The serial8250_do_startup() function unconditionally clears the
interrupts and for that it reads from the RX-FIFO without checking if
there is a byte in the FIFO or not. This works fine on OMAP4+ HW like
AM335x or DRA7.
OMAP3630 ES1.1 (which means probably all OMAP3 and earlier) does not like
this:

|Unhandled fault: external abort on non-linefetch (0x1028) at 0xfb020000
|Internal error: : 1028 [#1] ARM
|Modules linked in:
|CPU: 0 PID: 1 Comm: swapper Not tainted 3.16.0-00022-g7edcb57-dirty #1213
|task: de0572c0 ti: de058000 task.ti: de058000
|PC is at mem32_serial_in+0xc/0x1c
|LR is at serial8250_do_startup+0x220/0x85c
|Flags: nzCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment kernel
|Control: 10c5387d  Table: 80004019  DAC: 00000015
|[<c03051d4>] (mem32_serial_in) from [<c0307fe8>] (serial8250_do_startup+0x220/0x85c)
|[<c0307fe8>] (serial8250_do_startup) from [<c0309e00>] (omap_8250_startup+0x5c/0xe0)
|[<c0309e00>] (omap_8250_startup) from [<c030863c>] (serial8250_startup+0x18/0x2c)
|[<c030863c>] (serial8250_startup) from [<c030394c>] (uart_startup+0x78/0x1d8)
|[<c030394c>] (uart_startup) from [<c0304678>] (uart_open+0xe8/0x114)
|[<c0304678>] (uart_open) from [<c02e9e10>] (tty_open+0x1a8/0x5a4)
Reviewed-by: NTony Lindgren <tony@atomide.com>
Tested-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

While comparing the OMAP-serial and the 8250 part of this I noticed that
the latter does not use run time-pm. Here are the pieces. It is
basically a get before first register access and a last_busy + put after
last access. This has to be enabled from userland _and_ UART_CAP_RPM is
required for this.
The runtime PM can usually work transparently in the background however
there is one exception to this: After serial8250_tx_chars() completes
there still may be unsent bytes in the FIFO (depending on CPU speed vs
baud rate + flow control). Even if the TTY-buffer is empty we do not
want RPM to disable the device because it won't send the remaining
bytes. Instead we leave serial8250_tx_chars() with RPM enabled and wait
for the FIFO empty interrupt. Once we enter serial8250_tx_chars() with
an empty buffer we know that the FIFO is empty and since we are not going
to send anything, we can disable the device.
That xchg() is to ensure that serial8250_tx_chars() can be called
multiple times and only the first invocation will actually invoke the
runtime PM function. So that the last invocation of __stop_tx() will
disable runtime pm.

NOTE: do not enable RPM on the device unless you know what you do! If
the device goes idle, it won't be woken up by incomming RX data _unless_
there is a wakeup irq configured which is usually the RX pin configure
for wakeup via the reset module. The RX activity will then wake up the
device from idle. However the first character is garbage and lost. The
following bytes will be received once the device is up in time. On the
beagle board xm (omap3) it takes approx 13ms from the first wakeup byte
until the first byte that is received properly if the device was in
core-off.

v5…v8:
	- drop RPM from serial8250_set_mctrl() it will be used in
	  restore path which already has RPM active and holds
	  dev->power.lock
v4…v5:
	- add a wrapper around rpm function and introduce UART_CAP_RPM
	  to ensure RPM put is invoked after the TX FIFO is empty.
v3…v4:
	- added runtime to the console code
	- removed device_may_wakeup() from serial8250_set_sleep()

Cc: mika.westerberg@linux.intel.com
Reviewed-by: NTony Lindgren <tony@atomide.com>
Tested-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The OMAP UART provides support for HW assisted flow control. What is
missing is the support to throttle / unthrottle callbacks which are used
by the omap-serial driver at the moment.
This patch adds the callbacks. It should be safe to add them since they
are only invoked from the serial_core (uart_throttle()) if the feature
flags are set.
Reviewed-by: NTony Lindgren <tony@atomide.com>
Tested-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Commit c545b66c,
'tty: Serialize tcflow() with other tty flow control changes' and
commit 99416322,
'tty: Workaround Alpha non-atomic byte storage in tty_struct' work around
compiler bugs and non-atomic storage on multiple arches by padding
bitfields out to the declared type which is unsigned long. However, the
width varies by arch.

