Skip to contentSkip to conversation

5.0.17-stable review

PATCH In review Greg Kroah-Hartman 7 participants 144 messages May 15, 2019 – May 17, 2019

Subsystems (6)
lore
Share & tools
lore

Greg Kroah-Hartman

author maintainer
#0 lore raw

cover 5.0.17-stable review

This is the start of the stable review cycle for the 5.0.17 release. There are 137 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.

Responses should be made by Fri 17 May 2019 09:04:31 AM UTC. Anything received after that time might be too late.

The whole patch series can be found in one patch at:

	https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.0.17-rc1.gz

or in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-5.0.y
and the diffstat can be found below.

thanks,

greg k-h

-------------
Pseudo-Shortlog of commits:

Greg Kroah-Hartman <[email protected]>

    Linux 5.0.17-rc1

Damien Le Moal <[email protected]>
f2fs: Fix use of number of devices

Dexuan Cui <[email protected]>
PCI: hv: Add pci_destroy_slot() in pci_devices_present_work(), if necessary

Dexuan Cui <[email protected]>
PCI: hv: Add hv_pci_remove_slots() when we unload the driver

Dexuan Cui <[email protected]>
PCI: hv: Fix a memory leak in hv_eject_device_work()

YueHaibing <[email protected]>
virtio_ring: Fix potential mem leak in virtqueue_add_indirect_packed

Laurentiu Tudor <[email protected]>
powerpc/booke64: set RI in default MSR

Russell Currey <[email protected]>
powerpc/powernv/idle: Restore IAMR after idle

Rick Lindsley <[email protected]>
powerpc/book3s/64: check for NULL pointer in pgd_alloc()

Dan Carpenter <[email protected]>
drivers/virt/fsl_hypervisor.c: prevent integer overflow in ioctl

Dan Carpenter <[email protected]>
drivers/virt/fsl_hypervisor.c: dereferencing error pointers in ioctl

Paul Bolle <[email protected]>
isdn: bas_gigaset: use usb_fill_int_urb() properly

Eric Dumazet <[email protected]>
flow_dissector: disable preemption around BPF calls

Heiner Kallweit <[email protected]>
net: phy: fix phy_validate_pause

Jason Wang <[email protected]>

    tuntap: synchronize through tfiles array instead of tun->numqueues

Jason Wang <[email protected]>
tuntap: fix dividing by zero in ebpf queue selection

Oliver Neukum <[email protected]>
aqc111: fix double endianness swap on BE

Oliver Neukum <[email protected]>
aqc111: fix writing to the phy on BE

Oliver Neukum <[email protected]>
aqc111: fix endianness issue in aqc111_change_mtu

Stephen Suryaputra <[email protected]>
vrf: sit mtu should not be updated when vrf netdev is the link

Hangbin Liu <[email protected]>
vlan: disable SIOCSHWTSTAMP in container

Parthasarathy Bhuvaragan <[email protected]>
tipc: fix hanging clients using poll with EPOLLOUT flag

Paolo Abeni <[email protected]>
selinux: do not report error on connect(AF_UNSPEC)

YueHaibing <[email protected]>
packet: Fix error path in packet_init

Christophe Leroy <[email protected]>
net: ucc_geth - fix Oops when changing number of buffers in the ring

Thomas Bogendoerfer <[email protected]>
net: seeq: fix crash caused by not set dev.parent

Harini Katakam <[email protected]>
net: macb: Change interrupt and napi enable order in open

Corentin Labbe <[email protected]>
net: ethernet: stmmac: dwmac-sun8i: enable support of unicast filtering

YueHaibing <[email protected]>
net: dsa: Fix error cleanup path in dsa_init_module

David Ahern <[email protected]>
ipv4: Fix raw socket lookup for local traffic

Hangbin Liu <[email protected]>
fib_rules: return 0 directly if an exactly same rule exists when NLM_F_EXCL not supplied

Laurentiu Tudor <[email protected]>
dpaa_eth: fix SG frame cleanup

Tobin C. Harding <[email protected]>
bridge: Fix error path for kobject_init_and_add()

Jarod Wilson <[email protected]>
bonding: fix arp_validate toggling in active-backup mode

Nigel Croxon <[email protected]>
Don't jump to compute_result state from check_result state

Gustavo A. R. Silva <[email protected]>
rtlwifi: rtl8723ae: Fix missing break in switch statement

Petr Štetiar <[email protected]>
mwl8k: Fix rate_idx underflow

Wei Yongjun <[email protected]>
cw1200: fix missing unlock on error in cw1200_hw_scan()

Damian Kos <[email protected]>
drm/rockchip: fix for mailbox read validation.

Antoine Tenart <[email protected]>
net: mvpp2: fix validate for PPv2.1

John Hurley <[email protected]>
net: sched: fix cleanup NULL pointer exception in act_mirr

Willem de Bruijn <[email protected]>
bpf: only test gso type on gso packets

Andrey Ryabinin <[email protected]>
mm/page_alloc.c: avoid potential NULL pointer dereference

David Hildenbrand <[email protected]>
mm/memory_hotplug.c: drop memory device reference after find_memory_block()

Lijun Ou <[email protected]>
RDMA/hns: Bugfix for mapping user db

Geert Uytterhoeven <[email protected]>
gpio: Fix gpiochip_add_data_with_key() error path

Miaohe Lin <[email protected]>
net: vrf: Fix operation not supported when set vrf mac

Pan Bian <[email protected]>
Input: synaptics-rmi4 - fix possible double free

Jacky Bai <[email protected]>
Input: snvs_pwrkey - make it depend on ARCH_MXC

Paul Kocialkowski <[email protected]>
drm/sun4i: Unbind components before releasing DRM and memory

Dave Airlie <[email protected]>
Revert "drm/virtio: drop prime import/export callbacks"

Jeff Layton <[email protected]>
ceph: handle the case where a dentry has been renamed on outstanding req

Daniel Gomez <[email protected]>
spi: ST ST95HF NFC: declare missing of table

Daniel Gomez <[email protected]>
spi: Micrel eth switch: declare missing of table

Tigran Tadevosyan <[email protected]>
ARM: 8856/1: NOMMU: Fix CCR register faulty initialization when MPU is disabled

Russell King <[email protected]>
ARM: fix function graph tracer and unwinder dependencies

Lucas Stach <[email protected]>
drm/imx: don't skip DP channel disable for background plane

Lucas Stach <[email protected]>
gpu: ipu-v3: dp: fix CSC handling

Ard Biesheuvel <[email protected]>
arm64/module: ftrace: deal with place relative nature of PLTs

Stefan Wahren <[email protected]>
dmaengine: bcm2835: Avoid GFP_KERNEL in device_prep_slave_sg

Andrei Vagin <[email protected]>
netfilter: fix nf_l4proto_log_invalid to log invalid packets

Florian Westphal <[email protected]>

    netfilter: never get/set skb->tstamp

Po-Hsu Lin <[email protected]>
selftests/net: correct the return value for run_afpackettests

Po-Hsu Lin <[email protected]>
selftests/net: correct the return value for run_netsocktests

Petr Štetiar <[email protected]>
of_net: Fix residues after of_get_nvmem_mac_address removal

Paul Kocialkowski <[email protected]>
drm/sun4i: Fix component unbinding and component master deletion

Paul Kocialkowski <[email protected]>
drm/sun4i: Set device driver data at bind time for use in unbind

Arnd Bergmann <[email protected]>
s390: ctcm: fix ctcm_new_device error return code

Guy Levi <[email protected]>
IB/mlx5: Fix scatter to CQE in DCT QP creation

Petr Štetiar <[email protected]>
MIPS: perf: ath79: Fix perfcount IRQ assignment

Florian Westphal <[email protected]>
netfilter: nat: fix icmp id randomization

Dan Carpenter <[email protected]>
netfilter: nf_tables: prevent shift wrap in nft_chain_parse_hook()

Florian Westphal <[email protected]>
netfilter: ctnetlink: don't use conntrack/expect object addresses as id

Julian Anastasov <[email protected]>
ipvs: do not schedule icmp errors from tunnels

Florian Westphal <[email protected]>
selftests: netfilter: check icmp pkttoobig errors are set as related

Jonas Karlman <[email protected]>
drm: bridge: dw-hdmi: Fix overflow workaround for Rockchip SoCs

Dan Williams <[email protected]>
init: initialize jump labels before command line option parsing

Johannes Weiner <[email protected]>
mm: fix inactive list balancing between NUMA nodes and cgroups

Qian Cai <[email protected]>
mm/hotplug: treat CMA pages as unmovable

Qian Cai <[email protected]>
slab: store tagged freelist for off-slab slabmgmt

Christoph Hellwig <[email protected]>
scsi: aic7xxx: fix EISA support

Jiri Olsa <[email protected]>
perf tools: Fix map reference counting

Claudiu Manoil <[email protected]>
ocelot: Don't sleep in atomic context (irqs_disabled())

Tony Camuso <[email protected]>
ipmi: ipmi_si_hardcode.c: init si_type array to fix a crash

Jiri Olsa <[email protected]>
perf top: Always sample time to satisfy needs of use of ordered queuing

Rikard Falkeborn <[email protected]>
tools lib traceevent: Fix missing equality check for strcmp

Vitaly Kuznetsov <[email protected]>
KVM: x86: avoid misreporting level-triggered irqs as edge-triggered in tracing

Paolo Bonzini <[email protected]>
KVM: fix spectrev1 gadgets

Paolo Bonzini <[email protected]>
KVM: nVMX: always use early vmcs check when EPT is disabled

Jian-Hong Pan <[email protected]>
x86/reboot, efi: Use EFI reboot for Acer TravelMate X514-51T

Thomas Gleixner <[email protected]>
x86/mm: Prevent bogus warnings with "noexec=off"

Sami Tolvanen <[email protected]>
x86/build/lto: Fix truncated .bss with -fdata-sections

Harald Freudenberger <[email protected]>
s390/pkey: add one more argument space for debug feature entry

David Francis <[email protected]>
drm/amd/display: If one stream full updates, full update all planes

Denis Bolotin <[email protected]>
qed: Fix the DORQ's attentions handling

Denis Bolotin <[email protected]>
qed: Fix missing DORQ attentions

Denis Bolotin <[email protected]>
qed: Fix the doorbell address sanity check

Denis Bolotin <[email protected]>
qed: Delete redundant doorbell recovery types

David Howells <[email protected]>
afs: Fix in-progess ops to ignore server-level callback invalidation

Marc Dionne <[email protected]>
afs: Unlock pages for __pagevec_release()

Colin Ian King <[email protected]>
qede: fix write to free'd pointer error and double free of ptp

Colin Ian King <[email protected]>
vxge: fix return of a free'd memblock on a failed dma mapping

Tetsuo Handa <[email protected]>
mISDN: Check address length before reading address family

wentalou <[email protected]>
drm/amdgpu: shadow in shadow_list without tbo.mem.start cause page fault in sriov TDR

David Ahern <[email protected]>
selftests: fib_tests: Fix 'Command line is not complete' errors

Neil Armstrong <[email protected]>
clocksource/drivers/oxnas: Fix OX820 compatible

Arnd Bergmann <[email protected]>
clocksource/drivers/npcm: select TIMER_OF

Martin Leung <[email protected]>
drm/amd/display: extending AUX SW Timeout

Lin Yi <[email protected]>
drm/ttm: fix dma_fence refcount imbalance on error path

Martin Schwidefsky <[email protected]>

    s390/3270: fix lockdep false positive on view->lock

Dave Jiang <[email protected]>
tools/testing/nvdimm: Retain security state after overwrite

Li RongQing <[email protected]>
libnvdimm/pmem: fix a possible OOB access when read and write pmem

Dave Jiang <[email protected]>
libnvdimm/security: provide fix for secure-erase to use zero-key

Sunil Dutt <[email protected]>
nl80211: Add NL80211_FLAG_CLEAR_SKB flag for other NL commands

Felix Fietkau <[email protected]>
mac80211: fix memory accounting with A-MSDU aggregation

Ilan Peer <[email protected]>
cfg80211: Handle WMM rules in regulatory domain intersection

Andrei Otcheretianski <[email protected]>

    mac80211: Increase MAX_MSG_LEN

Felix Fietkau <[email protected]>
mac80211: fix unaligned access in mesh table hash function

Peter Oberparleiter <[email protected]>
s390/dasd: Fix capacity calculation for large volumes

Aditya Pakki <[email protected]>
libnvdimm/btt: Fix a kmemdup failure check

Dmitry Torokhov <[email protected]>
HID: input: add mapping for "Toggle Display" key

Dmitry Torokhov <[email protected]>
HID: input: add mapping for keyboard Brightness Up/Down/Toggle keys

Dmitry Torokhov <[email protected]>
HID: input: add mapping for Expose/Overview key

Kangjie Lu <[email protected]>
libnvdimm/namespace: Fix a potential NULL pointer dereference

Dan Williams <[email protected]>
acpi/nfit: Always dump _DSM output payload

Sven Van Asbroeck <[email protected]>
iio: adc: xilinx: prevent touching unclocked h/w on remove

Sven Van Asbroeck <[email protected]>
iio: adc: xilinx: fix potential use-after-free on probe

Sven Van Asbroeck <[email protected]>
iio: adc: xilinx: fix potential use-after-free on remove

Johan Hovold <[email protected]>
USB: serial: fix unthrottle races

Hans de Goede <[email protected]>
virt: vbox: Sanity-check parameter types for hgcm-calls coming from userspace

Andrea Parri <[email protected]>
kernfs: fix barrier usage in __kernfs_new_node()

Kees Cook <[email protected]>
selftests/seccomp: Handle namespace failures gracefully

Lei YU <[email protected]>
hwmon: (occ) Fix extended status bits

Stefan Wahren <[email protected]>
hwmon: (pwm-fan) Disable PWM if fetching cooling data fails

Mario Limonciello <[email protected]>
platform/x86: dell-laptop: fix rfkill functionality

Jiaxun Yang <[email protected]>
platform/x86: thinkpad_acpi: Disable Bluetooth for some machines

Gustavo A. R. Silva <[email protected]>
platform/x86: sony-laptop: Fix unintentional fall-through

Jens Axboe <[email protected]>
bfq: update internal depth state when queue depth changes

-------------

Diffstat:

 .../devicetree/bindings/net/davinci_emac.txt       |   2 +
 Documentation/devicetree/bindings/net/ethernet.txt |   2 -
 Documentation/devicetree/bindings/net/macb.txt     |   4 +
 Makefile                                           |   4 +-
 arch/arm/Kconfig                                   |   2 +-
 arch/arm/Kconfig.debug                             |   6 +-
 arch/arm/kernel/head-nommu.S                       |   2 +-
 arch/arm64/kernel/ftrace.c                         |   9 +-
 arch/mips/ath79/setup.c                            |   6 -
 arch/powerpc/include/asm/book3s/64/pgalloc.h       |   3 +
 arch/powerpc/include/asm/reg_booke.h               |   2 +-
 arch/powerpc/kernel/idle_book3s.S                  |  20 ++
 arch/x86/include/uapi/asm/vmx.h                    |   1 +
 arch/x86/kernel/reboot.c                           |  21 ++
 arch/x86/kernel/vmlinux.lds.S                      |   2 +-
 arch/x86/kvm/lapic.c                               |   4 +-
 arch/x86/kvm/trace.h                               |   4 +-
 arch/x86/kvm/vmx/nested.c                          |  22 +-
 arch/x86/mm/dump_pagetables.c                      |   3 +-
 arch/x86/mm/ioremap.c                              |   2 +-
 block/bfq-iosched.c                                |   8 +-
 block/blk-mq.c                                     |   2 +
 drivers/acpi/nfit/core.c                           |  12 +-
 drivers/char/ipmi/ipmi_si_hardcode.c               |   2 +
 drivers/clocksource/Kconfig                        |   1 +
 drivers/clocksource/timer-oxnas-rps.c              |   2 +-
 drivers/dma/bcm2835-dma.c                          |   2 +-
 drivers/gpio/gpiolib.c                             |  12 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c         |   1 +
 drivers/gpu/drm/amd/display/dc/core/dc.c           |  19 ++
 drivers/gpu/drm/amd/display/dc/dc.h                |   3 +
 drivers/gpu/drm/amd/display/dc/dce/dce_aux.c       |   9 +-
 drivers/gpu/drm/amd/display/dc/dce/dce_aux.h       |   6 +-
 drivers/gpu/drm/bridge/synopsys/dw-hdmi.c          |   4 +
 drivers/gpu/drm/imx/ipuv3-crtc.c                   |   2 +-
 drivers/gpu/drm/rockchip/cdn-dp-reg.c              |   2 +-
 drivers/gpu/drm/sun4i/sun4i_drv.c                  |   7 +
 drivers/gpu/drm/ttm/ttm_bo.c                       |   4 +-
 drivers/gpu/drm/virtio/virtgpu_drv.c               |   4 +
 drivers/gpu/drm/virtio/virtgpu_drv.h               |   4 +
 drivers/gpu/drm/virtio/virtgpu_prime.c             |  12 +
 drivers/gpu/ipu-v3/ipu-dp.c                        |  12 +-
 drivers/hid/hid-input.c                            |  14 +
 drivers/hwmon/occ/sysfs.c                          |   8 +-
 drivers/hwmon/pwm-fan.c                            |   2 +-
 drivers/iio/adc/xilinx-xadc-core.c                 |   3 +-
 drivers/infiniband/hw/hns/hns_roce_qp.c            |   2 +-
 drivers/infiniband/hw/mlx5/main.c                  |   2 +
 drivers/infiniband/hw/mlx5/qp.c                    |  11 +-
 drivers/input/keyboard/Kconfig                     |   2 +-
 drivers/input/rmi4/rmi_driver.c                    |   6 +-
 drivers/irqchip/irq-ath79-misc.c                   |  11 +
 drivers/isdn/gigaset/bas-gigaset.c                 |   9 +-
 drivers/isdn/mISDN/socket.c                        |   4 +-
 drivers/md/raid5.c                                 |  19 +-
 drivers/net/bonding/bond_options.c                 |   7 -
 drivers/net/ethernet/cadence/macb_main.c           |   6 +-
 drivers/net/ethernet/freescale/dpaa/dpaa_eth.c     |   2 +-
 drivers/net/ethernet/freescale/ucc_geth_ethtool.c  |   8 +-
 drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c    |   2 +-
 drivers/net/ethernet/mscc/ocelot.c                 |   2 +-
 drivers/net/ethernet/neterion/vxge/vxge-config.c   |   1 +
 drivers/net/ethernet/qlogic/qed/qed.h              |   7 +-
 drivers/net/ethernet/qlogic/qed/qed_dev.c          |  85 +++----
 drivers/net/ethernet/qlogic/qed/qed_int.c          |  83 ++++--
 drivers/net/ethernet/qlogic/qed/qed_int.h          |   4 +-
 drivers/net/ethernet/qlogic/qed/qed_main.c         |   2 +-
 drivers/net/ethernet/qlogic/qede/qede_ptp.c        |   7 +-
 drivers/net/ethernet/seeq/sgiseeq.c                |   1 +
 drivers/net/ethernet/stmicro/stmmac/dwmac-sun8i.c  |   2 +
 drivers/net/phy/phy_device.c                       |  11 +-
 drivers/net/phy/spi_ks8995.c                       |   9 +
 drivers/net/tun.c                                  |  14 +-
 drivers/net/usb/aqc111.c                           |  31 ++-
 drivers/net/vrf.c                                  |   2 +
 drivers/net/wireless/marvell/mwl8k.c               |  13 +-
 .../net/wireless/realtek/rtlwifi/rtl8723ae/hw.c    |   1 +
 drivers/net/wireless/st/cw1200/scan.c              |   5 +-
 drivers/nfc/st95hf/core.c                          |   7 +
 drivers/nvdimm/btt_devs.c                          |  18 +-
 drivers/nvdimm/namespace_devs.c                    |   5 +-
 drivers/nvdimm/pmem.c                              |   8 +-
 drivers/nvdimm/security.c                          |  17 +-
 drivers/of/of_net.c                                |   1 -
 drivers/pci/controller/pci-hyperv.c                |  23 ++
 drivers/platform/x86/dell-laptop.c                 |   6 +-
 drivers/platform/x86/sony-laptop.c                 |   8 +-
 drivers/platform/x86/thinkpad_acpi.c               |  72 +++++-
 drivers/s390/block/dasd_eckd.c                     |   6 +-
 drivers/s390/char/con3270.c                        |   2 +-
 drivers/s390/char/fs3270.c                         |   3 +-
 drivers/s390/char/raw3270.c                        |   3 +-
 drivers/s390/char/raw3270.h                        |   4 +-
 drivers/s390/char/tty3270.c                        |   3 +-
 drivers/s390/crypto/pkey_api.c                     |   3 +-
 drivers/s390/net/ctcm_main.c                       |   1 +
 drivers/scsi/aic7xxx/aic7770_osm.c                 |   1 +
 drivers/scsi/aic7xxx/aic7xxx.h                     |   1 +
 drivers/scsi/aic7xxx/aic7xxx_osm.c                 |  10 +-
 drivers/scsi/aic7xxx/aic7xxx_osm_pci.c             |   1 +
 drivers/usb/serial/generic.c                       |  39 ++-
 drivers/virt/fsl_hypervisor.c                      |  29 ++-
 drivers/virt/vboxguest/vboxguest_core.c            |  31 +++
 drivers/virtio/virtio_ring.c                       |   1 +
 fs/afs/callback.c                                  |   3 +-
 fs/afs/internal.h                                  |   4 +-
 fs/afs/server.c                                    |   1 -
 fs/afs/write.c                                     |   1 +
 fs/ceph/inode.c                                    |  16 +-
 fs/f2fs/data.c                                     |  17 +-
 fs/f2fs/f2fs.h                                     |  13 +-
 fs/f2fs/file.c                                     |   2 +-
 fs/f2fs/gc.c                                       |   2 +-
 fs/f2fs/segment.c                                  |  13 +-
 fs/kernfs/dir.c                                    |   5 +-
 include/linux/efi.h                                |   7 +-
 include/linux/elevator.h                           |   1 +
 include/linux/kvm_host.h                           |  10 +-
 include/linux/skbuff.h                             |   4 +-
 include/net/netfilter/nf_conntrack.h               |   2 +
 include/uapi/rdma/mlx5-abi.h                       |   1 +
 init/main.c                                        |   4 +-
 mm/memory_hotplug.c                                |   1 +
 mm/page_alloc.c                                    |  33 ++-
 mm/slab.c                                          |   1 -
 mm/vmscan.c                                        |  29 +--
 net/8021q/vlan_dev.c                               |   4 +-
 net/bridge/br_if.c                                 |  13 +-
 net/core/fib_rules.c                               |   6 +-
 net/core/filter.c                                  |   8 +-
 net/core/flow_dissector.c                          |   3 +
 net/dsa/dsa.c                                      |  11 +-
 net/ipv4/raw.c                                     |   4 +-
 net/ipv6/sit.c                                     |   2 +-
 net/mac80211/mesh_pathtbl.c                        |   2 +-
 net/mac80211/trace_msg.h                           |   7 +-
 net/mac80211/tx.c                                  |   3 +
 net/netfilter/ipvs/ip_vs_core.c                    |   2 +-
 net/netfilter/nf_conntrack_core.c                  |  42 ++-
 net/netfilter/nf_conntrack_netlink.c               |  34 ++-
 net/netfilter/nf_conntrack_proto.c                 |   2 +-
 net/netfilter/nf_nat_core.c                        |  11 +-
 net/netfilter/nf_tables_api.c                      |   2 +-
 net/netfilter/nfnetlink_log.c                      |   2 +-
 net/netfilter/nfnetlink_queue.c                    |   2 +-
 net/netfilter/xt_time.c                            |  23 +-
 net/packet/af_packet.c                             |  25 +-
 net/sched/act_mirred.c                             |   3 +
 net/tipc/socket.c                                  |   4 +-
 net/wireless/nl80211.c                             |  18 +-
 net/wireless/reg.c                                 |  39 +++
 security/selinux/hooks.c                           |   8 +-
 tools/lib/traceevent/event-parse.c                 |   2 +-
 tools/perf/builtin-top.c                           |   1 +
 tools/perf/util/map.c                              |   4 +-
 tools/testing/nvdimm/test/nfit.c                   |  17 +-
 tools/testing/selftests/net/fib_tests.sh           |  94 +++----
 tools/testing/selftests/net/run_afpackettests      |   5 +
 tools/testing/selftests/net/run_netsocktests       |   2 +-
 tools/testing/selftests/netfilter/Makefile         |   2 +-
 .../selftests/netfilter/conntrack_icmp_related.sh  | 283 +++++++++++++++++++++
 tools/testing/selftests/netfilter/nft_nat.sh       |  36 ++-
 tools/testing/selftests/seccomp/seccomp_bpf.c      |  43 ++--
 virt/kvm/irqchip.c                                 |   5 +-
 virt/kvm/kvm_main.c                                |   6 +-
 165 files changed, 1407 insertions(+), 495 deletions(-)
author maintainer
#1 lore raw

PATCH 1/137 bfq: update internal depth state when queue depth changes

[ Upstream commit 77f1e0a52d26242b6c2dba019f6ebebfb9ff701e ]

A previous commit moved the shallow depth and BFQ depth map calculations to be done at init time, moving it outside of the hotter IO path. This potentially causes hangs if the users changes the depth of the scheduler map, by writing to the 'nr_requests' sysfs file for that device.

Add a blk-mq-sched hook that allows blk-mq to inform the scheduler if the depth changes, so that the scheduler can update its internal state.

Fixes f0635b8a416e ("bfq: calculate shallow depths at init time")
---
 block/bfq-iosched.c      | 8 +++++++-
 block/blk-mq.c           | 2 ++
 include/linux/elevator.h | 1 +
 3 files changed, 10 insertions(+), 1 deletion(-)
block/bfq-iosched.c +7 -1
index 72510c4700014..356620414cf94 100644
@@ -5353,7 +5353,7 @@ static unsigned int bfq_update_depths(struct bfq_data *bfqd,
 	return min_shallow;
 }
 
-static int bfq_init_hctx(struct blk_mq_hw_ctx *hctx, unsigned int index)
+static void bfq_depth_updated(struct blk_mq_hw_ctx *hctx)
 {
 	struct bfq_data *bfqd = hctx->queue->elevator->elevator_data;
 	struct blk_mq_tags *tags = hctx->sched_tags;
@@ -5361,6 +5361,11 @@ static int bfq_init_hctx(struct blk_mq_hw_ctx *hctx, unsigned int index)
 
 	min_shallow = bfq_update_depths(bfqd, &tags->bitmap_tags);
 	sbitmap_queue_min_shallow_depth(&tags->bitmap_tags, min_shallow);
+}
+
+static int bfq_init_hctx(struct blk_mq_hw_ctx *hctx, unsigned int index)
+{
+	bfq_depth_updated(hctx);
 	return 0;
 }
 
@@ -5783,6 +5788,7 @@ static struct elevator_type iosched_bfq_mq = {
 		.requests_merged	= bfq_requests_merged,
 		.request_merged		= bfq_request_merged,
 		.has_work		= bfq_has_work,
+		.depth_updated		= bfq_depth_updated,
 		.init_hctx		= bfq_init_hctx,
 		.init_sched		= bfq_init_queue,
 		.exit_sched		= bfq_exit_queue,
block/blk-mq.c +2 -0
index 6930c82ab75fc..5b920a82bfe60 100644
@@ -3131,6 +3131,8 @@ int blk_mq_update_nr_requests(struct request_queue *q, unsigned int nr)
 		}
 		if (ret)
 			break;
+		if (q->elevator && q->elevator->type->ops.depth_updated)
+			q->elevator->type->ops.depth_updated(hctx);
 	}
 
 	if (!ret)
include/linux/elevator.h +1 -0
index 2e9e2763bf47d..6e8bc53740f05 100644
@@ -31,6 +31,7 @@ struct elevator_mq_ops {
 	void (*exit_sched)(struct elevator_queue *);
 	int (*init_hctx)(struct blk_mq_hw_ctx *, unsigned int);
 	void (*exit_hctx)(struct blk_mq_hw_ctx *, unsigned int);
+	void (*depth_updated)(struct blk_mq_hw_ctx *);
 
 	bool (*allow_merge)(struct request_queue *, struct request *, struct bio *);
 	bool (*bio_merge)(struct blk_mq_hw_ctx *, struct bio *);
Signature · 1 line
2.20.1
author maintainer
#2 lore raw

PATCH 2/137 platform/x86: sony-laptop: Fix unintentional fall-through

In reply to Greg Kroah-Hartman

From: Gustavo A. R. Silva <[email protected]>

commit 1cbd7a64959d33e7a2a1fa2bf36a62b350a9fcbd upstream.

It seems that the default case should return AE_CTRL_TERMINATE, instead of falling through to case ACPI_RESOURCE_TYPE_END_TAG and returning AE_OK; otherwise the line of code at the end of the function is unreachable and makes no sense:

return AE_CTRL_TERMINATE;

This fix is based on the following thread of discussion:

https://lore.kernel.org/patchwork/patch/959782/

Fixes 33a04454527e ("sony-laptop: Add SNY6001 device handling (sonypi reimplementation)")
---
 drivers/platform/x86/sony-laptop.c |    8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
drivers/platform/x86/sony-laptop.c +5 -3
@@ -4424,14 +4424,16 @@ sony_pic_read_possible_resource(struct a
 			}
 			return AE_OK;
 		}
+
+	case ACPI_RESOURCE_TYPE_END_TAG:
+		return AE_OK;
+
 	default:
 		dprintk("Resource %d isn't an IRQ nor an IO port\n",
 			resource->type);
+		return AE_CTRL_TERMINATE;
 
-	case ACPI_RESOURCE_TYPE_END_TAG:
-		return AE_OK;
 	}
-	return AE_CTRL_TERMINATE;
 }
 
 static int sony_pic_possible_resources(struct acpi_device *device)
author maintainer
#3 lore raw

PATCH 3/137 platform/x86: thinkpad_acpi: Disable Bluetooth for some machines

In reply to Greg Kroah-Hartman

From: Jiaxun Yang <[email protected]>

commit f7db839fccf087664e5587966220821289b6a9cb upstream.

Some AMD based ThinkPads have a firmware bug that calling "GBDC" will cause Bluetooth on Intel wireless cards blocked.

Probe these models by DMI match and disable Bluetooth subdriver if specified Intel wireless card exist.

---
 drivers/platform/x86/thinkpad_acpi.c |   72 ++++++++++++++++++++++++++++++++++-
 1 file changed, 70 insertions(+), 2 deletions(-)
drivers/platform/x86/thinkpad_acpi.c +70 -2
@@ -79,7 +79,7 @@
 #include <linux/jiffies.h>
 #include <linux/workqueue.h>
 #include <linux/acpi.h>
-#include <linux/pci_ids.h>
+#include <linux/pci.h>
 #include <linux/power_supply.h>
 #include <sound/core.h>
 #include <sound/control.h>
@@ -4501,6 +4501,74 @@ static void bluetooth_exit(void)
 	bluetooth_shutdown();
 }
 
+static const struct dmi_system_id bt_fwbug_list[] __initconst = {
+	{
+		.ident = "ThinkPad E485",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
+			DMI_MATCH(DMI_BOARD_NAME, "20KU"),
+		},
+	},
+	{
+		.ident = "ThinkPad E585",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
+			DMI_MATCH(DMI_BOARD_NAME, "20KV"),
+		},
+	},
+	{
+		.ident = "ThinkPad A285 - 20MW",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
+			DMI_MATCH(DMI_BOARD_NAME, "20MW"),
+		},
+	},
+	{
+		.ident = "ThinkPad A285 - 20MX",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
+			DMI_MATCH(DMI_BOARD_NAME, "20MX"),
+		},
+	},
+	{
+		.ident = "ThinkPad A485 - 20MU",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
+			DMI_MATCH(DMI_BOARD_NAME, "20MU"),
+		},
+	},
+	{
+		.ident = "ThinkPad A485 - 20MV",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
+			DMI_MATCH(DMI_BOARD_NAME, "20MV"),
+		},
+	},
+	{}
+};
+
+static const struct pci_device_id fwbug_cards_ids[] __initconst = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x24F3) },
+	{ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x24FD) },
+	{ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x2526) },
+	{}
+};
+
+
+static int __init have_bt_fwbug(void)
+{
+	/*
+	 * Some AMD based ThinkPads have a firmware bug that calling
+	 * "GBDC" will cause bluetooth on Intel wireless cards blocked
+	 */
+	if (dmi_check_system(bt_fwbug_list) && pci_dev_present(fwbug_cards_ids)) {
+		vdbg_printk(TPACPI_DBG_INIT | TPACPI_DBG_RFKILL,
+			FW_BUG "disable bluetooth subdriver for Intel cards\n");
+		return 1;
+	} else
+		return 0;
+}
+
 static int __init bluetooth_init(struct ibm_init_struct *iibm)
 {
 	int res;
@@ -4513,7 +4581,7 @@ static int __init bluetooth_init(struct
 
 	/* bluetooth not supported on 570, 600e/x, 770e, 770x, A21e, A2xm/p,
 	   G4x, R30, R31, R40e, R50e, T20-22, X20-21 */
-	tp_features.bluetooth = hkey_handle &&
+	tp_features.bluetooth = !have_bt_fwbug() && hkey_handle &&
 	    acpi_evalf(hkey_handle, &status, "GBDC", "qd");
 
 	vdbg_printk(TPACPI_DBG_INIT | TPACPI_DBG_RFKILL,
author maintainer
#4 lore raw

PATCH 4/137 platform/x86: dell-laptop: fix rfkill functionality

In reply to Greg Kroah-Hartman

From: Mario Limonciello <[email protected]>

commit 6cc13c28da5beee0f706db6450e190709700b34a upstream.

When converting the driver two arguments were transposed leading to rfkill not working.

---
 drivers/platform/x86/dell-laptop.c |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/platform/x86/dell-laptop.c +3 -3
@@ -531,7 +531,7 @@ static void dell_rfkill_query(struct rfk
 		return;
 	}
 
-	dell_fill_request(&buffer, 0, 0x2, 0, 0);
+	dell_fill_request(&buffer, 0x2, 0, 0, 0);
 	ret = dell_send_request(&buffer, CLASS_INFO, SELECT_RFKILL);
 	hwswitch = buffer.output[1];
 
@@ -562,7 +562,7 @@ static int dell_debugfs_show(struct seq_
 		return ret;
 	status = buffer.output[1];
 
-	dell_fill_request(&buffer, 0, 0x2, 0, 0);
+	dell_fill_request(&buffer, 0x2, 0, 0, 0);
 	hwswitch_ret = dell_send_request(&buffer, CLASS_INFO, SELECT_RFKILL);
 	if (hwswitch_ret)
 		return hwswitch_ret;
@@ -647,7 +647,7 @@ static void dell_update_rfkill(struct wo
 	if (ret != 0)
 		return;
 
-	dell_fill_request(&buffer, 0, 0x2, 0, 0);
+	dell_fill_request(&buffer, 0x2, 0, 0, 0);
 	ret = dell_send_request(&buffer, CLASS_INFO, SELECT_RFKILL);
 
 	if (ret == 0 && (status & BIT(0)))
author maintainer
#5 lore raw

PATCH 5/137 hwmon: (pwm-fan) Disable PWM if fetching cooling data fails

In reply to Greg Kroah-Hartman

From: Stefan Wahren <[email protected]>

commit 53f1647da3e8fb3e89066798f0fdc045064d353d upstream.

In case pwm_fan_of_get_cooling_data() fails we should disable the PWM just like in the other error cases.

Fixes 2e5219c77183 ("hwmon: (pwm-fan) Read PWM FAN configuration from device tree")
Cc <[email protected]> # 4.14+
---
 drivers/hwmon/pwm-fan.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/hwmon/pwm-fan.c +1 -1
@@ -254,7 +254,7 @@ static int pwm_fan_probe(struct platform
 
 	ret = pwm_fan_of_get_cooling_data(&pdev->dev, ctx);
 	if (ret)
-		return ret;
+		goto err_pwm_disable;
 
 	ctx->pwm_fan_state = ctx->pwm_fan_max_state;
 	if (IS_ENABLED(CONFIG_THERMAL)) {
author maintainer
#6 lore raw

PATCH 6/137 hwmon: (occ) Fix extended status bits

In reply to Greg Kroah-Hartman

From: Lei YU <[email protected]>

commit b88c5049219a7f322bb1fd65fc30d17472a23563 upstream.

The occ's extended status is checked and shown as sysfs attributes. But the code was incorrectly checking the "status" bits.
Fix it by checking the "ext_status" bits.

Fixes df04ced684d4 ("hwmon (occ): Add sysfs attributes for additional OCC data")
---
 drivers/hwmon/occ/sysfs.c |    8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/hwmon/occ/sysfs.c +4 -4
@@ -51,16 +51,16 @@ static ssize_t occ_sysfs_show(struct dev
 		val = !!(header->status & OCC_STAT_ACTIVE);
 		break;
 	case 2:
-		val = !!(header->status & OCC_EXT_STAT_DVFS_OT);
+		val = !!(header->ext_status & OCC_EXT_STAT_DVFS_OT);
 		break;
 	case 3:
-		val = !!(header->status & OCC_EXT_STAT_DVFS_POWER);
+		val = !!(header->ext_status & OCC_EXT_STAT_DVFS_POWER);
 		break;
 	case 4:
-		val = !!(header->status & OCC_EXT_STAT_MEM_THROTTLE);
+		val = !!(header->ext_status & OCC_EXT_STAT_MEM_THROTTLE);
 		break;
 	case 5:
-		val = !!(header->status & OCC_EXT_STAT_QUICK_DROP);
+		val = !!(header->ext_status & OCC_EXT_STAT_QUICK_DROP);
 		break;
 	case 6:
 		val = header->occ_state;
author maintainer
#7 lore raw

PATCH 7/137 selftests/seccomp: Handle namespace failures gracefully

In reply to Greg Kroah-Hartman

From: Kees Cook <[email protected]>

commit 9dd3fcb0ab73cb1e00b8562ef027a38521aaff87 upstream.

When running without USERNS or PIDNS the seccomp test would hang since it was waiting forever for the child to trigger the user notification since it seems the glibc() abort handler makes a call to getpid(), which would trap again. This changes the getpid filter to getppid, and makes sure ASSERTs execute to stop from spawning the listener.

