Add litmus-test naming scheme
Paul E. McKenney [Wed, 26 Sep 2018 18:29:16 +0000 (11:29 -0700)]
tools/memory-model: Add litmus-test naming scheme

This commit documents the scheme used to generate the names for the
litmus tests.

[ paulmck: Apply feedback from Andrea Parri and Will Deacon. ]
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Will Deacon <will.deacon@arm.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
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: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: akiyks@gmail.com
Cc: boqun.feng@gmail.com
Cc: dhowells@redhat.com
Cc: j.alglave@ucl.ac.uk
Cc: linux-arch@vger.kernel.org
Cc: luc.maranget@inria.fr
Cc: npiggin@gmail.com
Cc: parri.andrea@gmail.com
Cc: stern@rowland.harvard.edu
Link: http://lkml.kernel.org/r/20180926182920.27644-1-paulmck@linux.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agolocking/spinlocks: Remove an instruction from spin and write locks
Matthew Wilcox [Mon, 20 Aug 2018 14:19:14 +0000 (10:19 -0400)]
locking/spinlocks: Remove an instruction from spin and write locks

Both spin locks and write locks currently do:

 f0 0f b1 17             lock cmpxchg %edx,(%rdi)
 85 c0                   test   %eax,%eax
 75 05                   jne    [slowpath]

This 'test' insn is superfluous; the cmpxchg insn sets the Z flag
appropriately.  Peter pointed out that using atomic_try_cmpxchg_acquire()
will let the compiler know this is true.  Comparing before/after
disassemblies show the only effect is to remove this insn.

Take this opportunity to make the spin & write lock code resemble each
other more closely and have similar likely() hints.

Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Matthew Wilcox <willy@infradead.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Will Deacon <will.deacon@arm.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Waiman Long <longman@redhat.com>
Link: http://lkml.kernel.org/r/20180820162639.GC25153@bombadil.infradead.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
6 years agojump_label: Fix NULL dereference bug in __jump_label_mod_update()
Ard Biesheuvel [Mon, 1 Oct 2018 08:13:24 +0000 (10:13 +0200)]
jump_label: Fix NULL dereference bug in __jump_label_mod_update()

Commit 19483677684b ("jump_label: Annotate entries that operate on
__init code earlier") refactored the code that manages runtime
patching of jump labels in modules that are tied to static keys
defined in other modules or in the core kernel.

In the latter case, we may iterate over the static_key_mod linked
list until we hit the entry for the core kernel, whose 'mod' field
will be NULL, and attempt to dereference it to get at its 'state'
member.

So let's add a non-NULL check: this forces the 'init' argument of
__jump_label_update() to false for static keys that are defined in
the core kernel, which is appropriate given that __init annotated
jump_label entries in the core kernel should no longer be active
at this point (i.e., when loading modules).

Fixes: 19483677684b ("jump_label: Annotate entries that operate on ...")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Kees Cook <keescook@chromium.org>
Cc: Jessica Yu <jeyu@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20181001081324.11553-1-ard.biesheuvel@linaro.org
6 years agos390/vmlinux.lds: Move JUMP_TABLE_DATA into output section
Ard Biesheuvel [Sun, 30 Sep 2018 16:49:50 +0000 (18:49 +0200)]
s390/vmlinux.lds: Move JUMP_TABLE_DATA into output section

Commit e872267b8bcbb179 ("jump_table: move entries into ro_after_init
region") moved the __jump_table input section into the __ro_after_init
output section, but inadvertently put the macro in the wrong place in
the s390 linker script. Let's fix that.

Fixes: e872267b8bcbb179 ("jump_table: move entries into ro_after_init region")
Reported-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Kees Cook <keescook@chromium.org>
Acked-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Tested-by: Guenter Roeck <linux@roeck-us.net>
Cc: linux-s390@vger.kernel.org
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Jessica Yu <jeyu@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20180930164950.3841-1-ard.biesheuvel@linaro.org
6 years agoMerge tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux
Greg Kroah-Hartman [Tue, 2 Oct 2018 00:24:20 +0000 (17:24 -0700)]
Merge tag 'arm64-fixes' of git://git./linux/kernel/git/arm64/linux

Will writes:
  "Late arm64 fixes

   - Fix handling of young contiguous ptes for hugetlb mappings

   - Fix livelock when taking access faults on contiguous hugetlb mappings

   - Tighten up register accesses via KVM SET_ONE_REG ioctl()s"

* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
  arm64: KVM: Sanitize PSTATE.M when being set from userspace
  arm64: KVM: Tighten guest core register access from userspace
  arm64: hugetlb: Avoid unnecessary clearing in huge_ptep_set_access_flags
  arm64: hugetlb: Fix handling of young ptes

6 years agoMerge tag 'armsoc-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc
Greg Kroah-Hartman [Tue, 2 Oct 2018 00:23:27 +0000 (17:23 -0700)]
Merge tag 'armsoc-fixes' of git://git./linux/kernel/git/arm/arm-soc

Olof writes:
  "ARM: SoC fixes

   A handful of fixes that have been coming in the last couple of weeks:

   - Freescale fixes for on-chip accellerators
   - A DT fix for stm32 to avoid fallback to non-DMA SPI mode
   - Fixes for badly specified interrupts on BCM63xx SoCs
   - Allwinner A64 HDMI was incorrectly specified as fully compatble with R40
   - Drive strength fix for SAMA5D2 NAND pins on one board"

* tag 'armsoc-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:
  ARM: dts: stm32: update SPI6 dmas property on stm32mp157c
  soc: fsl: qe: Fix copy/paste bug in ucc_get_tdm_sync_shift()
  soc: fsl: qbman: qman: avoid allocating from non existing gen_pool
  ARM: dts: BCM63xx: Fix incorrect interrupt specifiers
  MAINTAINERS: update the Annapurna Labs maintainer email
  ARM: dts: sun8i: drop A64 HDMI PHY fallback compatible from R40 DT
  ARM: dts: at91: sama5d2_ptc_ek: fix nand pinctrl

6 years agoMerge tag 'pstore-v4.19-rc7' of https://git.kernel.org/pub/scm/linux/kernel/git/kees...
Greg Kroah-Hartman [Tue, 2 Oct 2018 00:22:36 +0000 (17:22 -0700)]
Merge tag 'pstore-v4.19-rc7' of https://git./linux/kernel/git/kees/linux

Kees writes:
  "Pstore fixes for v4.19-rc7

   - Fix failure-path memory leak in ramoops_init (nixiaoming)"

* tag 'pstore-v4.19-rc7' of https://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
  pstore/ram: Fix failure-path memory leak in ramoops_init

6 years agoarm64: KVM: Sanitize PSTATE.M when being set from userspace
Marc Zyngier [Thu, 27 Sep 2018 15:53:22 +0000 (16:53 +0100)]
arm64: KVM: Sanitize PSTATE.M when being set from userspace

Not all execution modes are valid for a guest, and some of them
depend on what the HW actually supports. Let's verify that what
userspace provides is compatible with both the VM settings and
the HW capabilities.

