When hardware supports APIC/x2APIC virtualization we don't need to use
pirqs for MSI handling and instead use APIC since most APIC accesses
(MMIO or MSR) will now be processed without VMEXITs.

As an example, netperf on the original code produces this profile
(collected wih 'xentrace -e 0x0008ffff -T 5'):

    342 cpu_change
    260 CPUID
  34638 HLT
  64067 INJ_VIRQ
  28374 INTR
  82733 INTR_WINDOW
     10 NPF
  24337 TRAP
 370610 vlapic_accept_pic_intr
 307528 VMENTRY
 307527 VMEXIT
 140998 VMMCALL
    127 wrap_buffer

After applying this patch the same test shows

    230 cpu_change
    260 CPUID
  36542 HLT
    174 INJ_VIRQ
  27250 INTR
    222 INTR_WINDOW
     20 NPF
  24999 TRAP
 381812 vlapic_accept_pic_intr
 166480 VMENTRY
 166479 VMEXIT
  77208 VMMCALL
     81 wrap_buffer

ApacheBench results (ab -n 10000 -c 200) improve by about 10%
Signed-off-by: NBoris Ostrovsky <boris.ostrovsky@oracle.com>
Reviewed-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Reviewed-by: NAndrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: NDavid Vrabel <david.vrabel@citrix.com>

If the hardware supports APIC virtualization we may decide not to use
pirqs and instead use APIC/x2APIC directly, meaning that we don't want
to set x86_msi.setup_msi_irqs and x86_msi.teardown_msi_irq to
Xen-specific routines.  However, x2APIC is not set up by the time
pci_xen_hvm_init() is called so we need to postpone setting these ops
until later, when we know which APIC mode is used.

(Note that currently x2APIC is never initialized on HVM guests. This
may change in the future)
Signed-off-by: NBoris Ostrovsky <boris.ostrovsky@oracle.com>
Acked-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: NDavid Vrabel <david.vrabel@citrix.com>

When a PF driver unloads, it may find it necessary to leave the VFs
around simply because of pciback having marked them as assigned to a
guest. Utilize a suitable notification to let go of the VFs, thus
allowing the PF to go back into the state it was before its driver
loaded (which in particular allows the driver to be loaded again with
it being able to create the VFs anew, but which also allows to then
pass through the PF instead of the VFs).

Don't do this however for any VFs currently in active use by a guest.
Signed-off-by: NJan Beulich <jbeulich@suse.com>
[v2: Removed the switch statement, moved it about]
[v3: Redid it a bit differently]
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: NDavid Vrabel <david.vrabel@citrix.com>

The commit "xen/pciback: Don't deadlock when unbinding." was using
the version of pci_reset_function which would lock the device lock.
That is no good as we can dead-lock. As such we swapped to using
the lock-less version and requiring that the callers
of 'pcistub_put_pci_dev' take the device lock. And as such
this bug got exposed.

Using the lock-less version is  OK, except that we tried to
use 'pci_restore_state' after the lock-less version of
__pci_reset_function_locked - which won't work as 'state_saved'
is set to false. Said 'state_saved' is a toggle boolean that
is to be used by the sequence of a) pci_save_state/pci_restore_state
or b) pci_load_and_free_saved_state/pci_restore_state. We don't
want to use a) as the guest might have messed up the PCI
configuration space and we want it to revert to the state
when the PCI device was binded to us. Therefore we pick
b) to restore the configuration space.

We restore from our 'golden' version of PCI configuration space, when an:
 - Device is unbinded from pciback
 - Device is detached from a guest.
Reported-by: NSander Eikelenboom <linux@eikelenboom.it>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: NDavid Vrabel <david.vrabel@citrix.com>

We have the pci_load_and_free_saved_state, and pci_store_saved_state
but are missing the functionality to just load the state
multiple times in the PCI device without having to free/save
the state.