Fixes 6a21cc50f0c7 ("seccomp: add a return code to trap to userspace")
---
 tools/testing/selftests/seccomp/seccomp_bpf.c |   43 +++++++++++++-------------
 1 file changed, 23 insertions(+), 20 deletions(-)
tools/testing/selftests/seccomp/seccomp_bpf.c +23 -20
@@ -3089,9 +3089,9 @@ TEST(user_notification_basic)
 
 	/* Check that we get -ENOSYS with no listener attached */
 	if (pid == 0) {
-		if (user_trap_syscall(__NR_getpid, 0) < 0)
+		if (user_trap_syscall(__NR_getppid, 0) < 0)
 			exit(1);
-		ret = syscall(__NR_getpid);
+		ret = syscall(__NR_getppid);
 		exit(ret >= 0 || errno != ENOSYS);
 	}
 
@@ -3106,12 +3106,12 @@ TEST(user_notification_basic)
 	EXPECT_EQ(seccomp(SECCOMP_SET_MODE_FILTER, 0, &prog), 0);
 
 	/* Check that the basic notification machinery works */
-	listener = user_trap_syscall(__NR_getpid,
+	listener = user_trap_syscall(__NR_getppid,
 				     SECCOMP_FILTER_FLAG_NEW_LISTENER);
 	ASSERT_GE(listener, 0);
 
 	/* Installing a second listener in the chain should EBUSY */
-	EXPECT_EQ(user_trap_syscall(__NR_getpid,
+	EXPECT_EQ(user_trap_syscall(__NR_getppid,
 				    SECCOMP_FILTER_FLAG_NEW_LISTENER),
 		  -1);
 	EXPECT_EQ(errno, EBUSY);
@@ -3120,7 +3120,7 @@ TEST(user_notification_basic)
 	ASSERT_GE(pid, 0);
 
 	if (pid == 0) {
-		ret = syscall(__NR_getpid);
+		ret = syscall(__NR_getppid);
 		exit(ret != USER_NOTIF_MAGIC);
 	}
 
@@ -3138,7 +3138,7 @@ TEST(user_notification_basic)
 	EXPECT_GT(poll(&pollfd, 1, -1), 0);
 	EXPECT_EQ(pollfd.revents, POLLOUT);
 
-	EXPECT_EQ(req.data.nr,  __NR_getpid);
+	EXPECT_EQ(req.data.nr,  __NR_getppid);
 
 	resp.id = req.id;
 	resp.error = 0;
@@ -3165,7 +3165,7 @@ TEST(user_notification_kill_in_middle)
 	struct seccomp_notif req = {};
 	struct seccomp_notif_resp resp = {};
 
-	listener = user_trap_syscall(__NR_getpid,
+	listener = user_trap_syscall(__NR_getppid,
 				     SECCOMP_FILTER_FLAG_NEW_LISTENER);
 	ASSERT_GE(listener, 0);
 
@@ -3177,7 +3177,7 @@ TEST(user_notification_kill_in_middle)
 	ASSERT_GE(pid, 0);
 
 	if (pid == 0) {
-		ret = syscall(__NR_getpid);
+		ret = syscall(__NR_getppid);
 		exit(ret != USER_NOTIF_MAGIC);
 	}
 
@@ -3277,7 +3277,7 @@ TEST(user_notification_closed_listener)
 	long ret;
 	int status, listener;
 
-	listener = user_trap_syscall(__NR_getpid,
+	listener = user_trap_syscall(__NR_getppid,
 				     SECCOMP_FILTER_FLAG_NEW_LISTENER);
 	ASSERT_GE(listener, 0);
 
@@ -3288,7 +3288,7 @@ TEST(user_notification_closed_listener)
 	ASSERT_GE(pid, 0);
 	if (pid == 0) {
 		close(listener);
-		ret = syscall(__NR_getpid);
+		ret = syscall(__NR_getppid);
 		exit(ret != -1 && errno != ENOSYS);
 	}
 
@@ -3311,14 +3311,15 @@ TEST(user_notification_child_pid_ns)
 
 	ASSERT_EQ(unshare(CLONE_NEWPID), 0);
 
-	listener = user_trap_syscall(__NR_getpid, SECCOMP_FILTER_FLAG_NEW_LISTENER);
+	listener = user_trap_syscall(__NR_getppid,
+				     SECCOMP_FILTER_FLAG_NEW_LISTENER);
 	ASSERT_GE(listener, 0);
 
 	pid = fork();
 	ASSERT_GE(pid, 0);
 
 	if (pid == 0)
-		exit(syscall(__NR_getpid) != USER_NOTIF_MAGIC);
+		exit(syscall(__NR_getppid) != USER_NOTIF_MAGIC);
 
 	EXPECT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_RECV, &req), 0);
 	EXPECT_EQ(req.pid, pid);
@@ -3346,7 +3347,8 @@ TEST(user_notification_sibling_pid_ns)
 	struct seccomp_notif req = {};
 	struct seccomp_notif_resp resp = {};
 
-	listener = user_trap_syscall(__NR_getpid, SECCOMP_FILTER_FLAG_NEW_LISTENER);
+	listener = user_trap_syscall(__NR_getppid,
+				     SECCOMP_FILTER_FLAG_NEW_LISTENER);
 	ASSERT_GE(listener, 0);
 
 	pid = fork();
@@ -3359,7 +3361,7 @@ TEST(user_notification_sibling_pid_ns)
 		ASSERT_GE(pid2, 0);
 
 		if (pid2 == 0)
-			exit(syscall(__NR_getpid) != USER_NOTIF_MAGIC);
+			exit(syscall(__NR_getppid) != USER_NOTIF_MAGIC);
 
 		EXPECT_EQ(waitpid(pid2, &status, 0), pid2);
 		EXPECT_EQ(true, WIFEXITED(status));
@@ -3368,11 +3370,11 @@ TEST(user_notification_sibling_pid_ns)
 	}
 
 	/* Create the sibling ns, and sibling in it. */
-	EXPECT_EQ(unshare(CLONE_NEWPID), 0);
-	EXPECT_EQ(errno, 0);
+	ASSERT_EQ(unshare(CLONE_NEWPID), 0);
+	ASSERT_EQ(errno, 0);
 
 	pid2 = fork();
-	EXPECT_GE(pid2, 0);
+	ASSERT_GE(pid2, 0);
 
 	if (pid2 == 0) {
 		ASSERT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_RECV, &req), 0);
@@ -3380,7 +3382,7 @@ TEST(user_notification_sibling_pid_ns)
 		 * The pid should be 0, i.e. the task is in some namespace that
 		 * we can't "see".
 		 */
-		ASSERT_EQ(req.pid, 0);
+		EXPECT_EQ(req.pid, 0);
 
 		resp.id = req.id;
 		resp.error = 0;
@@ -3408,14 +3410,15 @@ TEST(user_notification_fault_recv)
 	struct seccomp_notif req = {};
 	struct seccomp_notif_resp resp = {};
 
-	listener = user_trap_syscall(__NR_getpid, SECCOMP_FILTER_FLAG_NEW_LISTENER);
+	listener = user_trap_syscall(__NR_getppid,
+				     SECCOMP_FILTER_FLAG_NEW_LISTENER);
 	ASSERT_GE(listener, 0);
 
 	pid = fork();
 	ASSERT_GE(pid, 0);
 
 	if (pid == 0)
-		exit(syscall(__NR_getpid) != USER_NOTIF_MAGIC);
+		exit(syscall(__NR_getppid) != USER_NOTIF_MAGIC);
 
 	/* Do a bad recv() */
 	EXPECT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_RECV, NULL), -1);
author maintainer
#8 lore raw

PATCH 8/137 kernfs: fix barrier usage in __kernfs_new_node()

In reply to Greg Kroah-Hartman

From: Andrea Parri <[email protected]>

commit 998267900cee901c5d1dfa029a6304d00acbc29f upstream.

smp_mb__before_atomic() can not be applied to atomic_set(). Remove the barrier and rely on RELEASE synchronization.

Fixes ba16b2846a8c6 ("kernfs: add an API to get kernfs node from inode number")
---
 fs/kernfs/dir.c |    5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
fs/kernfs/dir.c +2 -3
@@ -650,11 +650,10 @@ static struct kernfs_node *__kernfs_new_
 	kn->id.generation = gen;
 
 	/*
-	 * set ino first. This barrier is paired with atomic_inc_not_zero in
+	 * set ino first. This RELEASE is paired with atomic_inc_not_zero in
 	 * kernfs_find_and_get_node_by_ino
 	 */
-	smp_mb__before_atomic();
-	atomic_set(&kn->count, 1);
+	atomic_set_release(&kn->count, 1);
 	atomic_set(&kn->active, KN_DEACTIVATED_BIAS);
 	RB_CLEAR_NODE(&kn->rb);
 
author maintainer
#9 lore raw

PATCH 9/137 virt: vbox: Sanity-check parameter types for hgcm-calls coming from userspace

In reply to Greg Kroah-Hartman

From: Hans de Goede <[email protected]>

commit cf4f2ad6b87dda2dbe0573b1ebeb0273f8d4aac6 upstream.

Userspace can make host function calls, called hgcm-calls through the /dev/vboxguest device.

In this case we should not accept all hgcm-function-parameter-types, some are only valid for in kernel calls.

This commit adds proper hgcm-function-parameter-type validation to the ioctl for doing a hgcm-call from userspace.

---
 drivers/virt/vboxguest/vboxguest_core.c |   31 +++++++++++++++++++++++++++++++
 1 file changed, 31 insertions(+)
drivers/virt/vboxguest/vboxguest_core.c +31 -0
@@ -1263,6 +1263,20 @@ static int vbg_ioctl_hgcm_disconnect(str
 	return ret;
 }
 
+static bool vbg_param_valid(enum vmmdev_hgcm_function_parameter_type type)
+{
+	switch (type) {
+	case VMMDEV_HGCM_PARM_TYPE_32BIT:
+	case VMMDEV_HGCM_PARM_TYPE_64BIT:
+	case VMMDEV_HGCM_PARM_TYPE_LINADDR:
+	case VMMDEV_HGCM_PARM_TYPE_LINADDR_IN:
+	case VMMDEV_HGCM_PARM_TYPE_LINADDR_OUT:
+		return true;
+	default:
+		return false;
+	}
+}
+
 static int vbg_ioctl_hgcm_call(struct vbg_dev *gdev,
 			       struct vbg_session *session, bool f32bit,
 			       struct vbg_ioctl_hgcm_call *call)
@@ -1298,6 +1312,23 @@ static int vbg_ioctl_hgcm_call(struct vb
 	}
 	call->hdr.size_out = actual_size;
 
+	/* Validate parameter types */
+	if (f32bit) {
+		struct vmmdev_hgcm_function_parameter32 *parm =
+			VBG_IOCTL_HGCM_CALL_PARMS32(call);
+
+		for (i = 0; i < call->parm_count; i++)
+			if (!vbg_param_valid(parm[i].type))
+				return -EINVAL;
+	} else {
+		struct vmmdev_hgcm_function_parameter *parm =
+			VBG_IOCTL_HGCM_CALL_PARMS(call);
+
+		for (i = 0; i < call->parm_count; i++)
+			if (!vbg_param_valid(parm[i].type))
+				return -EINVAL;
+	}
+
 	/*
 	 * Validate the client id.
 	 */
author maintainer
#10 lore raw

PATCH 10/137 USB: serial: fix unthrottle races

In reply to Greg Kroah-Hartman

From: Johan Hovold <[email protected]>

commit 3f5edd58d040bfa4b74fb89bc02f0bc6b9cd06ab upstream.

Fix two long-standing bugs which could potentially lead to memory corruption or leave the port throttled until it is reopened (on weakly ordered systems), respectively, when read-URB completion races with unthrottle().

First, the URB must not be marked as free before processing is complete to prevent it from being submitted by unthrottle() on another CPU.

	CPU 1				CPU 2
	================		================
	complete()			unthrottle()
	  process_urb();
	  smp_mb__before_atomic();
	  set_bit(i, free);		  if (test_and_clear_bit(i, free))
	  					  submit_urb();

Second, the URB must be marked as free before checking the throttled flag to prevent unthrottle() on another CPU from failing to observe that the URB needs to be submitted if complete() sees that the throttled flag is set.

	CPU 1				CPU 2
	================		================
	complete()			unthrottle()
	  set_bit(i, free);		  throttled = 0;
	  smp_mb__after_atomic();	  smp_mb();
	  if (throttled)		  if (test_and_clear_bit(i, free))
	  	  return;			  submit_urb();

Note that test_and_clear_bit() only implies barriers when the test is successful. To handle the case where the URB is still in use an explicit barrier needs to be added to unthrottle() for the second race condition.

Fixes d83b405383c9 ("USB: serial: add support for multiple read urbs")
---
 drivers/usb/serial/generic.c |   39 ++++++++++++++++++++++++++++++++-------
 1 file changed, 32 insertions(+), 7 deletions(-)
drivers/usb/serial/generic.c +32 -7
@@ -376,6 +376,7 @@ void usb_serial_generic_read_bulk_callba
 	struct usb_serial_port *port = urb->context;
 	unsigned char *data = urb->transfer_buffer;
 	unsigned long flags;
+	bool stopped = false;
 	int status = urb->status;
 	int i;
 
@@ -383,33 +384,51 @@ void usb_serial_generic_read_bulk_callba
 		if (urb == port->read_urbs[i])
 			break;
 	}
-	set_bit(i, &port->read_urbs_free);
 
 	dev_dbg(&port->dev, "%s - urb %d, len %d\n", __func__, i,
 							urb->actual_length);
 	switch (status) {
 	case 0:
+		usb_serial_debug_data(&port->dev, __func__, urb->actual_length,
+							data);
+		port->serial->type->process_read_urb(urb);
 		break;
 	case -ENOENT:
 	case -ECONNRESET:
 	case -ESHUTDOWN:
 		dev_dbg(&port->dev, "%s - urb stopped: %d\n",
 							__func__, status);
-		return;
+		stopped = true;
+		break;
 	case -EPIPE:
 		dev_err(&port->dev, "%s - urb stopped: %d\n",
 							__func__, status);
-		return;
+		stopped = true;
+		break;
 	default:
 		dev_dbg(&port->dev, "%s - nonzero urb status: %d\n",
 							__func__, status);
-		goto resubmit;
+		break;
 	}
 
-	usb_serial_debug_data(&port->dev, __func__, urb->actual_length, data);
-	port->serial->type->process_read_urb(urb);
+	/*
+	 * Make sure URB processing is done before marking as free to avoid
+	 * racing with unthrottle() on another CPU. Matches the barriers
+	 * implied by the test_and_clear_bit() in
+	 * usb_serial_generic_submit_read_urb().
+	 */
+	smp_mb__before_atomic();
+	set_bit(i, &port->read_urbs_free);
+	/*
+	 * Make sure URB is marked as free before checking the throttled flag
+	 * to avoid racing with unthrottle() on another CPU. Matches the
+	 * smp_mb() in unthrottle().
+	 */
+	smp_mb__after_atomic();
+
+	if (stopped)
+		return;
 
-resubmit:
 	/* Throttle the device if requested by tty */
 	spin_lock_irqsave(&port->lock, flags);
 	port->throttled = port->throttle_req;
@@ -484,6 +503,12 @@ void usb_serial_generic_unthrottle(struc
 	port->throttled = port->throttle_req = 0;
 	spin_unlock_irq(&port->lock);
 
+	/*
+	 * Matches the smp_mb__after_atomic() in
+	 * usb_serial_generic_read_bulk_callback().
+	 */
+	smp_mb();
+
 	if (was_throttled)
 		usb_serial_generic_submit_read_urbs(port, GFP_KERNEL);
 }
author maintainer
#11 lore raw

PATCH 11/137 iio: adc: xilinx: fix potential use-after-free on remove

In reply to Greg Kroah-Hartman

[ Upstream commit 62039b6aef63380ba7a37c113bbaeee8a55c5342 ]

When cancel_delayed_work() returns, the delayed work may still be running. This means that the core could potentially free the private structure (struct xadc) while the delayed work is still using it. This is a potential use-after-free.

Fix by calling cancel_delayed_work_sync(), which waits for any residual work to finish before returning.

---
 drivers/iio/adc/xilinx-xadc-core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iio/adc/xilinx-xadc-core.c +1 -1
index 3f6be5ac049a8..1960694e80076 100644
@@ -1320,7 +1320,7 @@ static int xadc_remove(struct platform_device *pdev)
 	}
 	free_irq(xadc->irq, indio_dev);
 	clk_disable_unprepare(xadc->clk);
-	cancel_delayed_work(&xadc->zynq_unmask_work);
+	cancel_delayed_work_sync(&xadc->zynq_unmask_work);
 	kfree(xadc->data);
 	kfree(indio_dev->channels);
 
Signature · 1 line
2.20.1
author maintainer
#12 lore raw

PATCH 12/137 iio: adc: xilinx: fix potential use-after-free on probe

In reply to Greg Kroah-Hartman

[ Upstream commit 862e4644fd2d7df8998edc65e0963ea2f567bde9 ]

If probe errors out after request_irq(), its error path does not explicitly cancel the delayed work, which may have been scheduled by the interrupt handler.

This means the delayed work may still be running when the core frees the private structure (struct xadc). This is a potential use-after-free.

Fix by inserting cancel_delayed_work_sync() in the probe error path.

---
 drivers/iio/adc/xilinx-xadc-core.c | 1 +
 1 file changed, 1 insertion(+)
drivers/iio/adc/xilinx-xadc-core.c +1 -0
index 1960694e80076..15e1a103f37da 100644
@@ -1290,6 +1290,7 @@ static int xadc_probe(struct platform_device *pdev)
 
 err_free_irq:
 	free_irq(xadc->irq, indio_dev);
+	cancel_delayed_work_sync(&xadc->zynq_unmask_work);
 err_clk_disable_unprepare:
 	clk_disable_unprepare(xadc->clk);
 err_free_samplerate_trigger:
Signature · 1 line
2.20.1
author maintainer
#13 lore raw

PATCH 13/137 iio: adc: xilinx: prevent touching unclocked h/w on remove

In reply to Greg Kroah-Hartman

[ Upstream commit 2e4b88f73966adead360e47621df0183586fac32 ]

In remove, the clock is disabled before canceling the delayed work. This means that the delayed work may be touching unclocked hardware.

Fix by disabling the clock after the delayed work is fully canceled. This is consistent with the probe error path order.

---
 drivers/iio/adc/xilinx-xadc-core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iio/adc/xilinx-xadc-core.c +1 -1
index 15e1a103f37da..1ae86e7359f73 100644
@@ -1320,8 +1320,8 @@ static int xadc_remove(struct platform_device *pdev)
 		iio_triggered_buffer_cleanup(indio_dev);
 	}
 	free_irq(xadc->irq, indio_dev);
-	clk_disable_unprepare(xadc->clk);
 	cancel_delayed_work_sync(&xadc->zynq_unmask_work);
+	clk_disable_unprepare(xadc->clk);
 	kfree(xadc->data);
 	kfree(indio_dev->channels);
 
Signature · 1 line
2.20.1
author maintainer
#14 lore raw

PATCH 14/137 acpi/nfit: Always dump _DSM output payload

In reply to Greg Kroah-Hartman

[ Upstream commit 351f339faa308c1c1461314a18c832239a841ca0 ]

The dynamic-debug statements for command payload output only get emitted when the command is not ND_CMD_CALL. Move the output payload dumping ahead of the early return path for ND_CMD_CALL.