Cc: <stable@vger.kernel.org>
Fixes: 0d854a60b1d7 ("arm64: KVM: enable initialization of a 32bit vcpu")
Reviewed-by: Christoffer Dall <christoffer.dall@arm.com>
Reviewed-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Dave Martin <Dave.Martin@arm.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
6 years agoarm64: KVM: Tighten guest core register access from userspace
Dave Martin [Thu, 27 Sep 2018 15:53:21 +0000 (16:53 +0100)]
arm64: KVM: Tighten guest core register access from userspace

We currently allow userspace to access the core register file
in about any possible way, including straddling multiple
registers and doing unaligned accesses.

This is not the expected use of the ABI, and nobody is actually
using it that way. Let's tighten it by explicitly checking
the size and alignment for each field of the register file.

Cc: <stable@vger.kernel.org>
Fixes: 2f4a07c5f9fe ("arm64: KVM: guest one-reg interface")
Reviewed-by: Christoffer Dall <christoffer.dall@arm.com>
Reviewed-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
[maz: rewrote Dave's initial patch to be more easily backported]
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
6 years agox86/build: Remove unused CONFIG_AS_CRC32
Masahiro Yamada [Mon, 1 Oct 2018 10:24:03 +0000 (19:24 +0900)]
x86/build: Remove unused CONFIG_AS_CRC32

CONFIG_AS_CRC32 is not used anywhere. Its last user was removed by

  0cb6c969ed9d ("net, lib: kill arch_fast_hash library bits")

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Link: https://lkml.kernel.org/r/1538389443-28514-1-git-send-email-yamada.masahiro@socionext.com
6 years agopstore/ram: Fix failure-path memory leak in ramoops_init
Kees Cook [Fri, 28 Sep 2018 22:17:50 +0000 (15:17 -0700)]
pstore/ram: Fix failure-path memory leak in ramoops_init

As reported by nixiaoming, with some minor clarifications:

1) memory leak in ramoops_register_dummy():
   dummy_data = kzalloc(sizeof(*dummy_data), GFP_KERNEL);
   but no kfree() if platform_device_register_data() fails.

2) memory leak in ramoops_init():
   Missing platform_device_unregister(dummy) and kfree(dummy_data)
   if platform_driver_register(&ramoops_driver) fails.

I've clarified the purpose of ramoops_register_dummy(), and added a
common cleanup routine for all three failure paths to call.

Reported-by: nixiaoming <nixiaoming@huawei.com>
Cc: stable@vger.kernel.org
Cc: Anton Vorontsov <anton@enomsg.org>
Cc: Colin Cross <ccross@android.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Geliang Tang <geliangtang@gmail.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
6 years agoLinux 4.19-rc6
Greg Kroah-Hartman [Sun, 30 Sep 2018 14:15:35 +0000 (07:15 -0700)]
Linux 4.19-rc6

6 years agoMerge tag 'auxdisplay-for-greg-v4.19-rc6' of https://github.com/ojeda/linux
Greg Kroah-Hartman [Sun, 30 Sep 2018 13:20:33 +0000 (06:20 -0700)]
Merge tag 'auxdisplay-for-greg-v4.19-rc6' of https://github.com/ojeda/linux

Miguel writes:
  "A trivial fix for auxdisplay

    - MAINTAINERS reference fix for moved file
      Reported by Joe Perches"

* tag 'auxdisplay-for-greg-v4.19-rc6' of https://github.com/ojeda/linux:
  MAINTAINERS: fix reference to moved drivers/{misc => auxdisplay}/panel.c

6 years agoMerge tag 'libnvdimm-fixes2-4.19-rc6' of git://git.kernel.org/pub/scm/linux/kernel...
Greg Kroah-Hartman [Sun, 30 Sep 2018 13:19:38 +0000 (06:19 -0700)]
Merge tag 'libnvdimm-fixes2-4.19-rc6' of git://git./linux/kernel/git/nvdimm/nvdimm

Dan writes:
  "filesystem-dax for 4.19-rc6

   Fix a deadlock in the new for 4.19 dax_lock_mapping_entry() routine."

* tag 'libnvdimm-fixes2-4.19-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm:
  dax: Fix deadlock in dax_lock_mapping_entry()

6 years agoMAINTAINERS: fix reference to moved drivers/{misc => auxdisplay}/panel.c
Miguel Ojeda [Sun, 30 Sep 2018 11:50:05 +0000 (13:50 +0200)]
MAINTAINERS: fix reference to moved drivers/{misc => auxdisplay}/panel.c

Commit 51c1e9b554c9 ("auxdisplay: Move panel.c to drivers/auxdisplay folder")
moved the file, but the MAINTAINERS reference was not updated.

Link: https://lore.kernel.org/lkml/20180928220131.31075-1-joe@perches.com/
Reported-by: Joe Perches <joe@perches.com>
Signed-off-by: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
6 years agoMerge tag 'for-linus-20180929' of git://git.kernel.dk/linux-block
Greg Kroah-Hartman [Sat, 29 Sep 2018 21:52:14 +0000 (14:52 -0700)]
Merge tag 'for-linus-20180929' of git://git.kernel.dk/linux-block

Jens writes:
  "Block fixes for 4.19-rc6

   A set of fixes that should go into this release. This pull request
   contains:

   - A fix (hopefully) for the persistent grants for xen-blkfront. A
     previous fix from this series wasn't complete, hence reverted, and
     this one should hopefully be it. (Boris Ostrovsky)

   - Fix for an elevator drain warning with SMR devices, which is
     triggered when you switch schedulers (Damien)

   - bcache deadlock fix (Guoju Fang)

   - Fix for the block unplug tracepoint, which has had the
     timer/explicit flag reverted since 4.11 (Ilya)

   - Fix a regression in this series where the blk-mq timeout hook is
     invoked with the RCU read lock held, hence preventing it from
     blocking (Keith)

   - NVMe pull from Christoph, with a single multipath fix (Susobhan Dey)"

* tag 'for-linus-20180929' of git://git.kernel.dk/linux-block:
  xen/blkfront: correct purging of persistent grants
  Revert "xen/blkfront: When purging persistent grants, keep them in the buffer"
  blk-mq: I/O and timer unplugs are inverted in blktrace
  bcache: add separate workqueue for journal_write to avoid deadlock
  xen/blkfront: When purging persistent grants, keep them in the buffer
  block: fix deadline elevator drain for zoned block devices
  blk-mq: Allow blocking queue tag iter callbacks
  nvme: properly propagate errors in nvme_mpath_init

6 years agoMerge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Greg Kroah-Hartman [Sat, 29 Sep 2018 21:34:06 +0000 (14:34 -0700)]
Merge branch 'x86-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Thomas writes:
  "A single fix for the AMD memory encryption boot code so it does not
   read random garbage instead of the cached encryption bit when a kexec
   kernel is allocated above the 32bit address limit."

* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/boot: Fix kexec booting failure in the SEV bit detection code

6 years agoMerge branch 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Greg Kroah-Hartman [Sat, 29 Sep 2018 21:32:49 +0000 (14:32 -0700)]
Merge branch 'timers-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Thomas writes:
  "Three small fixes for clocksource drivers:
   - Proper error handling in the Atmel PIT driver
   - Add CLOCK_SOURCE_SUSPEND_NONSTOP for TI SoCs so suspend works again
   - Fix the next event function for Facebook Backpack-CMM BMC chips so
     usleep(100) doesnt sleep several milliseconds"

* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  clocksource/drivers/timer-atmel-pit: Properly handle error cases
  clocksource/drivers/fttmr010: Fix set_next_event handler
  clocksource/drivers/ti-32k: Add CLOCK_SOURCE_SUSPEND_NONSTOP flag for non-am43 SoCs

6 years agoMerge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Greg Kroah-Hartman [Sat, 29 Sep 2018 18:32:03 +0000 (11:32 -0700)]
Merge branch 'perf-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Thomas writes:
  "A single fix for a missing sanity check when a pinned event is tried
  to be read on the wrong CPU due to a legit event scheduling failure."

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf/core: Add sanity check to deal with pinned event failure

6 years agoMerge tag 'pm-4.19-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Greg Kroah-Hartman [Sat, 29 Sep 2018 13:50:36 +0000 (06:50 -0700)]
Merge tag 'pm-4.19-rc6' of git://git./linux/kernel/git/rafael/linux-pm

Rafael writes:
  "Power management fix for 4.19-rc6

   Fix incorrect __init and __exit annotations in the Qualcomm
   Kryo cpufreq driver (Nathan Chancellor)."

* tag 'pm-4.19-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  cpufreq: qcom-kryo: Fix section annotations

6 years agocpufreq: qcom-kryo: Fix section annotations
Nathan Chancellor [Thu, 20 Sep 2018 00:22:21 +0000 (17:22 -0700)]
cpufreq: qcom-kryo: Fix section annotations

There is currently a warning when building the Kryo cpufreq driver into
the kernel image:

WARNING: vmlinux.o(.text+0x8aa424): Section mismatch in reference from
the function qcom_cpufreq_kryo_probe() to the function
.init.text:qcom_cpufreq_kryo_get_msm_id()
The function qcom_cpufreq_kryo_probe() references
the function __init qcom_cpufreq_kryo_get_msm_id().
This is often because qcom_cpufreq_kryo_probe lacks a __init
annotation or the annotation of qcom_cpufreq_kryo_get_msm_id is wrong.

Remove the '__init' annotation from qcom_cpufreq_kryo_get_msm_id
so that there is no more mismatch warning.

Additionally, Nick noticed that the remove function was marked as
'__init' when it should really be marked as '__exit'.

Fixes: 46e2856b8e18 (cpufreq: Add Kryo CPU scaling driver)
Fixes: 5ad7346b4ae2 (cpufreq: kryo: Add module remove and exit)
Reported-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Cc: 4.18+ <stable@vger.kernel.org> # 4.18+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
6 years agoMerge tag 'dma-mapping-4.19-3' of git://git.infradead.org/users/hch/dma-mapping
Greg Kroah-Hartman [Sat, 29 Sep 2018 09:52:24 +0000 (02:52 -0700)]
Merge tag 'dma-mapping-4.19-3' of git://git.infradead.org/users/hch/dma-mapping

Christoph writes:
  "dma mapping fix for 4.19-rc6

   fix a missing Kconfig symbol for commits introduced in 4.19-rc"

* tag 'dma-mapping-4.19-3' of git://git.infradead.org/users/hch/dma-mapping:
  dma-mapping: add the missing ARCH_HAS_SYNC_DMA_FOR_CPU_ALL declaration

6 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input
Greg Kroah-Hartman [Sat, 29 Sep 2018 01:04:50 +0000 (18:04 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/dtor/input

Dmitry writes:
  "Input updates for v4.19-rc5

   Just a few driver fixes"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
  Input: uinput - allow for max == min during input_absinfo validation
  Input: elantech - enable middle button of touchpad on ThinkPad P72
  Input: atakbd - fix Atari CapsLock behaviour
  Input: atakbd - fix Atari keymap
  Input: egalax_ts - add system wakeup support
  Input: gpio-keys - fix a documentation index issue

6 years agoMerge tag 'spi-fix-v4.19-rc5' of https://git.kernel.org/pub/scm/linux/kernel/git...
Greg Kroah-Hartman [Sat, 29 Sep 2018 01:04:06 +0000 (18:04 -0700)]
Merge tag 'spi-fix-v4.19-rc5' of https://git./linux/kernel/git/broonie/spi

Mark writes:
  "spi: Fixes for v4.19

   Quite a few fixes for the Renesas drivers in here, plus a fix for the
   Tegra driver and some documentation fixes for the recently added
   spi-mem code.  The Tegra fix is relatively large but fairly
   straightforward and mechanical, it runs on probe so it's been
   reasonably well covered in -next testing."

* tag 'spi-fix-v4.19-rc5' of https://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:
  spi: spi-mem: Move the DMA-able constraint doc to the kerneldoc header
  spi: spi-mem: Add missing description for data.nbytes field
  spi: rspi: Fix interrupted DMA transfers
  spi: rspi: Fix invalid SPI use during system suspend
  spi: sh-msiof: Fix handling of write value for SISTR register
  spi: sh-msiof: Fix invalid SPI use during system suspend
  spi: gpio: Fix copy-and-paste error
  spi: tegra20-slink: explicitly enable/disable clock

6 years agoMerge tag 'regulator-v4.19-rc5' of https://git.kernel.org/pub/scm/linux/kernel/git...
Greg Kroah-Hartman [Sat, 29 Sep 2018 01:02:25 +0000 (18:02 -0700)]
Merge tag 'regulator-v4.19-rc5' of https://git./linux/kernel/git/broonie/regulator

Mark writes:
  "regulator: Fixes for 4.19

   A collection of fairly minor bug fixes here, a couple of driver
   specific ones plus two core fixes.  There's one fix for the new
   suspend state code which fixes some confusion with constant values
   that are supposed to indicate noop operation and another fixing a
   race condition with the creation of sysfs files on new regulators."