This patch makes it possible to use this function.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Acked-by: NBjorn Helgaas <bhelgaas@google.com>
Signed-off-by: NDavid Vrabel <david.vrabel@citrix.com>

A little cleanup. No functional difference.
Reviewed-by: NBoris Ostrovsky <boris.ostrovsky@oracle.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: NDavid Vrabel <david.vrabel@citrix.com>

We had been printing it only if the device was built with
debug enabled. But this information is useful in the field
to troubleshoot.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Reviewed-by: NDavid Vrabel <david.vrabel@citrix.com>
Signed-off-by: NDavid Vrabel <david.vrabel@citrix.com>

Cleanup the function a bit - also include the id of the
domain that is using the device.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Reviewed-by: NDavid Vrabel <david.vrabel@citrix.com>
Signed-off-by: NDavid Vrabel <david.vrabel@citrix.com>

Instead of open-coding it in drivers that want to double check
that their functions are indeed holding the device lock.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Suggested-by: NDavid Vrabel <david.vrabel@citrix.com>
Acked-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: NDavid Vrabel <david.vrabel@citrix.com>

As commit 0a9fd015
'xen/pciback: Document the entry points for 'pcistub_put_pci_dev''
explained there are four entry points in this function.
Two of them are when the user fiddles in the SysFS to
unbind a device which might be in use by a guest or not.

Both 'unbind' states will cause a deadlock as the the PCI lock has
already been taken, which then pci_device_reset tries to take.

We can simplify this by requiring that all callers of
pcistub_put_pci_dev MUST hold the device lock. And then
we can just call the lockless version of pci_device_reset.

To make it even simpler we will modify xen_pcibk_release_pci_dev
to quality whether it should take a lock or not - as it ends
up calling xen_pcibk_release_pci_dev and needs to hold the lock.
Reviewed-by: NBoris Ostrovsky <boris.ostrovsky@oracle.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: NDavid Vrabel <david.vrabel@citrix.com>

Need to pass the pointer within the swiotlb internal buffer to the
swiotlb library, that in the case of xen_unmap_single is dev_addr, not
paddr.
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
CC: stable@vger.kernel.org

In xen_swiotlb_sync_single we always call xen_dma_sync_single_for_cpu,
even when we should call xen_dma_sync_single_for_device. Fix that.
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
CC: stable@vger.kernel.org

On x86 truncation cannot occur because config XEN depends on X86_64 ||
(X86_32 && X86_PAE).

On ARM truncation can occur without CONFIG_ARM_LPAE, when the dma
operation involves foreign grants. However in that case the physical
address returned by xen_bus_to_phys is actually invalid (there is no mfn
to pfn tracking for foreign grants on ARM) and it is not used.
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Reviewed-by: NCatalin Marinas <catalin.marinas@arm.com>
Acked-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
CC: stable@vger.kernel.org

xen_dma_unmap_page and xen_dma_sync_single_for_cpu take a dma_addr_t
handle as argument, not a physical address.
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Reviewed-by: NCatalin Marinas <catalin.marinas@arm.com>
Acked-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
CC: stable@vger.kernel.org

Introduce support for new hypercall GNTTABOP_cache_flush.
Use it to perform cache flashing on pages used for dma when necessary.

If GNTTABOP_cache_flush is supported by the hypervisor, we don't need to
bounce dma map operations that involve foreign grants and non-coherent
devices.
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Reviewed-by: NCatalin Marinas <catalin.marinas@arm.com>
Acked-by: NIan Campbell <ian.campbell@citrix.com>

Introduce an arch specific function to find out whether a particular dma
mapping operation needs to bounce on the swiotlb buffer.

On ARM and ARM64, if the page involved is a foreign page and the device
is not coherent, we need to bounce because at unmap time we cannot
execute any required cache maintenance operations (we don't know how to
find the pfn from the mfn).

No change of behaviour for x86.
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Reviewed-by: NDavid Vrabel <david.vrabel@citrix.com>
Reviewed-by: NCatalin Marinas <catalin.marinas@arm.com>
Acked-by: NIan Campbell <ian.campbell@citrix.com>
Acked-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

Merge xen/mm32.c into xen/mm.c.
As a consequence the code gets compiled on arm64 too.
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Reviewed-by: NCatalin Marinas <catalin.marinas@arm.com>

In xen_dma_map_page, if the page is a local page, call the native
map_page dma_ops. If the page is foreign, call __xen_dma_map_page that
issues any required cache maintenane operations via hypercall.

The reason for doing this is that the native dma_ops map_page could
allocate buffers than need to be freed. If the page is foreign we don't
call the native unmap_page dma_ops function, resulting in a memory leak.
Suggested-by: NCatalin Marinas <catalin.marinas@arm.com>
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Reviewed-by: NCatalin Marinas <catalin.marinas@arm.com>

dev_addr is the machine address of the page.

