linux-2.6-microblaze.git
3 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/livepatchin...
Linus Torvalds [Thu, 15 Oct 2020 22:07:57 +0000 (15:07 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/livepatching/livepatching

Pull livepatching update from Jiri Kosina:
 "livepatching kselftest output fix from Miroslav Benes"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/livepatching/livepatching:
  selftests/livepatch: Do not check order when using "comm" for dmesg checking

3 years agoMerge tag 'dio_for_v5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jack...
Linus Torvalds [Thu, 15 Oct 2020 22:03:10 +0000 (15:03 -0700)]
Merge tag 'dio_for_v5.10-rc1' of git://git./linux/kernel/git/jack/linux-fs

Pull direct-io fix from Jan Kara:
 "Fix for unaligned direct IO read past EOF in legacy DIO code"

* tag 'dio_for_v5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs:
  direct-io: defer alignment check until after the EOF check
  direct-io: don't force writeback for reads beyond EOF
  direct-io: clean up error paths of do_blockdev_direct_IO

3 years agoMerge tag 'fs_for_v5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jack...
Linus Torvalds [Thu, 15 Oct 2020 21:56:15 +0000 (14:56 -0700)]
Merge tag 'fs_for_v5.10-rc1' of git://git./linux/kernel/git/jack/linux-fs

Pull UDF, reiserfs, ext2, quota fixes from Jan Kara:

 - a couple of UDF fixes for issues found by syzbot fuzzing

 - a couple of reiserfs fixes for issues found by syzbot fuzzing

 - some minor ext2 cleanups

 - quota patches to support grace times beyond year 2038 for XFS quota
   APIs

* tag 'fs_for_v5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs:
  reiserfs: Fix oops during mount
  udf: Limit sparing table size
  udf: Remove pointless union in udf_inode_info
  udf: Avoid accessing uninitialized data on failed inode read
  quota: clear padding in v2r1_mem2diskdqb()
  reiserfs: Initialize inode keys properly
  udf: Fix memory leak when mounting
  udf: Remove redundant initialization of variable ret
  reiserfs: only call unlock_new_inode() if I_NEW
  ext2: Fix some kernel-doc warnings in balloc.c
  quota: Expand comment describing d_itimer
  quota: widen timestamps for the fs_disk_quota structure
  reiserfs: Fix memory leak in reiserfs_parse_options()
  udf: Use kvzalloc() in udf_sb_alloc_bitmap()
  ext2: remove duplicate include

3 years agoMerge tag 'configfs-5.10' of git://git.infradead.org/users/hch/configfs
Linus Torvalds [Thu, 15 Oct 2020 21:52:45 +0000 (14:52 -0700)]
Merge tag 'configfs-5.10' of git://git.infradead.org/users/hch/configfs

Pull configfs updates from Christoph Hellwig:
 "Various cleanups for the configfs samples (Bartosz Golaszewski)"

* tag 'configfs-5.10' of git://git.infradead.org/users/hch/configfs:
  samples: configfs: prefer pr_err() over bare printk(KERN_ERR
  samples: configfs: don't use spaces before tabs
  samples: configfs: consolidate local variables of the same type
  samples: configfs: don't reinitialize variables which are already zeroed
  samples: configfs: replace simple_strtoul() with kstrtoint()
  samples: configfs: fix alignment in item struct
  samples: configfs: drop unnecessary ternary operators
  samples: configfs: remove redundant newlines
  MAINTAINERS: add the sample directory to the configfs entry

3 years agoMerge tag 'dma-mapping-5.10' of git://git.infradead.org/users/hch/dma-mapping
Linus Torvalds [Thu, 15 Oct 2020 21:43:29 +0000 (14:43 -0700)]
Merge tag 'dma-mapping-5.10' of git://git.infradead.org/users/hch/dma-mapping

Pull dma-mapping updates from Christoph Hellwig:

 - rework the non-coherent DMA allocator

 - move private definitions out of <linux/dma-mapping.h>

 - lower CMA_ALIGNMENT (Paul Cercueil)

 - remove the omap1 dma address translation in favor of the common code

 - make dma-direct aware of multiple dma offset ranges (Jim Quinlan)

 - support per-node DMA CMA areas (Barry Song)

 - increase the default seg boundary limit (Nicolin Chen)

 - misc fixes (Robin Murphy, Thomas Tai, Xu Wang)

 - various cleanups

* tag 'dma-mapping-5.10' of git://git.infradead.org/users/hch/dma-mapping: (63 commits)
  ARM/ixp4xx: add a missing include of dma-map-ops.h
  dma-direct: simplify the DMA_ATTR_NO_KERNEL_MAPPING handling
  dma-direct: factor out a dma_direct_alloc_from_pool helper
  dma-direct check for highmem pages in dma_direct_alloc_pages
  dma-mapping: merge <linux/dma-noncoherent.h> into <linux/dma-map-ops.h>
  dma-mapping: move large parts of <linux/dma-direct.h> to kernel/dma
  dma-mapping: move dma-debug.h to kernel/dma/
  dma-mapping: remove <asm/dma-contiguous.h>
  dma-mapping: merge <linux/dma-contiguous.h> into <linux/dma-map-ops.h>
  dma-contiguous: remove dma_contiguous_set_default
  dma-contiguous: remove dev_set_cma_area
  dma-contiguous: remove dma_declare_contiguous
  dma-mapping: split <linux/dma-mapping.h>
  cma: decrease CMA_ALIGNMENT lower limit to 2
  firewire-ohci: use dma_alloc_pages
  dma-iommu: implement ->alloc_noncoherent
  dma-mapping: add new {alloc,free}_noncoherent dma_map_ops methods
  dma-mapping: add a new dma_alloc_pages API
  dma-mapping: remove dma_cache_sync
  53c700: convert to dma_alloc_noncoherent
  ...

3 years agoMerge tag 'dmaengine-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul...
Linus Torvalds [Thu, 15 Oct 2020 21:33:52 +0000 (14:33 -0700)]
Merge tag 'dmaengine-5.10-rc1' of git://git./linux/kernel/git/vkoul/dmaengine

Pull dmaengine updates from Vinod Koul:
 "Core:

   - Mark dma_request_slave_channel() deprecated in favour of
     dma_request_chan()

   - subsystem conversion for tasklet_setup() API

   - subsystem removal of local dma_parms for arm drivers

  Also updates to bunch of driver notably TI, DW and AXI-DMAC"

* tag 'dmaengine-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine: (104 commits)
  dmaengine: owl-dma: fix kernel-doc style for enum
  dmaengine: zynqmp_dma: fix kernel-doc style for tasklet
  dmaengine: xilinx_dma: fix kernel-doc style for tasklet
  dmaengine: qcom: bam_dma: fix kernel-doc style for tasklet
  dmaengine: altera-msgdma: fix kernel-doc style for tasklet
  dmaengine: xilinx: dpdma: convert tasklets to use new tasklet_setup() API
  dmaengine: sf-pdma: convert tasklets to use new tasklet_setup() API
  dt-bindings: Fix 'reg' size issues in zynqmp examples
  dmaengine: rcar-dmac: drop double zeroing
  dmaengine: sh: drop double zeroing
  dmaengine: ioat: Allocate correct size for descriptor chunk
  dmaengine: ti: k3-udma: use devm_platform_ioremap_resource_byname
  dmaengine: fsl: remove bad channel update
  dmaengine: dma-jz4780: Fix race in jz4780_dma_tx_status
  dmaengine: pl330: fix argument for tasklet
  dmaengine: dmatest: Return boolean result directly in filter()
  dmaengine: dmatest: Check list for emptiness before access its last entry
  dmaengine: ti: k3-udma-glue: fix channel enable functions
  dmaengine: iop-adma: Fix pointer cast warnings
  dmaengine: dw-edma: Fix Using plain integer as NULL pointer in dw-edma-v0-debugfs.c
  ...

3 years agoMerge tag 'sound-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai...
Linus Torvalds [Thu, 15 Oct 2020 18:07:44 +0000 (11:07 -0700)]
Merge tag 'sound-5.10-rc1' of git://git./linux/kernel/git/tiwai/sound

Pull sound updates from Takashi Iwai:
 "The amount of changes is smaller at this round (what a surprise), but
  lots of activity is seen. Most of changes are about ASoC driver
  development, especially Intel platforms. Here are some highlights:

  General:
   - Replace all tasklet usages with other alternatives
   - Cleanup of the ASoC error unwinding code
   - Fixes for trivial issues caught by static checker
   - Spell fixes allover the places

  ALSA Core:
   - Lockdep fix for control devices
   - Fix for potential OSS sequencer mutex stalls

  HD-audio and USB-audio:
   - SoundBlaster AE-7 support
   - Changes in quirk table for the rename handling
   - Quirks for HP and ASUS machines, Pioneer DJ DJM-250MK2.

  ASoC:
   - Lots of updates for Intel SOF and SoundWire enablement
   - Replacement of the DSP driver for some older x86 systems; the new
     code was written from scratch, better maintenance expected
   - Helpers for parsing auxiluary devices from the device tree
   - New support for AllWinner A64, Cirrus Logic CS4234, Mediatek MT6359
     Microchip S/PDIF TX and RX controllers, Realtek RT1015P, and Texas
     Instruments J721E, TAS2110, TAS2564 and TAS2764"

* tag 'sound-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound: (498 commits)
  ALSA: hda/hdmi: fix incorrect locking in hdmi_pcm_close
  ALSA: hda: fix jack detection with Realtek codecs when in D3
  ALSA: fireworks: use semicolons rather than commas to separate statements
  ALSA: hda: use semicolons rather than commas to separate statements
  ALSA: hda/i915 - fix list corruption with concurrent probes
  ASoC: dmaengine: Document support for TX only or RX only streams
  ASoC: mchp-spdiftx: remove 'TX' from playback stream name
  ASoC: ti: davinci-mcasp: Use &pdev->dev for early dev_warn
  ASoC: tas2764: Add the driver for the TAS2764
  dt-bindings: tas2764: Add the TAS2764 binding doc
  ASoC: Intel: catpt: Add explicit DMADEVICES kconfig dependency
  ASoC: Intel: catpt: Fix compilation when CONFIG_MODULES is disabled
  ASoC: stm32: dfsdm: add actual resolution trace
  ASoC: stm32: dfsdm: change rate limits
  ASoC: qcom: sc7180: Add support for audio over DP
  Asoc: qcom: lpass-platform : Increase buffer size
  ASoC: qcom: Add support for lpass hdmi driver
  Asoc: qcom: lpass:Update lpaif_dmactl members order
  Asoc:qcom:lpass-cpu:Update dts property read API
  ASoC: dt-bindings: Add dt binding for lpass hdmi
  ...

3 years agoMerge tag 'drm-next-2020-10-15' of git://anongit.freedesktop.org/drm/drm
Linus Torvalds [Thu, 15 Oct 2020 17:46:16 +0000 (10:46 -0700)]
Merge tag 'drm-next-2020-10-15' of git://anongit.freedesktop.org/drm/drm

Pull drm updates from Dave Airlie:
 "Not a major amount of change, the i915 trees got split into display
  and gt trees to better facilitate higher level review, and there's a
  major refactoring of i915 GEM locking to use more core kernel concepts
  (like ww-mutexes). msm gets per-process pagetables, older AMD SI cards
  get DC support, nouveau got a bump in displayport support with common
  code extraction from i915.

  Outside of drm this contains a couple of patches for hexint
  moduleparams which you've acked, and a virtio common code tree that
  you should also get via it's regular path.

  New driver:
   - Cadence MHDP8546 DisplayPort bridge driver

  core:
   - cross-driver scatterlist cleanups
   - devm_drm conversions
   - remove drm_dev_init
   - devm_drm_dev_alloc conversion

  ttm:
   - lots of refactoring and cleanups

  bridges:
   - chained bridge support in more drivers

  panel:
   - misc new panels

  scheduler:
   - cleanup priority levels

  displayport:
   - refactor i915 code into helpers for nouveau

  i915:
   - split into display and GT trees
   - WW locking refactoring in GEM
   - execbuf2 extension mechanism
   - syncobj timeline support
   - GEN 12 HOBL display powersaving
   - Rocket Lake display additions
   - Disable FBC on Tigerlake
   - Tigerlake Type-C + DP improvements
   - Hotplug interrupt refactoring

  amdgpu:
   - Sienna Cichlid updates
   - Navy Flounder updates
   - DCE6 (SI) support for DC
   - Plane rotation enabled
   - TMZ state info ioctl
   - PCIe DPC recovery support
   - DC interrupt handling refactor
   - OLED panel fixes

  amdkfd:
   - add SMI events for thermal throttling
   - SMI interface events ioctl update
   - process eviction counters

  radeon:
   - move to dma_ for allocations
   - expose sclk via sysfs

  msm:
   - DSI support for sm8150/sm8250
   - per-process GPU pagetable support
   - Displayport support

  mediatek:
   - move HDMI phy driver to PHY
   - convert mtk-dpi to bridge API
   - disable mt2701 tmds

  tegra:
   - bridge support

  exynos:
   - misc cleanups

  vc4:
   - dual display cleanups

  ast:
   - cleanups

  gma500:
   - conversion to GPIOd API

  hisilicon:
   - misc reworks

  ingenic:
   - clock handling and format improvements

  mcde:
   - DSI support

  mgag200:
   - desktop g200 support

  mxsfb:
   - i.MX7 + i.MX8M
   - alpha plane support

  panfrost:
   - devfreq support
   - amlogic SoC support

  ps8640:
   - EDID from eDP retrieval

  tidss:
   - AM65xx YUV workaround

  virtio:
   - virtio-gpu exported resources

  rcar-du:
   - R8A7742, R8A774E1 and R8A77961 support
   - YUV planar format fixes
   - non-visible plane handling
   - VSP device reference count fix
   - Kconfig fix to avoid displaying disabled options in .config"

* tag 'drm-next-2020-10-15' of git://anongit.freedesktop.org/drm/drm: (1494 commits)
  drm/ingenic: Fix bad revert
  drm/amdgpu: Fix invalid number of character '{' in amdgpu_acpi_init
  drm/amdgpu: Remove warning for virtual_display
  drm/amdgpu: kfd_initialized can be static
  drm/amd/pm: setup APU dpm clock table in SMU HW initialization
  drm/amdgpu: prevent spurious warning
  drm/amdgpu/swsmu: fix ARC build errors
  drm/amd/display: Fix OPTC_DATA_FORMAT programming
  drm/amd/display: Don't allow pstate if no support in blank
  drm/panfrost: increase readl_relaxed_poll_timeout values
  MAINTAINERS: Update entry for st7703 driver after the rename
  Revert "gpu/drm: ingenic: Add option to mmap GEM buffers cached"
  drm/amd/display: HDMI remote sink need mode validation for Linux
  drm/amd/display: Change to correct unit on audio rate
  drm/amd/display: Avoid set zero in the requested clk
  drm/amdgpu: align frag_end to covered address space
  drm/amdgpu: fix NULL pointer dereference for Renoir
  drm/vmwgfx: fix regression in thp code due to ttm init refactor.
  drm/amdgpu/swsmu: add interrupt work handler for smu11 parts
  drm/amdgpu/swsmu: add interrupt work function
  ...

3 years agoMerge tag 'char-misc-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregk...
Linus Torvalds [Thu, 15 Oct 2020 17:01:51 +0000 (10:01 -0700)]
Merge tag 'char-misc-5.10-rc1' of git://git./linux/kernel/git/gregkh/char-misc

Pull char/misc driver updates from Greg KH:
 "Here is the big set of char, misc, and other assorted driver subsystem
  patches for 5.10-rc1.