* tag 'regulator-v4.19-rc5' of https://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator:
  regulator: fix crash caused by null driver data
  regulator: Fix 'do-nothing' value for regulators without suspend state
  regulator: da9063: fix DT probing with constraints
  regulator: bd71837: Disable voltage monitoring for LDO3/4

6 years agoMerge tag 'powerpc-4.19-3' of https://git.kernel.org/pub/scm/linux/kernel/git/powerpc...
Greg Kroah-Hartman [Sat, 29 Sep 2018 00:43:32 +0000 (17:43 -0700)]
Merge tag 'powerpc-4.19-3' of https://git./linux/kernel/git/powerpc/linux

Michael writes:
  "powerpc fixes for 4.19 #3

   A reasonably big batch of fixes due to me being away for a few weeks.

   A fix for the TM emulation support on Power9, which could result in
   corrupting the guest r11 when running under KVM.

   Two fixes to the TM code which could lead to userspace GPR corruption
   if we take an SLB miss at exactly the wrong time.

   Our dynamic patching code had a bug that meant we could patch freed
   __init text, which could lead to corrupting userspace memory.

   csum_ipv6_magic() didn't work on little endian platforms since we
   optimised it recently.

   A fix for an endian bug when reading a device tree property telling
   us how many storage keys the machine has available.

   Fix a crash seen on some configurations of PowerVM when migrating the
   partition from one machine to another.

   A fix for a regression in the setup of our CPU to NUMA node mapping
   in KVM guests.

   A fix to our selftest Makefiles to make them work since a recent
   change to the shared Makefile logic."

* tag 'powerpc-4.19-3' of https://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
  selftests/powerpc: Fix Makefiles for headers_install change
  powerpc/numa: Use associativity if VPHN hcall is successful
  powerpc/tm: Avoid possible userspace r1 corruption on reclaim
  powerpc/tm: Fix userspace r13 corruption
  powerpc/pseries: Fix unitialized timer reset on migration
  powerpc/pkeys: Fix reading of ibm, processor-storage-keys property
  powerpc: fix csum_ipv6_magic() on little endian platforms
  powerpc/powernv/ioda2: Reduce upper limit for DMA window size (again)
  powerpc: Avoid code patching freed init sections
  KVM: PPC: Book3S HV: Fix guest r11 corruption with POWER9 TM workarounds

6 years agoMerge tag 'pinctrl-v4.19-4' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw...
Greg Kroah-Hartman [Sat, 29 Sep 2018 00:42:44 +0000 (17:42 -0700)]
Merge tag 'pinctrl-v4.19-4' of git://git./linux/kernel/git/linusw/linux-pinctrl

Linus writes:
  "Pin control fixes for v4.19:
   - Fixes to x86 hardware:
   - AMD interrupt debounce issues
   - Faulty Intel cannonlake register offset
   - Revert pin translation IRQ locking"

* tag 'pinctrl-v4.19-4' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl:
  Revert "pinctrl: intel: Do pin translation when lock IRQ"
  pinctrl: cannonlake: Fix HOSTSW_OWN register offset of H variant
  pinctrl/amd: poll InterruptEnable bits in amd_gpio_irq_set_type

6 years agoperf/core: Add sanity check to deal with pinned event failure
Reinette Chatre [Wed, 19 Sep 2018 17:29:06 +0000 (10:29 -0700)]
perf/core: Add sanity check to deal with pinned event failure

It is possible that a failure can occur during the scheduling of a
pinned event. The initial portion of perf_event_read_local() contains
the various error checks an event should pass before it can be
considered valid. Ensure that the potential scheduling failure
of a pinned event is checked for and have a credible error.

Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: fenghua.yu@intel.com
Cc: tony.luck@intel.com
Cc: acme@kernel.org
Cc: gavin.hindman@intel.com
Cc: jithu.joseph@intel.com
Cc: dave.hansen@intel.com
Cc: hpa@zytor.com
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/6486385d1f30336e9973b24c8c65f5079543d3d3.1537377064.git.reinette.chatre@intel.com
6 years agoMerge tag 'drm-fixes-2018-09-28' of git://anongit.freedesktop.org/drm/drm
Greg Kroah-Hartman [Fri, 28 Sep 2018 16:55:17 +0000 (18:55 +0200)]
Merge tag 'drm-fixes-2018-09-28' of git://anongit.freedesktop.org/drm/drm

Dave writes:
  "drm fixes for 4.19-rc6

   Looks like a pretty normal week for graphics,

   core: syncobj fix, panel link regression revert
   amd: suspend/resume fixes, EDID emulation fix
   mali-dp: NV12 writeback and vblank reset fixes
   etnaviv: DMA setup fix"

* tag 'drm-fixes-2018-09-28' of git://anongit.freedesktop.org/drm/drm:
  drm/amd/display: Fix Edid emulation for linux
  drm/amd/display: Fix Vega10 lightup on S3 resume
  drm/amdgpu: Fix vce work queue was not cancelled when suspend
  Revert "drm/panel: Add device_link from panel device to DRM device"
  drm/syncobj: Don't leak fences when WAIT_FOR_SUBMIT is set
  drm/malidp: Fix writeback in NV12
  drm: mali-dp: Call drm_crtc_vblank_reset on device init
  drm/etnaviv: add DMA configuration for etnaviv platform device

6 years agoMerge tag 'riscv-for-linus-4.19-rc6' of git://git.kernel.org/pub/scm/linux/kernel...
Greg Kroah-Hartman [Fri, 28 Sep 2018 16:53:22 +0000 (18:53 +0200)]
Merge tag 'riscv-for-linus-4.19-rc6' of git://git./linux/kernel/git/palmer/riscv-linux

Palmer writes:
  "A Single RISC-V Update for 4.19-rc6

   The Debian guys have been pushing on our port and found some
   unversioned symbols leaking into modules.  This PR contains a single
   fix for that issue."

* tag 'riscv-for-linus-4.19-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/palmer/riscv-linux:
  RISC-V: include linux/ftrace.h in asm-prototypes.h

6 years agoMerge tag 'pci-v4.19-fixes-2' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git...
Greg Kroah-Hartman [Fri, 28 Sep 2018 16:20:41 +0000 (18:20 +0200)]
Merge tag 'pci-v4.19-fixes-2' of ssh://gitolite./linux/kernel/git/helgaas/pci

Bjorn writes:
  "PCI fixes:

  - Fix ACPI hotplug issue that causes black screen crash at boot (Mika
    Westerberg)

  - Fix DesignWare "scheduling while atomic" issues (Jisheng Zhang)

  - Add PPC contacts to MAINTAINERS for PCI core error handling (Bjorn
    Helgaas)

  - Sort Mobiveil MAINTAINERS entry (Lorenzo Pieralisi)"

* tag 'pci-v4.19-fixes-2' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:
  ACPI / hotplug / PCI: Don't scan for non-hotplug bridges if slot is not bridge
  PCI: dwc: Fix scheduling while atomic issues
  MAINTAINERS: Move mobiveil PCI driver entry where it belongs
  MAINTAINERS: Update PPC contacts for PCI core error handling

6 years agommc: slot-gpio: Fix debounce time to use miliseconds again
Marek Szyprowski [Fri, 28 Sep 2018 12:20:40 +0000 (14:20 +0200)]
mmc: slot-gpio: Fix debounce time to use miliseconds again

The debounce value passed to mmc_gpiod_request_cd() function is in
microseconds, but msecs_to_jiffies() requires the value to be in
miliseconds to properly calculate the delay, so adjust the value stored
in cd_debounce_delay_ms context entry.