Pad bitfields to actual width of unsigned long (which is BITS_PER_LONG).
Reported-by: NFengguang Wu <fengguang.wu@intel.com>
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The Alpha EV4/EV5 cpus can corrupt adjacent byte and short data because
those cpus use RMW to store byte and short data. Thus, concurrent adjacent
byte stores could become corrupted, if serialized by a different lock.
tty_struct uses different locks to protect certain fields within the
structure, and thus is vulnerable to byte stores which are not atomic.

Merge the ->ctrl_status byte and packet mode bit, both protected by the
->ctrl_lock, into an unsigned long.

The padding bits are necessary to force the compiler to allocate the
type specified; otherwise, gcc will ignore the type specifier and
allocate the minimum number of bytes required to store the bitfield.
In turn, this would allow Alpha EV4/EV5 cpus to corrupt adjacent byte
or short storage (because those cpus use RMW to store byte and short data).

gcc versions < 4.7.2 will also corrupt storage adjacent to bitfields
smaller than unsigned long on ia64, ppc64, hppa64, and sparc64, thus
requiring more than unsigned int storage (which would otherwise be
sufficient to fix the Alpha non-atomic storage problem).
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

While transmitting a START/STOP char for tcflow(TCION/TCIOFF), prevent
a termios change. Otherwise, a garbage in-band flow control char
may be sent, if the termios change overlaps the transmission setup.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Relocate the file-scope function, send_prio_char(), as a global
helper tty_send_xchar(). Remove the global declarations for
tty_write_lock()/tty_write_unlock(), as these are file-scope only now.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Use newly-introduced tty->flow_lock to serialize updates to
tty->flow_stopped (via tcflow()) and with concurrent tty flow
control changes from other sources.

Merge the storage for ->stopped and ->flow_stopped, now that both
flags are serialized by ->flow_lock.

The padding bits are necessary to force the compiler to allocate the
type specified; otherwise, gcc will ignore the type specifier and
allocate the minimum number of bytes necessary to store the bitfield.
In turn, this would allow Alpha EV4 and EV5 cpus to corrupt adjacent
byte storage because those cpus use RMW to store byte and short data.

gcc versions < 4.7.2 will also corrupt storage adjacent to bitfields
smaller than unsigned long on ia64, ppc64, hppa64 and sparc64, thus
requiring more than unsigned int storage (which would otherwise be
sufficient to workaround the Alpha non-atomic byte/short storage problem).
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

When a master pty is set to packet mode, flow control changes to
the slave pty cause notifications to the master pty via reads and
polls. However, these tests are occurring for all ttys, not
just ptys.

Implement flow control packet mode notifications in the pty driver.
Only the slave side implements the flow control handlers since
packet mode is asymmetric; the master pty receives notifications
for slave-side changes, but not vice versa.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Without serialization, the flow control state can become inverted
wrt. the actual hardware state. For example,

CPU 0                          | CPU 1
stop_tty()                     |
  lock ctrl_lock               |
  tty->stopped = 1             |
  unlock ctrl_lock             |
                               | start_tty()
                               |   lock ctrl_lock
                               |   tty->stopped = 0
                               |   unlock ctrl_lock
                               |   driver->start()
  driver->stop()               |

In this case, the flow control state now indicates the tty has
been started, but the actual hardware state has actually been stopped.

Introduce tty->flow_lock spinlock to serialize tty flow control changes.
Split out unlocked __start_tty()/__stop_tty() flavors for use by
ioctl(TCXONC) in follow-on patch.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The stopped, hw_stopped, flow_stopped and packet bits are smp-unsafe
and interrupt-unsafe. For example,

CPU 0                         | CPU 1
                              |
tty->flow_stopped = 1         | tty->hw_stopped = 0

One of these updates will be corrupted, as the bitwise operation
on the bitfield is non-atomic.

Ensure each flag has a separate memory location, so concurrent
updates do not corrupt orthogonal states. Because DEC Alpha EV4 and EV5
cpus (from 1995) perform RMW on smaller-than-machine-word storage,
"separate memory location" must be int instead of byte.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

uart_set_termios() is called with interrupts enabled; no need to
save and restore the interrupt state when taking the uart port lock.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Commit 64851636,
serial: bfin-uart: Remove ASYNC_CTS_FLOW flag for hardware automatic CTS,
open-codes uart_handle_cts_change() when CONFIG_SERIAL_BFIN_HARD_CTSRTS
to skip start and stop tx.