Fixes 31eca76ba2fc9 ("...whitelisted dimm command marshaling mechanism")
---
 drivers/acpi/nfit/core.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
drivers/acpi/nfit/core.c +6 -6
index 4be4dc3e8aa62..38ec79bb3edde 100644
@@ -563,6 +563,12 @@ int acpi_nfit_ctl(struct nvdimm_bus_descriptor *nd_desc, struct nvdimm *nvdimm,
 		goto out;
 	}
 
+	dev_dbg(dev, "%s cmd: %s output length: %d\n", dimm_name,
+			cmd_name, out_obj->buffer.length);
+	print_hex_dump_debug(cmd_name, DUMP_PREFIX_OFFSET, 4, 4,
+			out_obj->buffer.pointer,
+			min_t(u32, 128, out_obj->buffer.length), true);
+
 	if (call_pkg) {
 		call_pkg->nd_fw_size = out_obj->buffer.length;
 		memcpy(call_pkg->nd_payload + call_pkg->nd_size_in,
@@ -581,12 +587,6 @@ int acpi_nfit_ctl(struct nvdimm_bus_descriptor *nd_desc, struct nvdimm *nvdimm,
 		return 0;
 	}
 
-	dev_dbg(dev, "%s cmd: %s output length: %d\n", dimm_name,
-			cmd_name, out_obj->buffer.length);
-	print_hex_dump_debug(cmd_name, DUMP_PREFIX_OFFSET, 4, 4,
-			out_obj->buffer.pointer,
-			min_t(u32, 128, out_obj->buffer.length), true);
-
 	for (i = 0, offset = 0; i < desc->out_num; i++) {
 		u32 out_size = nd_cmd_out_size(nvdimm, cmd, desc, i, buf,
 				(u32 *) out_obj->buffer.pointer,
Signature · 1 line
2.20.1
author maintainer
#15 lore raw

PATCH 15/137 libnvdimm/namespace: Fix a potential NULL pointer dereference

In reply to Greg Kroah-Hartman

[ Upstream commit 55c1fc0af29a6c1b92f217b7eb7581a882e0c07c ]

In case kmemdup fails, the fix goes to blk_err to avoid NULL pointer dereference.

---
 drivers/nvdimm/namespace_devs.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/nvdimm/namespace_devs.c +4 -1
index 33a3b23b3db71..e761b29f71606 100644
@@ -2249,9 +2249,12 @@ static struct device *create_namespace_blk(struct nd_region *nd_region,
 	if (!nsblk->uuid)
 		goto blk_err;
 	memcpy(name, nd_label->name, NSLABEL_NAME_LEN);
-	if (name[0])
+	if (name[0]) {
 		nsblk->alt_name = kmemdup(name, NSLABEL_NAME_LEN,
 				GFP_KERNEL);
+		if (!nsblk->alt_name)
+			goto blk_err;
+	}
 	res = nsblk_add_resource(nd_region, ndd, nsblk,
 			__le64_to_cpu(nd_label->dpa));
 	if (!res)
Signature · 1 line
2.20.1
author maintainer
#16 lore raw

PATCH 16/137 HID: input: add mapping for Expose/Overview key

In reply to Greg Kroah-Hartman

[ Upstream commit 96dd86871e1fffbc39e4fa61c9c75ec54ee9af0f ]

According to HUTRR77 usage 0x29f from the consumer page is reserved for the Desktop application to present all running user’s application windows. Linux defines KEY_SCALE to request Compiz Scale (Expose) mode, so let's add the mapping.

---
 drivers/hid/hid-input.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/hid/hid-input.c +2 -0
index ff92a7b2fc897..468da6f6765db 100644
@@ -1042,6 +1042,8 @@ static void hidinput_configure_usage(struct hid_input *hidinput, struct hid_fiel
 		case 0x2cb: map_key_clear(KEY_KBDINPUTASSIST_ACCEPT);	break;
 		case 0x2cc: map_key_clear(KEY_KBDINPUTASSIST_CANCEL);	break;
 
+		case 0x29f: map_key_clear(KEY_SCALE);		break;
+
 		default: map_key_clear(KEY_UNKNOWN);
 		}
 		break;
Signature · 1 line
2.20.1
author maintainer
#17 lore raw

PATCH 17/137 HID: input: add mapping for keyboard Brightness Up/Down/Toggle keys

In reply to Greg Kroah-Hartman

[ Upstream commit 7975a1d6a7afeb3eb61c971a153d24dd8fa032f3 ]

According to HUTRR73 usages 0x79, 0x7a and 0x7c from the consumer page correspond to Brightness Up/Down/Toggle keys, so let's add the mappings.

---
 drivers/hid/hid-input.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/hid/hid-input.c +4 -0
index 468da6f6765db..290efac7e6bfd 100644
@@ -908,6 +908,10 @@ static void hidinput_configure_usage(struct hid_input *hidinput, struct hid_fiel
 		case 0x074: map_key_clear(KEY_BRIGHTNESS_MAX);		break;
 		case 0x075: map_key_clear(KEY_BRIGHTNESS_AUTO);		break;
 
+		case 0x079: map_key_clear(KEY_KBDILLUMUP);	break;
+		case 0x07a: map_key_clear(KEY_KBDILLUMDOWN);	break;
+		case 0x07c: map_key_clear(KEY_KBDILLUMTOGGLE);	break;
+
 		case 0x082: map_key_clear(KEY_VIDEO_NEXT);	break;
 		case 0x083: map_key_clear(KEY_LAST);		break;
 		case 0x084: map_key_clear(KEY_ENTER);		break;
Signature · 1 line
2.20.1
author maintainer
#18 lore raw

PATCH 18/137 HID: input: add mapping for "Toggle Display" key

In reply to Greg Kroah-Hartman

[ Upstream commit c01908a14bf735b871170092807c618bb9dae654 ]

According to HUT 1.12 usage 0xb5 from the generic desktop page is reserved for switching between external and internal display, so let's add the mapping.

---
 drivers/hid/hid-input.c | 8 ++++++++
 1 file changed, 8 insertions(+)
drivers/hid/hid-input.c +8 -0
index 290efac7e6bfd..4f119300ce3f5 100644
@@ -677,6 +677,14 @@ static void hidinput_configure_usage(struct hid_input *hidinput, struct hid_fiel
 			break;
 		}
 
+		if ((usage->hid & 0xf0) == 0xb0) {	/* SC - Display */
+			switch (usage->hid & 0xf) {
+			case 0x05: map_key_clear(KEY_SWITCHVIDEOMODE); break;
+			default: goto ignore;
+			}
+			break;
+		}
+
 		/*
 		 * Some lazy vendors declare 255 usages for System Control,
 		 * leading to the creation of ABS_X|Y axis and too many others.
Signature · 1 line
2.20.1
author maintainer
#19 lore raw

PATCH 19/137 libnvdimm/btt: Fix a kmemdup failure check

In reply to Greg Kroah-Hartman

[ Upstream commit 486fa92df4707b5df58d6508728bdb9321a59766 ]

In case kmemdup fails, the fix releases resources and returns to avoid the NULL pointer dereference.

---
 drivers/nvdimm/btt_devs.c | 18 +++++++++++++-----
 1 file changed, 13 insertions(+), 5 deletions(-)
drivers/nvdimm/btt_devs.c +13 -5
index 795ad4ff35caf..e341498876cad 100644
@@ -190,14 +190,15 @@ static struct device *__nd_btt_create(struct nd_region *nd_region,
 		return NULL;
 
 	nd_btt->id = ida_simple_get(&nd_region->btt_ida, 0, 0, GFP_KERNEL);
-	if (nd_btt->id < 0) {
-		kfree(nd_btt);
-		return NULL;
-	}
+	if (nd_btt->id < 0)
+		goto out_nd_btt;
 
 	nd_btt->lbasize = lbasize;
-	if (uuid)
+	if (uuid) {
 		uuid = kmemdup(uuid, 16, GFP_KERNEL);
+		if (!uuid)
+			goto out_put_id;
+	}
 	nd_btt->uuid = uuid;
 	dev = &nd_btt->dev;
 	dev_set_name(dev, "btt%d.%d", nd_region->id, nd_btt->id);
@@ -212,6 +213,13 @@ static struct device *__nd_btt_create(struct nd_region *nd_region,
 		return NULL;
 	}
 	return dev;
+
+out_put_id:
+	ida_simple_remove(&nd_region->btt_ida, nd_btt->id);
+
+out_nd_btt:
+	kfree(nd_btt);
+	return NULL;
 }
 
 struct device *nd_btt_create(struct nd_region *nd_region)
Signature · 1 line
2.20.1
author maintainer
#20 lore raw

PATCH 20/137 s390/dasd: Fix capacity calculation for large volumes

In reply to Greg Kroah-Hartman

[ Upstream commit 2cc9637ce825f3a9f51f8f78af7474e9e85bfa5f ]

The DASD driver incorrectly limits the maximum number of blocks of ECKD DASD volumes to 32 bit numbers. Volumes with a capacity greater than 2^32-1 blocks are incorrectly recognized as smaller volumes.

This results in the following volume capacity limits depending on the formatted block size:

  BLKSIZE  MAX_GB   MAX_CYL
      512    2047   5843492
     1024    4095   8676701
     2048    8191  13634816
     4096   16383  23860929

The same problem occurs when a volume with more than 17895697 cylinders is accessed in raw-track-access mode.

Fix this problem by adding an explicit type cast when calculating the maximum number of blocks.

---
 drivers/s390/block/dasd_eckd.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/s390/block/dasd_eckd.c +3 -3
index 6e294b4d3635f..f89f9d02e7884 100644
@@ -2004,14 +2004,14 @@ static int dasd_eckd_end_analysis(struct dasd_block *block)
 	blk_per_trk = recs_per_track(&private->rdc_data, 0, block->bp_block);
 
 raw:
-	block->blocks = (private->real_cyl *
+	block->blocks = ((unsigned long) private->real_cyl *
 			  private->rdc_data.trk_per_cyl *
 			  blk_per_trk);
 
 	dev_info(&device->cdev->dev,
-		 "DASD with %d KB/block, %d KB total size, %d KB/track, "
+		 "DASD with %u KB/block, %lu KB total size, %u KB/track, "
 		 "%s\n", (block->bp_block >> 10),
-		 ((private->real_cyl *
+		 (((unsigned long) private->real_cyl *
 		   private->rdc_data.trk_per_cyl *
 		   blk_per_trk * (block->bp_block >> 9)) >> 1),
 		 ((blk_per_trk * block->bp_block) >> 10),
Signature · 1 line
2.20.1
author maintainer
#21 lore raw

PATCH 21/137 mac80211: fix unaligned access in mesh table hash function

In reply to Greg Kroah-Hartman

[ Upstream commit 40586e3fc400c00c11151804dcdc93f8c831c808 ]

The pointer to the last four bytes of the address is not guaranteed to be aligned, so we need to use __get_unaligned_cpu32 here

---
 net/mac80211/mesh_pathtbl.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/mac80211/mesh_pathtbl.c +1 -1
index 88a6d5e18ccc9..ac1f5db529945 100644
@@ -23,7 +23,7 @@ static void mesh_path_free_rcu(struct mesh_table *tbl, struct mesh_path *mpath);
 static u32 mesh_table_hash(const void *addr, u32 len, u32 seed)
 {
 	/* Use last four bytes of hw addr as hash index */
-	return jhash_1word(*(u32 *)(addr+2), seed);
+	return jhash_1word(__get_unaligned_cpu32((u8 *)addr + 2), seed);
 }
 
 static const struct rhashtable_params mesh_rht_params = {
Signature · 1 line
2.20.1
author maintainer
#22 lore raw

PATCH 22/137 mac80211: Increase MAX_MSG_LEN

In reply to Greg Kroah-Hartman

[ Upstream commit 78be2d21cc1cd3069c6138dcfecec62583130171 ]

Looks that 100 chars isn't enough for messages, as we keep getting warnings popping from different places due to message shortening. Instead of trying to shorten the prints, just increase the buffer size.

---
 net/mac80211/trace_msg.h | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
net/mac80211/trace_msg.h +6 -1
index 366b9e6f043e2..40141df09f255 100644
@@ -1,4 +1,9 @@
 /* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Portions of this file
+ * Copyright (C) 2019 Intel Corporation
+ */
+
 #ifdef CONFIG_MAC80211_MESSAGE_TRACING
 
 #if !defined(__MAC80211_MSG_DRIVER_TRACE) || defined(TRACE_HEADER_MULTI_READ)
@@ -11,7 +16,7 @@
 #undef TRACE_SYSTEM
 #define TRACE_SYSTEM mac80211_msg
 
-#define MAX_MSG_LEN	100
+#define MAX_MSG_LEN	120
 
 DECLARE_EVENT_CLASS(mac80211_msg_event,
 	TP_PROTO(struct va_format *vaf),
Signature · 1 line
2.20.1
author maintainer
#23 lore raw

PATCH 23/137 cfg80211: Handle WMM rules in regulatory domain intersection

In reply to Greg Kroah-Hartman

[ Upstream commit 08a75a887ee46828b54600f4bb7068d872a5edd5 ]

The support added for regulatory WMM rules did not handle the case of regulatory domain intersections. Fix it.

Fixes 230ebaa189af ("cfg80211: read wmm rules from regulatory database")
---
 net/wireless/reg.c | 39 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 39 insertions(+)
net/wireless/reg.c +39 -0
index dd58b9909ac99..649c89946dec1 100644
@@ -1298,6 +1298,16 @@ reg_intersect_dfs_region(const enum nl80211_dfs_regions dfs_region1,
 	return dfs_region1;
 }
 
+static void reg_wmm_rules_intersect(const struct ieee80211_wmm_ac *wmm_ac1,
+				    const struct ieee80211_wmm_ac *wmm_ac2,
+				    struct ieee80211_wmm_ac *intersect)
+{
+	intersect->cw_min = max_t(u16, wmm_ac1->cw_min, wmm_ac2->cw_min);
+	intersect->cw_max = max_t(u16, wmm_ac1->cw_max, wmm_ac2->cw_max);
+	intersect->cot = min_t(u16, wmm_ac1->cot, wmm_ac2->cot);
+	intersect->aifsn = max_t(u8, wmm_ac1->aifsn, wmm_ac2->aifsn);
+}
+
 /*
  * Helper for regdom_intersect(), this does the real
  * mathematical intersection fun
@@ -1312,6 +1322,8 @@ static int reg_rules_intersect(const struct ieee80211_regdomain *rd1,
 	struct ieee80211_freq_range *freq_range;
 	const struct ieee80211_power_rule *power_rule1, *power_rule2;
 	struct ieee80211_power_rule *power_rule;
+	const struct ieee80211_wmm_rule *wmm_rule1, *wmm_rule2;
+	struct ieee80211_wmm_rule *wmm_rule;
 	u32 freq_diff, max_bandwidth1, max_bandwidth2;
 
 	freq_range1 = &rule1->freq_range;
@@ -1322,6 +1334,10 @@ static int reg_rules_intersect(const struct ieee80211_regdomain *rd1,
 	power_rule2 = &rule2->power_rule;
 	power_rule = &intersected_rule->power_rule;
 
+	wmm_rule1 = &rule1->wmm_rule;
+	wmm_rule2 = &rule2->wmm_rule;
+	wmm_rule = &intersected_rule->wmm_rule;
+
 	freq_range->start_freq_khz = max(freq_range1->start_freq_khz,
 					 freq_range2->start_freq_khz);
 	freq_range->end_freq_khz = min(freq_range1->end_freq_khz,
@@ -1365,6 +1381,29 @@ static int reg_rules_intersect(const struct ieee80211_regdomain *rd1,
 	intersected_rule->dfs_cac_ms = max(rule1->dfs_cac_ms,
 					   rule2->dfs_cac_ms);
 
+	if (rule1->has_wmm && rule2->has_wmm) {
+		u8 ac;
+
+		for (ac = 0; ac < IEEE80211_NUM_ACS; ac++) {
+			reg_wmm_rules_intersect(&wmm_rule1->client[ac],
+						&wmm_rule2->client[ac],
+						&wmm_rule->client[ac]);
+			reg_wmm_rules_intersect(&wmm_rule1->ap[ac],
+						&wmm_rule2->ap[ac],
+						&wmm_rule->ap[ac]);
+		}
+
+		intersected_rule->has_wmm = true;
+	} else if (rule1->has_wmm) {
+		*wmm_rule = *wmm_rule1;
+		intersected_rule->has_wmm = true;
+	} else if (rule2->has_wmm) {
+		*wmm_rule = *wmm_rule2;
+		intersected_rule->has_wmm = true;
+	} else {
+		intersected_rule->has_wmm = false;
+	}
+
 	if (!is_valid_reg_rule(intersected_rule))
 		return -EINVAL;
 
Signature · 1 line
2.20.1
author maintainer
#24 lore raw

PATCH 24/137 mac80211: fix memory accounting with A-MSDU aggregation

In reply to Greg Kroah-Hartman

[ Upstream commit eb9b64e3a9f8483e6e54f4e03b2ae14ae5db2690 ]

skb->truesize can change due to memory reallocation or when adding extra fragments. Adjust fq->memory_usage accordingly

---
 net/mac80211/tx.c | 3 +++
 1 file changed, 3 insertions(+)
net/mac80211/tx.c +3 -0
index 928f13a208b05..714d80e48a102 100644
@@ -3214,6 +3214,7 @@ static bool ieee80211_amsdu_aggregate(struct ieee80211_sub_if_data *sdata,
 	u8 max_subframes = sta->sta.max_amsdu_subframes;
 	int max_frags = local->hw.max_tx_fragments;
 	int max_amsdu_len = sta->sta.max_amsdu_len;
+	int orig_truesize;
 	__be16 len;
 	void *data;
 	bool ret = false;
@@ -3254,6 +3255,7 @@ static bool ieee80211_amsdu_aggregate(struct ieee80211_sub_if_data *sdata,
 	if (!head || skb_is_gso(head))
 		goto out;
 
+	orig_truesize = head->truesize;
 	orig_len = head->len;
 
 	if (skb->len + head->len > max_amsdu_len)
@@ -3311,6 +3313,7 @@ static bool ieee80211_amsdu_aggregate(struct ieee80211_sub_if_data *sdata,
 	*frag_tail = skb;
 
 out_recalc:
+	fq->memory_usage += head->truesize - orig_truesize;
 	if (head->len != orig_len) {
 		flow->backlog += head->len - orig_len;
 		tin->backlog_bytes += head->len - orig_len;
Signature · 1 line
2.20.1
author maintainer
#25 lore raw

PATCH 25/137 nl80211: Add NL80211_FLAG_CLEAR_SKB flag for other NL commands

In reply to Greg Kroah-Hartman

[ Upstream commit d6db02a88a4aaa1cd7105137c67ddec7f3bdbc05 ]

This commit adds NL80211_FLAG_CLEAR_SKB flag to other NL commands that carry key data to ensure they do not stick around on heap after the SKB is freed.

Also introduced this flag for NL80211_CMD_VENDOR as there are sub commands which configure the keys.