Fixes: 1d71926bbd59 ("mmc: core: Fix debounce time to use microseconds")
Fixes: bfd694d5e21c ("mmc: core: Add tunable delay before detecting card
after card is inserted")
Cc: stable@vger.kernel.org # v4.18+
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
6 years agoMerge branch 'nvme-4.19' of git://git.infradead.org/nvme into for-linus
Jens Axboe [Fri, 28 Sep 2018 15:41:40 +0000 (09:41 -0600)]
Merge branch 'nvme-4.19' of git://git.infradead.org/nvme into for-linus

Pull NVMe fix from Christoph.

* 'nvme-4.19' of git://git.infradead.org/nvme:
  nvme: properly propagate errors in nvme_mpath_init

6 years agoxen/blkfront: correct purging of persistent grants
Juergen Gross [Fri, 28 Sep 2018 07:28:27 +0000 (09:28 +0200)]
xen/blkfront: correct purging of persistent grants

Commit a46b53672b2c2e3770b38a4abf90d16364d2584b ("xen/blkfront: cleanup
stale persistent grants") introduced a regression as purged persistent
grants were not pu into the list of free grants again. Correct that.

Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Signed-off-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
6 years agoRevert "xen/blkfront: When purging persistent grants, keep them in the buffer"
Jens Axboe [Fri, 28 Sep 2018 15:40:17 +0000 (09:40 -0600)]
Revert "xen/blkfront: When purging persistent grants, keep them in the buffer"

Fix didn't work for all cases, reverting to add a (hopefully)
better fix.

This reverts commit f151ba989d149bbdfc90e5405724bbea094f9b17.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
6 years agoselftests/powerpc: Fix Makefiles for headers_install change
Michael Ellerman [Fri, 28 Sep 2018 04:53:18 +0000 (14:53 +1000)]
selftests/powerpc: Fix Makefiles for headers_install change

Commit b2d35fa5fc80 ("selftests: add headers_install to lib.mk")
introduced a requirement that Makefiles more than one level below the
selftests directory need to define top_srcdir, but it didn't update
any of the powerpc Makefiles.

This broke building all the powerpc selftests with eg:

  make[1]: Entering directory '/src/linux/tools/testing/selftests/powerpc'
  BUILD_TARGET=/src/linux/tools/testing/selftests/powerpc/alignment; mkdir -p $BUILD_TARGET; make OUTPUT=$BUILD_TARGET -k -C alignment all
  make[2]: Entering directory '/src/linux/tools/testing/selftests/powerpc/alignment'
  ../../lib.mk:20: ../../../../scripts/subarch.include: No such file or directory
  make[2]: *** No rule to make target '../../../../scripts/subarch.include'.
  make[2]: Failed to remake makefile '../../../../scripts/subarch.include'.
  Makefile:38: recipe for target 'alignment' failed

Fix it by setting top_srcdir in the affected Makefiles.

Fixes: b2d35fa5fc80 ("selftests: add headers_install to lib.mk")
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
6 years agoMerge branch 'drm-fixes-4.19' of git://people.freedesktop.org/~agd5f/linux into drm...
Dave Airlie [Thu, 27 Sep 2018 23:30:11 +0000 (09:30 +1000)]
Merge branch 'drm-fixes-4.19' of git://people.freedesktop.org/~agd5f/linux into drm-fixes

Just a few fixes for 4.19:
- Couple of suspend/resume fixes
- Fix EDID emulation with DC

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexdeucher@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180927155418.2813-1-alexander.deucher@amd.com
6 years agoMerge tag 'drm-misc-fixes-2018-09-27-1' of git://anongit.freedesktop.org/drm/drm...
Dave Airlie [Thu, 27 Sep 2018 23:25:26 +0000 (09:25 +1000)]
Merge tag 'drm-misc-fixes-2018-09-27-1' of git://anongit.freedesktop.org/drm/drm-misc into drm-fixes

- Revert adding device-link to panels
- Don't leak fences in drm/syncobj

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Sean Paul <sean@poorly.run>
Link: https://patchwork.freedesktop.org/patch/msgid/20180927152712.GA53076@art_vandelay
6 years agoMerge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma
Greg Kroah-Hartman [Thu, 27 Sep 2018 19:53:55 +0000 (21:53 +0200)]
Merge tag 'for-linus' of git://git./linux/kernel/git/rdma/rdma

Jason writes:
  "Second RDMA rc pull request

   - Fix a long standing race bug when destroying comp_event file descriptors

   - srp, hfi1, bnxt_re: Various driver crashes from missing validation
     and other cases

   - Fixes for regressions in patches merged this window in the gid
     cache, devx, ucma and uapi."

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
  RDMA/core: Set right entry state before releasing reference
  IB/mlx5: Destroy the DEVX object upon error flow
  IB/uverbs: Free uapi on destroy
  RDMA/bnxt_re: Fix system crash during RDMA resource initialization
  IB/hfi1: Fix destroy_qp hang after a link down
  IB/hfi1: Fix context recovery when PBC has an UnsupportedVL
  IB/hfi1: Invalid user input can result in crash
  IB/hfi1: Fix SL array bounds check
  RDMA/uverbs: Fix validity check for modify QP
  IB/srp: Avoid that sg_reset -d ${srp_device} triggers an infinite loop
  ucma: fix a use-after-free in ucma_resolve_ip()
  RDMA/uverbs: Atomically flush and mark closed the comp event queue
  cxgb4: fix abort_req_rss6 struct

6 years agoMerge tag 'for_v4.19-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs
Greg Kroah-Hartman [Thu, 27 Sep 2018 19:16:24 +0000 (21:16 +0200)]
Merge tag 'for_v4.19-rc6' of git://git./linux/kernel/git/jack/linux-fs

Jan writes:
  "an ext2 patch fixing fsync(2) for DAX mounts."

* tag 'for_v4.19-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs:
  ext2, dax: set ext2_dax_aops for dax files

6 years agoblk-mq: I/O and timer unplugs are inverted in blktrace
Ilya Dryomov [Wed, 26 Sep 2018 12:35:50 +0000 (14:35 +0200)]
blk-mq: I/O and timer unplugs are inverted in blktrace

trace_block_unplug() takes true for explicit unplugs and false for
implicit unplugs.  schedule() unplugs are implicit and should be
reported as timer unplugs.  While correct in the legacy code, this has
been inverted in blk-mq since 4.11.