But the CTS interrupt handler _still_ calls uart_handle_cts_change();
only call uart_handle_cts_change() if !CONFIG_SERIAL_BFIN_HARD_CTSRTS.

cc: Sonic Zhang <sonic.zhang@analog.com>
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The tty core does not test tty->hw_stopped; remove from drivers
which don't test it themselves.
Acked-by: NJohan Hovold <johan@kernel.org>
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

tty->hw_stopped is not used by the tty core and is thread-unsafe;
hw_stopped is a member of a bitfield whose fields are updated
non-atomically and no lock is suitable for serializing updates.

Replace serial core usage of tty->hw_stopped with uport->hw_stopped.
Use int storage which works around Alpha EV4/5 non-atomic byte storage,
since uart_port uses different locks to protect certain fields within the
structure.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

ISDN4Linux always enables CTS flow control and does not use the
tty_port_cts_enabled() helper function; remove ASYNC_CTS_FLOW
state enable/disable.

cc: Karsten Keil <isdn@linux-pingi.de>
cc: <netdev@vger.kernel.org>
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The serial core uses the tty port flags, ASYNC_CTS_FLOW and
ASYNC_CD_CHECK, to track whether CTS and DCD changes should be
ignored or handled. However, the tty port flags are not safe for
atomic bit operations and no lock provides serialized updates.

Introduce the struct uart_port status field to track CTS and DCD
enable states, and serialize access with uart port lock. Substitute
uart_cts_enabled() helper for tty_port_cts_enabled().
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The serial core provides two helper functions, uart_handle_dcd_change()
and uart_handle_cts_change(), for UART drivers to use at interrupt
time. The serial core expects the UART driver to hold the uart port lock
when calling these helpers to prevent state corruption.

If lockdep enabled, trigger a warning if the uart port lock is not held
when calling these helper functions.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

An interface may need to assert a lock invariant and not flood the
system logs; add a lockdep helper macro equivalent to
lockdep_assert_held() which only WARNs once.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Acked-by: NIngo Molnar <mingo@kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This patch fixes COMPILE_TEST build of serial_mctrl_gpio module for
architectures with custom termios.h header.

sparc64:allmodconfig:

In file included from drivers/tty/serial/serial_mctrl_gpio.c:21:0:
include/uapi/asm-generic/termios.h:22:8: error: redefinition of 'struct termio'
./arch/sparc/include/uapi/asm/termbits.h:16:8: note: originally defined here
make[3]: *** [drivers/tty/serial/serial_mctrl_gpio.o] Error 1
Reported-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NAlexander Shiyan <shc_work@mail.ru>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

We want those fixes in here as well.
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Pull vfs fixes from Al Viro:
 "double iput() on failure exit in lustre, racy removal of spliced
  dentries from ->s_anon in __d_materialise_dentry() plus a bunch of
  assorted RCU pathwalk fixes"

The RCU pathwalk fixes end up fixing a couple of cases where we
incorrectly dropped out of RCU walking, due to incorrect initialization
and testing of the sequence locks in some corner cases.  Since dropping
out of RCU walk mode forces the slow locked accesses, those corner cases
slowed down quite dramatically.

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  be careful with nd->inode in path_init() and follow_dotdot_rcu()
  don't bugger nd->seq on set_root_rcu() from follow_dotdot_rcu()
  fix bogus read_seqretry() checks introduced in b37199e6
  move the call of __d_drop(anon) into __d_materialise_unique(dentry, anon)
  [fix] lustre: d_make_root() does iput() on dentry allocation failure

The performance regression that Josef Bacik reported in the pathname
lookup (see commit 99d263d4 "vfs: fix bad hashing of dentries") made
me look at performance stability of the dcache code, just to verify that
the problem was actually fixed.  That turned up a few other problems in
this area.

There are a few cases where we exit RCU lookup mode and go to the slow
serializing case when we shouldn't, Al has fixed those and they'll come
in with the next VFS pull.

But my performance verification also shows that link_path_walk() turns
out to have a very unfortunate 32-bit store of the length and hash of
the name we look up, followed by a 64-bit read of the combined hash_len
field.  That screws up the processor store to load forwarding, causing
an unnecessary hickup in this critical routine.