  There's a lot of different things in here, all over the drivers/
  directory. Some summaries:

   - soundwire driver updates

   - habanalabs driver updates

   - extcon driver updates

   - nitro_enclaves new driver

   - fsl-mc driver and core updates

   - mhi core and bus updates

   - nvmem driver updates

   - eeprom driver updates

   - binder driver updates and fixes

   - vbox minor bugfixes

   - fsi driver updates

   - w1 driver updates

   - coresight driver updates

   - interconnect driver updates

   - misc driver updates

   - other minor driver updates

  All of these have been in linux-next for a while with no reported
  issues"

* tag 'char-misc-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc: (396 commits)
  binder: fix UAF when releasing todo list
  docs: w1: w1_therm: Fix broken xref, mistakes, clarify text
  misc: Kconfig: fix a HISI_HIKEY_USB dependency
  LSM: Fix type of id parameter in kernel_post_load_data prototype
  misc: Kconfig: add a new dependency for HISI_HIKEY_USB
  firmware_loader: fix a kernel-doc markup
  w1: w1_therm: make w1_poll_completion static
  binder: simplify the return expression of binder_mmap
  test_firmware: Test partial read support
  firmware: Add request_partial_firmware_into_buf()
  firmware: Store opt_flags in fw_priv
  fs/kernel_file_read: Add "offset" arg for partial reads
  IMA: Add support for file reads without contents
  LSM: Add "contents" flag to kernel_read_file hook
  module: Call security_kernel_post_load_data()
  firmware_loader: Use security_post_load_data()
  LSM: Introduce kernel_post_load_data() hook
  fs/kernel_read_file: Add file_size output argument
  fs/kernel_read_file: Switch buffer size arg to size_t
  fs/kernel_read_file: Remove redundant size argument
  ...

3 years agoMerge tag 'usb-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Linus Torvalds [Thu, 15 Oct 2020 16:51:18 +0000 (09:51 -0700)]
Merge tag 'usb-5.10-rc1' of git://git./linux/kernel/git/gregkh/usb

Pull USB/PHY/Thunderbolt driver updates from Greg KH:
 "Here is the big set of USB, PHY, and Thunderbolt driver updates for
  5.10-rc1.

  Lots of tiny different things for these subsystems are in here,
  including:

   - phy driver updates

   - thunderbolt / USB 4 updates and additions

   - USB gadget driver updates

   - xhci fixes and updates

   - typec driver additions and updates

   - api conversions to various drivers for core kernel api changes

   - new USB control message functions to make it harder to get wrong,
     as found by syzbot (took 2 tries to get it right)

   - lots of tiny USB driver fixes and updates all over the place

  All of these have been in linux-next for a while, with the exception
  of the last "obviously correct" patch that updated a FALLTHROUGH
  comment that got merged last weekend"

* tag 'usb-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb: (374 commits)
  usb: musb: gadget: Use fallthrough pseudo-keyword
  usb: typec: Add QCOM PMIC typec detection driver
  USB: serial: option: add Cellient MPL200 card
  usb: typec: tcpci_maxim: Add support for Sink FRS
  usb: typec: tcpci: Implement callbacks for FRS
  usb: typec: tcpm: Add support for Sink Fast Role SWAP(FRS)
  usb: typec: tcpci_maxim: Chip level TCPC driver
  usb: typec: tcpci: Add set_vbus tcpci callback
  usb: typec: tcpci: Add a getter method to retrieve tcpm_port reference
  usbip: vhci_hcd: fix calling usb_hcd_giveback_urb() with irqs enabled
  usb: cdc-acm: add quirk to blacklist ETAS ES58X devices
  USB: serial: ftdi_sio: use cur_altsetting for consistency
  USB: serial: option: Add Telit FT980-KS composition
  USB: core: remove polling for /sys/kernel/debug/usb/devices
  usb: typec: add support for STUSB160x Type-C controller family
  usb: typec: add typec_find_pwr_opmode
  usb: typec: hd3ss3220: Use OF graph API to get the connector fwnode
  dt-bindings: usb: renesas,usb3-peri: Document HS and SS data bus
  dt-bindings: usb: convert ti,hd3ss3220 bindings to json-schema
  usb: dwc2: Fix INTR OUT transfers in DDMA mode.
  ...

3 years agoMerge tag 'staging-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh...
Linus Torvalds [Thu, 15 Oct 2020 16:46:23 +0000 (09:46 -0700)]
Merge tag 'staging-5.10-rc1' of git://git./linux/kernel/git/gregkh/staging

Pull staging / IIO driver updates from Greg KH:
 "Here is the large set of staging and IIO driver updates for 5.10-rc1.

  Included in here are:

   - new IIO drivers

   - new IIO driver frameworks

   - various IIO driver fixes and updates

   - IIO device tree conversions to yaml

   - so many minor staging driver coding style cleanups

   - most cdev driver moved out of staging

   - no staging drivers added or removed

  Full details are in the shortlog.

  All of these have been in linux-next for a while with no reported
  issues"

* tag 'staging-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging: (476 commits)
  staging: comedi: check validity of wMaxPacketSize of usb endpoints found
  staging: wfx: improve robustness of wfx_get_hw_rate()
  staging: wfx: drop unicode characters from strings
  staging: wfx: gpiod_get_value() can return an error
  staging: wfx: increase robustness of hif_generic_confirm()
  staging: wfx: wfx_init_common() returns NULL on error
  staging: wfx: standardize the error when vif does not exist
  staging: wfx: check memory allocation
  staging: wfx: improve error handling of hif_join()
  staging: dpaa2-switch: add a dpaa2_switch prefix to all functions in ethsw.c
  staging: dpaa2-switch: add a dpaa2_switch_ prefix to all functions in ethsw-ethtool.c
  staging: rtl8188eu: Fix long lines
  dt-bindings: staging: wfx: silabs,wfx yaml conversion
  staging: wfx: update copyrights dates
  staging: wfx: fix QoS priority for slow buses
  staging: wfx: fix BA sessions for older firmwares
  staging: wfx: remove remaining code of 'secure link' feature
  staging: wfx: fix handling of MMIC error
  staging: vchiq: Fix list_for_each exit tests
  staging: greybus: use __force when assigning __u8 value to snd_ctl_elem_type_t
  ...

3 years agoMerge tag 'spdx-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh...
Linus Torvalds [Wed, 14 Oct 2020 23:19:42 +0000 (16:19 -0700)]
Merge tag 'spdx-5.10-rc1' of git://git./linux/kernel/git/gregkh/spdx

Pull SPDX updates from Greg KH:
 "Here are some SPDX-specific changes for 5.10-rc1.

  They include:

   - driver fixes to make spdxcheck.pl work properly

   - add GFDL licenses as "deprecated" but required due to some of our
     documentation using them

   - add Zlib license as "deprecated" but required because we have code
     with this license in the tree.

   - convert some drivers to have SPDX identifiers that previously
     didn't have them.

  All have been in linux-next for a very long time with no reported
  issues"

* tag 'spdx-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/spdx:
  scripts/spdxcheck.py: handle license identifiers in XML comments
  net/mlx5: IPsec: make spdxcheck.py happy
  LICENSES/deprecated: add Zlib license text
  LICENSE: add GFDL deprecated licenses
  net/qla3xxx: Convert to SPDX license identifiers
  net/qlge: Convert to SPDX license identifiers
  net/qlcnic: Convert to SPDX license identifiers
  scsi/qla2xxx: Convert to SPDX license identifiers
  scsi/qla4xxx: Convert to SPDX license identifiers

3 years agoMerge tag 'driver-core-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Wed, 14 Oct 2020 23:09:32 +0000 (16:09 -0700)]
Merge tag 'driver-core-5.10-rc1' of git://git./linux/kernel/git/gregkh/driver-core

Pull driver core updates from Greg KH:
 "Here is the "big" set of driver core patches for 5.10-rc1

  They include a lot of different things, all related to the driver core
  and/or some driver logic:

   - sysfs common write functions to make it easier to audit sysfs
     attributes

   - device connection cleanups and fixes

   - devm helpers for a few functions

   - NOIO allocations for when devices are being removed

   - minor cleanups and fixes

  All have been in linux-next for a while with no reported issues"

* tag 'driver-core-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core: (31 commits)
  regmap: debugfs: use semicolons rather than commas to separate statements
  platform/x86: intel_pmc_core: do not create a static struct device
  drivers core: node: Use a more typical macro definition style for ACCESS_ATTR
  drivers core: Use sysfs_emit for shared_cpu_map_show and shared_cpu_list_show
  mm: and drivers core: Convert hugetlb_report_node_meminfo to sysfs_emit
  drivers core: Miscellaneous changes for sysfs_emit
  drivers core: Reindent a couple uses around sysfs_emit
  drivers core: Remove strcat uses around sysfs_emit and neaten
  drivers core: Use sysfs_emit and sysfs_emit_at for show(device *...) functions
  sysfs: Add sysfs_emit and sysfs_emit_at to format sysfs output
  dyndbg: use keyword, arg varnames for query term pairs
  driver core: force NOIO allocations during unplug
  platform_device: switch to simpler IDA interface
  driver core: platform: Document return type of more functions
  Revert "driver core: Annotate dev_err_probe() with __must_check"
  Revert "test_firmware: Test platform fw loading on non-EFI systems"
  iio: adc: xilinx-xadc: use devm_krealloc()
  hwmon: pmbus: use more devres helpers
  devres: provide devm_krealloc()
  syscore: Use pm_pr_dbg() for syscore_{suspend,resume}()
  ...

3 years agoMerge tag 'tty-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty
Linus Torvalds [Wed, 14 Oct 2020 23:05:52 +0000 (16:05 -0700)]
Merge tag 'tty-5.10-rc1' of git://git./linux/kernel/git/gregkh/tty

Pull tty/serial updates from Greg KH:
 "Here is the big set of tty and serial driver patches for 5.10-rc1.

  Lots of little things in here, including:

   - tasklet_setup api conversions

   - sysrq support for capital letters

   - vt and vc cleanups and unwinding the mess some more

   - serial driver updates and minor tweaks

   - new device ids

   - rs485 support for some drivers

   - serial binding documentation updates

   - lots of small serial driver changes for reported issues

  All have been in linux-next for a while with no reported issues"

* tag 'tty-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty: (79 commits)
  serial: mcf: add sysrq capability
  serial: fsl_lpuart: add sysrq support when using dma
  fbcon: remove no-op fbcon_set_origin()
  tty/sysrq: Extend the sysrq_key_table to cover capital letters
  serial: max310x: rework RX interrupt handling
  serial: 8250_dw: Fix clk-notifier/port suspend deadlock
  serial: 8250: Skip uninitialized TTY port baud rate update
  serial: 8250: Discard RTS/DTS setting from clock update method
  tty: serial: imx: disable TXDC IRQ in imx_uart_shutdown() to avoid IRQ storm
  serial: 8250_fsl: Fix TX interrupt handling condition
  serial: pl011: Fix lockdep splat when handling magic-sysrq interrupt
  tty: serial: fsl_lpuart: fix lpuart32_poll_get_char
  tty: serial: lpuart: fix lpuart32_write usage
  serial: qcom_geni_serial: To correct QUP Version detection logic
  serial: mvebu-uart: fix unused variable warning
  vt_ioctl: make VT_RESIZEX behave like VT_RESIZE
  serial: mvebu-uart: simplify the return expression of mvebu_uart_probe()
  tty: serial: imx: fix link error with CONFIG_SERIAL_CORE_CONSOLE=n
  tty: hvc: fix link error with CONFIG_SERIAL_CORE_CONSOLE=n
  pch_uart: drop double zeroing
  ...

3 years agopowerpc32: don't adjust unmoved stack pointer in csum_partial_copy_generic() epilogue
Jason A. Donenfeld [Wed, 14 Oct 2020 23:02:09 +0000 (01:02 +0200)]
powerpc32: don't adjust unmoved stack pointer in csum_partial_copy_generic() epilogue

A recent change to the checksum code removed usage of some extra
arguments, alongside with storage on the stack for those, and the stack
pointer no longer needed to be adjusted in the function prologue.

But a left over subtraction wasn't removed in the function epilogue,
causing the function to return with the stack pointer moved 16 bytes
away from where it should have.  This corrupted local state and lead to
weird crashes.

This simply removes the leftover instruction from the epilogue.

Fixes: 70d65cd555c5 ("ppc: propagate the calling conventions change down to csum_partial_copy_generic()")
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoMerge tag 'backlight-next-5.10' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Wed, 14 Oct 2020 22:59:42 +0000 (15:59 -0700)]
Merge tag 'backlight-next-5.10' of git://git./linux/kernel/git/lee/backlight

Pull backlight updates from Lee Jones:
 "New Drivers:
   - Add support for KTD253

  Fix-ups:
   - Add Device Tree documentation; common, kinetic,ktd253
   - Use correct header(s); tosa_lcd, tosa_bl

  Bug Fixes:
   - Fix refcount imbalance; sky81452-backlight"

* tag 'backlight-next-5.10' of git://git.kernel.org/pub/scm/linux/kernel/git/lee/backlight:
  backlight: tosa_bl: Include the right header
  backlight: tosa_lcd: Include the right header
  backlight: Add Kinetic KTD253 backlight driver
  dt-bindings: backlight: Add Kinetic KTD253 bindings
  dt-bindings: backlight: Add some common backlight properties
  backlight: sky81452-backlight: Fix refcount imbalance on error

3 years agoMerge tag 'mfd-next-5.10' of git://git.kernel.org/pub/scm/linux/kernel/git/lee/mfd
Linus Torvalds [Wed, 14 Oct 2020 22:56:58 +0000 (15:56 -0700)]
Merge tag 'mfd-next-5.10' of git://git./linux/kernel/git/lee/mfd

Pull MFD updates from Lee Jones:
 "New Drivers:
   - Add support for initialising shared (between children) Regmaps
   - Add support for Kontron SL28CPLD
   - Add support for ENE KB3930 Embedded Controller
   - Add support for Intel FPGA PAC MAX 10 BMC

  New Device Support:
   - Add support for Power to Ricoh RN5T618
   - Add support for UART to Intel Lakefield
   - Add support for LP87524_Q1 to Texas Instruments LP87565

  New Functionality:
   - Device Tree; ene-kb3930, sl28cpld, syscon, lp87565, lp87524-q1
   - Use new helper dev_err_probe(); madera-core, stmfx, wcd934x
   - Use new GPIOD API; dm355evm_msp
   - Add wake-up capability; sprd-sc27xx-spi
   - Add ACPI support; kempld-core

  Fix-ups:
   - Trivial (spelling/whitespace); Kconfig, ab8500
   - Fix for unused variables; khadas-mcu, kempld-core
   - Remove unused header file(s); mt6360-core
   - Use correct IRQ flags in docs; act8945a, gateworks-gsc, rohm,bd70528-pmic
   - Add COMPILE_TEST support; asic3, tmio_core
   - Add dependency on I2C; SL28CPLD

  Bug Fixes:
   - Fix memory leak(s); sm501
   - Do not free regmap_config's 'name' until exit; syscon"

* tag 'mfd-next-5.10' of git://git.kernel.org/pub/scm/linux/kernel/git/lee/mfd: (34 commits)
  mfd: kempld-core: Fix unused variable 'kempld_acpi_table' when !ACPI
  mfd: sl28cpld: Depend on I2C
  mfd: asic3: Build if COMPILE_TEST=y
  dt-bindings: mfd: Correct interrupt flags in examples
  mfd: Add ACPI support to Kontron PLD driver
  mfd: intel-m10-bmc: Add Intel MAX 10 BMC chip support for Intel FPGA PAC
  mfd: lp87565: Add LP87524-Q1 variant
  dt-bindings: mfd: Add LP87524-Q1
  dt-bindings: mfd: lp87565: Convert to yaml
  mfd: mt6360: Remove unused include <linux/version.h>
  mfd: sm501: Fix leaks in probe()
  mfd: syscon: Don't free allocated name for regmap_config
  dt-bindings: mfd: syscon: Document Exynos3 and Exynos5433 compatibles
  dt-bindings: mfd: syscon: Merge Samsung Exynos Sysreg bindings
  dt-bindings: mfd: ab8500: Remove weird Unicode characters
  mfd: sprd: Add wakeup capability for PMIC IRQ
  mfd: intel-lpss: Add device IDs for UART ports for Lakefield
  mfd: dm355evm_msp: Convert LEDs to GPIO descriptor table
  mfd: wcd934x: Simplify with dev_err_probe()
  mfd: stmfx: Simplify with dev_err_probe()
  ...