Cc: stable@vger.kernel.org
Fixes: bd166ef183c2 ("blk-mq-sched: add framework for MQ capable IO schedulers")
Reviewed-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
6 years agodax: Fix deadlock in dax_lock_mapping_entry()
Jan Kara [Thu, 27 Sep 2018 11:23:32 +0000 (13:23 +0200)]
dax: Fix deadlock in dax_lock_mapping_entry()

When dax_lock_mapping_entry() has to sleep to obtain entry lock, it will
fail to unlock mapping->i_pages spinlock and thus immediately deadlock
against itself when retrying to grab the entry lock again. Fix the
problem by unlocking mapping->i_pages before retrying.

Fixes: c2a7d2a11552 ("filesystem-dax: Introduce dax_lock_mapping_entry()")
Reported-by: Barret Rhoden <brho@google.com>
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
6 years agox86/boot: Fix kexec booting failure in the SEV bit detection code
Kairui Song [Thu, 27 Sep 2018 12:38:45 +0000 (20:38 +0800)]
x86/boot: Fix kexec booting failure in the SEV bit detection code

Commit

  1958b5fc4010 ("x86/boot: Add early boot support when running with SEV active")

can occasionally cause system resets when kexec-ing a second kernel even
if SEV is not active.

That's because get_sev_encryption_bit() uses 32-bit rIP-relative
addressing to read the value of enc_bit - a variable which caches a
previously detected encryption bit position - but kexec may allocate
the early boot code to a higher location, beyond the 32-bit addressing
limit.

In this case, garbage will be read and get_sev_encryption_bit() will
return the wrong value, leading to accessing memory with the wrong
encryption setting.

Therefore, remove enc_bit, and thus get rid of the need to do 32-bit
rIP-relative addressing in the first place.

 [ bp: massage commit message heavily. ]

Fixes: 1958b5fc4010 ("x86/boot: Add early boot support when running with SEV active")
Suggested-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Kairui Song <kasong@redhat.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Tom Lendacky <thomas.lendacky@amd.com>
Cc: linux-kernel@vger.kernel.org
Cc: tglx@linutronix.de
Cc: mingo@redhat.com
Cc: hpa@zytor.com
Cc: brijesh.singh@amd.com
Cc: kexec@lists.infradead.org
Cc: dyoung@redhat.com
Cc: bhe@redhat.com
Cc: ghook@redhat.com
Link: https://lkml.kernel.org/r/20180927123845.32052-1-kasong@redhat.com
6 years agos390/jump_label: Switch to relative references
Heiko Carstens [Wed, 19 Sep 2018 06:51:44 +0000 (23:51 -0700)]
s390/jump_label: Switch to relative references

Enable support for relative references in jump_label entries.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-s390@vger.kernel.org
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Kees Cook <keescook@chromium.org>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Jessica Yu <jeyu@kernel.org>
Link: https://lkml.kernel.org/r/20180919065144.25010-10-ard.biesheuvel@linaro.org
6 years agojump_table: Move entries into ro_after_init region
Ard Biesheuvel [Wed, 19 Sep 2018 06:51:43 +0000 (23:51 -0700)]
jump_table: Move entries into ro_after_init region

The __jump_table sections emitted into the core kernel and into
each module consist of statically initialized references into
other parts of the code, and with the exception of entries that
point into init code, which are defused at post-init time, these
data structures are never modified.

So let's move them into the ro_after_init section, to prevent them
from being corrupted inadvertently by buggy code, or deliberately
by an attacker.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Kees Cook <keescook@chromium.org>
Acked-by: Jessica Yu <jeyu@kernel.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-s390@vger.kernel.org
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Link: https://lkml.kernel.org/r/20180919065144.25010-9-ard.biesheuvel@linaro.org
6 years agojump_label: Annotate entries that operate on __init code earlier
Ard Biesheuvel [Wed, 19 Sep 2018 06:51:42 +0000 (23:51 -0700)]
jump_label: Annotate entries that operate on __init code earlier

Jump table entries are mostly read-only, with the exception of the
init and module loader code that defuses entries that point into init
code when the code being referred to is freed.

For robustness, it would be better to move these entries into the
ro_after_init section, but clearing the 'code' member of each jump
table entry referring to init code at module load time races with the
module_enable_ro() call that remaps the ro_after_init section read
only, so we'd like to do it earlier.

So given that whether such an entry refers to init code can be decided
much earlier, we can pull this check forward. Since we may still need
the code entry at this point, let's switch to setting a low bit in the
'key' member just like we do to annotate the default state of a jump
table entry.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Kees Cook <keescook@chromium.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-s390@vger.kernel.org
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Jessica Yu <jeyu@kernel.org>
Link: https://lkml.kernel.org/r/20180919065144.25010-8-ard.biesheuvel@linaro.org
6 years agox86/jump_table: Use relative references
Ard Biesheuvel [Wed, 19 Sep 2018 06:51:41 +0000 (23:51 -0700)]
x86/jump_table: Use relative references

Similar to the arm64 case, 64-bit x86 can benefit from using relative
references rather than absolute ones when emitting struct jump_entry
instances. Not only does this reduce the memory footprint of the entries
themselves by 33%, it also removes the need for carrying relocation
metadata on relocatable builds (i.e., for KASLR) which saves a fair
chunk of .init space as well (although the savings are not as dramatic
as on arm64)

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-s390@vger.kernel.org
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Jessica Yu <jeyu@kernel.org>
Link: https://lkml.kernel.org/r/20180919065144.25010-7-ard.biesheuvel@linaro.org
6 years agox86/jump_label: Switch to jump_entry accessors
Ard Biesheuvel [Wed, 19 Sep 2018 06:51:40 +0000 (23:51 -0700)]
x86/jump_label: Switch to jump_entry accessors

In preparation of switching x86 to use place-relative references for
the code, target and key members of struct jump_entry, replace direct
references to the struct members with invocations of the new accessors.
This will allow us to make the switch by modifying the accessors only.