It's caused by the ugly calling convention for the "hash_name()"
function, and easily fixed by just making hash_name() fill in the whole
'struct qstr' rather than passing it a pointer to just the hash value.

With that, the profile for this function looks much smoother.
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Pull parisc updates from Helge Deller:
 "The most important patch is a new Light Weigth Syscall (LWS) for 8,
  16, 32 and 64 bit atomic CAS operations which is required in order to
  be able to implement the atomic gcc builtins on our platform.

  Other than that, we wire up the seccomp, getrandom and memfd_create
  syscalls, fixes a minor off-by-one bug and a wrong printk string"

* 'parisc-3.17-1' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux:
  parisc: Implement new LWS CAS supporting 64 bit operations.
  parisc: Wire up seccomp, getrandom and memfd_create syscalls
  parisc: dino: fix %d confusingly prefixed with 0x in format string
  parisc: sys_hpux: NUL terminator is one past the end

in the former we simply check if dentry is still valid after picking
its ->d_inode; in the latter we fetch ->d_inode in the same places
where we fetch dentry and its ->d_seq, under the same checks.

Cc: stable@vger.kernel.org # 2.6.38+
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

return the value instead, and have path_init() do the assignment.  Broken by
"vfs: Fix absolute RCU path walk failures due to uninitialized seq number",
which was Cc-stable with 2.6.38+ as destination.  This one should go where
it went.

To avoid dummy value returned in case when root is already set (it would do
no harm, actually, since the only caller that doesn't ignore the return value
is guaranteed to have nd->root *not* set, but it's more obvious that way),
lift the check into callers.  And do the same to set_root(), to keep them
in sync.

Cc: stable@vger.kernel.org # 2.6.38+
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Pull ntb driver bugfixes from Jon Mason:
 "NTB driver fixes for queue spread and buffer alignment.  Also, update
  to MAINTAINERS to reflect new e-mail address"

* tag 'ntb-3.17' of git://github.com/jonmason/ntb:
  ntb: Add alignment check to meet hardware requirement
  MAINTAINERS: update NTB info
  NTB: correct the spread of queues over mw's

Pull ARM irq chip fixes from Thomas Gleixner:
 "Another pile of ARM specific irq chip fixlets:

   - off by one bugs in the crossbar driver
   - missing annotations
   - a bunch of "make it compile" updates

  I pulled the lot today from Jason, but it has been in -next for at
  least a week"

* 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  irqchip: gic-v3: Declare rdist as __percpu pointer to __iomem pointer
  irqchip: gic: Make gic_default_routable_irq_domain_ops static
  irqchip: exynos-combiner: Fix compilation error on ARM64
  irqchip: crossbar: Off by one bugs in init
  irqchip: gic-v3: Tag all low level accessors __maybe_unused
  irqchip: gic-v3: Only define gic_peek_irq() when building SMP

irqchip fixes for v3.17 from Jason Cooper

 - GIC/GICV3: Various fixlets
 - crossbar: Fix off-by-one bug
 - exynos-combiner: Fix arm64 build error

The NTB translate register must have the value to be BAR size aligned.
This alignment check make sure that the DMA memory allocated has the
proper alignment. Another requirement for NTB to function properly with
memory window BAR size greater or equal to 4M is to use the CMA feature
in 3.16 kernel with the appropriate CONFIG_CMA_ALIGNMENT and
CONFIG_CMA_SIZE_MBYTES set.
Signed-off-by: NDave Jiang <dave.jiang@intel.com>
Signed-off-by: NJon Mason <jdmason@kudzu.us>

Update my contact info to my personal email address and add Dave Jiang.
Signed-off-by: NJon Mason <jon.mason@intel.com>
Signed-off-by: NDave Jiang <dave.jiang@intel.com>

The detection of an uneven number of queues on the given memory windows
was not correct.  The mw_num is zero based and the mod should be
division to spread them evenly over the mw's.
Signed-off-by: NJon Mason <jon.mason@intel.com>

read_seqretry() returns true on mismatch, not on match...