The new parameter can be used by the ARM and ARM64 implementations of
xen_dma_map_page to find out if the page is a local page (pfn == mfn) or
a foreign page (pfn != mfn).

dev_addr could be retrieved again from the physical address, using
pfn_to_mfn, but it requires accessing an rbtree. Since we already have
the dev_addr in our hands at the call site there is no need to get the
mfn twice.
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Reviewed-by: NCatalin Marinas <catalin.marinas@arm.com>
Acked-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

Use is_device_dma_coherent to check whether we need to issue cache
maintenance operations rather than checking on the existence of a
particular dma_ops function for the device.

This is correct because coherent devices don't need cache maintenance
operations - arm_coherent_dma_ops does not set the hooks that we
were previously checking for existance.
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Reviewed-by: NCatalin Marinas <catalin.marinas@arm.com>
Acked-by: NIan Campbell <ian.campbell@citrix.com>

Introduce a boolean flag and an accessor function to check whether a
device is dma_coherent. Set the flag from set_arch_dma_coherent_ops.
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: NCatalin Marinas <catalin.marinas@arm.com>
Reviewed-by: NCatalin Marinas <catalin.marinas@arm.com>
Acked-by: NRussell King <linux@arm.linux.org.uk>

Introduce a boolean flag and an accessor function to check whether a
device is dma_coherent. Set the flag from set_arch_dma_coherent_ops.
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: NCatalin Marinas <catalin.marinas@arm.com>
Reviewed-by: NCatalin Marinas <catalin.marinas@arm.com>

Remove code duplication in mm32.c by calling the native dma_ops if the
page is a local page (not a foreign page). Use a simple pfn_valid(pfn)
check to figure out if the page is local, exploiting the fact that dom0
is mapped 1:1, therefore pfn_valid always returns false when called on a
foreign mfn.
Suggested-by: NCatalin Marinas <catalin.marinas@arm.com>
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Reviewed-by: NCatalin Marinas <catalin.marinas@arm.com>

Dom0 is not actually capable of issuing outer_inv_range or
outer_clean_range calls.
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: NIan Campbell <ian.campbell@citrix.com>
Reviewed-by: NCatalin Marinas <catalin.marinas@arm.com>

The feature has been removed from Xen. Also Linux cannot use it on ARM32
without CONFIG_ARM_LPAE.
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: NIan Campbell <ian.campbell@citrix.com>
Reviewed-by: NDavid Vrabel <david.vrabel@citrix.com>
Reviewed-by: NCatalin Marinas <catalin.marinas@arm.com>

Pull drm fixes from Dave Airlie:
 "Two i915 regressions and one dual-gpu laptop radeon fix"

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
  drm/radeon: report disconnected for LVDS/eDP with PX if ddc fails
  drm/i915: Cancel vdd off work before suspend
  drm/i915: Ignore SURFLIVE and flip counter when the GPU gets reset

Don Bailey noticed that our page zeroing for compression at end-io time
isn't complete.  This reworks a patch from Linus to push the zeroing
into the zlib and lzo specific functions instead of trying to handle the
corners inside btrfs_decompress_buf2page
Signed-off-by: NChris Mason <clm@fb.com>
Reviewed-by: NJosef Bacik <jbacik@fb.com>
Reported-by: NDon A. Bailey <donb@securitymouse.com>
cc: stable@vger.kernel.org
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Pull irq fixes from Thomas Gleixner:
 "Three fixlets from the ARM SoC camp:
   - correct irqdomain initialization for atmel-aic
   - correct error handling for device tree parsing in bcm controllers"

* 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  irqchip: brcmstb-l2: Fix error handling of irq_of_parse_and_map
  irqchip: bcm7120-l2: Fix error handling of irq_of_parse_and_map
  irqchip: atmel-aic: Fix irqdomain initialization

Pull SCSI fixes from James Bottomley:
 "This is a set of ten fixes: 8 for UFS including four static checker
  warnings, a potential null deref in the voltage regulator code, a race
  on module unload, a ref counting fix on the well known LUNs which made
  it impossible to remove the ufs module and fix to correct the
  information in pwr_info.