This incorporates a cleanup of __jump_label_transform() proposed by
Peter.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-s390@vger.kernel.org
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Jessica Yu <jeyu@kernel.org>
Link: https://lkml.kernel.org/r/20180919065144.25010-6-ard.biesheuvel@linaro.org
6 years agox86: Add support for 64-bit place relative relocations
Ard Biesheuvel [Wed, 19 Sep 2018 06:51:39 +0000 (23:51 -0700)]
x86: Add support for 64-bit place relative relocations

Add support for R_X86_64_PC64 relocations, which operate on 64-bit
quantities holding a relative symbol reference. Also remove the
definition of R_X86_64_NUM: given that it is currently unused, it
is unclear what the new value should be.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-s390@vger.kernel.org
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Jessica Yu <jeyu@kernel.org>
Link: https://lkml.kernel.org/r/20180919065144.25010-5-ard.biesheuvel@linaro.org
6 years agoarm64/kernel: jump_label: Switch to relative references
Ard Biesheuvel [Wed, 19 Sep 2018 06:51:38 +0000 (23:51 -0700)]
arm64/kernel: jump_label: Switch to relative references

On a randomly chosen distro kernel build for arm64, vmlinux.o shows the
following sections, containing jump label entries, and the associated
RELA relocation records, respectively:

  ...
  [38088] __jump_table      PROGBITS         0000000000000000  00e19f30
       000000000002ea10  0000000000000000  WA       0     0     8
  [38089] .rela__jump_table RELA             0000000000000000  01fd8bb0
       000000000008be30  0000000000000018   I      38178   38088     8
  ...

In other words, we have 190 KB worth of 'struct jump_entry' instances,
and 573 KB worth of RELA entries to relocate each entry's code, target
and key members. This means the RELA section occupies 10% of the .init
segment, and the two sections combined represent 5% of vmlinux's entire
memory footprint.

So let's switch from 64-bit absolute references to 32-bit relative
references for the code and target field, and a 64-bit relative
reference for the 'key' field (which may reside in another module or the
core kernel, which may be more than 4 GB way on arm64 when running with
KASLR enable): this reduces the size of the __jump_table by 33%, and
gets rid of the RELA section entirely.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Will Deacon <will.deacon@arm.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-s390@vger.kernel.org
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Jessica Yu <jeyu@kernel.org>
Link: https://lkml.kernel.org/r/20180919065144.25010-4-ard.biesheuvel@linaro.org
6 years agojump_label: Implement generic support for relative references
Ard Biesheuvel [Wed, 19 Sep 2018 06:51:37 +0000 (23:51 -0700)]
jump_label: Implement generic support for relative references

To reduce the size taken up by absolute references in jump label
entries themselves and the associated relocation records in the
.init segment, add support for emitting them as relative references
instead.

Note that this requires some extra care in the sorting routine, given
that the offsets change when entries are moved around in the jump_entry
table.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-s390@vger.kernel.org
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Jessica Yu <jeyu@kernel.org>
Link: https://lkml.kernel.org/r/20180919065144.25010-3-ard.biesheuvel@linaro.org
6 years agojump_label: Abstract jump_entry member accessors
Ard Biesheuvel [Wed, 19 Sep 2018 06:51:36 +0000 (23:51 -0700)]
jump_label: Abstract jump_entry member accessors

In preparation of allowing architectures to use relative references
in jump_label entries [which can dramatically reduce the memory
footprint], introduce abstractions for references to the 'code' and
'key' members of struct jump_entry.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-s390@vger.kernel.org
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Jessica Yu <jeyu@kernel.org>
Link: https://lkml.kernel.org/r/20180919065144.25010-2-ard.biesheuvel@linaro.org
6 years agobcache: add separate workqueue for journal_write to avoid deadlock
Guoju Fang [Thu, 27 Sep 2018 15:41:46 +0000 (23:41 +0800)]
bcache: add separate workqueue for journal_write to avoid deadlock

After write SSD completed, bcache schedules journal_write work to
system_wq, which is a public workqueue in system, without WQ_MEM_RECLAIM
flag. system_wq is also a bound wq, and there may be no idle kworker on
current processor. Creating a new kworker may unfortunately need to
reclaim memory first, by shrinking cache and slab used by vfs, which
depends on bcache device. That's a deadlock.

This patch create a new workqueue for journal_write with WQ_MEM_RECLAIM
flag. It's rescuer thread will work to avoid the deadlock.

Signed-off-by: Guoju Fang <fangguoju@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: Coly Li <colyli@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
6 years agodrm/amd/display: Fix Edid emulation for linux
Bhawanpreet Lakha [Wed, 26 Sep 2018 17:42:10 +0000 (13:42 -0400)]
drm/amd/display: Fix Edid emulation for linux

[Why]
EDID emulation didn't work properly for linux, as we stop programming
if nothing is connected physically.

[How]
We get a flag from DRM when we want to do edid emulation. We check if
this flag is true and nothing is connected physically, if so we only
program the front end using VIRTUAL_SIGNAL.

Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Acked-by: Leo Li <sunpeng.li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
6 years agodrm/amd/display: Fix Vega10 lightup on S3 resume
Roman Li [Wed, 26 Sep 2018 17:42:16 +0000 (13:42 -0400)]
drm/amd/display: Fix Vega10 lightup on S3 resume

[Why]
There have been a few reports of Vega10 display remaining blank
after S3 resume. The regression is caused by workaround for mode
change on Vega10 - skip set_bandwidth if stream count is 0.
As a result we skipped dispclk reset on suspend, thus on resume
we may skip the clock update assuming it hasn't been changed.
On some systems it causes display blank or 'out of range'.

[How]
Revert "drm/amd/display: Fix Vega10 black screen after mode change"
Verified that it hadn't cause mode change regression.

Signed-off-by: Roman Li <Roman.Li@amd.com>
Reviewed-by: Sun peng Li <Sunpeng.Li@amd.com>
Acked-by: Leo Li <sunpeng.li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
6 years agodrm/amdgpu: Fix vce work queue was not cancelled when suspend
Rex Zhu [Thu, 27 Sep 2018 12:48:39 +0000 (20:48 +0800)]
drm/amdgpu: Fix vce work queue was not cancelled when suspend

The vce cancel_delayed_work_sync never be called.
driver call the function in error path.

This caused the A+A suspend hang when runtime pm enebled.
As we will visit the smu in the idle queue. this will cause
smu hang because the dgpu has been suspend, and the dgpu also
will be waked up. As the smu has been hang, so the dgpu resume
will failed.

Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
6 years agoRevert "drm/panel: Add device_link from panel device to DRM device"
Linus Walleij [Thu, 27 Sep 2018 12:41:30 +0000 (14:41 +0200)]
Revert "drm/panel: Add device_link from panel device to DRM device"

This reverts commit 0c08754b59da5557532d946599854e6df28edc22.

commit 0c08754b59da
("drm/panel: Add device_link from panel device to DRM device")
creates a circular dependency under these circumstances:

1. The panel depends on dsi-host because it is MIPI-DSI child
   device.
2. dsi-host depends on the drm parent device (connector->dev->dev)
   this should be allowed.
3. drm parent dev (connector->dev->dev) depends on the panel
   after this patch.

This makes the dependency circular and while it appears it
does not affect any in-tree drivers (they do not seem to have
dsi hosts depending on the same parent device) this does not
seem right.