Cc: stable@vger.kernel.org # 3.15+
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

and lock the right list there

Cc: stable@vger.kernel.org
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

double-free is a bad thing
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Merge branches 'locking-urgent-for-linus' and 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull futex and timer fixes from Thomas Gleixner:
 "A oneliner bugfix for the jinxed futex code:

   - Drop hash bucket lock in the error exit path.  I really could slap
     myself for intruducing that bug while fixing all the other horror
     in that code three month ago ...

  and the timer department is not too proud about the following fixes:

   - Deal with a long standing rounding bug in the timeval to jiffies
     conversion.  It's a real issue and this fix fell through the cracks
     for quite some time.

   - Another round of alarmtimer fixes.  Finally this code gets used
     more widely and the subtle issues hidden for quite some time are
     noticed and fixed.  Nothing really exciting, just the itty bitty
     details which bite the serious users here and there"

* 'locking-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  futex: Unlock hb->lock in futex_wait_requeue_pi() error path

* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  alarmtimer: Lock k_itimer during timer callback
  alarmtimer: Do not signal SIGEV_NONE timers
  alarmtimer: Return relative times in timer_gettime
  jiffies: Fix timeval conversion to jiffies

The current LWS cas only works correctly for 32bit. The new LWS allows
for CAS operations of variable size.
Signed-off-by: NGuy Martin <gmsoft@tuxicoman.be>
Cc: <stable@vger.kernel.org> # 3.13+
Signed-off-by: NHelge Deller <deller@gmx.de>

Josef Bacik found a performance regression between 3.2 and 3.10 and
narrowed it down to commit bfcfaa77 ("vfs: use 'unsigned long'
accesses for dcache name comparison and hashing"). He reports:

 "The test case is essentially

      for (i = 0; i < 1000000; i++)
              mkdir("a$i");

  On xfs on a fio card this goes at about 20k dir/sec with 3.2, and 12k
  dir/sec with 3.10.  This is because we spend waaaaay more time in
  __d_lookup on 3.10 than in 3.2.

  The new hashing function for strings is suboptimal for <
  sizeof(unsigned long) string names (and hell even > sizeof(unsigned
  long) string names that I've tested).  I broke out the old hashing
  function and the new one into a userspace helper to get real numbers
  and this is what I'm getting:

      Old hash table had 1000000 entries, 0 dupes, 0 max dupes
      New hash table had 12628 entries, 987372 dupes, 900 max dupes
      We had 11400 buckets with a p50 of 30 dupes, p90 of 240 dupes, p99 of 567 dupes for the new hash

  My test does the hash, and then does the d_hash into a integer pointer
  array the same size as the dentry hash table on my system, and then
  just increments the value at the address we got to see how many
  entries we overlap with.

  As you can see the old hash function ended up with all 1 million
  entries in their own bucket, whereas the new one they are only
  distributed among ~12.5k buckets, which is why we're using so much
  more CPU in __d_lookup".

The reason for this hash regression is two-fold:

 - On 64-bit architectures the down-mixing of the original 64-bit
   word-at-a-time hash into the final 32-bit hash value is very
   simplistic and suboptimal, and just adds the two 32-bit parts
   together.

   In particular, because there is no bit shuffling and the mixing
   boundary is also a byte boundary, similar character patterns in the
   low and high word easily end up just canceling each other out.

 - the old byte-at-a-time hash mixed each byte into the final hash as it
   hashed the path component name, resulting in the low bits of the hash
   generally being a good source of hash data.  That is not true for the
   word-at-a-time case, and the hash data is distributed among all the
   bits.

The fix is the same in both cases: do a better job of mixing the bits up
and using as much of the hash data as possible.  We already have the
"hash_32|64()" functions to do that.
Reported-by: NJosef Bacik <jbacik@fb.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Chris Mason <clm@fb.com>
Cc: linux-fsdevel@vger.kernel.org
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The hash_64() function historically does the multiply by the
GOLDEN_RATIO_PRIME_64 number with explicit shifts and adds, because
unlike the 32-bit case, gcc seems unable to turn the constant multiply
into the more appropriate shift and adds when required.

However, that means that we generate those shifts and adds even when the
architecture has a fast multiplier, and could just do it better in
hardware.

Use the now-cleaned-up CONFIG_ARCH_HAS_FAST_MULTIPLIER (together with
"is it a 64-bit architecture") to decide whether to use an integer
multiply or the explicit sequence of shift/add instructions.
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>