  In addition to UFS, there's a blacklist for the Intel Multi-Flex array
  which chokes on report supported operation codes and a fix to an oops
  in bnx2fc caused by shared skbs"

[ For us non-SCSI people: "UFS" here is "Universal Flash Storage" not
  the filesystem.  - Linus ]

* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
  ufs: fix NULL dereference when no regulators are defined
  ufs: ensure clk gating work is finished before module unloading
  scsi: ufs: fix static checker warning in ufshcd_parse_clock_info
  scsi: ufs: fix static checker warning in __ufshcd_setup_clocks
  scsi: ufs: fix static checker warning in ufshcd_populate_vreg
  scsi: ufs: fix static checker errors in ufshcd_system_suspend
  ufs: fix power info after link start-up
  ufs: fix reference counting of W-LUs
  scsi: add Intel Multi-Flex to scsi scan blacklist
  bnx2fc: do not add shared skbs to the fcoe_rx_list

Pull staging/IIO driver fixes from Greg KH:
 "Here are some staging and IIO driver fixes for 3.18-rc7 that resolve a
  number of reported issues, and a new device id for a staging wireless
  driver.

  All of these have been in linux-next"

* tag 'staging-3.18-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging:
  staging: r8188eu: Add new device ID for DLink GO-USB-N150
  staging: r8188eu: Fix scheduling while atomic error introduced in commit fadbe0cd
  iio: accel: bmc150: set low default thresholds
  iio: accel: bmc150: Fix iio_event_spec direction
  iio: accel: bmc150: Send x, y and z motion separately
  iio: accel: bmc150: Error handling when mode set fails
  iio: gyro: bmg160: Fix iio_event_spec direction
  iio: gyro: bmg160: Send x, y and z motion separately
  iio: gyro: bmg160: Don't let interrupt mode to be open drain
  iio: gyro: bmg160: Error handling when mode set fails
  iio: adc: men_z188_adc: Add terminating entry for men_z188_ids
  iio: accel: kxcjk-1013: Fix kxcjk10013_set_range
  iio: Fix IIO_EVENT_CODE_EXTRACT_DIR bit mask

Pull tty/serial fix from Greg KH:
 "Here is a single revert for the of-serial driver that resolves a
  reported issue.

  This revert has been in linux-next for a while"

* tag 'tty-3.18-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty:
  Revert "serial: of-serial: add PM suspend/resume support"

Pull USB fixes from Greg KH:
 "Here are some USB driver fixes and new device ids for 3.18-rc7.

  Full details are in the shortlog, and all of these have been in the
  linux-next tree for a while"

* tag 'usb-3.18-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:
  usb-quirks: Add reset-resume quirk for MS Wireless Laser Mouse 6000
  usb: xhci: rework root port wake bits if controller isn't allowed to wakeup
  USB: xhci: Reset a halted endpoint immediately when we encounter a stall.
  Revert "xhci: clear root port wake on bits if controller isn't wake-up capable"
  USB: xhci: don't start a halted endpoint before its new dequeue is set
  USB: uas: Add no-uas quirk for Hitachi usb-3 enclosures 4971:1012
  USB: ssu100: fix overrun-error reporting
  USB: keyspan: fix overrun-error reporting
  USB: keyspan: fix tty line-status reporting
  usb: serial: ftdi_sio: add PIDs for Matrix Orbital products
  usb: dwc3: ep0: fix for dead code
  USB: serial: cp210x: add IDs for CEL MeshConnect USB Stick

Pull thermal fixes from Eduardo Valentin:
 "In this -rc still very minor changes:

   - Lee Jones fixes compilation warning in sti thermal driver
   - Marjus Elfring removes unnecessary checks in exynos thermal driver
     (as per coccinelle)
   - Now we always update cpufreq policies, and thus get (hopefully)
     always in sync with cpufreq, thanks to Yadwinder"

* 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/evalenti/linux-soc-thermal:
  thermal: Exynos: Deletion of unnecessary checks before two function calls
  thermal: sti: Ignore suspend/resume functions when !PM_SLEEP
  thermal: cpu_cooling: Update always cpufreq policy with thermal constraints

Pull sound fixes from Takashi Iwai:
 "No excitement, here are only minor fixes: an endian fix for the new
  DSD format we added in 3.18, a fix for HP mute LED, and a fix for
  Native Instrument quirk"

* tag 'sound-3.18-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
  ALSA: pcm: Add big-endian DSD sample formats and fix XMOS DSD sample format
  ALSA: hda - One more HP machine needs to change mute led quirk
  ALSA: usb-audio: Use snd_usb_ctl_msg() for Native Instruments quirk

Pull ARM SoC fixes from Arnd Bergmann:
 "Not much interesting going on fixes-wise for us this week, as it
  should be for an -rc7.  I'm not expecting Olof to work much over
  Thanksgiving weekend, so I decided to take over again and push these
  out to you.