---
 net/wireless/nl80211.c | 18 ++++++++++++------
 1 file changed, 12 insertions(+), 6 deletions(-)
net/wireless/nl80211.c +12 -6
index d91a408db113e..156ce708b5330 100644
@@ -13596,7 +13596,8 @@ static const struct genl_ops nl80211_ops[] = {
 		.policy = nl80211_policy,
 		.flags = GENL_UNS_ADMIN_PERM,
 		.internal_flags = NL80211_FLAG_NEED_NETDEV_UP |
-				  NL80211_FLAG_NEED_RTNL,
+				  NL80211_FLAG_NEED_RTNL |
+				  NL80211_FLAG_CLEAR_SKB,
 	},
 	{
 		.cmd = NL80211_CMD_DEAUTHENTICATE,
@@ -13647,7 +13648,8 @@ static const struct genl_ops nl80211_ops[] = {
 		.policy = nl80211_policy,
 		.flags = GENL_UNS_ADMIN_PERM,
 		.internal_flags = NL80211_FLAG_NEED_NETDEV_UP |
-				  NL80211_FLAG_NEED_RTNL,
+				  NL80211_FLAG_NEED_RTNL |
+				  NL80211_FLAG_CLEAR_SKB,
 	},
 	{
 		.cmd = NL80211_CMD_UPDATE_CONNECT_PARAMS,
@@ -13655,7 +13657,8 @@ static const struct genl_ops nl80211_ops[] = {
 		.policy = nl80211_policy,
 		.flags = GENL_ADMIN_PERM,
 		.internal_flags = NL80211_FLAG_NEED_NETDEV_UP |
-				  NL80211_FLAG_NEED_RTNL,
+				  NL80211_FLAG_NEED_RTNL |
+				  NL80211_FLAG_CLEAR_SKB,
 	},
 	{
 		.cmd = NL80211_CMD_DISCONNECT,
@@ -13684,7 +13687,8 @@ static const struct genl_ops nl80211_ops[] = {
 		.policy = nl80211_policy,
 		.flags = GENL_UNS_ADMIN_PERM,
 		.internal_flags = NL80211_FLAG_NEED_NETDEV_UP |
-				  NL80211_FLAG_NEED_RTNL,
+				  NL80211_FLAG_NEED_RTNL |
+				  NL80211_FLAG_CLEAR_SKB,
 	},
 	{
 		.cmd = NL80211_CMD_DEL_PMKSA,
@@ -14036,7 +14040,8 @@ static const struct genl_ops nl80211_ops[] = {
 		.policy = nl80211_policy,
 		.flags = GENL_UNS_ADMIN_PERM,
 		.internal_flags = NL80211_FLAG_NEED_WIPHY |
-				  NL80211_FLAG_NEED_RTNL,
+				  NL80211_FLAG_NEED_RTNL |
+				  NL80211_FLAG_CLEAR_SKB,
 	},
 	{
 		.cmd = NL80211_CMD_SET_QOS_MAP,
@@ -14091,7 +14096,8 @@ static const struct genl_ops nl80211_ops[] = {
 		.doit = nl80211_set_pmk,
 		.policy = nl80211_policy,
 		.internal_flags = NL80211_FLAG_NEED_NETDEV_UP |
-				  NL80211_FLAG_NEED_RTNL,
+				  NL80211_FLAG_NEED_RTNL |
+				  NL80211_FLAG_CLEAR_SKB,
 	},
 	{
 		.cmd = NL80211_CMD_DEL_PMK,
Signature · 1 line
2.20.1
author maintainer
#26 lore raw

PATCH 26/137 libnvdimm/security: provide fix for secure-erase to use zero-key

In reply to Greg Kroah-Hartman

[ Upstream commit 037c8489ade669e0f09ad40d5b91e5e1159a14b1 ]

Add a zero key in order to standardize hardware that want a key of 0's to be passed. Some platforms defaults to a zero-key with security enabled rather than allow the OS to enable the security. The zero key would allow us to manage those platform as well. This also adds a fix to secure erase so it can use the zero key to do crypto erase. Some other security commands already use zero keys. This introduces a standard zero-key to allow unification of semantics cross nvdimm security commands.

---
 drivers/nvdimm/security.c        | 17 ++++++++++++-----
 tools/testing/nvdimm/test/nfit.c | 11 +++++++++--
 2 files changed, 21 insertions(+), 7 deletions(-)
drivers/nvdimm/security.c +12 -5
index f8bb746a549f7..6bea6852bf278 100644
@@ -22,6 +22,8 @@ static bool key_revalidate = true;
 module_param(key_revalidate, bool, 0444);
 MODULE_PARM_DESC(key_revalidate, "Require key validation at init.");
 
+static const char zero_key[NVDIMM_PASSPHRASE_LEN];
+
 static void *key_data(struct key *key)
 {
 	struct encrypted_key_payload *epayload = dereference_key_locked(key);
@@ -286,8 +288,9 @@ int nvdimm_security_erase(struct nvdimm *nvdimm, unsigned int keyid,
 {
 	struct device *dev = &nvdimm->dev;
 	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);
-	struct key *key;
+	struct key *key = NULL;
 	int rc;
+	const void *data;
 
 	/* The bus lock should be held at the top level of the call stack */
 	lockdep_assert_held(&nvdimm_bus->reconfig_mutex);
@@ -319,11 +322,15 @@ int nvdimm_security_erase(struct nvdimm *nvdimm, unsigned int keyid,
 		return -EOPNOTSUPP;
 	}
 
-	key = nvdimm_lookup_user_key(nvdimm, keyid, NVDIMM_BASE_KEY);
-	if (!key)
-		return -ENOKEY;
+	if (keyid != 0) {
+		key = nvdimm_lookup_user_key(nvdimm, keyid, NVDIMM_BASE_KEY);
+		if (!key)
+			return -ENOKEY;
+		data = key_data(key);
+	} else
+		data = zero_key;
 
-	rc = nvdimm->sec.ops->erase(nvdimm, key_data(key), pass_type);
+	rc = nvdimm->sec.ops->erase(nvdimm, data, pass_type);
 	dev_dbg(dev, "key: %d erase%s: %s\n", key_serial(key),
 			pass_type == NVDIMM_MASTER ? "(master)" : "(user)",
 			rc == 0 ? "success" : "fail");
tools/testing/nvdimm/test/nfit.c +9 -2
index b579f962451d6..cad719876ef45 100644
@@ -225,6 +225,8 @@ static struct workqueue_struct *nfit_wq;
 
 static struct gen_pool *nfit_pool;
 
+static const char zero_key[NVDIMM_PASSPHRASE_LEN];
+
 static struct nfit_test *to_nfit_test(struct device *dev)
 {
 	struct platform_device *pdev = to_platform_device(dev);
@@ -1059,8 +1061,7 @@ static int nd_intel_test_cmd_secure_erase(struct nfit_test *t,
 	struct device *dev = &t->pdev.dev;
 	struct nfit_test_sec *sec = &dimm_sec_info[dimm];
 
-	if (!(sec->state & ND_INTEL_SEC_STATE_ENABLED) ||
-			(sec->state & ND_INTEL_SEC_STATE_FROZEN)) {
+	if (sec->state & ND_INTEL_SEC_STATE_FROZEN) {
 		nd_cmd->status = ND_INTEL_STATUS_INVALID_STATE;
 		dev_dbg(dev, "secure erase: wrong security state\n");
 	} else if (memcmp(nd_cmd->passphrase, sec->passphrase,
@@ -1068,6 +1069,12 @@ static int nd_intel_test_cmd_secure_erase(struct nfit_test *t,
 		nd_cmd->status = ND_INTEL_STATUS_INVALID_PASS;
 		dev_dbg(dev, "secure erase: wrong passphrase\n");
 	} else {
+		if (!(sec->state & ND_INTEL_SEC_STATE_ENABLED)
+				&& (memcmp(nd_cmd->passphrase, zero_key,
+					ND_INTEL_PASSPHRASE_SIZE) != 0)) {
+			dev_dbg(dev, "invalid zero key\n");
+			return 0;
+		}
 		memset(sec->passphrase, 0, ND_INTEL_PASSPHRASE_SIZE);
 		memset(sec->master_passphrase, 0, ND_INTEL_PASSPHRASE_SIZE);
 		sec->state = 0;
Signature · 1 line
2.20.1
author maintainer
#27 lore raw

PATCH 27/137 libnvdimm/pmem: fix a possible OOB access when read and write pmem

In reply to Greg Kroah-Hartman

[ Upstream commit 9dc6488e84b0f64df17672271664752488cd6a25 ]

If offset is not zero and length is bigger than PAGE_SIZE, this will cause to out of boundary access to a page memory

Fixes 98cc093cba1e ("block, THP: make block_device_operations.rw_page support THP")
---
 drivers/nvdimm/pmem.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/nvdimm/pmem.c +4 -4
index bc2f700feef8a..0279eb1da3ef5 100644
@@ -113,13 +113,13 @@ static void write_pmem(void *pmem_addr, struct page *page,
 
 	while (len) {
 		mem = kmap_atomic(page);
-		chunk = min_t(unsigned int, len, PAGE_SIZE);
+		chunk = min_t(unsigned int, len, PAGE_SIZE - off);
 		memcpy_flushcache(pmem_addr, mem + off, chunk);
 		kunmap_atomic(mem);
 		len -= chunk;
 		off = 0;
 		page++;
-		pmem_addr += PAGE_SIZE;
+		pmem_addr += chunk;
 	}
 }
 
@@ -132,7 +132,7 @@ static blk_status_t read_pmem(struct page *page, unsigned int off,
 
 	while (len) {
 		mem = kmap_atomic(page);
-		chunk = min_t(unsigned int, len, PAGE_SIZE);
+		chunk = min_t(unsigned int, len, PAGE_SIZE - off);
 		rem = memcpy_mcsafe(mem + off, pmem_addr, chunk);
 		kunmap_atomic(mem);
 		if (rem)
@@ -140,7 +140,7 @@ static blk_status_t read_pmem(struct page *page, unsigned int off,
 		len -= chunk;
 		off = 0;
 		page++;
-		pmem_addr += PAGE_SIZE;
+		pmem_addr += chunk;
 	}
 	return BLK_STS_OK;
 }
Signature · 1 line
2.20.1
author maintainer
#28 lore raw

PATCH 28/137 tools/testing/nvdimm: Retain security state after overwrite

In reply to Greg Kroah-Hartman

[ Upstream commit 2170a0d53bee1a6c1a4ebd042f99d85aafc6c0ea ]

Overwrite retains the security state after completion of operation. Fix nfit_test to reflect this so that the kernel can test the behavior it is more likely to see in practice.

Fixes 926f74802cb1 ("tools/testing/nvdimm: Add overwrite support for nfit_test")
---
 tools/testing/nvdimm/test/nfit.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
tools/testing/nvdimm/test/nfit.c +4 -2
index cad719876ef45..85ffdcfa596b5 100644
@@ -146,6 +146,7 @@ static int dimm_fail_cmd_code[ARRAY_SIZE(handle)];
 struct nfit_test_sec {
 	u8 state;
 	u8 ext_state;
+	u8 old_state;
 	u8 passphrase[32];
 	u8 master_passphrase[32];
 	u64 overwrite_end_time;
@@ -1100,7 +1101,7 @@ static int nd_intel_test_cmd_overwrite(struct nfit_test *t,
 		return 0;
 	}
 
-	memset(sec->passphrase, 0, ND_INTEL_PASSPHRASE_SIZE);
+	sec->old_state = sec->state;
 	sec->state = ND_INTEL_SEC_STATE_OVERWRITE;
 	dev_dbg(dev, "overwrite progressing.\n");
 	sec->overwrite_end_time = get_jiffies_64() + 5 * HZ;
@@ -1122,7 +1123,8 @@ static int nd_intel_test_cmd_query_overwrite(struct nfit_test *t,
 
 	if (time_is_before_jiffies64(sec->overwrite_end_time)) {
 		sec->overwrite_end_time = 0;
-		sec->state = 0;
+		sec->state = sec->old_state;
+		sec->old_state = 0;
 		sec->ext_state = ND_INTEL_SEC_ESTATE_ENABLED;
 		dev_dbg(dev, "overwrite is complete\n");
 	} else
Signature · 1 line
2.20.1
author maintainer
#29 lore raw

PATCH 29/137 s390/3270: fix lockdep false positive on view->lock

In reply to Greg Kroah-Hartman

[ Upstream commit 5712f3301a12c0c3de9cc423484496b0464f2faf ]

The spinlock in the raw3270_view structure is used by con3270, tty3270 and fs3270 in different ways. For con3270 the lock can be acquired in irq context, for tty3270 and fs3270 the highest context is bh.

Lockdep sees the view->lock as a single class and if the 3270 driver is used for the console the following message is generated:

WARNING: inconsistent lock state
5.1.0-rc3-05157-g5c168033979d #12 Not tainted
--------------------------------
inconsistent {IN-HARDIRQ-W} -> {HARDIRQ-ON-W} usage.
swapper/0/1 [HC0[0]:SC1[1]:HE1:SE0] takes:
(____ptrval____) (&(&view->lock)->rlock){?.-.}, at: tty3270_update+0x7c/0x330

Introduce a lockdep subclass for the view lock to distinguish bh from irq locks.

---
 drivers/s390/char/con3270.c | 2 +-
 drivers/s390/char/fs3270.c  | 3 ++-
 drivers/s390/char/raw3270.c | 3 ++-
 drivers/s390/char/raw3270.h | 4 +++-
 drivers/s390/char/tty3270.c | 3 ++-
 5 files changed, 10 insertions(+), 5 deletions(-)
drivers/s390/char/con3270.c +1 -1
index fd2146bcc0add..e17364e13d2f7 100644
@@ -629,7 +629,7 @@ con3270_init(void)
 		     (void (*)(unsigned long)) con3270_read_tasklet,
 		     (unsigned long) condev->read);
 
-	raw3270_add_view(&condev->view, &con3270_fn, 1);
+	raw3270_add_view(&condev->view, &con3270_fn, 1, RAW3270_VIEW_LOCK_IRQ);
 
 	INIT_LIST_HEAD(&condev->freemem);
 	for (i = 0; i < CON3270_STRING_PAGES; i++) {
drivers/s390/char/fs3270.c +2 -1
index 8f3a2eeb28dca..8b48ba9c598ec 100644
@@ -463,7 +463,8 @@ fs3270_open(struct inode *inode, struct file *filp)
 
 	init_waitqueue_head(&fp->wait);
 	fp->fs_pid = get_pid(task_pid(current));
-	rc = raw3270_add_view(&fp->view, &fs3270_fn, minor);
+	rc = raw3270_add_view(&fp->view, &fs3270_fn, minor,
+			      RAW3270_VIEW_LOCK_BH);
 	if (rc) {
 		fs3270_free_view(&fp->view);
 		goto out;
drivers/s390/char/raw3270.c +2 -1
index f8cd2935fbfd4..63a41b1687610 100644
@@ -920,7 +920,7 @@ raw3270_deactivate_view(struct raw3270_view *view)
  * Add view to device with minor "minor".
  */
 int
-raw3270_add_view(struct raw3270_view *view, struct raw3270_fn *fn, int minor)
+raw3270_add_view(struct raw3270_view *view, struct raw3270_fn *fn, int minor, int subclass)
 {
 	unsigned long flags;
 	struct raw3270 *rp;
@@ -942,6 +942,7 @@ raw3270_add_view(struct raw3270_view *view, struct raw3270_fn *fn, int minor)
 		view->cols = rp->cols;
 		view->ascebc = rp->ascebc;
 		spin_lock_init(&view->lock);
+		lockdep_set_subclass(&view->lock, subclass);
 		list_add(&view->list, &rp->view_list);
 		rc = 0;
 		spin_unlock_irqrestore(get_ccwdev_lock(rp->cdev), flags);
drivers/s390/char/raw3270.h +3 -1
index 114ca7cbf8897..3afaa35f73513 100644
@@ -150,6 +150,8 @@ struct raw3270_fn {
 struct raw3270_view {
 	struct list_head list;
 	spinlock_t lock;
+#define RAW3270_VIEW_LOCK_IRQ	0
+#define RAW3270_VIEW_LOCK_BH	1
 	atomic_t ref_count;
 	struct raw3270 *dev;
 	struct raw3270_fn *fn;
@@ -158,7 +160,7 @@ struct raw3270_view {
 	unsigned char *ascebc;		/* ascii -> ebcdic table */
 };
 
-int raw3270_add_view(struct raw3270_view *, struct raw3270_fn *, int);
+int raw3270_add_view(struct raw3270_view *, struct raw3270_fn *, int, int);
 int raw3270_activate_view(struct raw3270_view *);
 void raw3270_del_view(struct raw3270_view *);
 void raw3270_deactivate_view(struct raw3270_view *);
drivers/s390/char/tty3270.c +2 -1
index 2b0c36c2c5688..98d7fc152e32f 100644
@@ -980,7 +980,8 @@ static int tty3270_install(struct tty_driver *driver, struct tty_struct *tty)
 		return PTR_ERR(tp);
 
 	rc = raw3270_add_view(&tp->view, &tty3270_fn,
-			      tty->index + RAW3270_FIRSTMINOR);
+			      tty->index + RAW3270_FIRSTMINOR,
+			      RAW3270_VIEW_LOCK_BH);
 	if (rc) {
 		tty3270_free_view(tp);
 		return rc;
Signature · 1 line
2.20.1
author maintainer
#30 lore raw

PATCH 30/137 drm/ttm: fix dma_fence refcount imbalance on error path

In reply to Greg Kroah-Hartman

[ Upstream commit 543c364d8eeeb42c0edfaac9764f4e9f3d777ec1 ]

the ttm_bo_add_move_fence takes a reference to the struct dma_fence, but failed to release it on the error path, leading to a memory leak. add dma_fence_put before return when error occur.

---
 drivers/gpu/drm/ttm/ttm_bo.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/ttm/ttm_bo.c +3 -1
index 996cadd83f244..d8e1b3f129046 100644
@@ -881,8 +881,10 @@ static int ttm_bo_add_move_fence(struct ttm_buffer_object *bo,
 		reservation_object_add_shared_fence(bo->resv, fence);
 
 		ret = reservation_object_reserve_shared(bo->resv, 1);
-		if (unlikely(ret))
+		if (unlikely(ret)) {
+			dma_fence_put(fence);
 			return ret;
+		}
 
 		dma_fence_put(bo->moving);
 		bo->moving = fence;
Signature · 1 line
2.20.1
author maintainer
#31 lore raw

PATCH 31/137 drm/amd/display: extending AUX SW Timeout

In reply to Greg Kroah-Hartman

[ Upstream commit f4bbebf8e7eb4d294b040ab2d2ba71e70e69b930 ]

[Why]
AUX takes longer to reply when using active DP-DVI dongle on some asics resulting in up to 2000+ us edid read (timeout).

[How]
1. Adjust AUX poll to match spec
2. Extend the SW timeout. This does not affect normal operation since we exit the loop as soon as AUX acks.