3 years agoMerge tag 'devicetree-for-5.10' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Wed, 14 Oct 2020 22:31:58 +0000 (15:31 -0700)]
Merge tag 'devicetree-for-5.10' of git://git./linux/kernel/git/robh/linux

Pull devicetree updates from Rob Herring:

 - Update dtc to upstream version v1.6.0-31-gcbca977ea121

 - dtx_diff help text reformatting

 - Speed-up validation time for binding and dtb checks using json for
   intermediate files

 - Add support for running yamllint on DT schema files

 - Remove old booting-without-of.rst

 - Extend the example schema to address common issues

 - Cleanup handling of additionalProperties/unevaluatedProperties

 - Ensure all DSI controller schemas reference dsi-controller.yaml

 - Vendor prefixes for Zealz, Wandbord/Technexion, Embest RIoT, Rex,
   DFI, and Cisco Meraki

 - Convert at25, SPMI bus, TI hwlock, HiSilicon Hi3660 USB3 PHY, Arm
   SP805 watchdog, Arm SP804, and Samsung 11-pin USB connector to DT
   schema

 - Convert HiSilicon SoC and syscon bindings to DT schema

 - Convert SiFive Risc-V L2 cache, PLIC, PRCI, and PWM to DT schema

 - Convert i.MX bindings for w1, crypto, rng, SIM, PM, DDR, SATA, vf610
   GPIO, and UART to DT schema

 - Add i.MX 8M compatible strings

 - Add LM81 and DS1780 as trivial devices

 - Various missing properties added to fix dtb validation warnings

* tag 'devicetree-for-5.10' of git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux: (111 commits)
  dt-bindings: misc: explicitly add #address-cells for slave mode
  spi: dt-bindings: spi-controller: explicitly require #address-cells=<0> for slave mode
  dt: Remove booting-without-of.rst
  dt-bindings: update usb-c-connector example
  dt-bindings: arm: hisilicon: add missing properties into cpuctrl.yaml
  dt-bindings: arm: hisilicon: add missing properties into sysctrl.yaml
  dt-bindings: pwm: imx: document i.MX compatibles
  scripts/dtc: Update to upstream version v1.6.0-31-gcbca977ea121
  dt-bindings: Add running yamllint to dt_binding_check
  dt-bindings: powerpc: Add a schema for the 'sleep' property
  dt-bindings: pinctrl: sirf: Fix typo abitrary
  dt-bindings: pinctrl: qcom: Fix typo abitrary
  dt-bindings: Explicitly allow additional properties in common schemas
  dt-bindings: Use 'additionalProperties' instead of 'unevaluatedProperties'
  dt-bindings: Add missing 'unevaluatedProperties'
  Docs: Fixing spelling errors in Documentation/devicetree/bindings/
  dt-bindings: arm: hisilicon: convert Hi6220 domain controller bindings to json-schema
  dt-bindings: riscv: convert pwm bindings to json-schema
  dt-bindings: riscv: convert plic bindings to json-schema
  dt-bindings: fu540: prci: convert PRCI bindings to json-schema
  ...

3 years agoMerge tag 'pinctrl-v5.10-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw...
Linus Torvalds [Wed, 14 Oct 2020 22:25:04 +0000 (15:25 -0700)]
Merge tag 'pinctrl-v5.10-1' of git://git./linux/kernel/git/linusw/linux-pinctrl

Pull pin control updates from Linus Walleij:
 "Core changes:

   - NONE whatsoever, we don't even touch the core files this time
     around.

  New drivers:

   - New driver for the Toshiba Visconti SoC.

   - New subdriver for the Qualcomm MSM8226 SoC.

   - New subdriver for the Actions Semiconductor S500 SoC.

   - New subdriver for the Mediatek MT8192 SoC.

   - New subdriver for the Microchip SAMA7G5 SoC.

  Driver enhancements:

   - Intel Cherryview and Baytrail cleanups and refactorings.

   - Enhanced support for the Renesas R8A7790, more pins and groups.

   - Some optimizations for the MCP23S08 MCP23x17 variant.

   - Some cleanups around the Actions Semiconductor subdrivers.

   - A bunch of cleanups around the SH-PFC and Emma Mobile drivers.

   - The "SH-PFC" (literally SuperH pin function controller, I think)
     subdirectory is now renamed to the more neutral "renesas", as these
     are not very much centered around SuperH anymore.

   - Non-critical fixes for the Aspeed driver.

   - Non-critical fixes for the Ingenic (MIPS!) driver.

   - Fix a bunch of missing pins on the AMD pinctrl driver"

* tag 'pinctrl-v5.10-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl: (78 commits)
  pinctrl: amd: Add missing pins to the pin group list
  dt-bindings: pinctrl: sunxi: Allow pinctrl with more interrupt banks
  pinctrl: visconti: PINCTRL_TMPV7700 should depend on ARCH_VISCONTI
  pinctrl: mediatek: Free eint data on failure
  pinctrl: single: fix debug output when #pinctrl-cells = 2
  pinctrl: single: fix pinctrl_spec.args_count bounds check
  pinctrl: sunrisepoint: Modify COMMUNITY macros to be consistent
  pinctrl: cannonlake: Modify COMMUNITY macros to be consistent
  pinctrl: tigerlake: Fix register offsets for TGL-H variant
  pinctrl: Document pinctrl-single,pins when #pinctrl-cells = 2
  pinctrl: mediatek: use devm_platform_ioremap_resource_byname()
  pinctrl: nuvoton: npcm7xx: Constify static ops structs
  pinctrl: mediatek: mt7622: add antsel pins/groups
  pinctrl: ocelot: simplify the return expression of ocelot_gpiochip_register()
  pinctrl: at91-pio4: add support for sama7g5 SoC
  dt-bindings: pinctrl: at91-pio4: add microchip,sama7g5
  pinctrl: spear: simplify the return expression of tvc_connect()
  pinctrl: spear: simplify the return expression of spear310_pinctrl_probe
  pinctrl: sprd: use module_platform_driver to simplify the code
  pinctrl: Ingenic: Add I2S pins support for Ingenic SoCs.
  ...

3 years agoMerge tag 'leds-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/pavel...
Linus Torvalds [Wed, 14 Oct 2020 22:22:07 +0000 (15:22 -0700)]
Merge tag 'leds-5.10-rc1' of git://git./linux/kernel/git/pavel/linux-leds

Pull LED updates from Pavel Machek:
 "Quite a lot of stuff is going on here. Great cleanups/fixes from Marek
  and others are biggest part.

  I limited CPU LED trigger to 8 LEDs, because it was willing to
  register 1024 'triggers' on machine with 1024 CPUs. I don't believe it
  will cause any problems, but we can raise the limit if it does"

* tag 'leds-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/pavel/linux-leds: (84 commits)
  leds: pwm: Remove platform_data support
  leds: lm3697: Fix out-of-bound access
  leds: ns2: do not guard OF match pointer with of_match_ptr
  leds: ns2: convert to fwnode API
  leds: tlc591xx: fix leak of device node iterator
  leds: pca963x: use struct led_init_data when registering
  leds: pca963x: register LEDs immediately after parsing, get rid of platdata
  leds: tca6507: remove binding comment
  leds: tca6507: cosmetic change: use helper variable
  leds: tca6507: do not set GPIO names
  dt-bindings: leds: tca6507: convert to YAML
  ledtrig-cpu: Limit to 8 CPUs
  leds: TODO: Add documentation about possible subsystem improvements
  leds: pca9532: read pwm settings from device tree
  leds: pca9532: correct shift computation in pca9532_getled
  leds: lm36274: Fix warning for undefined parameters
  leds: lm3532: Fix warnings for undefined parameters
  leds: pca963x: use flexible array
  leds: pca963x: cosmetic: rename variables
  leds: pca963x: cosmetic: rename variables
  ...

3 years agoMerge tag 'scsi-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi
Linus Torvalds [Wed, 14 Oct 2020 22:15:35 +0000 (15:15 -0700)]
Merge tag 'scsi-misc' of git://git./linux/kernel/git/jejb/scsi

Pull SCSI updates from James Bottomley:
 "The usual driver updates (ufs, qla2xxx, tcmu, ibmvfc, lpfc, smartpqi,
  hisi_sas, qedi, qedf, mpt3sas) and minor bug fixes.

  There are only three core changes: adding sense codes, cleaning up
  noretry and adding an option for limitless retries"

* tag 'scsi-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi: (226 commits)
  scsi: hisi_sas: Recover PHY state according to the status before reset
  scsi: hisi_sas: Filter out new PHY up events during suspend
  scsi: hisi_sas: Add device link between SCSI devices and hisi_hba
  scsi: hisi_sas: Add check for methods _PS0 and _PR0
  scsi: hisi_sas: Add controller runtime PM support for v3 hw
  scsi: hisi_sas: Switch to new framework to support suspend and resume
  scsi: hisi_sas: Use hisi_hba->cq_nvecs for calling calling synchronize_irq()
  scsi: qedf: Remove redundant assignment to variable 'rc'
  scsi: lpfc: Remove unneeded variable 'status' in lpfc_fcp_cpu_map_store()
  scsi: snic: Convert to use DEFINE_SEQ_ATTRIBUTE macro
  scsi: qla4xxx: Delete unneeded variable 'status' in qla4xxx_process_ddb_changed
  scsi: sun_esp: Use module_platform_driver to simplify the code
  scsi: sun3x_esp: Use module_platform_driver to simplify the code
  scsi: sni_53c710: Use module_platform_driver to simplify the code
  scsi: qlogicpti: Use module_platform_driver to simplify the code
  scsi: mac_esp: Use module_platform_driver to simplify the code
  scsi: jazz_esp: Use module_platform_driver to simplify the code
  scsi: mvumi: Fix error return in mvumi_io_attach()
  scsi: lpfc: Drop nodelist reference on error in lpfc_gen_req()
  scsi: be2iscsi: Fix a theoretical leak in beiscsi_create_eqs()
  ...

3 years agoMerge tag 'for-5.10/dm-changes' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Wed, 14 Oct 2020 22:05:38 +0000 (15:05 -0700)]
Merge tag 'for-5.10/dm-changes' of git://git./linux/kernel/git/device-mapper/linux-dm

Pull device mapper updates from Mike Snitzer:

 - Improve DM core's bio splitting to use blk_max_size_offset(). Also
   fix bio splitting for bios that were deferred to the worker thread
   due to a DM device being suspended.

 - Remove DM core's special handling of NVMe devices now that block core
   has internalized efficiencies drivers previously needed to be
   concerned about (via now removed direct_make_request).

 - Fix request-based DM to not bounce through indirect dm_submit_bio;
   instead have block core make direct call to blk_mq_submit_bio().

 - Various DM core cleanups to simplify and improve code.

 - Update DM cryot to not use drivers that set
   CRYPTO_ALG_ALLOCATES_MEMORY.

 - Fix DM raid's raid1 and raid10 discard limits for the purposes of
   linux-stable. But then remove DM raid's discard limits settings now
   that MD raid can efficiently handle large discards.

 - A couple small cleanups across various targets.

* tag 'for-5.10/dm-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm:
  dm: fix request-based DM to not bounce through indirect dm_submit_bio
  dm: remove special-casing of bio-based immutable singleton target on NVMe
  dm: export dm_copy_name_and_uuid
  dm: fix comment in __dm_suspend()
  dm: fold dm_process_bio() into dm_submit_bio()
  dm: fix missing imposition of queue_limits from dm_wq_work() thread
  dm snap persistent: simplify area_io()
  dm thin metadata: Remove unused local variable when create thin and snap
  dm raid: remove unnecessary discard limits for raid10
  dm raid: fix discard limits for raid1 and raid10
  dm crypt: don't use drivers that have CRYPTO_ALG_ALLOCATES_MEMORY
  dm: use dm_table_get_device_name() where appropriate in targets
  dm table: make 'struct dm_table' definition accessible to all of DM core
  dm: eliminate need for start_io_acct() forward declaration
  dm: simplify __process_abnormal_io()
  dm: push use of on-stack flush_bio down to __send_empty_flush()
  dm: optimize max_io_len() by inlining max_io_len_target_boundary()
  dm: push md->immutable_target optimization down to __process_bio()
  dm: change max_io_len() to use blk_max_size_offset()
  dm table: stack 'chunk_sectors' limit to account for target-specific splitting

3 years agoMerge tag 'for-linus-5.10-1' of git://github.com/cminyard/linux-ipmi
Linus Torvalds [Wed, 14 Oct 2020 22:00:20 +0000 (15:00 -0700)]
Merge tag 'for-linus-5.10-1' of git://github.com/cminyard/linux-ipmi

Pull IPMI updates from Corey Minyard:
 "Some minor bug fixes, return values, cleanups of prints, conversion of
  tasklets to the new API.

  The biggest change is retrying the initial information fetch from the
  management controller. If that fails, the iterface is not operational,
  and one group was having trouble with the management controller not
  being ready when the OS started up. So a retry was added"

* tag 'for-linus-5.10-1' of git://github.com/cminyard/linux-ipmi:
  ipmi_si: Fix wrong return value in try_smi_init()
  ipmi: msghandler: Fix a signedness bug
  ipmi: add retry in try_get_dev_id()
  ipmi: Clean up some printks
  ipmi:msghandler: retry to get device id on an error
  ipmi:sm: Print current state when the state is invalid
  ipmi: Reset response handler when failing to send the command
  ipmi: add a newline when printing parameter 'panic_op' by sysfs
  char: ipmi: convert tasklets to use new tasklet_setup() API

3 years agoMerge branch 'for-5.10' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup
Linus Torvalds [Wed, 14 Oct 2020 21:58:32 +0000 (14:58 -0700)]
Merge branch 'for-5.10' of git://git./linux/kernel/git/tj/cgroup

Pull cgroup updates from Tejun Heo:
 "Two minor changes.

  One makes cgroup interface files ignore zero-sized writes rather than
  triggering -EINVAL on them. The other change is a cleanup which
  doesn't cause any behavior changes"

* 'for-5.10' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:
  cgroup: Zero sized write should be no-op
  cgroup: remove redundant kernfs_activate in cgroup_setup_root()

3 years agofs: fix NULL dereference due to data race in prepend_path()
Andrii Nakryiko [Wed, 14 Oct 2020 20:45:28 +0000 (13:45 -0700)]
fs: fix NULL dereference due to data race in prepend_path()

Fix data race in prepend_path() with re-reading mnt->mnt_ns twice
without holding the lock.

is_mounted() does check for NULL, but is_anon_ns(mnt->mnt_ns) might
re-read the pointer again which could be NULL already, if in between
reads one of kern_unmount()/kern_unmount_array()/umount_tree() sets
mnt->mnt_ns to NULL.

This is seen in production with the following stack trace:

  BUG: kernel NULL pointer dereference, address: 0000000000000048
  ...
  RIP: 0010:prepend_path.isra.4+0x1ce/0x2e0
  Call Trace:
    d_path+0xe6/0x150
    proc_pid_readlink+0x8f/0x100
    vfs_readlink+0xf8/0x110
    do_readlinkat+0xfd/0x120
    __x64_sys_readlinkat+0x1a/0x20
    do_syscall_64+0x42/0x110
    entry_SYSCALL_64_after_hwframe+0x44/0xa9

Fixes: f2683bd8d5bd ("[PATCH] fix d_absolute_path() interplay with fsmount()")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoMerge tag 'threads-v5.10' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner...
Linus Torvalds [Wed, 14 Oct 2020 21:39:20 +0000 (14:39 -0700)]
Merge tag 'threads-v5.10' of git://git./linux/kernel/git/brauner/linux

Pull pidfd updates from Christian Brauner:
 "This introduces a new extension to the pidfd_open() syscall. Users can
  now raise the new PIDFD_NONBLOCK flag to support non-blocking pidfd
  file descriptors. This has been requested for uses in async process
  management libraries such as async-pidfd in Rust.

  Ever since the introduction of pidfds and more advanced async io
  various programming languages such as Rust have grown support for
  async event libraries. These libraries are created to help build
  epoll-based event loops around file descriptors. A common pattern is
  to automatically make all file descriptors they manage to O_NONBLOCK.

  For such libraries the EAGAIN error code is treated specially. When a
  function is called that returns EAGAIN the function isn't called again
  until the event loop indicates the the file descriptor is ready.
  Supporting EAGAIN when waiting on pidfds makes such libraries just
  work with little effort.

  This introduces a new flag PIDFD_NONBLOCK that is equivalent to
  O_NONBLOCK. This follows the same patterns we have for other (anon
  inode) file descriptors such as EFD_NONBLOCK, IN_NONBLOCK,
  SFD_NONBLOCK, TFD_NONBLOCK and the same for close-on-exec flags.