  Just four simple fixes this week:

   - one missing of_node_put() on armv7 based mvebu
   - forcing the USB host into the right mode on Chromebook
     (exynos5-snow)
   - enabling two important drivers for exynos_defconfig
   - fixing a noncritical bug for tegra that would cause a regression
     with common code patches queued for 3.19"

* tag 'armsoc-for-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:
  ARM: tegra: irq: fix buggy usage of irq_data irq field
  ARM: exynos_defconfig: Enable max77802 rtc and clock drivers
  ARM: dts: Explicitly set dr_mode on exynos5250-snow
  ARM: mvebu: add missing of_node_put() call in coherency.c

Pull ARM fixes from Russell King:
 "Another round of relatively small ARM fixes.

  Thomas spotted that the strex backoff delay bit was a disable bit, so
  it needed to be clear for this to work.  Vladimir spotted that using a
  restart block for the cache flush operation would return -EINTR, which
  userspace was not expecting.  Dmitry spotted that the auxiliary
  control register accesses for Xscale were not correct"

* 'fixes' of git://ftp.arm.linux.org.uk/~rmk/linux-arm:
  ARM: 8226/1: cacheflush: get rid of restarting block
  ARM: 8222/1: mvebu: enable strex backoff delay
  ARM: 8216/1: xscale: correct auxiliary register in suspend/resume

Two regression fixes from Ville.

* tag 'drm-intel-fixes-2014-11-27' of git://anongit.freedesktop.org/drm-intel:
  drm/i915: Cancel vdd off work before suspend
  drm/i915: Ignore SURFLIVE and flip counter when the GPU gets reset

Pull mips fixes from Ralf Baechle:
 "The hopefully final round of fixes for 3.18:

   - Fix a number of build errors affecting particular configurations.
   - Handle EVA correctly when flushing a signal trampoline and dcache
     lines.
   - Fix printks printing jibberish.
   - Handle 64 bit memory addresses correctly when adding memory chunk
     on 32 bit kernels.
   - Fix a race condition in the hardware tablewalker code"

* 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus:
  MIPS: tlbex: Fix potential HTW race on TLBL/M/S handlers
  MIPS: Fix address type used for early memory detection.
  MIPS: Kconfig: Don't allow both microMIPS and SmartMIPS to be selected.
  MIPS: kernel: cps-vec: Set ISA level to mips32r2 for the MIPS MT ASE
  MIPS: Netlogic: handle modular AHCI builds
  MIPS: Netlogic: handle modular USB case
  MIPS: Loongson: Make platform serial setup always built-in.
  MIPS: fix EVA & non-SMP non-FPU FP context signal handling
  MIPS: cpu-probe: Set the FTLB probability bit on supported cores
  MIPS: BMIPS: Fix ".previous without corresponding .section" warnings
  MIPS: uaccess.h: Fix strnlen_user comment.
  MIPS: r4kcache: Add EVA case for protected_writeback_dcache_line
  MIPS: Fix info about plat_setup in arch_mem_init comment
  MIPS: rtlx: Remove KERN_DEBUG from pr_debug() arguments in rtlx.c
  MIPS: SEAD3: Fix LED device registration.
  MIPS: Fix a copy & paste error in unistd.h

Pull powerpc fixes from Michael Ellerman:
 "Here are five fixes for you to pull please.

  They're all CC'ed to stable except the "Fix PE state format" one which
  went in this release"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mpe/linux:
  powerpc: 32 bit getcpu VDSO function uses 64 bit instructions
  powerpc/powernv: Replace OPAL_DEASSERT_RESET with EEH_RESET_DEACTIVATE
  powerpc/eeh: Fix PE state format
  powerpc/pseries: Fix endiannes issue in RTAS call from xmon
  powerpc/powernv: Fix the hmi event version check.