---
 drivers/gpu/drm/amd/display/dc/dce/dce_aux.c | 9 ++++++---
 drivers/gpu/drm/amd/display/dc/dce/dce_aux.h | 6 +++---
 2 files changed, 9 insertions(+), 6 deletions(-)
drivers/gpu/drm/amd/display/dc/dce/dce_aux.c +6 -3
index aaeb7faac0c43..e0fff5744b5f6 100644
@@ -189,6 +189,12 @@ static void submit_channel_request(
 				1,
 				0);
 	}
+
+	REG_UPDATE(AUX_INTERRUPT_CONTROL, AUX_SW_DONE_ACK, 1);
+
+	REG_WAIT(AUX_SW_STATUS, AUX_SW_DONE, 0,
+				10, aux110->timeout_period/10);
+
 	/* set the delay and the number of bytes to write */
 
 	/* The length include
@@ -241,9 +247,6 @@ static void submit_channel_request(
 		}
 	}
 
-	REG_UPDATE(AUX_INTERRUPT_CONTROL, AUX_SW_DONE_ACK, 1);
-	REG_WAIT(AUX_SW_STATUS, AUX_SW_DONE, 0,
-				10, aux110->timeout_period/10);
 	REG_UPDATE(AUX_SW_CONTROL, AUX_SW_GO, 1);
 }
 
drivers/gpu/drm/amd/display/dc/dce/dce_aux.h +3 -3
index f7caab85dc801..2c6f50b4245a4 100644
@@ -69,11 +69,11 @@ enum {	/* This is the timeout as defined in DP 1.2a,
 	 * at most within ~240usec. That means,
 	 * increasing this timeout will not affect normal operation,
 	 * and we'll timeout after
-	 * SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD = 1600usec.
+	 * SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD = 2400usec.
 	 * This timeout is especially important for
-	 * resume from S3 and CTS.
+	 * converters, resume from S3, and CTS.
 	 */
-	SW_AUX_TIMEOUT_PERIOD_MULTIPLIER = 4
+	SW_AUX_TIMEOUT_PERIOD_MULTIPLIER = 6
 };
 struct aux_engine_dce110 {
 	struct aux_engine base;
Signature · 1 line
2.20.1
author maintainer
#32 lore raw

PATCH 32/137 clocksource/drivers/npcm: select TIMER_OF

In reply to Greg Kroah-Hartman

[ Upstream commit 99834eead2a04e93a120abb112542b87c42ff5e1 ]

When this is disabled, we get a link failure:

drivers/clocksource/timer-npcm7xx.o: In function `npcm7xx_timer_init':
timer-npcm7xx.c:(.init.text+0xf): undefined reference to `timer_of_init'

Fixes 1c00289ecd12 ("clocksource/drivers/npcm: Add NPCM7xx timer driver")
---
 drivers/clocksource/Kconfig | 1 +
 1 file changed, 1 insertion(+)
drivers/clocksource/Kconfig +1 -0
index 8dfd3bc448d04..9df90daa9c030 100644
@@ -144,6 +144,7 @@ config VT8500_TIMER
 config NPCM7XX_TIMER
 	bool "NPCM7xx timer driver" if COMPILE_TEST
 	depends on HAS_IOMEM
+	select TIMER_OF
 	select CLKSRC_MMIO
 	help
 	  Enable 24-bit TIMER0 and TIMER1 counters in the NPCM7xx architecture,
Signature · 1 line
2.20.1
author maintainer
#33 lore raw

PATCH 33/137 clocksource/drivers/oxnas: Fix OX820 compatible

In reply to Greg Kroah-Hartman

[ Upstream commit fbc87aa0f7c429999dc31f1bac3b2615008cac32 ]

The OX820 compatible is wrong is the driver, fix it.

Fixes 2ea3401e2a84 ("clocksource/drivers/oxnas: Add OX820 compatible")
---
 drivers/clocksource/timer-oxnas-rps.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/clocksource/timer-oxnas-rps.c +1 -1
index eed6feff8b5f2..30c6f4ce672b3 100644
@@ -296,4 +296,4 @@ static int __init oxnas_rps_timer_init(struct device_node *np)
 TIMER_OF_DECLARE(ox810se_rps,
 		       "oxsemi,ox810se-rps-timer", oxnas_rps_timer_init);
 TIMER_OF_DECLARE(ox820_rps,
-		       "oxsemi,ox820se-rps-timer", oxnas_rps_timer_init);
+		       "oxsemi,ox820-rps-timer", oxnas_rps_timer_init);
Signature · 1 line
2.20.1
author maintainer
#34 lore raw

PATCH 34/137 selftests: fib_tests: Fix Command line is not complete errors

In reply to Greg Kroah-Hartman

[ Upstream commit a5f622984a623df9a84cf43f6b098d8dd76fbe05 ]

A couple of tests are verifying a route has been removed. The helper expects the prefix as the first part of the expected output. When checking that a route has been deleted the prefix is empty leading to an invalid ip command:

  $ ip ro ls match
  Command line is not complete. Try option "help"

Fix by moving the comparison of expected output and output to a new function that is used by both check_route and check_route6. Use the new helper for the 2 checks on route removal.

Also, remove the reset of 'set -x' in route_setup which overrides the user managed setting.

Fixes d69faad76584c ("selftests: fib_tests: Add prefix route tests with metric")
---
 tools/testing/selftests/net/fib_tests.sh | 94 ++++++++++--------------
 1 file changed, 40 insertions(+), 54 deletions(-)
tools/testing/selftests/net/fib_tests.sh +40 -54
index 1080ff55a788f..0d2a5f4f1e638 100755
@@ -605,6 +605,39 @@ run_cmd()
 	return $rc
 }
 
+check_expected()
+{
+	local out="$1"
+	local expected="$2"
+	local rc=0
+
+	[ "${out}" = "${expected}" ] && return 0
+
+	if [ -z "${out}" ]; then
+		if [ "$VERBOSE" = "1" ]; then
+			printf "\nNo route entry found\n"
+			printf "Expected:\n"
+			printf "    ${expected}\n"
+		fi
+		return 1
+	fi
+
+	# tricky way to convert output to 1-line without ip's
+	# messy '\'; this drops all extra white space
+	out=$(echo ${out})
+	if [ "${out}" != "${expected}" ]; then
+		rc=1
+		if [ "${VERBOSE}" = "1" ]; then
+			printf "    Unexpected route entry. Have:\n"
+			printf "        ${out}\n"
+			printf "    Expected:\n"
+			printf "        ${expected}\n\n"
+		fi
+	fi
+
+	return $rc
+}
+
 # add route for a prefix, flushing any existing routes first
 # expected to be the first step of a test
 add_route6()
@@ -652,31 +685,7 @@ check_route6()
 	pfx=$1
 
 	out=$($IP -6 ro ls match ${pfx} | sed -e 's/ pref medium//')
-	[ "${out}" = "${expected}" ] && return 0
-
-	if [ -z "${out}" ]; then
-		if [ "$VERBOSE" = "1" ]; then
-			printf "\nNo route entry found\n"
-			printf "Expected:\n"
-			printf "    ${expected}\n"
-		fi
-		return 1
-	fi
-
-	# tricky way to convert output to 1-line without ip's
-	# messy '\'; this drops all extra white space
-	out=$(echo ${out})
-	if [ "${out}" != "${expected}" ]; then
-		rc=1
-		if [ "${VERBOSE}" = "1" ]; then
-			printf "    Unexpected route entry. Have:\n"
-			printf "        ${out}\n"
-			printf "    Expected:\n"
-			printf "        ${expected}\n\n"
-		fi
-	fi
-
-	return $rc
+	check_expected "${out}" "${expected}"
 }
 
 route_cleanup()
@@ -725,7 +734,7 @@ route_setup()
 	ip -netns ns2 addr add 172.16.103.2/24 dev veth4
 	ip -netns ns2 addr add 172.16.104.1/24 dev dummy1
 
-	set +ex
+	set +e
 }
 
 # assumption is that basic add of a single path route works
@@ -960,7 +969,8 @@ ipv6_addr_metric_test()
 	run_cmd "$IP li set dev dummy2 down"
 	rc=$?
 	if [ $rc -eq 0 ]; then
-		check_route6 ""
+		out=$($IP -6 ro ls match 2001:db8:104::/64)
+		check_expected "${out}" ""
 		rc=$?
 	fi
 	log_test $rc 0 "Prefix route removed on link down"
@@ -1091,38 +1101,13 @@ check_route()
 	local pfx
 	local expected="$1"
 	local out
-	local rc=0
 
 	set -- $expected
 	pfx=$1
 	[ "${pfx}" = "unreachable" ] && pfx=$2
 
 	out=$($IP ro ls match ${pfx})
-	[ "${out}" = "${expected}" ] && return 0
-
-	if [ -z "${out}" ]; then
-		if [ "$VERBOSE" = "1" ]; then
-			printf "\nNo route entry found\n"
-			printf "Expected:\n"
-			printf "    ${expected}\n"
-		fi
-		return 1
-	fi
-
-	# tricky way to convert output to 1-line without ip's
-	# messy '\'; this drops all extra white space
-	out=$(echo ${out})
-	if [ "${out}" != "${expected}" ]; then
-		rc=1
-		if [ "${VERBOSE}" = "1" ]; then
-			printf "    Unexpected route entry. Have:\n"
-			printf "        ${out}\n"
-			printf "    Expected:\n"
-			printf "        ${expected}\n\n"
-		fi
-	fi
-
-	return $rc
+	check_expected "${out}" "${expected}"
 }
 
 # assumption is that basic add of a single path route works
@@ -1387,7 +1372,8 @@ ipv4_addr_metric_test()
 	run_cmd "$IP li set dev dummy2 down"
 	rc=$?
 	if [ $rc -eq 0 ]; then
-		check_route ""
+		out=$($IP ro ls match 172.16.104.0/24)
+		check_expected "${out}" ""
 		rc=$?
 	fi
 	log_test $rc 0 "Prefix route removed on link down"
Signature · 1 line
2.20.1
author maintainer
#35 lore raw

PATCH 35/137 drm/amdgpu: shadow in shadow_list without tbo.mem.start cause page fault in sriov TDR

In reply to Greg Kroah-Hartman

[ Upstream commit b575f10dbd6f84c2c8744ff1f486bfae1e4f6f38 ]

shadow was added into shadow_list by amdgpu_bo_create_shadow.
meanwhile, shadow->tbo.mem was not fully configured.
tbo.mem would be fully configured by amdgpu_vm_sdma_map_table until calling amdgpu_vm_clear_bo. If sriov TDR occurred between amdgpu_bo_create_shadow and amdgpu_vm_sdma_map_table, amdgpu_device_recover_vram would deal with shadow without tbo.mem.start.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 1 +
 1 file changed, 1 insertion(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +1 -0
index d55dd570a7023..27baac26d8e9c 100644
@@ -3150,6 +3150,7 @@ static int amdgpu_device_recover_vram(struct amdgpu_device *adev)
 
 		/* No need to recover an evicted BO */
 		if (shadow->tbo.mem.mem_type != TTM_PL_TT ||
+		    shadow->tbo.mem.start == AMDGPU_BO_INVALID_OFFSET ||
 		    shadow->parent->tbo.mem.mem_type != TTM_PL_VRAM)
 			continue;
 
Signature · 1 line
2.20.1
author maintainer
#36 lore raw

PATCH 36/137 mISDN: Check address length before reading address family

In reply to Greg Kroah-Hartman

[ Upstream commit 238ffdc49ef98b15819cfd5e3fb23194e3ea3d39 ]

KMSAN will complain if valid address length passed to bind() is shorter than sizeof("struct sockaddr_mISDN"->family) bytes.

---
 drivers/isdn/mISDN/socket.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/isdn/mISDN/socket.c +2 -2
index 15d3ca37669a4..04da3a17cd950 100644
@@ -710,10 +710,10 @@ base_sock_bind(struct socket *sock, struct sockaddr *addr, int addr_len)
 	struct sock *sk = sock->sk;
 	int err = 0;
 
-	if (!maddr || maddr->family != AF_ISDN)
+	if (addr_len < sizeof(struct sockaddr_mISDN))
 		return -EINVAL;
 
-	if (addr_len < sizeof(struct sockaddr_mISDN))
+	if (!maddr || maddr->family != AF_ISDN)
 		return -EINVAL;
 
 	lock_sock(sk);
Signature · 1 line
2.20.1
author maintainer
#37 lore raw

PATCH 37/137 vxge: fix return of a freed memblock on a failed dma mapping

In reply to Greg Kroah-Hartman

[ Upstream commit 0a2c34f18c94b596562bf3d019fceab998b8b584 ]

Currently if a pci dma mapping failure is detected a free'd memblock address is returned rather than a NULL (that indicates an error). Fix this by ensuring NULL is returned on this error case.

Addresses-Coverity: ("Use after free")

Fixes 528f727279ae ("vxge: code cleanup and reorganization")
---
 drivers/net/ethernet/neterion/vxge/vxge-config.c | 1 +
 1 file changed, 1 insertion(+)
drivers/net/ethernet/neterion/vxge/vxge-config.c +1 -0
index 7cde387e5ec62..51cd57ab3d958 100644
@@ -2366,6 +2366,7 @@ static void *__vxge_hw_blockpool_malloc(struct __vxge_hw_device *devh, u32 size,
 				dma_object->addr))) {
 			vxge_os_dma_free(devh->pdev, memblock,
 				&dma_object->acc_handle);
+			memblock = NULL;
 			goto exit;
 		}
 
Signature · 1 line
2.20.1
author maintainer
#38 lore raw

PATCH 38/137 qede: fix write to freed pointer error and double free of ptp

In reply to Greg Kroah-Hartman

[ Upstream commit 1dc2b3d65523780ed1972d446c76e62e13f3e8f5 ]

The err2 error return path calls qede_ptp_disable that cleans up on an error and frees ptp. After this, the free'd ptp is dereferenced when ptp->clock is set to NULL and the code falls-through to error path err1 that frees ptp again.

Fix this by calling qede_ptp_disable and exiting via an error return path that does not set ptp->clock or kfree ptp.

Addresses-Coverity: ("Write to pointer after free")

Fixes 035744975aec ("qede: Add support for PTP resource locking.")
---
 drivers/net/ethernet/qlogic/qede/qede_ptp.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)
drivers/net/ethernet/qlogic/qede/qede_ptp.c +3 -4
index 5f3f42a253616..bddb2b5982dcf 100644
@@ -490,18 +490,17 @@ int qede_ptp_enable(struct qede_dev *edev, bool init_tc)
 
 	ptp->clock = ptp_clock_register(&ptp->clock_info, &edev->pdev->dev);
 	if (IS_ERR(ptp->clock)) {
-		rc = -EINVAL;
 		DP_ERR(edev, "PTP clock registration failed\n");
+		qede_ptp_disable(edev);
+		rc = -EINVAL;
 		goto err2;
 	}
 
 	return 0;
 
-err2:
-	qede_ptp_disable(edev);
-	ptp->clock = NULL;
 err1:
 	kfree(ptp);
+err2:
 	edev->ptp = NULL;
 
 	return rc;
Signature · 1 line
2.20.1
author maintainer
#39 lore raw

PATCH 39/137 afs: Unlock pages for __pagevec_release()

In reply to Greg Kroah-Hartman

[ Upstream commit 21bd68f196ca91fc0f3d9bd1b32f6e530e8c1c88 ]

__pagevec_release() complains loudly if any page in the vector is still locked. The pages need to be locked for generic_error_remove_page(), but that function doesn't actually unlock them.

Unlock the pages afterwards.

---
 fs/afs/write.c | 1 +
 1 file changed, 1 insertion(+)
fs/afs/write.c +1 -0
index 72efcfcf9f95e..0122d7445fba1 100644
@@ -264,6 +264,7 @@ static void afs_kill_pages(struct address_space *mapping,
 				first = page->index + 1;
 			lock_page(page);
 			generic_error_remove_page(mapping, page);
+			unlock_page(page);
 		}
 
 		__pagevec_release(&pv);
Signature · 1 line
2.20.1
author maintainer
#40 lore raw

PATCH 40/137 afs: Fix in-progess ops to ignore server-level callback invalidation

In reply to Greg Kroah-Hartman

[ Upstream commit eeba1e9cf31d064284dd1fa7bd6cfe01395bd03d ]

The in-kernel afs filesystem client counts the number of server-level callback invalidation events (CB.InitCallBackState* RPC operations) that it receives from the server. This is stored in cb_s_break in various structures, including afs_server and afs_vnode.

If an inode is examined by afs_validate(), say, the afs_server copy is compared, along with other break counters, to those in afs_vnode, and if one or more of the counters do not match, it is considered that the server's callback promise is broken. At points where this happens, AFS_VNODE_CB_PROMISED is cleared to indicate that the status must be refetched from the server.

afs_validate() issues an FS.FetchStatus operation to get updated metadata - and based on the updated data_version may invalidate the pagecache too.

However, the break counters are also used to determine whether to note a new callback in the vnode (which would set the AFS_VNODE_CB_PROMISED flag) and whether to cache the permit data included in the YFSFetchStatus record by the server.

The problem comes when the server sends us a CB.InitCallBackState op. The first such instance doesn't cause cb_s_break to be incremented, but rather causes AFS_SERVER_FL_NEW to be cleared - but thereafter, say some hours after last use and all the volumes have been automatically unmounted and the server has forgotten about the client[*], this *will* likely cause an increment.

[*] There are other circumstances too, such as the server restarting or
needing to make space in its callback table.

Note that the server won't send us a CB.InitCallBackState op until we talk to it again.

So what happens is:

(1) A mount for a new volume is attempted, a inode is created for the root vnode and vnode->cb_s_break and AFS_VNODE_CB_PROMISED aren't set immediately, as we don't have a nominated server to talk to yet - and we may iterate through a few to find one.

(2) Before the operation happens, afs_fetch_status(), say, notes in the cursor (fc.cb_break) the break counter sum from the vnode, volume and server counters, but the server->cb_s_break is currently 0.

(3) We send FS.FetchStatus to the server. The server sends us back

     CB.InitCallBackState.  We increment server->cb_s_break.

(4) Our FS.FetchStatus completes. The reply includes a callback record.

(5) xdr_decode_AFSCallBack()/xdr_decode_YFSCallBack() check to see whether the callback promise was broken by checking the break counter sum from step (2) against the current sum.

This fails because of step (3), so we don't set the callback record and, importantly, don't set AFS_VNODE_CB_PROMISED on the vnode.

This does not preclude the syscall from progressing, and we don't loop here rechecking the status, but rather assume it's good enough for one round only and will need to be rechecked next time.

(6) afs_validate() it triggered on the vnode, probably called from d_revalidate() checking the parent directory.

(7) afs_validate() notes that AFS_VNODE_CB_PROMISED isn't set, so doesn't

     update vnode->cb_s_break and assumes the vnode to be invalid.

(8) afs_validate() needs to calls afs_fetch_status(). Go back to step (2) and repeat, every time the vnode is validated.

This primarily affects volume root dir vnodes. Everything subsequent to those inherit an already incremented cb_s_break upon mounting.

The issue is that we assume that the callback record and the cached permit information in a reply from the server can't be trusted after getting a server break - but this is wrong since the server makes sure things are done in the right order, holding up our ops if necessary[*].

[*] There is an extremely unlikely scenario where a reply from before the
CB.InitCallBackState could get its delivery deferred till after - at which point we think we have a promise when we don't. This, however, requires unlucky mass packet loss to one call.

AFS_SERVER_FL_NEW tries to paper over the cracks for the initial mount from a server we've never contacted before, but this should be unnecessary. It's also further insulated from the problem on an initial mount by querying the server first with FS.GetCapabilities, which triggers the CB.InitCallBackState.

Fix this by

(1) Remove AFS_SERVER_FL_NEW.

(2) In afs_calc_vnode_cb_break(), don't include cb_s_break in the calculation.

(3) In afs_cb_is_broken(), don't include cb_s_break in the check.