  Passing a non-blocking pidfd to waitid() currently has no effect, i.e.
  is not supported. There are users which would like to use waitid() on
  pidfds that are O_NONBLOCK and mix it with pidfds that are blocking
  and both pass them to waitid().

  The expected behavior is to have waitid() return -EAGAIN for
  non-blocking pidfds and to block for blocking pidfds without needing
  to perform any additional checks for flags set on the pidfd before
  passing it to waitid(). Non-blocking pidfds will return EAGAIN from
  waitid() when no child process is ready yet. Returning -EAGAIN for
  non-blocking pidfds makes it easier for event loops that handle EAGAIN
  specially.

  It also makes the API more consistent and uniform. In essence,
  waitid() is treated like a read on a non-blocking pidfd or a recvmsg()
  on a non-blocking socket.

  With the addition of support for non-blocking pidfds we support the
  same functionality that sockets do. For sockets() recvmsg() supports
  MSG_DONTWAIT for pidfds waitid() supports WNOHANG. Both flags are
  per-call options. In contrast non-blocking pidfds and non-blocking
  sockets are a setting on an open file description affecting all
  threads in the calling process as well as other processes that hold
  file descriptors referring to the same open file description. Both
  behaviors, per call and per open file description, have genuine
  use-cases.

  The interaction with the WNOHANG flag is documented as follows:

   - If a non-blocking pidfd is passed and WNOHANG is not raised we
     simply raise the WNOHANG flag internally. When do_wait() returns
     indicating that there are eligible child processes but none have
     exited yet we set EAGAIN. If no child process exists we continue
     returning ECHILD.

   - If a non-blocking pidfd is passed and WNOHANG is raised waitid()
     will continue returning 0, i.e. it will not set EAGAIN. This ensure
     backwards compatibility with applications passing WNOHANG
     explicitly with pidfds"

* tag 'threads-v5.10' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux:
  tests: remove O_NONBLOCK before waiting for WSTOPPED
  tests: add waitid() tests for non-blocking pidfds
  tests: port pidfd_wait to kselftest harness
  pidfd: support PIDFD_NONBLOCK in pidfd_open()
  exit: support non-blocking pidfds

3 years agoMerge tag 'kernel-clone-v5.9' of git://git.kernel.org/pub/scm/linux/kernel/git/braune...
Linus Torvalds [Wed, 14 Oct 2020 21:32:52 +0000 (14:32 -0700)]
Merge tag 'kernel-clone-v5.9' of git://git./linux/kernel/git/brauner/linux

Pull kernel_clone() updates from Christian Brauner:
 "During the v5.9 merge window we reworked the process creation
  codepaths across multiple architectures. After this work we were only
  left with the _do_fork() helper based on the struct kernel_clone_args
  calling convention. As was pointed out _do_fork() isn't valid
  kernelese especially for a helper that isn't just static.

  This series removes the _do_fork() helper and introduces the new
  kernel_clone() helper. The process creation cleanup didn't change the
  name to something more reasonable mainly because _do_fork() was used
  in quite a few places. So sending this as a separate series seemed the
  better strategy.

  I originally intended to send this early in the v5.9 development cycle
  after the merge window had closed but given that this was touching
  quite a few places I decided to defer this until the v5.10 merge
  window"

* tag 'kernel-clone-v5.9' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux:
  sched: remove _do_fork()
  tracing: switch to kernel_clone()
  kgdbts: switch to kernel_clone()
  kprobes: switch to kernel_clone()
  x86: switch to kernel_clone()
  sparc: switch to kernel_clone()
  nios2: switch to kernel_clone()
  m68k: switch to kernel_clone()
  ia64: switch to kernel_clone()
  h8300: switch to kernel_clone()
  fork: introduce kernel_clone()

3 years agoMerge tag 'linux-kselftest-fixes-5.10-rc1' of git://git.kernel.org/pub/scm/linux...
Linus Torvalds [Wed, 14 Oct 2020 21:23:51 +0000 (14:23 -0700)]
Merge tag 'linux-kselftest-fixes-5.10-rc1' of git://git./linux/kernel/git/shuah/linux-kselftest

Pull kselftest updates from Shuah Khan:

 - a selftests harness fix to flush stdout before forking to avoid
   parent and child printing duplicates messages. This is evident when
   test output is redirected to a file.

 - a tools/ wide change to avoid comma separated statements from Joe
   Perches. This fix spans tools/lib, tools/power/cpupower, and
   selftests.

* tag 'linux-kselftest-fixes-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest:
  tools: Avoid comma separated statements
  selftests/harness: Flush stdout before forking

3 years agoMerge tag 'xfs-5.10-merge-2' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux
Linus Torvalds [Wed, 14 Oct 2020 21:06:06 +0000 (14:06 -0700)]
Merge tag 'xfs-5.10-merge-2' of git://git./fs/xfs/xfs-linux

Pull xfs updates from Darrick Wong:
 "The biggest changes are two new features for the ondisk metadata: one
  to record the sizes of the inode btrees in the AG to increase
  redundancy checks and to improve mount times; and a second new feature
  to support timestamps until the year 2486.

  We also fixed a problem where reflinking into a file that requires
  synchronous writes wouldn't actually flush the updates to disk; clean
  up a fair amount of cruft; and started fixing some bugs in the
  realtime volume code.

  Summary:

   - Clean up the buffer ioend calling path so that the retry strategy
     isn't quite so scattered everywhere.

   - Clean up m_sb_bp handling.

   - New feature: storing inode btree counts in the AGI to speed up
     certain mount time per-AG block reservation operatoins and add a
     little more metadata redundancy.

   - New feature: Widen inode timestamps and quota grace expiration
     timestamps to support dates through the year 2486.

   - Get rid of more of our custom buffer allocation API wrappers.

   - Use a proper VLA for shortform xattr structure namevals.

   - Force the log after reflinking or deduping into a file that is
     opened with O_SYNC or O_DSYNC.

   - Fix some math errors in the realtime allocator"

* tag 'xfs-5.10-merge-2' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux: (42 commits)
  xfs: ensure that fpunch, fcollapse, and finsert operations are aligned to rt extent size
  xfs: make sure the rt allocator doesn't run off the end
  xfs: Remove unneeded semicolon
  xfs: force the log after remapping a synchronous-writes file
  xfs: Convert xfs_attr_sf macros to inline functions
  xfs: Use variable-size array for nameval in xfs_attr_sf_entry
  xfs: Remove typedef xfs_attr_shortform_t
  xfs: remove typedef xfs_attr_sf_entry_t
  xfs: Remove kmem_zalloc_large()
  xfs: enable big timestamps
  xfs: trace timestamp limits
  xfs: widen ondisk quota expiration timestamps to handle y2038+
  xfs: widen ondisk inode timestamps to deal with y2038+
  xfs: redefine xfs_ictimestamp_t
  xfs: redefine xfs_timestamp_t
  xfs: move xfs_log_dinode_to_disk to the log recovery code
  xfs: refactor quota timestamp coding
  xfs: refactor default quota grace period setting code
  xfs: refactor quota expiration timer modification
  xfs: explicitly define inode timestamp range
  ...

3 years agoMerge tag 'iomap-5.10-merge-4' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux
Linus Torvalds [Wed, 14 Oct 2020 19:23:00 +0000 (12:23 -0700)]
Merge tag 'iomap-5.10-merge-4' of git://git./fs/xfs/xfs-linux

Pull iomap updates from Darrick Wong:
 "There's not a lot of new stuff going on here -- a little bit of code
  refactoring to make iomap workable with btrfs' fsync locking model,
  cleanups in preparation for adding THP support for filesystems, and
  fixing a data corruption issue for blocksize < pagesize filesystems.

  Summary:

   - Don't WARN_ON weird states that unprivileged users can create.

   - Don't invalidate page cache when direct writes want to fall back to
     buffered.

   - Fix some problems when readahead ios fail.

   - Fix a problem where inline data pages weren't getting flushed
     during an unshare operation.

   - Rework iomap to support arbitrarily many blocks per page in
     preparation to support THP for the page cache.

   - Fix a bug in the blocksize < pagesize buffered io path where we
     could fail to initialize the many-blocks-per-page uptodate bitmap
     correctly when the backing page is actually up to date. This could
     cause us to forget to write out dirty pages.

   - Split out the generic_write_sync at the end of the directio write
     path so that btrfs can drop the inode lock before sync'ing the
     file.

   - Call inode_dio_end before trying to sync the file after a O_DSYNC
     direct write (instead of afterwards) to match the behavior of the
     old directio code"

* tag 'iomap-5.10-merge-4' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
  iomap: Call inode_dio_end() before generic_write_sync()
  iomap: Allow filesystem to call iomap_dio_complete without i_rwsem
  iomap: Set all uptodate bits for an Uptodate page
  iomap: Change calling convention for zeroing
  iomap: Convert iomap_write_end types
  iomap: Convert write_count to write_bytes_pending
  iomap: Convert read_count to read_bytes_pending
  iomap: Support arbitrarily many blocks per page
  iomap: Use bitmap ops to set uptodate bits
  iomap: Use kzalloc to allocate iomap_page
  fs: Introduce i_blocks_per_page
  iomap: Fix misplaced page flushing
  iomap: Use round_down/round_up macros in __iomap_write_begin
  iomap: Mark read blocks uptodate in write_begin
  iomap: Clear page error before beginning a write
  iomap: Fix direct I/O write consistency check
  iomap: fix WARN_ON_ONCE() from unprivileged users

3 years agoMerge tag 'iommu-updates-v5.10' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Wed, 14 Oct 2020 19:08:34 +0000 (12:08 -0700)]
Merge tag 'iommu-updates-v5.10' of git://git./linux/kernel/git/joro/iommu

Pull iommu updates from Joerg Roedel:

 - ARM-SMMU Updates from Will:

      - Continued SVM enablement, where page-table is shared with CPU

      - Groundwork to support integrated SMMU with Adreno GPU

      - Allow disabling of MSI-based polling on the kernel command-line

      - Minor driver fixes and cleanups (octal permissions, error
        messages, ...)

 - Secure Nested Paging Support for AMD IOMMU. The IOMMU will fault when
   a device tries DMA on memory owned by a guest. This needs new
   fault-types as well as a rewrite of the IOMMU memory semaphore for
   command completions.

 - Allow broken Intel IOMMUs (wrong address widths reported) to still be
   used for interrupt remapping.

 - IOMMU UAPI updates for supporting vSVA, where the IOMMU can access
   address spaces of processes running in a VM.

 - Support for the MT8167 IOMMU in the Mediatek IOMMU driver.

 - Device-tree updates for the Renesas driver to support r8a7742.

 - Several smaller fixes and cleanups all over the place.

* tag 'iommu-updates-v5.10' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu: (57 commits)
  iommu/vt-d: Gracefully handle DMAR units with no supported address widths
  iommu/vt-d: Check UAPI data processed by IOMMU core
  iommu/uapi: Handle data and argsz filled by users
  iommu/uapi: Rename uapi functions
  iommu/uapi: Use named union for user data
  iommu/uapi: Add argsz for user filled data
  docs: IOMMU user API
  iommu/qcom: add missing put_device() call in qcom_iommu_of_xlate()
  iommu/arm-smmu-v3: Add SVA device feature
  iommu/arm-smmu-v3: Check for SVA features
  iommu/arm-smmu-v3: Seize private ASID
  iommu/arm-smmu-v3: Share process page tables
  iommu/arm-smmu-v3: Move definitions to a header
  iommu/io-pgtable-arm: Move some definitions to a header
  iommu/arm-smmu-v3: Ensure queue is read after updating prod pointer
  iommu/amd: Re-purpose Exclusion range registers to support SNP CWWB
  iommu/amd: Add support for RMP_PAGE_FAULT and RMP_HW_ERR
  iommu/amd: Use 4K page for completion wait write-back semaphore
  iommu/tegra-smmu: Allow to group clients in same swgroup
  iommu/tegra-smmu: Fix iova->phys translation
  ...

3 years agoMerge branch 'stable/for-linus-5.10' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Wed, 14 Oct 2020 19:00:02 +0000 (12:00 -0700)]
Merge branch 'stable/for-linus-5.10' of git://git./linux/kernel/git/konrad/swiotlb

Pull swiotlb updates from Konrad Rzeszutek Wilk:
 "Minor enhancement of using %p to print phys_addr_r and also compiler
  warnings"

* 'stable/for-linus-5.10' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/swiotlb:
  swiotlb: Mark max_segment with static keyword
  swiotlb: Declare swiotlb_late_init_with_default_size() in header
  swiotlb: Use %pa to print phys_addr_t variables

3 years agoMerge tag 'pnp-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael...
Linus Torvalds [Wed, 14 Oct 2020 18:50:01 +0000 (11:50 -0700)]
Merge tag 'pnp-5.10-rc1' of git://git./linux/kernel/git/rafael/linux-pm

Pull PNP updates from Rafael Wysocki:
 "These clean the PNP code somewhat:

   - Remove the now unused pnp_find_card() function (Christoph Hellwig)

   - Drop duplicate pci.h include from the quirks code and add an
     "internal.h" include to acpi_pnp.c to fix a compiler warning (Tian
     Tao)"

* tag 'pnp-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  PNP: remove the now unused pnp_find_card() function
  PNP: ACPI: Fix missing-prototypes in acpi_pnp.c
  PNP: quirks: Fix duplicate included pci.h

3 years agoMerge tag 'acpi-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael...
Linus Torvalds [Wed, 14 Oct 2020 18:42:04 +0000 (11:42 -0700)]
Merge tag 'acpi-5.10-rc1' of git://git./linux/kernel/git/rafael/linux-pm

Pull ACPI updates from Rafael Wysocki:
 "These add support for generic initiator-only proximity domains to the
  ACPI NUMA code and the architectures using it, clean up some
  non-ACPICA code referring to debug facilities from ACPICA, reduce the
  overhead related to accessing GPE registers, add a new DPTF (Dynamic
  Power and Thermal Framework) participant driver, update the ACPICA
  code in the kernel to upstream revision 20200925, add a new ACPI
  backlight whitelist entry, fix a few assorted issues and clean up some
  code.

  Specifics:

   - Add support for generic initiator-only proximity domains to the
     ACPI NUMA code and the architectures using it (Jonathan Cameron)

   - Clean up some non-ACPICA code referring to debug facilities from
     ACPICA that are not actually used in there (Hanjun Guo)

   - Add new DPTF driver for the PCH FIVR participant (Srinivas
     Pandruvada)

   - Reduce overhead related to accessing GPE registers in ACPICA and
     the OS interface layer and make it possible to access GPE registers
     using logical addresses if they are memory-mapped (Rafael Wysocki)

   - Update the ACPICA code in the kernel to upstream revision 20200925
     including changes as follows:
      + Add predefined names from the SMBus sepcification (Bob Moore)
      + Update acpi_help UUID list (Bob Moore)
      + Return exceptions for string-to-integer conversions in iASL (Bob
        Moore)
      + Add a new "ALL <NameSeg>" debugger command (Bob Moore)
      + Add support for 64 bit risc-v compilation (Colin Ian King)
      + Do assorted cleanups (Bob Moore, Colin Ian King, Randy Dunlap)

   - Add new ACPI backlight whitelist entry for HP 635 Notebook (Alex
     Hung)

   - Move TPS68470 OpRegion driver to drivers/acpi/pmic/ and split out
     Kconfig and Makefile specific for ACPI PMIC (Andy Shevchenko)

   - Clean up the ACPI SoC driver for AMD SoCs (Hanjun Guo)

   - Add missing config_item_put() to fix refcount leak (Hanjun Guo)

   - Drop lefrover field from struct acpi_memory_device (Hanjun Guo)

   - Make the ACPI extlog driver check for RDMSR failures (Ben
     Hutchings)

   - Fix handling of lid state changes in the ACPI button driver when
     input device is closed (Dmitry Torokhov)

   - Fix several assorted build issues (Barnabás PÅ‘cze, John Garry,
     Nathan Chancellor, Tian Tao)

   - Drop unused inline functions and reduce code duplication by using
     kobj_to_dev() in the NFIT parsing code (YueHaibing, Wang Qing)

   - Serialize tools/power/acpi Makefile (Thomas Renninger)"

* tag 'acpi-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (64 commits)
  ACPICA: Update version to 20200925 Version 20200925
  ACPICA: Remove unnecessary semicolon
  ACPICA: Debugger: Add a new command: "ALL <NameSeg>"
  ACPICA: iASL: Return exceptions for string-to-integer conversions
  ACPICA: acpi_help: Update UUID list
  ACPICA: Add predefined names found in the SMBus sepcification
  ACPICA: Tree-wide: fix various typos and spelling mistakes
  ACPICA: Drop the repeated word "an" in a comment
  ACPICA: Add support for 64 bit risc-v compilation
  ACPI: button: fix handling lid state changes when input device closed
  tools/power/acpi: Serialize Makefile
  ACPI: scan: Replace ACPI_DEBUG_PRINT() with pr_debug()
  ACPI: memhotplug: Remove 'state' from struct acpi_memory_device
  ACPI / extlog: Check for RDMSR failure
  ACPI: Make acpi_evaluate_dsm() prototype consistent
  docs: mm: numaperf.rst Add brief description for access class 1.
  node: Add access1 class to represent CPU to memory characteristics
  ACPI: HMAT: Fix handling of changes from ACPI 6.2 to ACPI 6.3
  ACPI: Let ACPI know we support Generic Initiator Affinity Structures
  x86: Support Generic Initiator only proximity domains
  ...