As noted in a response from Andrzej Hajda, the intent is
likely to make the panel dependent on the DRM device
(connector->dev) not its parent. But we have no way of
doing that since the DRM device doesn't contain any
struct device on its own (arguably it should).

Revert this until a proper approach is figured out.

Cc: Jyri Sarha <jsarha@ti.com>
Cc: Eric Anholt <eric@anholt.net>
Cc: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20180927124130.9102-1-linus.walleij@linaro.org
6 years agoMerge branch 'clockevents/4.19-fixes' of https://git.linaro.org/people/daniel.lezcano...
Thomas Gleixner [Thu, 27 Sep 2018 14:50:22 +0000 (16:50 +0200)]
Merge branch 'clockevents/4.19-fixes' of https://git.linaro.org/people/daniel.lezcano/linux into timers/urgent

Pull another fix from Daniel Lezcano, which felt through the cracks:

 - Fix a potential memory leak reported by smatch in the atmel timer driver

6 years agoxen/blkfront: When purging persistent grants, keep them in the buffer
Boris Ostrovsky [Sat, 22 Sep 2018 19:55:49 +0000 (15:55 -0400)]
xen/blkfront: When purging persistent grants, keep them in the buffer

Commit a46b53672b2c ("xen/blkfront: cleanup stale persistent grants")
added support for purging persistent grants when they are not in use. As
part of the purge, the grants were removed from the grant buffer, This
eventually causes the buffer to become empty, with BUG_ON triggered in
get_free_grant(). This can be observed even on an idle system, within
20-30 minutes.

We should keep the grants in the buffer when purging, and only free the
grant ref.

Fixes: a46b53672b2c ("xen/blkfront: cleanup stale persistent grants")
Reviewed-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
6 years agoclocksource/drivers/timer-atmel-pit: Properly handle error cases
Alexandre Belloni [Wed, 25 Apr 2018 10:14:39 +0000 (12:14 +0200)]
clocksource/drivers/timer-atmel-pit: Properly handle error cases

The smatch utility reports a possible leak:

smatch warnings:
drivers/clocksource/timer-atmel-pit.c:183 at91sam926x_pit_dt_init() warn: possible memory leak of 'data'

Ensure data is freed before exiting with an error.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Cc: stable@vger.kernel.org
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
6 years agoblock: fix deadline elevator drain for zoned block devices
Damien Le Moal [Thu, 27 Sep 2018 01:55:13 +0000 (10:55 +0900)]
block: fix deadline elevator drain for zoned block devices

When the deadline scheduler is used with a zoned block device, writes
to a zone will be dispatched one at a time. This causes the warning
message:

deadline: forced dispatching is broken (nr_sorted=X), please report this

to be displayed when switching to another elevator with the legacy I/O
path while write requests to a zone are being retained in the scheduler
queue.

Prevent this message from being displayed when executing
elv_drain_elevator() for a zoned block device. __blk_drain_queue() will
loop until all writes are dispatched and completed, resulting in the
desired elevator queue drain without extensive modifications to the
deadline code itself to handle forced-dispatch calls.

Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Fixes: 8dc8146f9c92 ("deadline-iosched: Introduce zone locking support")
Cc: stable@vger.kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
6 years agoMerge branch 'for-upstream/malidp-fixes' of git://linux-arm.org/linux-ld into drm...
Dave Airlie [Thu, 27 Sep 2018 00:49:44 +0000 (10:49 +1000)]
Merge branch 'for-upstream/malidp-fixes' of git://linux-arm.org/linux-ld into drm-fixes

Fix NV12 writeback and fix vblank reset.

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Liviu Dudau <Liviu.Dudau@arm.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180921112354.GR936@e110455-lin.cambridge.arm.com
6 years agoMerge branch 'etnaviv/fixes' of https://git.pengutronix.de/git/lst/linux into drm...
Dave Airlie [Thu, 27 Sep 2018 00:19:26 +0000 (10:19 +1000)]
Merge branch 'etnaviv/fixes' of https://git.pengutronix.de/git/lst/linux into drm-fixes

one fix to get a proper DMA configuration in place for the etnaviv
virtual device. I'm sending this as a fix, as a dma-mapping change at
the ARC architecture side during the 4.19 cycle broke etnaviv on this
platform, which gets remedied with this patch, but it also enables
ARM64.

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Lucas Stach <l.stach@pengutronix.de>
Link: https://patchwork.freedesktop.org/patch/msgid/ea1f712bf09bf9439c6b092bf2c2bde7bb01cf5e.camel@pengutronix.de
6 years agommc: core: Fix debounce time to use microseconds
Tony Lindgren [Tue, 18 Sep 2018 23:16:56 +0000 (16:16 -0700)]
mmc: core: Fix debounce time to use microseconds

The debounce value in device tree is in milliseconds but needs to be in
microseconds for mmc_gpiod_request_cd().

Fixes: bfd694d5e21c ("mmc: core: Add tunable delay before detecting card
after card is inserted")
Cc: Shawn Lin <shawn.lin@rock-chips.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Cc: stable@vger.kernel.org # v4.18+
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
6 years agoACPI / hotplug / PCI: Don't scan for non-hotplug bridges if slot is not bridge
Mika Westerberg [Wed, 26 Sep 2018 20:39:28 +0000 (15:39 -0500)]
ACPI / hotplug / PCI: Don't scan for non-hotplug bridges if slot is not bridge

HP 6730b laptop has an ethernet NIC connected to one of the PCIe root
ports.  The root ports themselves are native PCIe hotplug capable.  Now,
during boot after PCI devices are scanned the BIOS triggers ACPI bus check
directly to the NIC:

  ACPI: \_SB_.PCI0.RP06.NIC_: Bus check in hotplug_event()

It is not clear why it is sending bus check but regardless the ACPI hotplug
notify handler calls enable_slot() directly (instead of going through
acpiphp_check_bridge() as there is no bridge), which ends up handling
special case for non-hotplug bridges with native PCIe hotplug.  This
results a crash of some kind but the reporter only sees black screen so it
is hard to figure out the exact spot and what actually happens.  Based on
a few fix proposals it was tracked to crash somewhere inside
pci_assign_unassigned_bridge_resources().

In any case we should not really be in that special branch at all because
the ACPI notify happened to a slot that is not a PCI bridge (it is just a
regular PCI device).

Fix this so that we only go to that special branch if we are calling
enable_slot() for a bridge (e.g., the ACPI notification was for the
bridge).

Link: https://bugzilla.kernel.org/show_bug.cgi?id=201127
Fixes: 84c8b58ed3ad ("ACPI / hotplug / PCI: Don't scan bridges managed by native hotplug")
Reported-by: Peter Anemone <peter.anemone@gmail.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
CC: stable@vger.kernel.org # v4.18+