---
 fs/afs/callback.c | 3 +--
 fs/afs/internal.h | 4 +---
 fs/afs/server.c   | 1 -
 3 files changed, 2 insertions(+), 6 deletions(-)
fs/afs/callback.c +1 -2
index 1c7955f5cdaf2..128f2dbe256a4 100644
@@ -203,8 +203,7 @@ void afs_put_cb_interest(struct afs_net *net, struct afs_cb_interest *cbi)
  */
 void afs_init_callback_state(struct afs_server *server)
 {
-	if (!test_and_clear_bit(AFS_SERVER_FL_NEW, &server->flags))
-		server->cb_s_break++;
+	server->cb_s_break++;
 }
 
 /*
fs/afs/internal.h +1 -3
index 8871b9e8645f1..465526f495b01 100644
@@ -475,7 +475,6 @@ struct afs_server {
 	time64_t		put_time;	/* Time at which last put */
 	time64_t		update_at;	/* Time at which to next update the record */
 	unsigned long		flags;
-#define AFS_SERVER_FL_NEW	0		/* New server, don't inc cb_s_break */
 #define AFS_SERVER_FL_NOT_READY	1		/* The record is not ready for use */
 #define AFS_SERVER_FL_NOT_FOUND	2		/* VL server says no such server */
 #define AFS_SERVER_FL_VL_FAIL	3		/* Failed to access VL server */
@@ -828,7 +827,7 @@ static inline struct afs_cb_interest *afs_get_cb_interest(struct afs_cb_interest
 
 static inline unsigned int afs_calc_vnode_cb_break(struct afs_vnode *vnode)
 {
-	return vnode->cb_break + vnode->cb_s_break + vnode->cb_v_break;
+	return vnode->cb_break + vnode->cb_v_break;
 }
 
 static inline bool afs_cb_is_broken(unsigned int cb_break,
@@ -836,7 +835,6 @@ static inline bool afs_cb_is_broken(unsigned int cb_break,
 				    const struct afs_cb_interest *cbi)
 {
 	return !cbi || cb_break != (vnode->cb_break +
-				    cbi->server->cb_s_break +
 				    vnode->volume->cb_v_break);
 }
 
fs/afs/server.c +0 -1
index 642afa2e9783c..65b33b6da48b9 100644
@@ -226,7 +226,6 @@ static struct afs_server *afs_alloc_server(struct afs_net *net,
 	RCU_INIT_POINTER(server->addresses, alist);
 	server->addr_version = alist->version;
 	server->uuid = *uuid;
-	server->flags = (1UL << AFS_SERVER_FL_NEW);
 	server->update_at = ktime_get_real_seconds() + afs_server_update_delay;
 	rwlock_init(&server->fs_lock);
 	INIT_HLIST_HEAD(&server->cb_volumes);
Signature · 1 line
2.20.1
author maintainer
#41 lore raw

PATCH 41/137 qed: Delete redundant doorbell recovery types

In reply to Greg Kroah-Hartman

[ Upstream commit 9ac6bb1414ac0d45fe9cefbd1f5b06f0e1a3c98a ]

DB_REC_DRY_RUN (running doorbell recovery without sending doorbells) is never used. DB_REC_ONCE (send a single doorbell from the doorbell recovery) is not needed anymore because by running the periodic handler we make sure we check the overflow status later instead.
This patch is needed because in the next patches, the only doorbell recovery type being used is DB_REC_REAL_DEAL, and the fixes are much cleaner without this enum.

---
 drivers/net/ethernet/qlogic/qed/qed.h     |  3 +-
 drivers/net/ethernet/qlogic/qed/qed_dev.c | 69 +++++++++--------------
 drivers/net/ethernet/qlogic/qed/qed_int.c |  6 +-
 drivers/net/ethernet/qlogic/qed/qed_int.h |  4 +-
 4 files changed, 31 insertions(+), 51 deletions(-)
drivers/net/ethernet/qlogic/qed/qed.h +1 -2
index 2d8a77cc156ba..d5fece7eb1698 100644
@@ -918,8 +918,7 @@ u16 qed_get_cm_pq_idx_llt_mtc(struct qed_hwfn *p_hwfn, u8 tc);
 
 /* doorbell recovery mechanism */
 void qed_db_recovery_dp(struct qed_hwfn *p_hwfn);
-void qed_db_recovery_execute(struct qed_hwfn *p_hwfn,
-			     enum qed_db_rec_exec db_exec);
+void qed_db_recovery_execute(struct qed_hwfn *p_hwfn);
 bool qed_edpm_enabled(struct qed_hwfn *p_hwfn);
 
 /* Other Linux specific common definitions */
drivers/net/ethernet/qlogic/qed/qed_dev.c +26 -43

95 lines not on this page · raw message

drivers/net/ethernet/qlogic/qed/qed_int.c +2 -4
index 92340919d8521..b994f81eb51c3 100644
@@ -409,10 +409,8 @@ int qed_db_rec_handler(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt)
 
 	overflow = qed_rd(p_hwfn, p_ptt, DORQ_REG_PF_OVFL_STICKY);
 	DP_NOTICE(p_hwfn, "PF Overflow sticky 0x%x\n", overflow);
-	if (!overflow) {
-		qed_db_recovery_execute(p_hwfn, DB_REC_ONCE);
+	if (!overflow)
 		return 0;
-	}
 
 	if (qed_edpm_enabled(p_hwfn)) {
 		rc = qed_db_rec_flush_queue(p_hwfn, p_ptt);
@@ -427,7 +425,7 @@ int qed_db_rec_handler(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt)
 	qed_wr(p_hwfn, p_ptt, DORQ_REG_PF_OVFL_STICKY, 0x0);
 
 	/* Repeat all last doorbells (doorbell drop recovery) */
-	qed_db_recovery_execute(p_hwfn, DB_REC_REAL_DEAL);
+	qed_db_recovery_execute(p_hwfn);
 
 	return 0;
 }
drivers/net/ethernet/qlogic/qed/qed_int.h +2 -2
index d81a62ebd5244..df26bf333893d 100644
@@ -192,8 +192,8 @@ void qed_int_disable_post_isr_release(struct qed_dev *cdev);
 
 /**
  * @brief - Doorbell Recovery handler.
- *          Run DB_REAL_DEAL doorbell recovery in case of PF overflow
- *          (and flush DORQ if needed), otherwise run DB_REC_ONCE.
+ *          Run doorbell recovery in case of PF overflow (and flush DORQ if
+ *          needed).
  *
  * @param p_hwfn
  * @param p_ptt
Signature · 1 line
2.20.1
author maintainer
#42 lore raw

PATCH 42/137 qed: Fix the doorbell address sanity check

In reply to Greg Kroah-Hartman

[ Upstream commit b61b04ad81d5f975349d66abbecabf96ba211140 ]

Fix the condition which verifies that doorbell address is inside the doorbell bar by checking that the end of the address is within range as well.

---
 drivers/net/ethernet/qlogic/qed/qed_dev.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)
drivers/net/ethernet/qlogic/qed/qed_dev.c +8 -8
index ff0bbf8d073d6..228891e459bc0 100644
@@ -102,11 +102,15 @@ static void qed_db_recovery_dp_entry(struct qed_hwfn *p_hwfn,
 
 /* Doorbell address sanity (address within doorbell bar range) */
 static bool qed_db_rec_sanity(struct qed_dev *cdev,
-			      void __iomem *db_addr, void *db_data)
+			      void __iomem *db_addr,
+			      enum qed_db_rec_width db_width,
+			      void *db_data)
 {
+	u32 width = (db_width == DB_REC_WIDTH_32B) ? 32 : 64;
+
 	/* Make sure doorbell address is within the doorbell bar */
 	if (db_addr < cdev->doorbells ||
-	    (u8 __iomem *)db_addr >
+	    (u8 __iomem *)db_addr + width >
 	    (u8 __iomem *)cdev->doorbells + cdev->db_size) {
 		WARN(true,
 		     "Illegal doorbell address: %p. Legal range for doorbell addresses is [%p..%p]\n",
@@ -159,7 +163,7 @@ int qed_db_recovery_add(struct qed_dev *cdev,
 	}
 
 	/* Sanitize doorbell address */
-	if (!qed_db_rec_sanity(cdev, db_addr, db_data))
+	if (!qed_db_rec_sanity(cdev, db_addr, db_width, db_data))
 		return -EINVAL;
 
 	/* Obtain hwfn from doorbell address */
@@ -205,10 +209,6 @@ int qed_db_recovery_del(struct qed_dev *cdev,
 		return 0;
 	}
 
-	/* Sanitize doorbell address */
-	if (!qed_db_rec_sanity(cdev, db_addr, db_data))
-		return -EINVAL;
-
 	/* Obtain hwfn from doorbell address */
 	p_hwfn = qed_db_rec_find_hwfn(cdev, db_addr);
 
@@ -317,7 +317,7 @@ static void qed_db_recovery_ring(struct qed_hwfn *p_hwfn,
 
 	/* Sanity */
 	if (!qed_db_rec_sanity(p_hwfn->cdev, db_entry->db_addr,
-			       db_entry->db_data))
+			       db_entry->db_width, db_entry->db_data))
 		return;
 
 	/* Flush the write combined buffer. Since there are multiple doorbelling
Signature · 1 line
2.20.1
author maintainer
#43 lore raw

PATCH 43/137 qed: Fix missing DORQ attentions

In reply to Greg Kroah-Hartman

[ Upstream commit d4476b8a6151b2dd86c09b5acec64f66430db55d ]

When the DORQ (doorbell block) is overflowed, all PFs get attentions at the same time. If one PF finished handling the attention before another PF even started, the second PF might miss the DORQ's attention bit and not handle the attention at all.
If the DORQ attention is missed and the issue is not resolved, another attention will not be sent, therefore each attention is treated as a potential DORQ attention.
As a result, the attention callback is called more frequently so the debug print was moved to reduce its quantity.
The number of periodic doorbell recovery handler schedules was reduced because it was the previous way to mitigating the missed attention issue.

---
 drivers/net/ethernet/qlogic/qed/qed.h      |  1 +
 drivers/net/ethernet/qlogic/qed/qed_int.c  | 20 ++++++++++++++++++--
 drivers/net/ethernet/qlogic/qed/qed_main.c |  2 +-
 3 files changed, 20 insertions(+), 3 deletions(-)
drivers/net/ethernet/qlogic/qed/qed.h +1 -0
index d5fece7eb1698..07ae600d0f357 100644
@@ -436,6 +436,7 @@ struct qed_db_recovery_info {
 
 	/* Lock to protect the doorbell recovery mechanism list */
 	spinlock_t lock;
+	bool dorq_attn;
 	u32 db_recovery_counter;
 };
 
drivers/net/ethernet/qlogic/qed/qed_int.c +18 -2
index b994f81eb51c3..00688f4c04645 100644
@@ -436,17 +436,19 @@ static int qed_dorq_attn_cb(struct qed_hwfn *p_hwfn)
 	struct qed_ptt *p_ptt = p_hwfn->p_dpc_ptt;
 	int rc;
 
-	int_sts = qed_rd(p_hwfn, p_ptt, DORQ_REG_INT_STS);
-	DP_NOTICE(p_hwfn->cdev, "DORQ attention. int_sts was %x\n", int_sts);
+	p_hwfn->db_recovery_info.dorq_attn = true;
 
 	/* int_sts may be zero since all PFs were interrupted for doorbell
 	 * overflow but another one already handled it. Can abort here. If
 	 * This PF also requires overflow recovery we will be interrupted again.
 	 * The masked almost full indication may also be set. Ignoring.
 	 */
+	int_sts = qed_rd(p_hwfn, p_ptt, DORQ_REG_INT_STS);
 	if (!(int_sts & ~DORQ_REG_INT_STS_DORQ_FIFO_AFULL))
 		return 0;
 
+	DP_NOTICE(p_hwfn->cdev, "DORQ attention. int_sts was %x\n", int_sts);
+
 	/* check if db_drop or overflow happened */
 	if (int_sts & (DORQ_REG_INT_STS_DB_DROP |
 		       DORQ_REG_INT_STS_DORQ_FIFO_OVFL_ERR)) {
@@ -503,6 +505,17 @@ static int qed_dorq_attn_cb(struct qed_hwfn *p_hwfn)
 	return -EINVAL;
 }
 
+static void qed_dorq_attn_handler(struct qed_hwfn *p_hwfn)
+{
+	if (p_hwfn->db_recovery_info.dorq_attn)
+		goto out;
+
+	/* Call DORQ callback if the attention was missed */
+	qed_dorq_attn_cb(p_hwfn);
+out:
+	p_hwfn->db_recovery_info.dorq_attn = false;
+}
+
 /* Instead of major changes to the data-structure, we have a some 'special'
  * identifiers for sources that changed meaning between adapters.
  */
@@ -1076,6 +1089,9 @@ static int qed_int_deassertion(struct qed_hwfn  *p_hwfn,
 		}
 	}
 
+	/* Handle missed DORQ attention */
+	qed_dorq_attn_handler(p_hwfn);
+
 	/* Clear IGU indication for the deasserted bits */
 	DIRECT_REG_WR((u8 __iomem *)p_hwfn->regview +
 				    GTT_BAR0_MAP_REG_IGU_CMD +
drivers/net/ethernet/qlogic/qed/qed_main.c +1 -1
index 6adf5bda9811e..26bfcbeebc4ca 100644
@@ -966,7 +966,7 @@ static void qed_update_pf_params(struct qed_dev *cdev,
 	}
 }
 
-#define QED_PERIODIC_DB_REC_COUNT		100
+#define QED_PERIODIC_DB_REC_COUNT		10
 #define QED_PERIODIC_DB_REC_INTERVAL_MS		100
 #define QED_PERIODIC_DB_REC_INTERVAL \
 	msecs_to_jiffies(QED_PERIODIC_DB_REC_INTERVAL_MS)
Signature · 1 line
2.20.1
author maintainer
#44 lore raw

PATCH 44/137 qed: Fix the DORQs attentions handling

In reply to Greg Kroah-Hartman

[ Upstream commit 0d72c2ac89185f179da1e8a91c40c82f3fa38f0b ]

Separate the overflow handling from the hardware interrupt status analysis. The interrupt status is a single register and is common for all PFs. The first PF reading the register is not necessarily the one who overflowed. All PFs must check their overflow status on every attention.
In this change we clear the sticky indication in the attention handler to allow doorbells to be processed again as soon as possible, but running the doorbell recovery is scheduled for the periodic handler to reduce the time spent in the attention handler.
Checking the need for DORQ flush was changed to "db_bar_no_edpm" because qed_edpm_enabled()'s result could change dynamically and might have prevented a needed flush.

---
 drivers/net/ethernet/qlogic/qed/qed.h     |  3 ++
 drivers/net/ethernet/qlogic/qed/qed_int.c | 61 +++++++++++++++++------
 2 files changed, 48 insertions(+), 16 deletions(-)
drivers/net/ethernet/qlogic/qed/qed.h +3 -0
index 07ae600d0f357..f458c9776a89c 100644
@@ -431,6 +431,8 @@ struct qed_qm_info {
 	u8 num_pf_rls;
 };
 
+#define QED_OVERFLOW_BIT	1
+
 struct qed_db_recovery_info {
 	struct list_head list;
 
@@ -438,6 +440,7 @@ struct qed_db_recovery_info {
 	spinlock_t lock;
 	bool dorq_attn;
 	u32 db_recovery_counter;
+	unsigned long overflow;
 };
 
 struct storm_stats {
drivers/net/ethernet/qlogic/qed/qed_int.c +45 -16

104 lines not on this page · raw message

Signature · 1 line
2.20.1
author maintainer
#45 lore raw

PATCH 45/137 drm/amd/display: If one stream full updates, full update all planes

In reply to Greg Kroah-Hartman

[ Upstream commit c238bfe0be9ef7420f7669a69e27c8c8f4d8a568 ]

[Why]
On some compositors, with two monitors attached, VT terminal switch can cause a graphical issue by the following means:

There are two streams, one for each monitor. Each stream has one plane

current state:

	M1:S1->P1
	M2:S2->P2

The user calls for a terminal switch and a commit is made to change both planes to linear swizzle mode. In atomic check, a new dc_state is constructed with new planes on each stream

new state:

	M1:S1->P3
	M2:S2->P4

In commit tail, each stream is committed, one at a time. The first stream (S1) updates properly, triggerring a full update and replacing the state

current state:

	M1:S1->P3
	M2:S2->P4

The update for S2 comes in, but dc detects that there is no difference between the stream and plane in the new and current states, and so triggers a fast update. The fast update does not program swizzle, so the second monitor is corrupted

[How]
Add a flag to dc_plane_state that forces full updates

When a stream undergoes a full update, set this flag on all changed planes, then clear it on the current stream

Subsequent streams will get full updates as a result

---
 drivers/gpu/drm/amd/display/dc/core/dc.c | 19 +++++++++++++++++++
 drivers/gpu/drm/amd/display/dc/dc.h      |  3 +++
 2 files changed, 22 insertions(+)
drivers/gpu/drm/amd/display/dc/core/dc.c +19 -0
index 1f92e7e8e3d38..5af2ea1f201d3 100644
@@ -1308,6 +1308,11 @@ static enum surface_update_type det_surface_update(const struct dc *dc,
 		return UPDATE_TYPE_FULL;
 	}
 
+	if (u->surface->force_full_update) {
+		update_flags->bits.full_update = 1;
+		return UPDATE_TYPE_FULL;
+	}
+
 	type = get_plane_info_update_type(u);
 	elevate_update_type(&overall_type, type);
 
@@ -1637,6 +1642,14 @@ void dc_commit_updates_for_stream(struct dc *dc,
 		}
 
 		dc_resource_state_copy_construct(state, context);
+
+		for (i = 0; i < dc->res_pool->pipe_count; i++) {
+			struct pipe_ctx *new_pipe = &context->res_ctx.pipe_ctx[i];
+			struct pipe_ctx *old_pipe = &dc->current_state->res_ctx.pipe_ctx[i];
+
+			if (new_pipe->plane_state && new_pipe->plane_state != old_pipe->plane_state)
+				new_pipe->plane_state->force_full_update = true;
+		}
 	}
 
 
@@ -1680,6 +1693,12 @@ void dc_commit_updates_for_stream(struct dc *dc,
 		dc->current_state = context;
 		dc_release_state(old);
 
+		for (i = 0; i < dc->res_pool->pipe_count; i++) {
+			struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
+
+			if (pipe_ctx->plane_state && pipe_ctx->stream == stream)
+				pipe_ctx->plane_state->force_full_update = false;
+		}
 	}
 	/*let's use current_state to update watermark etc*/
 	if (update_type >= UPDATE_TYPE_FULL)
drivers/gpu/drm/amd/display/dc/dc.h +3 -0
index 4b5bbb13ce7fe..7d5656d7e460d 100644
@@ -496,6 +496,9 @@ struct dc_plane_state {
 	struct dc_plane_status status;
 	struct dc_context *ctx;
 
+	/* HACK: Workaround for forcing full reprogramming under some conditions */
+	bool force_full_update;
+
 	/* private to dc_surface.c */
 	enum dc_irq_source irq_source;
 	struct kref refcount;
Signature · 1 line
2.20.1
author maintainer
#46 lore raw

PATCH 46/137 s390/pkey: add one more argument space for debug feature entry

In reply to Greg Kroah-Hartman

[ Upstream commit 6b1f16ba730d4c0cda1247568c3a1bf4fa3a2f2f ]

The debug feature entries have been used with up to 5 arguents (including the pointer to the format string) but there was only space reserved for 4 arguemnts. So now the registration does reserve space for 5 times a long value.

This fixes a sometime appearing weired value as the last value of an debug feature entry like this:

... pkey_sec2protkey zcrypt_send_cprb (cardnr=10 domain=12)
failed with errno -2143346254

---
 drivers/s390/crypto/pkey_api.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/s390/crypto/pkey_api.c +2 -1
index 2f92bbed4bf68..097e890e0d6d9 100644
@@ -51,7 +51,8 @@ static debug_info_t *debug_info;
 
 static void __init pkey_debug_init(void)
 {
-	debug_info = debug_register("pkey", 1, 1, 4 * sizeof(long));
+	/* 5 arguments per dbf entry (including the format string ptr) */
+	debug_info = debug_register("pkey", 1, 1, 5 * sizeof(long));
 	debug_register_view(debug_info, &debug_sprintf_view);
 	debug_set_level(debug_info, 3);
 }
Signature · 1 line
2.20.1
author maintainer
#47 lore raw

PATCH 47/137 x86/build/lto: Fix truncated .bss with -fdata-sections

In reply to Greg Kroah-Hartman

[ Upstream commit 6a03469a1edc94da52b65478f1e00837add869a3 ]

With CONFIG_LD_DEAD_CODE_DATA_ELIMINATION=y, we compile the kernel with -fdata-sections, which also splits the .bss section.