3 years agoMerge tag 'pm-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Linus Torvalds [Wed, 14 Oct 2020 17:45:41 +0000 (10:45 -0700)]
Merge tag 'pm-5.10-rc1' of git://git./linux/kernel/git/rafael/linux-pm

Pull power management updates from Rafael Wysocki:
 "These rework the collection of cpufreq statistics to allow it to take
  place if fast frequency switching is enabled in the governor, rework
  the frequency invariance handling in the cpufreq core and drivers, add
  new hardware support to a couple of cpufreq drivers, fix a number of
  assorted issues and clean up the code all over.

  Specifics:

   - Rework cpufreq statistics collection to allow it to take place when
     fast frequency switching is enabled in the governor (Viresh Kumar).

   - Make the cpufreq core set the frequency scale on behalf of the
     driver and update several cpufreq drivers accordingly (Ionela
     Voinescu, Valentin Schneider).

   - Add new hardware support to the STI and qcom cpufreq drivers and
     improve them (Alain Volmat, Manivannan Sadhasivam).

   - Fix multiple assorted issues in cpufreq drivers (Jon Hunter,
     Krzysztof Kozlowski, Matthias Kaehlcke, Pali Rohár, Stephan
     Gerhold, Viresh Kumar).

   - Fix several assorted issues in the operating performance points
     (OPP) framework (Stephan Gerhold, Viresh Kumar).

   - Allow devfreq drivers to fetch devfreq instances by DT enumeration
     instead of using explicit phandles and modify the devfreq core code
     to support driver-specific devfreq DT bindings (Leonard Crestez,
     Chanwoo Choi).

   - Improve initial hardware resetting in the tegra30 devfreq driver
     and clean up the tegra cpuidle driver (Dmitry Osipenko).

   - Update the cpuidle core to collect state entry rejection statistics
     and expose them via sysfs (Lina Iyer).

   - Improve the ACPI _CST code handling diagnostics (Chen Yu).

   - Update the PSCI cpuidle driver to allow the PM domain
     initialization to occur in the OSI mode as well as in the PC mode
     (Ulf Hansson).

   - Rework the generic power domains (genpd) core code to allow domain
     power off transition to be aborted in the absence of the "power
     off" domain callback (Ulf Hansson).

   - Fix two suspend-to-idle issues in the ACPI EC driver (Rafael
     Wysocki).

   - Fix the handling of timer_expires in the PM-runtime framework on
     32-bit systems and the handling of device links in it (Grygorii
     Strashko, Xiang Chen).

   - Add IO requests batching support to the hibernate image saving and
     reading code and drop a bogus get_gendisk() from there (Xiaoyi
     Chen, Christoph Hellwig).

   - Allow PCIe ports to be put into the D3cold power state if they are
     power-manageable via ACPI (Lukas Wunner).

   - Add missing header file include to a power capping driver (Pujin
     Shi).

   - Clean up the qcom-cpr AVS driver a bit (Liu Shixin).

   - Kevin Hilman steps down as designated reviwer of adaptive voltage
     scaling (AVS) drivers (Kevin Hilman)"

* tag 'pm-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (65 commits)
  cpufreq: stats: Fix string format specifier mismatch
  arm: disable frequency invariance for CONFIG_BL_SWITCHER
  cpufreq,arm,arm64: restructure definitions of arch_set_freq_scale()
  cpufreq: stats: Add memory barrier to store_reset()
  cpufreq: schedutil: Simplify sugov_fast_switch()
  ACPI: EC: PM: Drop ec_no_wakeup check from acpi_ec_dispatch_gpe()
  ACPI: EC: PM: Flush EC work unconditionally after wakeup
  PCI/ACPI: Whitelist hotplug ports for D3 if power managed by ACPI
  PM: hibernate: remove the bogus call to get_gendisk() in software_resume()
  cpufreq: Move traces and update to policy->cur to cpufreq core
  cpufreq: stats: Enable stats for fast-switch as well
  cpufreq: stats: Mark few conditionals with unlikely()
  cpufreq: stats: Remove locking
  cpufreq: stats: Defer stats update to cpufreq_stats_record_transition()
  PM: domains: Allow to abort power off when no ->power_off() callback
  PM: domains: Rename power state enums for genpd
  PM / devfreq: tegra30: Improve initial hardware resetting
  PM / devfreq: event: Change prototype of devfreq_event_get_edev_by_phandle function
  PM / devfreq: Change prototype of devfreq_get_devfreq_by_phandle function
  PM / devfreq: Add devfreq_get_devfreq_by_node function
  ...

3 years agoMerge tag 'platform-drivers-x86-v5.10-1' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Wed, 14 Oct 2020 17:43:24 +0000 (10:43 -0700)]
Merge tag 'platform-drivers-x86-v5.10-1' of git://git./linux/kernel/git/pdx86/platform-drivers-x86

Pull x86 platform driver updates from Hans de Goede:
 "Rather calm cycle for x86 platform drivers, all these have been in
  for-next for a couple of days with no bot complaints.

  Highlights:

   - PMC TigerLake fixes and new RocketLake support

   - various small fixes / updates in other drivers/tools"

* tag 'platform-drivers-x86-v5.10-1' of git://git.kernel.org/pub/scm/linux/kernel/git/pdx86/platform-drivers-x86:
  MAINTAINERS: update X86 PLATFORM DRIVERS entry with new kernel.org git repo
  platform/x86: mlx-platform: Add capability field to platform FAN description
  platform_data/mlxreg: Extend core platform structure
  platform_data/mlxreg: Update module license
  platform/x86: mlx-platform: Remove PSU EEPROM configuration
  MAINTAINERS: Update maintainers for pmc_core driver
  platform/x86: intel_pmc_core: fix: Replace dev_dbg macro with dev_info()
  platform/x86: intel_pmc_core: Add Intel RocketLake (RKL) support
  platform/x86: intel_pmc_core: Clean up: Remove the duplicate comments and reorganize
  platform/x86: intel_pmc_core: Fix the slp_s0 counter displayed value
  platform/x86: intel_pmc_core: Fix TigerLake power gating status map
  platform/x86: pmc_core: Use descriptive names for LPM registers
  tools/power/x86/intel-speed-select: Update version for v5.10
  tools/power/x86/intel-speed-select: Fix missing base-freq core IDs
  platform/x86: hp-wmi: add support for thermal policy

3 years agoMerge tag 'for-linus-5.10b-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Wed, 14 Oct 2020 17:34:45 +0000 (10:34 -0700)]
Merge tag 'for-linus-5.10b-rc1-tag' of git://git./linux/kernel/git/xen/tip

Pull xen updates from Juergen Gross:

 - two small cleanup patches

 - avoid error messages when initializing MCA banks in a Xen dom0

 - a small series for converting the Xen gntdev driver to use
   pin_user_pages*() instead of get_user_pages*()

 - intermediate fix for running as a Xen guest on Arm with KPTI enabled
   (the final solution will need new Xen functionality)

* tag 'for-linus-5.10b-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
  x86/xen: Fix typo in xen_pagetable_p2m_free()
  x86/xen: disable Firmware First mode for correctable memory errors
  xen/arm: do not setup the runstate info page if kpti is enabled
  xen: remove redundant initialization of variable ret
  xen/gntdev.c: Convert get_user_pages*() to pin_user_pages*()
  xen/gntdev.c: Mark pages as dirty

3 years agoMerge tag 'hyperv-next-signed' of git://git.kernel.org/pub/scm/linux/kernel/git/hyper...
Linus Torvalds [Wed, 14 Oct 2020 17:32:10 +0000 (10:32 -0700)]
Merge tag 'hyperv-next-signed' of git://git./linux/kernel/git/hyperv/linux

Pull Hyper-V updates from Wei Liu:

 - a series from Boqun Feng to support page size larger than 4K

 - a few miscellaneous clean-ups

* tag 'hyperv-next-signed' of git://git.kernel.org/pub/scm/linux/kernel/git/hyperv/linux:
  hv: clocksource: Add notrace attribute to read_hv_sched_clock_*() functions
  x86/hyperv: Remove aliases with X64 in their name
  PCI: hv: Document missing hv_pci_protocol_negotiation() parameter
  scsi: storvsc: Support PAGE_SIZE larger than 4K
  Driver: hv: util: Use VMBUS_RING_SIZE() for ringbuffer sizes
  HID: hyperv: Use VMBUS_RING_SIZE() for ringbuffer sizes
  Input: hyperv-keyboard: Use VMBUS_RING_SIZE() for ringbuffer sizes
  hv_netvsc: Use HV_HYP_PAGE_SIZE for Hyper-V communication
  hv: hyperv.h: Introduce some hvpfn helper functions
  Drivers: hv: vmbus: Move virt_to_hvpfn() to hyperv header
  Drivers: hv: Use HV_HYP_PAGE in hv_synic_enable_regs()
  Drivers: hv: vmbus: Introduce types of GPADL
  Drivers: hv: vmbus: Move __vmbus_open()
  Drivers: hv: vmbus: Always use HV_HYP_PAGE_SIZE for gpadl
  drivers: hv: remove cast from hyperv_die_event

3 years agoMerge tag 'x86_seves_for_v5.10' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Wed, 14 Oct 2020 17:21:34 +0000 (10:21 -0700)]
Merge tag 'x86_seves_for_v5.10' of git://git./linux/kernel/git/tip/tip

Pull x86 SEV-ES support from Borislav Petkov:
 "SEV-ES enhances the current guest memory encryption support called SEV
  by also encrypting the guest register state, making the registers
  inaccessible to the hypervisor by en-/decrypting them on world
  switches. Thus, it adds additional protection to Linux guests against
  exfiltration, control flow and rollback attacks.

  With SEV-ES, the guest is in full control of what registers the
  hypervisor can access. This is provided by a guest-host exchange
  mechanism based on a new exception vector called VMM Communication
  Exception (#VC), a new instruction called VMGEXIT and a shared
  Guest-Host Communication Block which is a decrypted page shared
  between the guest and the hypervisor.

  Intercepts to the hypervisor become #VC exceptions in an SEV-ES guest
  so in order for that exception mechanism to work, the early x86 init
  code needed to be made able to handle exceptions, which, in itself,
  brings a bunch of very nice cleanups and improvements to the early
  boot code like an early page fault handler, allowing for on-demand
  building of the identity mapping. With that, !KASLR configurations do
  not use the EFI page table anymore but switch to a kernel-controlled
  one.

  The main part of this series adds the support for that new exchange
  mechanism. The goal has been to keep this as much as possibly separate
  from the core x86 code by concentrating the machinery in two
  SEV-ES-specific files:

    arch/x86/kernel/sev-es-shared.c
    arch/x86/kernel/sev-es.c

  Other interaction with core x86 code has been kept at minimum and
  behind static keys to minimize the performance impact on !SEV-ES
  setups.

  Work by Joerg Roedel and Thomas Lendacky and others"

* tag 'x86_seves_for_v5.10' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (73 commits)
  x86/sev-es: Use GHCB accessor for setting the MMIO scratch buffer
  x86/sev-es: Check required CPU features for SEV-ES
  x86/efi: Add GHCB mappings when SEV-ES is active
  x86/sev-es: Handle NMI State
  x86/sev-es: Support CPU offline/online
  x86/head/64: Don't call verify_cpu() on starting APs
  x86/smpboot: Load TSS and getcpu GDT entry before loading IDT
  x86/realmode: Setup AP jump table
  x86/realmode: Add SEV-ES specific trampoline entry point
  x86/vmware: Add VMware-specific handling for VMMCALL under SEV-ES
  x86/kvm: Add KVM-specific VMMCALL handling under SEV-ES
  x86/paravirt: Allow hypervisor-specific VMMCALL handling under SEV-ES
  x86/sev-es: Handle #DB Events
  x86/sev-es: Handle #AC Events
  x86/sev-es: Handle VMMCALL Events
  x86/sev-es: Handle MWAIT/MWAITX Events
  x86/sev-es: Handle MONITOR/MONITORX Events
  x86/sev-es: Handle INVD Events
  x86/sev-es: Handle RDPMC Events
  x86/sev-es: Handle RDTSC(P) Events
  ...

3 years agoMerge tag 'objtool-core-2020-10-13' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Wed, 14 Oct 2020 17:13:37 +0000 (10:13 -0700)]
Merge tag 'objtool-core-2020-10-13' of git://git./linux/kernel/git/tip/tip

Pull objtool updates from Ingo Molnar:
 "Most of the changes are cleanups and reorganization to make the
  objtool code more arch-agnostic. This is in preparation for non-x86
  support.

  Other changes:

   - KASAN fixes

   - Handle unreachable trap after call to noreturn functions better

   - Ignore unreachable fake jumps

   - Misc smaller fixes & cleanups"

* tag 'objtool-core-2020-10-13' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (21 commits)
  perf build: Allow nested externs to enable BUILD_BUG() usage
  objtool: Allow nested externs to enable BUILD_BUG()
  objtool: Permit __kasan_check_{read,write} under UACCESS
  objtool: Ignore unreachable trap after call to noreturn functions
  objtool: Handle calling non-function symbols in other sections
  objtool: Ignore unreachable fake jumps
  objtool: Remove useless tests before save_reg()
  objtool: Decode unwind hint register depending on architecture
  objtool: Make unwind hint definitions available to other architectures
  objtool: Only include valid definitions depending on source file type
  objtool: Rename frame.h -> objtool.h
  objtool: Refactor jump table code to support other architectures
  objtool: Make relocation in alternative handling arch dependent
  objtool: Abstract alternative special case handling
  objtool: Move macros describing structures to arch-dependent code
  objtool: Make sync-check consider the target architecture
  objtool: Group headers to check in a single list
  objtool: Define 'struct orc_entry' only when needed
  objtool: Skip ORC entry creation for non-text sections
  objtool: Move ORC logic out of check()
  ...

3 years agoMerge branch 'akpm' (patches from Andrew)
Linus Torvalds [Wed, 14 Oct 2020 16:57:24 +0000 (09:57 -0700)]
Merge branch 'akpm' (patches from Andrew)

Merge misc updates from Andrew Morton:
 "181 patches.

  Subsystems affected by this patch series: kbuild, scripts, ntfs,
  ocfs2, vfs, mm (slab, slub, kmemleak, dax, debug, pagecache, fadvise,
  gup, swap, memremap, memcg, selftests, pagemap, mincore, hmm, dma,
  memory-failure, vmallo and migration)"

* emailed patches from Andrew Morton <akpm@linux-foundation.org>: (181 commits)
  mm/migrate: remove obsolete comment about device public
  mm/migrate: remove cpages-- in migrate_vma_finalize()
  mm, oom_adj: don't loop through tasks in __set_oom_adj when not necessary
  memblock: use separate iterators for memory and reserved regions
  memblock: implement for_each_reserved_mem_region() using __next_mem_region()
  memblock: remove unused memblock_mem_size()
  x86/setup: simplify reserve_crashkernel()
  x86/setup: simplify initrd relocation and reservation
  arch, drivers: replace for_each_membock() with for_each_mem_range()
  arch, mm: replace for_each_memblock() with for_each_mem_pfn_range()
  memblock: reduce number of parameters in for_each_mem_range()
  memblock: make memblock_debug and related functionality private
  memblock: make for_each_memblock_type() iterator private
  mircoblaze: drop unneeded NUMA and sparsemem initializations
  riscv: drop unneeded node initialization
  h8300, nds32, openrisc: simplify detection of memory extents
  arm64: numa: simplify dummy_numa_init()
  arm, xtensa: simplify initialization of high memory pages
  dma-contiguous: simplify cma_early_percent_memory()
  KVM: PPC: Book3S HV: simplify kvm_cma_reserve()
  ...

3 years agodt-bindings: misc: explicitly add #address-cells for slave mode
Zhen Lei [Tue, 13 Oct 2020 16:08:43 +0000 (00:08 +0800)]
dt-bindings: misc: explicitly add #address-cells for slave mode

Explicitly add "#address-cells = <0>" and "#size-cells = <0>" to
eliminate below warnings.

(spi_bus_bridge): /example-0/spi: incorrect #address-cells for SPI bus
(spi_bus_bridge): /example-0/spi: incorrect #size-cells for SPI bus
(spi_bus_reg): Failed prerequisite 'spi_bus_bridge'

Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Link: https://lore.kernel.org/r/20201013160845.1772-5-thunder.leizhen@huawei.com
Signed-off-by: Rob Herring <robh@kernel.org>
3 years agospi: dt-bindings: spi-controller: explicitly require #address-cells=<0> for slave...
Zhen Lei [Tue, 13 Oct 2020 16:08:42 +0000 (00:08 +0800)]
spi: dt-bindings: spi-controller: explicitly require #address-cells=<0> for slave mode

scripts/dtc/checks.c:
if (get_property(node, "spi-slave"))
spi_addr_cells = 0;
if (node_addr_cells(node) != spi_addr_cells)
FAIL(c, dti, node, "incorrect #address-cells for SPI bus");
if (node_size_cells(node) != 0)
FAIL(c, dti, node, "incorrect #size-cells for SPI bus");

The above code in check_spi_bus_bridge() require that the number of address
cells must be 0. So we should explicitly declare "#address-cells = <0>".

Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Link: https://lore.kernel.org/r/20201013160845.1772-4-thunder.leizhen@huawei.com
Signed-off-by: Rob Herring <robh@kernel.org>
3 years agomm/migrate: remove obsolete comment about device public
Ralph Campbell [Tue, 13 Oct 2020 23:58:42 +0000 (16:58 -0700)]
mm/migrate: remove obsolete comment about device public

Device public memory never had an in tree consumer and was removed in
commit 25b2995a35b6 ("mm: remove MEMORY_DEVICE_PUBLIC support").  Delete
the obsolete comment.

Signed-off-by: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Cc: Jerome Glisse <jglisse@redhat.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20200827190735.12752-2-rcampbell@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/migrate: remove cpages-- in migrate_vma_finalize()
Ralph Campbell [Tue, 13 Oct 2020 23:58:39 +0000 (16:58 -0700)]
mm/migrate: remove cpages-- in migrate_vma_finalize()

The variable struct migrate_vma->cpages is only used in
migrate_vma_setup().  There is no need to decrement it in
migrate_vma_finalize() since it is never checked.

Signed-off-by: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Jerome Glisse <jglisse@redhat.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20200827190735.12752-1-rcampbell@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm, oom_adj: don't loop through tasks in __set_oom_adj when not necessary
Suren Baghdasaryan [Tue, 13 Oct 2020 23:58:35 +0000 (16:58 -0700)]
mm, oom_adj: don't loop through tasks in __set_oom_adj when not necessary

Currently __set_oom_adj loops through all processes in the system to keep
oom_score_adj and oom_score_adj_min in sync between processes sharing
their mm.  This is done for any task with more that one mm_users, which
includes processes with multiple threads (sharing mm and signals).
However for such processes the loop is unnecessary because their signal
structure is shared as well.

Android updates oom_score_adj whenever a tasks changes its role
(background/foreground/...) or binds to/unbinds from a service, making it
more/less important.  Such operation can happen frequently.  We noticed
that updates to oom_score_adj became more expensive and after further
investigation found out that the patch mentioned in "Fixes" introduced a
regression.  Using Pixel 4 with a typical Android workload, write time to
oom_score_adj increased from ~3.57us to ~362us.  Moreover this regression
linearly depends on the number of multi-threaded processes running on the
system.

Mark the mm with a new MMF_MULTIPROCESS flag bit when task is created with
(CLONE_VM && !CLONE_THREAD && !CLONE_VFORK).  Change __set_oom_adj to use
MMF_MULTIPROCESS instead of mm_users to decide whether oom_score_adj
update should be synchronized between multiple processes.  To prevent
races between clone() and __set_oom_adj(), when oom_score_adj of the
process being cloned might be modified from userspace, we use
oom_adj_mutex.  Its scope is changed to global.

The combination of (CLONE_VM && !CLONE_THREAD) is rarely used except for
the case of vfork().  To prevent performance regressions of vfork(), we
skip taking oom_adj_mutex and setting MMF_MULTIPROCESS when CLONE_VFORK is
specified.  Clearing the MMF_MULTIPROCESS flag (when the last process
sharing the mm exits) is left out of this patch to keep it simple and
because it is believed that this threading model is rare.  Should there
ever be a need for optimizing that case as well, it can be done by hooking
into the exit path, likely following the mm_update_next_owner pattern.

With the combination of (CLONE_VM && !CLONE_THREAD && !CLONE_VFORK) being
quite rare, the regression is gone after the change is applied.

[surenb@google.com: v3]
Link: https://lkml.kernel.org/r/20200902012558.2335613-1-surenb@google.com
Fixes: 44a70adec910 ("mm, oom_adj: make sure processes sharing mm have same view of oom_score_adj")
Reported-by: Tim Murray <timmurray@google.com>
Suggested-by: Michal Hocko <mhocko@kernel.org>
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Acked-by: Oleg Nesterov <oleg@redhat.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Eugene Syromiatnikov <esyr@redhat.com>
Cc: Christian Kellner <christian@kellner.me>
Cc: Adrian Reber <areber@redhat.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Alexey Gladkov <gladkov.alexey@gmail.com>
Cc: Michel Lespinasse <walken@google.com>
Cc: Daniel Jordan <daniel.m.jordan@oracle.com>
Cc: Andrei Vagin <avagin@gmail.com>
Cc: Bernd Edlinger <bernd.edlinger@hotmail.de>
Cc: John Johansen <john.johansen@canonical.com>
Cc: Yafang Shao <laoar.shao@gmail.com>
Link: https://lkml.kernel.org/r/20200824153036.3201505-1-surenb@google.com
Debugged-by: Minchan Kim <minchan@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomemblock: use separate iterators for memory and reserved regions
Mike Rapoport [Tue, 13 Oct 2020 23:58:30 +0000 (16:58 -0700)]
memblock: use separate iterators for memory and reserved regions

for_each_memblock() is used to iterate over memblock.memory in a few
places that use data from memblock_region rather than the memory ranges.

Introduce separate for_each_mem_region() and
for_each_reserved_mem_region() to improve encapsulation of memblock
internals from its users.

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Acked-by: Ingo Molnar <mingo@kernel.org> [x86]
Acked-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> [MIPS]
Acked-by: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com> [.clang-format]
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-18-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomemblock: implement for_each_reserved_mem_region() using __next_mem_region()
Mike Rapoport [Tue, 13 Oct 2020 23:58:25 +0000 (16:58 -0700)]
memblock: implement for_each_reserved_mem_region() using __next_mem_region()

Iteration over memblock.reserved with for_each_reserved_mem_region() used
__next_reserved_mem_region() that implemented a subset of
__next_mem_region().

Use __for_each_mem_range() and, essentially, __next_mem_region() with
appropriate parameters to reduce code duplication.

While on it, rename for_each_reserved_mem_region() to
for_each_reserved_mem_range() for consistency.

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com> [.clang-format]
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Baoquan He <bhe@redhat.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-17-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomemblock: remove unused memblock_mem_size()
Mike Rapoport [Tue, 13 Oct 2020 23:58:20 +0000 (16:58 -0700)]
memblock: remove unused memblock_mem_size()

The only user of memblock_mem_size() was x86 setup code, it is gone now
and memblock_mem_size() funciton can be removed.

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-16-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agox86/setup: simplify reserve_crashkernel()
Mike Rapoport [Tue, 13 Oct 2020 23:58:16 +0000 (16:58 -0700)]
x86/setup: simplify reserve_crashkernel()

* Replace magic numbers with defines
* Replace memblock_find_in_range() + memblock_reserve() with
  memblock_phys_alloc_range()
* Stop checking for low memory size in reserve_crashkernel_low(). The
  allocation from limited range will anyway fail if there is no enough
  memory, so there is no need for extra traversal of memblock.memory

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Acked-by: Ingo Molnar <mingo@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-15-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agox86/setup: simplify initrd relocation and reservation
Mike Rapoport [Tue, 13 Oct 2020 23:58:12 +0000 (16:58 -0700)]
x86/setup: simplify initrd relocation and reservation

Currently, initrd image is reserved very early during setup and then it
might be relocated and re-reserved after the initial physical memory
mapping is created.  The "late" reservation of memblock verifies that
mapped memory size exceeds the size of initrd, then checks whether the
relocation required and, if yes, relocates inirtd to a new memory
allocated from memblock and frees the old location.

The check for memory size is excessive as memblock allocation will anyway
fail if there is not enough memory.  Besides, there is no point to
allocate memory from memblock using memblock_find_in_range() +
memblock_reserve() when there exists memblock_phys_alloc_range() with
required functionality.

Remove the redundant check and simplify memblock allocation.

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Acked-by: Ingo Molnar <mingo@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-14-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoarch, drivers: replace for_each_membock() with for_each_mem_range()
Mike Rapoport [Tue, 13 Oct 2020 23:58:08 +0000 (16:58 -0700)]
arch, drivers: replace for_each_membock() with for_each_mem_range()

There are several occurrences of the following pattern:

for_each_memblock(memory, reg) {
start = __pfn_to_phys(memblock_region_memory_base_pfn(reg);
end = __pfn_to_phys(memblock_region_memory_end_pfn(reg));

/* do something with start and end */
}

Using for_each_mem_range() iterator is more appropriate in such cases and
allows simpler and cleaner code.

[akpm@linux-foundation.org: fix arch/arm/mm/pmsa-v7.c build]
[rppt@linux.ibm.com: mips: fix cavium-octeon build caused by memblock refactoring]
Link: http://lkml.kernel.org/r/20200827124549.GD167163@linux.ibm.com
Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Baoquan He <bhe@redhat.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-13-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoarch, mm: replace for_each_memblock() with for_each_mem_pfn_range()
Mike Rapoport [Tue, 13 Oct 2020 23:58:03 +0000 (16:58 -0700)]
arch, mm: replace for_each_memblock() with for_each_mem_pfn_range()

There are several occurrences of the following pattern:

for_each_memblock(memory, reg) {
start_pfn = memblock_region_memory_base_pfn(reg);
end_pfn = memblock_region_memory_end_pfn(reg);

/* do something with start_pfn and end_pfn */
}

Rather than iterate over all memblock.memory regions and each time query
for their start and end PFNs, use for_each_mem_pfn_range() iterator to get
simpler and clearer code.

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Acked-by: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com> [.clang-format]
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-12-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomemblock: reduce number of parameters in for_each_mem_range()
Mike Rapoport [Tue, 13 Oct 2020 23:57:59 +0000 (16:57 -0700)]
memblock: reduce number of parameters in for_each_mem_range()

Currently for_each_mem_range() and for_each_mem_range_rev() iterators are
the most generic way to traverse memblock regions.  As such, they have 8
parameters and they are hardly convenient to users.  Most users choose to
utilize one of their wrappers and the only user that actually needs most
of the parameters is memblock itself.

To avoid yet another naming for memblock iterators, rename the existing
for_each_mem_range[_rev]() to __for_each_mem_range[_rev]() and add a new
for_each_mem_range[_rev]() wrappers with only index, start and end
parameters.

The new wrapper nicely fits into init_unavailable_mem() and will be used
in upcoming changes to simplify memblock traversals.

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> [MIPS]
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Baoquan He <bhe@redhat.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-11-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomemblock: make memblock_debug and related functionality private
Mike Rapoport [Tue, 13 Oct 2020 23:57:54 +0000 (16:57 -0700)]
memblock: make memblock_debug and related functionality private

The only user of memblock_dbg() outside memblock was s390 setup code and
it is converted to use pr_debug() instead.  This allows to stop exposing
memblock_debug and memblock_dbg() to the rest of the kernel.

[akpm@linux-foundation.org: make memblock_dbg() safer and neater]

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-10-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomemblock: make for_each_memblock_type() iterator private
Mike Rapoport [Tue, 13 Oct 2020 23:57:49 +0000 (16:57 -0700)]
memblock: make for_each_memblock_type() iterator private

for_each_memblock_type() is not used outside mm/memblock.c, move it there
from include/linux/memblock.h

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-9-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomircoblaze: drop unneeded NUMA and sparsemem initializations
Mike Rapoport [Tue, 13 Oct 2020 23:57:45 +0000 (16:57 -0700)]
mircoblaze: drop unneeded NUMA and sparsemem initializations

microblaze does not support neither NUMA not SPARSMEM, so there is no
point to call memblock_set_node() and
sparse_memory_present_with_active_regions() functions during microblaze
memory initialization.

Remove these calls and the surrounding code.

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Baoquan He <bhe@redhat.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-8-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoriscv: drop unneeded node initialization
Mike Rapoport [Tue, 13 Oct 2020 23:57:40 +0000 (16:57 -0700)]
riscv: drop unneeded node initialization

RISC-V does not (yet) support NUMA and for UMA architectures node 0 is
used implicitly during early memory initialization.

There is no need to call memblock_set_node(), remove this call and the
surrounding code.

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Baoquan He <bhe@redhat.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-7-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoh8300, nds32, openrisc: simplify detection of memory extents
Mike Rapoport [Tue, 13 Oct 2020 23:57:36 +0000 (16:57 -0700)]
h8300, nds32, openrisc: simplify detection of memory extents

Instead of traversing memblock.memory regions to find memory_start and
memory_end, simply query memblock_{start,end}_of_DRAM().

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Stafford Horne <shorne@gmail.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Baoquan He <bhe@redhat.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-6-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoarm64: numa: simplify dummy_numa_init()
Mike Rapoport [Tue, 13 Oct 2020 23:57:31 +0000 (16:57 -0700)]
arm64: numa: simplify dummy_numa_init()

dummy_numa_init() loops over memblock.memory and passes nid=0 to
numa_add_memblk() which essentially wraps memblock_set_node().  However,
memblock_set_node() can cope with entire memory span itself, so the loop
over memblock.memory regions is redundant.

Using a single call to memblock_set_node() rather than a loop also fixes
an issue with a buggy ACPI firmware in which the SRAT table covers some
but not all of the memory in the EFI memory map.

Jonathan Cameron says:

  This issue can be easily triggered by having an SRAT table which fails
  to cover all elements of the EFI memory map.

  This firmware error is detected and a warning printed. e.g.
  "NUMA: Warning: invalid memblk node 64 [mem 0x240000000-0x27fffffff]"
  At that point we fall back to dummy_numa_init().

  However, the failed ACPI init has left us with our memblocks all broken
  up as we split them when trying to assign them to NUMA nodes.

  We then iterate over the memblocks and add them to node 0.

  numa_add_memblk() calls memblock_set_node() which merges regions that
  were previously split up during the earlier attempt to add them to
  different nodes during parsing of SRAT.

  This means elements are moved in the memblock array and we can end up
  in a different memblock after the call to numa_add_memblk().
  Result is:

  Unable to handle kernel paging request at virtual address 0000000000003a40
  Mem abort info:
    ESR = 0x96000004
    EC = 0x25: DABT (current EL), IL = 32 bits
    SET = 0, FnV = 0
    EA = 0, S1PTW = 0
  Data abort info:
    ISV = 0, ISS = 0x00000004
    CM = 0, WnR = 0
  [0000000000003a40] user address but active_mm is swapper
  Internal error: Oops: 96000004 [#1] PREEMPT SMP

  ...

  Call trace:
    sparse_init_nid+0x5c/0x2b0
    sparse_init+0x138/0x170
    bootmem_init+0x80/0xe0
    setup_arch+0x2a0/0x5fc
    start_kernel+0x8c/0x648

Replace the loop with a single call to memblock_set_node() to the entire
memory.

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Baoquan He <bhe@redhat.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-5-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoarm, xtensa: simplify initialization of high memory pages
Mike Rapoport [Tue, 13 Oct 2020 23:57:26 +0000 (16:57 -0700)]
arm, xtensa: simplify initialization of high memory pages

free_highpages() in both arm and xtensa essentially open-code
for_each_free_mem_range() loop to detect high memory pages that were not
reserved and that should be initialized and passed to the buddy allocator.

Replace open-coded implementation of for_each_free_mem_range() with usage
of memblock API to simplify the code.

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Tested-by: Max Filippov <jcmvbkbc@gmail.com> [xtensa]
Reviewed-by: Max Filippov <jcmvbkbc@gmail.com> [xtensa]
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Baoquan He <bhe@redhat.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-4-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agodma-contiguous: simplify cma_early_percent_memory()
Mike Rapoport [Tue, 13 Oct 2020 23:57:22 +0000 (16:57 -0700)]
dma-contiguous: simplify cma_early_percent_memory()

The memory size calculation in cma_early_percent_memory() traverses
memblock.memory rather than simply call memblock_phys_mem_size().  The
comment in that function suggests that at some point there should have
been call to memblock_analyze() before memblock_phys_mem_size() could be
used.  As of now, there is no memblock_analyze() at all and
memblock_phys_mem_size() can be used as soon as cold-plug memory is
registered with memblock.

Replace loop over memblock.memory with a call to memblock_phys_mem_size().

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Baoquan He <bhe@redhat.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: https://lkml.kernel.org/r/20200818151634.14343-3-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoKVM: PPC: Book3S HV: simplify kvm_cma_reserve()
Mike Rapoport [Tue, 13 Oct 2020 23:57:17 +0000 (16:57 -0700)]
KVM: PPC: Book3S HV: simplify kvm_cma_reserve()

Patch series "memblock: seasonal cleaning^w cleanup", v3.

These patches simplify several uses of memblock iterators and hide some of
the memblock implementation details from the rest of the system.

This patch (of 17):

The memory size calculation in kvm_cma_reserve() traverses memblock.memory
rather than simply call memblock_phys_mem_size().  The comment in that
function suggests that at some point there should have been call to
memblock_analyze() before memblock_phys_mem_size() could be used.  As of
now, there is no memblock_analyze() at all and memblock_phys_mem_size()
can be used as soon as cold-plug memory is registered with memblock.

Replace loop over memblock.memory with a call to memblock_phys_mem_size().

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Baoquan He <bhe@redhat.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Axtens <dja@axtens.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Emil Renner Berthing <kernel@esmil.dk>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Link: https://lkml.kernel.org/r/20200818151634.14343-1-rppt@kernel.org
Link: https://lkml.kernel.org/r/20200818151634.14343-2-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/mempool: add 'else' to split mutually exclusive case
Miaohe Lin [Tue, 13 Oct 2020 23:57:14 +0000 (16:57 -0700)]
mm/mempool: add 'else' to split mutually exclusive case

Add else to split mutually exclusive case and avoid some unnecessary check.
It doesn't seem to change code generation (compiler is smart), but I think
it helps readability.

[akpm@linux-foundation.org: fix comment location]

Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Link: https://lkml.kernel.org/r/20200924111641.28922-1-linmiaohe@huawei.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm: remove unused alloc_page_vma_node()
Wei Yang [Tue, 13 Oct 2020 23:57:11 +0000 (16:57 -0700)]
mm: remove unused alloc_page_vma_node()

No one use this macro anymore.

Also fix code style of policy_node().

Signed-off-by: Wei Yang <richard.weiyang@linux.alibaba.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Link: https://lkml.kernel.org/r/20200921021401.84508-1-richard.weiyang@linux.alibaba.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/mempolicy: remove or narrow the lock on current
Wei Yang [Tue, 13 Oct 2020 23:57:08 +0000 (16:57 -0700)]
mm/mempolicy: remove or narrow the lock on current

It is not necessary to hold the lock of current when setting nodemask of
a new policy.

Signed-off-by: Wei Yang <richard.weiyang@linux.alibaba.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Link: https://lkml.kernel.org/r/20200921040416.86185-1-richard.weiyang@linux.alibaba.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoselftests/vm: 8x compaction_test speedup
John Hubbard [Tue, 13 Oct 2020 23:57:04 +0000 (16:57 -0700)]
selftests/vm: 8x compaction_test speedup

This patch reduces the running time for compaction_test from about 27 sec,
to 3.3 sec, which is about an 8x speedup.

These numbers are for an Intel x86_64 system with 32 GB of DRAM.

The compaction_test.c program was spending most of its time doing mmap(),
1 MB at a time, on about 25 GB of memory.

Instead, do the mmaps 100 MB at a time.  (Going past 100 MB doesn't make
things go much faster, because other parts of the program are using the
remaining time.)

Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Sri Jayaramappa <sjayaram@akamai.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Mel Gorman <mgorman@techsingularity.net>
Link: https://lkml.kernel.org/r/20201002080621.551044-2-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoinclude/linux/compaction.h: clean code by removing unused enum value
Mateusz Nosek [Tue, 13 Oct 2020 23:57:01 +0000 (16:57 -0700)]
include/linux/compaction.h: clean code by removing unused enum value

The enum value 'COMPACT_INACTIVE' is never used so can be removed.

Signed-off-by: Mateusz Nosek <mateusznosek0@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Link: https://lkml.kernel.org/r/20200917110750.12015-1-mateusznosek0@gmail.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/compaction.c: micro-optimization remove unnecessary branch
Mateusz Nosek [Tue, 13 Oct 2020 23:56:58 +0000 (16:56 -0700)]
mm/compaction.c: micro-optimization remove unnecessary branch

The same code can work both for 'zone->compact_considered > defer_limit'
and 'zone->compact_considered >= defer_limit'.  In the latter there is one
branch less which is more effective considering performance.

Signed-off-by: Mateusz Nosek <mateusznosek0@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Mel Gorman <mgorman@suse.de>
Cc: David Rientjes <rientjes@google.com>
Link: https://lkml.kernel.org/r/20200913190448.28649-1-mateusznosek0@gmail.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/zbud: remove redundant initialization
Xiang Chen [Tue, 13 Oct 2020 23:56:55 +0000 (16:56 -0700)]
mm/zbud: remove redundant initialization

zhdr is already initialized in the front of the function, so remove
redundant initialization here.

Signed-off-by: Xiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: David Hildenbrand <david@redhat.com>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Link: https://lkml.kernel.org/r/1600419885-191907-1-git-send-email-chenxiang66@hisilicon.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/z3fold.c: use xx_zalloc instead xx_alloc and memset
Hui Su [Tue, 13 Oct 2020 23:56:52 +0000 (16:56 -0700)]
mm/z3fold.c: use xx_zalloc instead xx_alloc and memset

alloc_slots() allocates memory for slots using kmem_cache_alloc(), then
memsets it.  We can just use kmem_cache_zalloc().

Signed-off-by: Hui Su <sh_def@163.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Link: https://lkml.kernel.org/r/20200926100834.GA184671@rlk
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/vmscan: fix comments for isolate_lru_page()
Hui Su [Tue, 13 Oct 2020 23:56:49 +0000 (16:56 -0700)]
mm/vmscan: fix comments for isolate_lru_page()

fix comments for isolate_lru_page():
s/fundamentnal/fundamental

Signed-off-by: Hui Su <sh_def@163.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Link: https://lkml.kernel.org/r/20200927173923.GA8058@rlk
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/vmscan: fix infinite loop in drop_slab_node
Chunxin Zang [Tue, 13 Oct 2020 23:56:46 +0000 (16:56 -0700)]
mm/vmscan: fix infinite loop in drop_slab_node

We have observed that drop_caches can take a considerable amount of
time (<put data here>).  Especially when there are many memcgs involved
because they are adding an additional overhead.

It is quite unfortunate that the operation cannot be interrupted by a
signal currently.  Add a check for fatal signals into the main loop so
that userspace can control early bailout.

There are two reasons:

1. We have too many memcgs, even though one object freed in one memcg,
   the sum of object is bigger than 10.

2. We spend a lot of time in traverse memcg once.  So, the memcg who
   traversed at the first have been freed many objects.  Traverse memcg
   next time, the freed count bigger than 10 again.

We can get the following info through 'ps':

  root:~# ps -aux | grep drop
  root  357956 ... R    Aug25 21119854:55 echo 3 > /proc/sys/vm/drop_caches
  root 1771385 ... R    Aug16 21146421:17 echo 3 > /proc/sys/vm/drop_caches
  root 1986319 ... R    18:56 117:27 echo 3 > /proc/sys/vm/drop_caches
  root 2002148 ... R    Aug24 5720:39 echo 3 > /proc/sys/vm/drop_caches
  root 2564666 ... R    18:59 113:58 echo 3 > /proc/sys/vm/drop_caches
  root 2639347 ... R    Sep03 2383:39 echo 3 > /proc/sys/vm/drop_caches
  root 3904747 ... R    03:35 993:31 echo 3 > /proc/sys/vm/drop_caches
  root 4016780 ... R    Aug21 7882:18 echo 3 > /proc/sys/vm/drop_caches

Use bpftrace follow 'freed' value in drop_slab_node:

  root:~# bpftrace -e 'kprobe:drop_slab_node+70 {@ret=hist(reg("bp")); }'
  Attaching 1 probe...
  ^B^C

  @ret:
  [64, 128)        1 |                                                    |
  [128, 256)      28 |                                                    |
  [256, 512)     107 |@                                                   |
  [512, 1K)      298 |@@@                                                 |
  [1K, 2K)       613 |@@@@@@@                                             |
  [2K, 4K)      4435 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@|
  [4K, 8K)       442 |@@@@@                                               |
  [8K, 16K)      299 |@@@                                                 |
  [16K, 32K)     100 |@                                                   |
  [32K, 64K)     139 |@                                                   |
  [64K, 128K)     56 |                                                    |
  [128K, 256K)    26 |                                                    |
  [256K, 512K)     2 |                                                    |

In the while loop, we can check whether the TASK_KILLABLE signal is set,
if so, we should break the loop.

Signed-off-by: Chunxin Zang <zangchunxin@bytedance.com>
Signed-off-by: Muchun Song <songmuchun@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Chris Down <chris@chrisdown.name>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Matthew Wilcox <willy@infradead.org>
Link: https://lkml.kernel.org/r/20200909152047.27905-1-zangchunxin@bytedance.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agohugetlb: add lockdep check for i_mmap_rwsem held in huge_pmd_share
Mike Kravetz [Tue, 13 Oct 2020 23:56:42 +0000 (16:56 -0700)]
hugetlb: add lockdep check for i_mmap_rwsem held in huge_pmd_share

As a debugging aid, huge_pmd_share should make sure i_mmap_rwsem is held
if necessary.  To clarify the 'if necessary', expand the comment block at
the beginning of huge_pmd_share.

No functional change.  The added i_mmap_assert_locked() call is only
enabled if CONFIG_LOCKDEP.

Ideally, this should have been included with commit 34ae204f1851
("hugetlbfs: remove call to huge_pte_alloc without i_mmap_rwsem").

Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: "Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Link: https://lkml.kernel.org/r/20200911201248.88537-1-mike.kravetz@oracle.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/hugetlb: take the free hpage during the iteration directly
Wei Yang [Tue, 13 Oct 2020 23:56:39 +0000 (16:56 -0700)]
mm/hugetlb: take the free hpage during the iteration directly

Function dequeue_huge_page_node_exact() iterates the free list and return
the first valid free hpage.

Instead of break and check the loop variant, we could return in the loop
directly.  This could reduce some redundant check.

[mike.kravetz@oracle.com: points out a logic error]
[richard.weiyang@linux.alibaba.com: v4]
Link: https://lkml.kernel.org/r/20200901014636.29737-8-richard.weiyang@linux.alibaba.com
Signed-off-by: Wei Yang <richard.weiyang@linux.alibaba.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Baoquan He <bhe@redhat.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: https://lkml.kernel.org/r/20200831022351.20916-8-richard.weiyang@linux.alibaba.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/hugetlb: narrow the hugetlb_lock protection area during preparing huge page
Wei Yang [Tue, 13 Oct 2020 23:56:36 +0000 (16:56 -0700)]
mm/hugetlb: narrow the hugetlb_lock protection area during preparing huge page

set_hugetlb_cgroup_[rsvd] just manipulate page local data, which is not
necessary to be protected by hugetlb_lock.

Let's take this out.

Signed-off-by: Wei Yang <richard.weiyang@linux.alibaba.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: https://lkml.kernel.org/r/20200831022351.20916-7-richard.weiyang@linux.alibaba.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/hugetlb: a page from buddy is not on any list
Wei Yang [Tue, 13 Oct 2020 23:56:33 +0000 (16:56 -0700)]
mm/hugetlb: a page from buddy is not on any list

The page allocated from buddy is not on any list, so just use list_add()
is enough.

Signed-off-by: Wei Yang <richard.weiyang@linux.alibaba.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: https://lkml.kernel.org/r/20200831022351.20916-6-richard.weiyang@linux.alibaba.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/hugetlb: count file_region to be added when regions_needed != NULL
Wei Yang [Tue, 13 Oct 2020 23:56:30 +0000 (16:56 -0700)]
mm/hugetlb: count file_region to be added when regions_needed != NULL

There are only two cases of function add_reservation_in_range()

    * count file_region and return the number in regions_needed
    * do the real list operation without counting

This means it is not necessary to have two parameters to classify these
two cases.

Just use regions_needed to separate them.

Signed-off-by: Wei Yang <richard.weiyang@linux.alibaba.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: https://lkml.kernel.org/r/20200831022351.20916-5-richard.weiyang@linux.alibaba.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/hugetlb: use list_splice to merge two list at once
Wei Yang [Tue, 13 Oct 2020 23:56:27 +0000 (16:56 -0700)]
mm/hugetlb: use list_splice to merge two list at once

Instead of add allocated file_region one by one to region_cache, we could
use list_splice to merge two list at once.

Also we know the number of entries in the list, increase the number
directly.

Signed-off-by: Wei Yang <richard.weiyang@linux.alibaba.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: https://lkml.kernel.org/r/20200831022351.20916-4-richard.weiyang@linux.alibaba.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/hugetlb: remove VM_BUG_ON(!nrg) in get_file_region_entry_from_cache()
Wei Yang [Tue, 13 Oct 2020 23:56:23 +0000 (16:56 -0700)]
mm/hugetlb: remove VM_BUG_ON(!nrg) in get_file_region_entry_from_cache()

We are sure to get a valid file_region, otherwise the
VM_BUG_ON(resv->region_cache_count <= 0) at the very beginning would be
triggered.

Let's remove the redundant one.

Signed-off-by: Wei Yang <richard.weiyang@linux.alibaba.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Baoquan He <bhe@redhat.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: https://lkml.kernel.org/r/20200831022351.20916-3-richard.weiyang@linux.alibaba.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/hugetlb: not necessary to coalesce regions recursively
Wei Yang [Tue, 13 Oct 2020 23:56:20 +0000 (16:56 -0700)]
mm/hugetlb: not necessary to coalesce regions recursively

Patch series "mm/hugetlb: code refine and simplification", v4.

Following are some cleanups for hugetlb.  Simple testing with
tools/testing/selftests/vm/map_hugetlb passes.

This patch (of 7):

Per my understanding, we keep the regions ordered and would always
coalesce regions properly.  So the task to keep this property is just to
coalesce its neighbour.

Let's simplify this.

Signed-off-by: Wei Yang <richard.weiyang@linux.alibaba.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: https://lkml.kernel.org/r/20200901014636.29737-1-richard.weiyang@linux.alibaba.com
Link: https://lkml.kernel.org/r/20200831022351.20916-1-richard.weiyang@linux.alibaba.com
Link: https://lkml.kernel.org/r/20200831022351.20916-2-richard.weiyang@linux.alibaba.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agodoc/vm: fix typo in the hugetlb admin documentation
Baoquan He [Tue, 13 Oct 2020 23:56:17 +0000 (16:56 -0700)]
doc/vm: fix typo in the hugetlb admin documentation

Change 'pecify' to 'Specify'.

Signed-off-by: Baoquan He <bhe@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Cc: Anshuman Khandual <anshuman.khandual@arm.com>
Link: https://lkml.kernel.org/r/20200723032248.24772-4-bhe@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/hugetlb.c: remove the unnecessary non_swap_entry()
Baoquan He [Tue, 13 Oct 2020 23:56:14 +0000 (16:56 -0700)]
mm/hugetlb.c: remove the unnecessary non_swap_entry()

If a swap entry tests positive for either is_[migration|hwpoison]_entry(),
then its swap_type() is among SWP_MIGRATION_READ, SWP_MIGRATION_WRITE and
SWP_HWPOISON.  All these types >= MAX_SWAPFILES, exactly what is asserted
with non_swap_entry().

So the checking non_swap_entry() in is_hugetlb_entry_migration() and
is_hugetlb_entry_hwpoisoned() is redundant.

Let's remove it to optimize code.

Signed-off-by: Baoquan He <bhe@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Link: https://lkml.kernel.org/r/20200723032248.24772-3-bhe@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/hugetlb.c: make is_hugetlb_entry_hwpoisoned return bool
Baoquan He [Tue, 13 Oct 2020 23:56:10 +0000 (16:56 -0700)]
mm/hugetlb.c: make is_hugetlb_entry_hwpoisoned return bool

Patch series "mm/hugetlb: Small cleanup and improvement", v2.

This patch (of 3):

Just like its neighbour is_hugetlb_entry_migration() has done.

Signed-off-by: Baoquan He <bhe@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Link: https://lkml.kernel.org/r/20200723032248.24772-1-bhe@redhat.com
Link: https://lkml.kernel.org/r/20200723032248.24772-2-bhe@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoinclude/linux/gfp.h: clarify usage of GFP_ATOMIC in !preemptible contexts
Michal Hocko [Tue, 13 Oct 2020 23:56:07 +0000 (16:56 -0700)]
include/linux/gfp.h: clarify usage of GFP_ATOMIC in !preemptible contexts

There is a general understanding that GFP_ATOMIC/GFP_NOWAIT are to be used
from atomic contexts.  E.g.  from within a spin lock or from the IRQ
context.  This is correct but there are some atomic contexts where the
above doesn't hold.  One of them would be an NMI context.  Page allocator
has never supported that and the general fear of this context didn't let
anybody to actually even try to use the allocator there.  Good, but let's
be more specific about that.

Another such a context, and that is where people seem to be more daring,
is raw_spin_lock.  Mostly because it simply resembles regular spin lock
which is supported by the allocator and there is not any implementation
difference with !RT kernels in the first place.  Be explicit that such a
context is not supported by the allocator.  The underlying reason is that
zone->lock would have to become raw_spin_lock as well and that has turned
out to be a problem for RT
(http://lkml.kernel.org/r/87mu305c1w.fsf@nanos.tec.linutronix.de).

Signed-off-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Uladzislau Rezki <urezki@gmail.com>
Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Link: https://lkml.kernel.org/r/20200929123010.5137-1-mhocko@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/page_alloc.c: fix freeing non-compound pages
Matthew Wilcox (Oracle) [Tue, 13 Oct 2020 23:56:04 +0000 (16:56 -0700)]
mm/page_alloc.c: fix freeing non-compound pages

Here is a very rare race which leaks memory:

Page P0 is allocated to the page cache.  Page P1 is free.

Thread A                Thread B                Thread C
find_get_entry():
xas_load() returns P0
Removes P0 from page cache
P0 finds its buddy P1
alloc_pages(GFP_KERNEL, 1) returns P0
P0 has refcount 1
page_cache_get_speculative(P0)
P0 has refcount 2
__free_pages(P0)
P0 has refcount 1
put_page(P0)
P1 is not freed

Fix this by freeing all the pages in __free_pages() that won't be freed
by the call to put_page().  It's usually not a good idea to split a page,
but this is a very unlikely scenario.

Fixes: e286781d5f2e ("mm: speculative page references")
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Mike Rapoport <rppt@linux.ibm.com>
Cc: Nick Piggin <npiggin@gmail.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200926213919.26642-1-willy@infradead.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm: move call to compound_head() in release_pages()
Ralph Campbell [Tue, 13 Oct 2020 23:56:00 +0000 (16:56 -0700)]
mm: move call to compound_head() in release_pages()

The function is_huge_zero_page() doesn't call compound_head() to make sure
the page pointer is a head page. The call to is_huge_zero_page() in
release_pages() is made before compound_head() is called so the test would
fail if release_pages() was called with a tail page of the huge_zero_page
and put_page_testzero() would be called releasing the page.
This is unlikely to be happening in normal use or we would be seeing all
sorts of process data corruption when accessing a THP zero page.

Looking at other places where is_huge_zero_page() is called, all seem to
only pass a head page so I think the right solution is to move the call
to compound_head() in release_pages() to a point before calling
is_huge_zero_page().

Signed-off-by: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Yu Zhao <yuzhao@google.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: https://lkml.kernel.org/r/20200917173938.16420-1-rcampbell@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agommzone: clean code by removing unused macro parameter
Mateusz Nosek [Tue, 13 Oct 2020 23:55:57 +0000 (16:55 -0700)]
mmzone: clean code by removing unused macro parameter

Previously 'for_next_zone_zonelist_nodemask' macro parameter 'zlist' was
unused so this patch removes it.

Signed-off-by: Mateusz Nosek <mateusznosek0@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Link: https://lkml.kernel.org/r/20200917211906.30059-1-mateusznosek0@gmail.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/page_alloc.c: __perform_reclaim should return 'unsigned long'
Yanfei Xu [Tue, 13 Oct 2020 23:55:54 +0000 (16:55 -0700)]
mm/page_alloc.c: __perform_reclaim should return 'unsigned long'

__perform_reclaim()'s single caller expects it to return 'unsigned long',
hence change its return value and a local variable to 'unsigned long'.

Suggested-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Yanfei Xu <yanfei.xu@windriver.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Link: https://lkml.kernel.org/r/20200916022138.16740-1-yanfei.xu@windriver.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/page_alloc.c: clean code by merging two functions
Mateusz Nosek [Tue, 13 Oct 2020 23:55:51 +0000 (16:55 -0700)]
mm/page_alloc.c: clean code by merging two functions

finalise_ac() is just 'epilogue' for 'prepare_alloc_pages'.  Therefore
there is no need to keep them both so 'finalise_ac' content can be merged
into prepare_alloc_pages() code.  It would make __alloc_pages_nodemask()
cleaner when it comes to readability.

Signed-off-by: Mateusz Nosek <mateusznosek0@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Mike Rapoport <rppt@kernel.org>
Link: https://lkml.kernel.org/r/20200916110118.6537-1-mateusznosek0@gmail.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/page_alloc.c: fix early params garbage value accesses
Mateusz Nosek [Tue, 13 Oct 2020 23:55:48 +0000 (16:55 -0700)]
mm/page_alloc.c: fix early params garbage value accesses

Previously in '__init early_init_on_alloc' and '__init early_init_on_free'
the return values from 'kstrtobool' were not handled properly.  That
caused potential garbage value read from variable 'bool_result'.
Introduced patch fixes error handling.

Signed-off-by: Mateusz Nosek <mateusznosek0@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Link: https://lkml.kernel.org/r/20200916214125.28271-1-mateusznosek0@gmail.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/page_alloc.c: micro-optimization remove unnecessary branch
Mateusz Nosek [Tue, 13 Oct 2020 23:55:45 +0000 (16:55 -0700)]
mm/page_alloc.c: micro-optimization remove unnecessary branch

Previously flags check was separated into two separated checks with two
separated branches. In case of presence of any of two mentioned flags,
the same effect on flow occurs. Therefore checks can be merged and one
branch can be avoided.

Signed-off-by: Mateusz Nosek <mateusznosek0@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Link: https://lkml.kernel.org/r/20200911092310.31136-1-mateusznosek0@gmail.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/page_alloc.c: clean code by removing unnecessary initialization
Mateusz Nosek [Tue, 13 Oct 2020 23:55:42 +0000 (16:55 -0700)]
mm/page_alloc.c: clean code by removing unnecessary initialization

Previously variable 'tmp' was initialized, but was not read later before
reassigning.  So the initialization can be removed.

[akpm@linux-foundation.org: remove `tmp' altogether]

Signed-off-by: Mateusz Nosek <mateusznosek0@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Link: https://lkml.kernel.org/r/20200904132422.17387-1-mateusznosek0@gmail.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm, isolation: avoid checking unmovable pages across pageblock boundary
Li Xinhai [Tue, 13 Oct 2020 23:55:39 +0000 (16:55 -0700)]
mm, isolation: avoid checking unmovable pages across pageblock boundary

In has_unmovable_pages(), the page parameter would not always be the first
page within a pageblock (see how the page pointer is passed in from
start_isolate_page_range() after call __first_valid_page()), so that would
cause checking unmovable pages span two pageblocks.

After this patch, the checking is enforced within one pageblock no matter
the page is first one or not, and obey the semantics of this function.

This issue is found by code inspection.

Michal said "this might lead to false negatives when an unrelated block
would cause an isolation failure".

Signed-off-by: Li Xinhai <lixinhai.lxh@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: David Hildenbrand <david@redhat.com>
Link: https://lkml.kernel.org/r/20200824065811.383266-1-lixinhai.lxh@gmail.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm: document semantics of ZONE_MOVABLE
David Hildenbrand [Tue, 13 Oct 2020 23:55:35 +0000 (16:55 -0700)]
mm: document semantics of ZONE_MOVABLE

Let's document what ZONE_MOVABLE means, how it's used, and which special
cases we have regarding unmovable pages (memory offlining vs.  migration /
allocations).

Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Mike Rapoport <rppt@linux.ibm.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Pankaj Gupta <pankaj.gupta.linux@gmail.com>
Cc: Baoquan He <bhe@redhat.com>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Qian Cai <cai@lca.pw>
Link: http://lkml.kernel.org/r/20200816125333.7434-7-david@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agovirtio-mem: don't special-case ZONE_MOVABLE
David Hildenbrand [Tue, 13 Oct 2020 23:55:31 +0000 (16:55 -0700)]
virtio-mem: don't special-case ZONE_MOVABLE

When introducing virtio-mem, the semantics of ZONE_MOVABLE were rather
unclear, which is why we special-cased ZONE_MOVABLE such that partially
plugged blocks would never end up in ZONE_MOVABLE.

Now that the semantics are much clearer (and will be documented in a
follow-up patch including the new virtio-mem behavior), let's allow to
online partially plugged memory blocks to ZONE_MOVABLE and also consider
memory blocks that were onlined to ZONE_MOVABLE when unplugging memory.
While unplugged memory pages are, in general, unmovable, they can be
skipped when offlining memory.

virtio-mem only unplugs fairly big chunks (in the megabyte range) and
rather tries to shrink the memory region than randomly choosing memory.
In theory, if all other pages in the movable zone would be movable,
virtio-mem would only shrink that zone and not create any kind of
fragmentation.

In the future, we might want to remember the zone again and use the
information when (un)plugging memory.  For now, let's keep it simple.

Note: Support for defragmentation is planned, to deal with fragmentation
after unplug due to memory chunks within memory blocks that could not get
unplugged before (e.g., somebody pinning pages within ZONE_MOVABLE for a
longer time).

Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Pankaj Gupta <pankaj.gupta.linux@gmail.com>
Cc: Baoquan He <bhe@redhat.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Qian Cai <cai@lca.pw>
Link: http://lkml.kernel.org/r/20200816125333.7434-6-david@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/page_isolation: cleanup set_migratetype_isolate()
David Hildenbrand [Tue, 13 Oct 2020 23:55:28 +0000 (16:55 -0700)]
mm/page_isolation: cleanup set_migratetype_isolate()

Let's clean it up a bit, simplifying the exit paths.

Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Reviewed-by: Pankaj Gupta <pankaj.gupta.linux@gmail.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Qian Cai <cai@lca.pw>
Link: http://lkml.kernel.org/r/20200816125333.7434-5-david@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/page_isolation: drop WARN_ON_ONCE() in set_migratetype_isolate()
David Hildenbrand [Tue, 13 Oct 2020 23:55:24 +0000 (16:55 -0700)]
mm/page_isolation: drop WARN_ON_ONCE() in set_migratetype_isolate()

Inside has_unmovable_pages(), we have a comment describing how unmovable
data could end up in ZONE_MOVABLE - via "movablecore".  Also, besides
checking if the first page in the pageblock is reserved, we don't perform
any further checks in case of ZONE_MOVABLE.

In case of memory offlining, we set REPORT_FAILURE, properly dump_page()
the page and handle the error gracefully.  alloc_contig_pages() users
currently never allocate from ZONE_MOVABLE.  E.g., hugetlb uses
alloc_contig_pages() for the allocation of gigantic pages only, which will
never end up on the MOVABLE zone (see htlb_alloc_mask()).

Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Pankaj Gupta <pankaj.gupta.linux@gmail.com>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Qian Cai <cai@lca.pw>
Link: http://lkml.kernel.org/r/20200816125333.7434-4-david@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/page_isolation: exit early when pageblock is isolated in set_migratetype_isolate()
David Hildenbrand [Tue, 13 Oct 2020 23:55:21 +0000 (16:55 -0700)]
mm/page_isolation: exit early when pageblock is isolated in set_migratetype_isolate()

Right now, if we have two isolations racing on a pageblock that's in the
MOVABLE zone, we would trigger the WARN_ON_ONCE().  Let's just return
directly, simplifying error handling.

The change was introduced in commit 3d680bdf60a5 ("mm/page_isolation: fix
potential warning from user").  As far as I can see, we currently don't
have alloc_contig_range() users that use the ZONE_MOVABLE (anymore), so
it's currently more a cleanup and a preparation for the future than a fix.

Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Reviewed-by: Pankaj Gupta <pankaj.gupta.linux@gmail.com>
Acked-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Qian Cai <cai@lca.pw>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Mike Rapoport <rppt@kernel.org>
Link: http://lkml.kernel.org/r/20200816125333.7434-3-david@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agomm/page_alloc: tweak comments in has_unmovable_pages()
David Hildenbrand [Tue, 13 Oct 2020 23:55:17 +0000 (16:55 -0700)]
mm/page_alloc: tweak comments in has_unmovable_pages()

Patch series "mm / virtio-mem: support ZONE_MOVABLE", v5.

When introducing virtio-mem, the semantics of ZONE_MOVABLE were rather
unclear, which is why we special-cased ZONE_MOVABLE such that partially
plugged blocks would never end up in ZONE_MOVABLE.

Now that the semantics are much clearer (and are documented in patch #6),
let's support partially plugged memory blocks in ZONE_MOVABLE, allowing
partially plugged memory blocks to be online to ZONE_MOVABLE and also
unplugging from such memory blocks.  This avoids surprises when onlining
of memory blocks suddenly fails, just because they are not completely
populated by virtio-mem (yet).

This is especially helpful for testing, but also paves the way for
virtio-mem optimizations, allowing more memory to get reliably unplugged.

Cleanup has_unmovable_pages() and set_migratetype_isolate(), providing
better documentation of how ZONE_MOVABLE interacts with different kind of
unmovable pages (memory offlining vs.  alloc_contig_range()).

This patch (of 6):

Let's move the split comment regarding bootmem allocations and memory
holes, especially in the context of ZONE_MOVABLE, to the PageReserved()
check.

Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Baoquan He <bhe@redhat.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Pankaj Gupta <pankaj.gupta.linux@gmail.com>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Qian Cai <cai@lca.pw>
Link: http://lkml.kernel.org/r/20200816125333.7434-1-david@redhat.com
Link: http://lkml.kernel.org/r/20200816125333.7434-2-david@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>