The new section, with a new .bss.* name, which pattern gets missed by the main x86 linker script which only expects the '.bss' name. This results in the discarding of the second part and a too small, truncated .bss section and an unhappy, non-working kernel.

Use the common BSS_MAIN macro in the linker script to properly capture and merge all the generated BSS sections.

Cc Borislav Petkov <[email protected]>
Cc Kees Cook <[email protected]>
Cc Linus Torvalds <[email protected]>
Cc Nicholas Piggin <[email protected]>
Cc Nick Desaulniers <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Thomas Gleixner <[email protected]>
[ Extended the changelog. ]
---
 arch/x86/kernel/vmlinux.lds.S | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kernel/vmlinux.lds.S +1 -1
index ee3b5c7d662e1..c45214c44e612 100644
@@ -362,7 +362,7 @@ SECTIONS
 	.bss : AT(ADDR(.bss) - LOAD_OFFSET) {
 		__bss_start = .;
 		*(.bss..page_aligned)
-		*(.bss)
+		*(BSS_MAIN)
 		BSS_DECRYPTED
 		. = ALIGN(PAGE_SIZE);
 		__bss_stop = .;
Signature · 1 line
2.20.1
author maintainer
#48 lore raw

PATCH 48/137 x86/mm: Prevent bogus warnings with "noexec=off"

In reply to Greg Kroah-Hartman

[ Upstream commit 510bb96fe5b3480b4b22d815786377e54cb701e7 ]

Xose Vazquez Perez reported boot warnings when NX is disabled on the kernel command line.

__early_set_fixmap() triggers this warning:

attempted to set unsupported pgprot: 8000000000000163

			       bits:      8000000000000000
			       supported: 7fffffffffffffff
  WARNING: CPU: 0 PID: 0 at arch/x86/include/asm/pgtable.h:537
			    __early_set_fixmap+0xa2/0xff

because it uses __default_kernel_pte_mask to mask out unsupported bits.

Use __supported_pte_mask instead.

Disabling NX on the command line also triggers the NX warning in the page table mapping check:

  WARNING: CPU: 1 PID: 1 at arch/x86/mm/dump_pagetables.c:262 note_page+0x2ae/0x650
  ....

Make the warning depend on NX set in __supported_pte_mask.

---
 arch/x86/mm/dump_pagetables.c | 3 ++-
 arch/x86/mm/ioremap.c         | 2 +-
 2 files changed, 3 insertions(+), 2 deletions(-)
arch/x86/mm/dump_pagetables.c +2 -1
index e3cdc85ce5b6e..84304626b1cb8 100644
@@ -259,7 +259,8 @@ static void note_wx(struct pg_state *st)
 #endif
 	/* Account the WX pages */
 	st->wx_pages += npages;
-	WARN_ONCE(1, "x86/mm: Found insecure W+X mapping at address %pS\n",
+	WARN_ONCE(__supported_pte_mask & _PAGE_NX,
+		  "x86/mm: Found insecure W+X mapping at address %pS\n",
 		  (void *)st->start_address);
 }
 
arch/x86/mm/ioremap.c +1 -1
index 5378d10f1d31d..3b76fe954978c 100644
@@ -825,7 +825,7 @@ void __init __early_set_fixmap(enum fixed_addresses idx,
 	pte = early_ioremap_pte(addr);
 
 	/* Sanitize 'prot' against any unsupported bits: */
-	pgprot_val(flags) &= __default_kernel_pte_mask;
+	pgprot_val(flags) &= __supported_pte_mask;
 
 	if (pgprot_val(flags))
 		set_pte(pte, pfn_pte(phys >> PAGE_SHIFT, flags));
Signature · 1 line
2.20.1
author maintainer
#49 lore raw

PATCH 49/137 x86/reboot, efi: Use EFI reboot for Acer TravelMate X514-51T

In reply to Greg Kroah-Hartman

[ Upstream commit 0082517fa4bce073e7cf542633439f26538a14cc ]

Upon reboot, the Acer TravelMate X514-51T laptop appears to complete the shutdown process, but then it hangs in BIOS POST with a black screen.

The problem is intermittent - at some points it has appeared related to Secure Boot settings or different kernel builds, but ultimately we have not been able to identify the exact conditions that trigger the issue to come and go.

Besides, the EFI mode cannot be disabled in the BIOS of this model.

However, after extensive testing, we observe that using the EFI reboot method reliably avoids the issue in all cases.

So add a boot time quirk to use EFI reboot on such systems.

Cc Ard Biesheuvel <[email protected]>
Cc Borislav Petkov <[email protected]>
Cc Linus Torvalds <[email protected]>
Cc Matt Fleming <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Thomas Gleixner <[email protected]>
[ Fix !CONFIG_EFI build failure, clarify the code and the changelog a bit. ]
---
 arch/x86/kernel/reboot.c | 21 +++++++++++++++++++++
 include/linux/efi.h      |  7 ++++++-
 2 files changed, 27 insertions(+), 1 deletion(-)
arch/x86/kernel/reboot.c +21 -0
index 725624b6c0c05..8fd3cedd9accd 100644
@@ -81,6 +81,19 @@ static int __init set_bios_reboot(const struct dmi_system_id *d)
 	return 0;
 }
 
+/*
+ * Some machines don't handle the default ACPI reboot method and
+ * require the EFI reboot method:
+ */
+static int __init set_efi_reboot(const struct dmi_system_id *d)
+{
+	if (reboot_type != BOOT_EFI && !efi_runtime_disabled()) {
+		reboot_type = BOOT_EFI;
+		pr_info("%s series board detected. Selecting EFI-method for reboot.\n", d->ident);
+	}
+	return 0;
+}
+
 void __noreturn machine_real_restart(unsigned int type)
 {
 	local_irq_disable();
@@ -166,6 +179,14 @@ static const struct dmi_system_id reboot_dmi_table[] __initconst = {
 			DMI_MATCH(DMI_PRODUCT_NAME, "AOA110"),
 		},
 	},
+	{	/* Handle reboot issue on Acer TravelMate X514-51T */
+		.callback = set_efi_reboot,
+		.ident = "Acer TravelMate X514-51T",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Acer"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "TravelMate X514-51T"),
+		},
+	},
 
 	/* Apple */
 	{	/* Handle problems with rebooting on Apple MacBook5 */
include/linux/efi.h +6 -1
index a86485ac7c878..de05a43025292 100644
@@ -1598,7 +1598,12 @@ efi_status_t efi_setup_gop(efi_system_table_t *sys_table_arg,
 			   struct screen_info *si, efi_guid_t *proto,
 			   unsigned long size);
 
-bool efi_runtime_disabled(void);
+#ifdef CONFIG_EFI
+extern bool efi_runtime_disabled(void);
+#else
+static inline bool efi_runtime_disabled(void) { return true; }
+#endif
+
 extern void efi_call_virt_check_flags(unsigned long flags, const char *call);
 
 enum efi_secureboot_mode {
Signature · 1 line
2.20.1
author maintainer
#50 lore raw

PATCH 50/137 KVM: nVMX: always use early vmcs check when EPT is disabled

In reply to Greg Kroah-Hartman

[ Upstream commit 2b27924bb1d48e3775f432b70bdad5e6dd4e7798 ]

The remaining failures of vmx.flat when EPT is disabled are caused by incorrectly reflecting VMfails to the L1 hypervisor. What happens is that nested_vmx_restore_host_state corrupts the guest CR3, reloading it with the host's shadow CR3 instead, because it blindly loads GUEST_CR3 from the vmcs01.

For simplicity let's just always use hardware VMCS checks when EPT is disabled. This way, nested_vmx_restore_host_state is not reached at all (or at least shouldn't be reached).

---
 arch/x86/include/uapi/asm/vmx.h |  1 +
 arch/x86/kvm/vmx/nested.c       | 22 ++++++++++++++++++++--
 2 files changed, 21 insertions(+), 2 deletions(-)
arch/x86/include/uapi/asm/vmx.h +1 -0
index f0b0c90dd3982..d213ec5c3766d 100644
@@ -146,6 +146,7 @@
 
 #define VMX_ABORT_SAVE_GUEST_MSR_FAIL        1
 #define VMX_ABORT_LOAD_HOST_PDPTE_FAIL       2
+#define VMX_ABORT_VMCS_CORRUPTED             3
 #define VMX_ABORT_LOAD_HOST_MSR_FAIL         4
 
 #endif /* _UAPIVMX_H */
arch/x86/kvm/vmx/nested.c +20 -2
index 8f8c42b048757..2a16bd8877297 100644
@@ -3790,8 +3790,18 @@ static void nested_vmx_restore_host_state(struct kvm_vcpu *vcpu)
 	vmx_set_cr4(vcpu, vmcs_readl(CR4_READ_SHADOW));
 
 	nested_ept_uninit_mmu_context(vcpu);
-	vcpu->arch.cr3 = vmcs_readl(GUEST_CR3);
-	__set_bit(VCPU_EXREG_CR3, (ulong *)&vcpu->arch.regs_avail);
+
+	/*
+	 * This is only valid if EPT is in use, otherwise the vmcs01 GUEST_CR3
+	 * points to shadow pages!  Fortunately we only get here after a WARN_ON
+	 * if EPT is disabled, so a VMabort is perfectly fine.
+	 */
+	if (enable_ept) {
+		vcpu->arch.cr3 = vmcs_readl(GUEST_CR3);
+		__set_bit(VCPU_EXREG_CR3, (ulong *)&vcpu->arch.regs_avail);
+	} else {
+		nested_vmx_abort(vcpu, VMX_ABORT_VMCS_CORRUPTED);
+	}
 
 	/*
 	 * Use ept_save_pdptrs(vcpu) to load the MMU's cached PDPTRs
@@ -5739,6 +5749,14 @@ __init int nested_vmx_hardware_setup(int (*exit_handlers[])(struct kvm_vcpu *))
 {
 	int i;
 
+	/*
+	 * Without EPT it is not possible to restore L1's CR3 and PDPTR on
+	 * VMfail, because they are not available in vmcs01.  Just always
+	 * use hardware checks.
+	 */
+	if (!enable_ept)
+		nested_early_check = 1;
+
 	if (!cpu_has_vmx_shadow_vmcs())
 		enable_shadow_vmcs = 0;
 	if (enable_shadow_vmcs) {
Signature · 1 line
2.20.1
author maintainer
#51 lore raw

PATCH 51/137 KVM: fix spectrev1 gadgets

In reply to Greg Kroah-Hartman

[ Upstream commit 1d487e9bf8ba66a7174c56a0029c54b1eca8f99c ]

These were found with smatch, and then generalized when applicable.

---
 arch/x86/kvm/lapic.c     |  4 +++-
 include/linux/kvm_host.h | 10 ++++++----
 virt/kvm/irqchip.c       |  5 +++--
 virt/kvm/kvm_main.c      |  6 ++++--
 4 files changed, 16 insertions(+), 9 deletions(-)
arch/x86/kvm/lapic.c +3 -1
index 3339697de6e52..235687f3388fa 100644
@@ -137,6 +137,7 @@ static inline bool kvm_apic_map_get_logical_dest(struct kvm_apic_map *map,
 		if (offset <= max_apic_id) {
 			u8 cluster_size = min(max_apic_id - offset + 1, 16U);
 
+			offset = array_index_nospec(offset, map->max_apic_id + 1);
 			*cluster = &map->phys_map[offset];
 			*mask = dest_id & (0xffff >> (16 - cluster_size));
 		} else {
@@ -899,7 +900,8 @@ static inline bool kvm_apic_map_get_dest_lapic(struct kvm *kvm,
 		if (irq->dest_id > map->max_apic_id) {
 			*bitmap = 0;
 		} else {
-			*dst = &map->phys_map[irq->dest_id];
+			u32 dest_id = array_index_nospec(irq->dest_id, map->max_apic_id + 1);
+			*dst = &map->phys_map[dest_id];
 			*bitmap = 1;
 		}
 		return true;
include/linux/kvm_host.h +6 -4
index cf761ff582248..e41503b2c5a16 100644
@@ -28,6 +28,7 @@
 #include <linux/irqbypass.h>
 #include <linux/swait.h>
 #include <linux/refcount.h>
+#include <linux/nospec.h>
 #include <asm/signal.h>
 
 #include <linux/kvm.h>
@@ -492,10 +493,10 @@ static inline struct kvm_io_bus *kvm_get_bus(struct kvm *kvm, enum kvm_bus idx)
 
 static inline struct kvm_vcpu *kvm_get_vcpu(struct kvm *kvm, int i)
 {
-	/* Pairs with smp_wmb() in kvm_vm_ioctl_create_vcpu, in case
-	 * the caller has read kvm->online_vcpus before (as is the case
-	 * for kvm_for_each_vcpu, for example).
-	 */
+	int num_vcpus = atomic_read(&kvm->online_vcpus);
+	i = array_index_nospec(i, num_vcpus);
+
+	/* Pairs with smp_wmb() in kvm_vm_ioctl_create_vcpu.  */
 	smp_rmb();
 	return kvm->vcpus[i];
 }
@@ -579,6 +580,7 @@ void kvm_put_kvm(struct kvm *kvm);
 
 static inline struct kvm_memslots *__kvm_memslots(struct kvm *kvm, int as_id)
 {
+	as_id = array_index_nospec(as_id, KVM_ADDRESS_SPACE_NUM);
 	return srcu_dereference_check(kvm->memslots[as_id], &kvm->srcu,
 			lockdep_is_held(&kvm->slots_lock) ||
 			!refcount_read(&kvm->users_count));
virt/kvm/irqchip.c +3 -2
index b1286c4e07122..0bd0683640bdf 100644
@@ -144,18 +144,19 @@ static int setup_routing_entry(struct kvm *kvm,
 {
 	struct kvm_kernel_irq_routing_entry *ei;
 	int r;
+	u32 gsi = array_index_nospec(ue->gsi, KVM_MAX_IRQ_ROUTES);
 
 	/*
 	 * Do not allow GSI to be mapped to the same irqchip more than once.
 	 * Allow only one to one mapping between GSI and non-irqchip routing.
 	 */
-	hlist_for_each_entry(ei, &rt->map[ue->gsi], link)
+	hlist_for_each_entry(ei, &rt->map[gsi], link)
 		if (ei->type != KVM_IRQ_ROUTING_IRQCHIP ||
 		    ue->type != KVM_IRQ_ROUTING_IRQCHIP ||
 		    ue->u.irqchip.irqchip == ei->irqchip.irqchip)
 			return -EINVAL;
 
-	e->gsi = ue->gsi;
+	e->gsi = gsi;
 	e->type = ue->type;
 	r = kvm_set_routing_entry(kvm, e, ue);
 	if (r)
virt/kvm/kvm_main.c +4 -2
index b4f2d892a1d36..ff68b07e94e97 100644
@@ -2974,12 +2974,14 @@ static int kvm_ioctl_create_device(struct kvm *kvm,
 	struct kvm_device_ops *ops = NULL;
 	struct kvm_device *dev;
 	bool test = cd->flags & KVM_CREATE_DEVICE_TEST;
+	int type;
 	int ret;
 
 	if (cd->type >= ARRAY_SIZE(kvm_device_ops_table))
 		return -ENODEV;
 
-	ops = kvm_device_ops_table[cd->type];
+	type = array_index_nospec(cd->type, ARRAY_SIZE(kvm_device_ops_table));
+	ops = kvm_device_ops_table[type];
 	if (ops == NULL)
 		return -ENODEV;
 
@@ -2994,7 +2996,7 @@ static int kvm_ioctl_create_device(struct kvm *kvm,
 	dev->kvm = kvm;
 
 	mutex_lock(&kvm->lock);
-	ret = ops->create(dev, cd->type);
+	ret = ops->create(dev, type);
 	if (ret < 0) {
 		mutex_unlock(&kvm->lock);
 		kfree(dev);
Signature · 1 line
2.20.1
author maintainer
#52 lore raw

PATCH 52/137 KVM: x86: avoid misreporting level-triggered irqs as edge-triggered in tracing

In reply to Greg Kroah-Hartman

[ Upstream commit 7a223e06b1a411cef6c4cd7a9b9a33c8d225b10e ]

In __apic_accept_irq() interface trig_mode is int and actually on some code paths it is set above u8:

kvm_apic_set_irq() extracts it from 'struct kvm_lapic_irq' where trig_mode is u16. This is done on purpose as e.g. kvm_set_msi_irq() sets it to (1 << 15) & e->msi.data

kvm_apic_local_deliver sets it to reg & (1 << 15).

Fix the immediate issue by making 'tm' into u16. We may also want to adjust __apic_accept_irq() interface and use proper sizes for vector, level, trig_mode but this is not urgent.

---
 arch/x86/kvm/trace.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/kvm/trace.h +2 -2
index 6432d08c7de79..4d47a2631d1fb 100644
@@ -438,13 +438,13 @@ TRACE_EVENT(kvm_apic_ipi,
 );
 
 TRACE_EVENT(kvm_apic_accept_irq,
-	    TP_PROTO(__u32 apicid, __u16 dm, __u8 tm, __u8 vec),
+	    TP_PROTO(__u32 apicid, __u16 dm, __u16 tm, __u8 vec),
 	    TP_ARGS(apicid, dm, tm, vec),
 
 	TP_STRUCT__entry(
 		__field(	__u32,		apicid		)
 		__field(	__u16,		dm		)
-		__field(	__u8,		tm		)
+		__field(	__u16,		tm		)
 		__field(	__u8,		vec		)
 	),
 
Signature · 1 line
2.20.1
author maintainer
#53 lore raw

PATCH 53/137 tools lib traceevent: Fix missing equality check for strcmp

In reply to Greg Kroah-Hartman

[ Upstream commit f32c2877bcb068a718bb70094cd59ccc29d4d082 ]

There was a missing comparison with 0 when checking if type is "s64" or "u64". Therefore, the body of the if-statement was entered if "type" was "u64" or not "s64", which made the first strcmp() redundant since if type is "u64", it's not "s64".

If type is "s64", the body of the if-statement is not entered but since the remainder of the function consists of if-statements which will not be entered if type is "s64", we will just return "val", which is correct, albeit at the cost of a few more calls to strcmp(), i.e., it will behave just as if the if-statement was entered.

If type is neither "s64" or "u64", the body of the if-statement will be entered incorrectly and "val" returned. This means that any type that is checked after "s64" and "u64" is handled the same way as "s64" and "u64", i.e., the limiting of "val" to fit in for example "s8" is never reached.

This was introduced in the kernel tree when the sources were copied from trace-cmd in commit f7d82350e597 ("tools/events: Add files to create libtraceevent.a"), and in the trace-cmd repo in 1cdbae6035cei ("Implement typecasting in parser") when the function was introduced, i.e., it has always behaved the wrong way.

Detected by cppcheck.

Cc Tzvetomir Stoyanov <[email protected]>
Fixes f7d82350e597 ("tools/events: Add files to create libtraceevent.a")
---
 tools/lib/traceevent/event-parse.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/lib/traceevent/event-parse.c +1 -1
index 87494c7c619d8..981c6ce2da2c7 100644
@@ -2233,7 +2233,7 @@ eval_type_str(unsigned long long val, const char *type, int pointer)
 		return val & 0xffffffff;
 
 	if (strcmp(type, "u64") == 0 ||
-	    strcmp(type, "s64"))
+	    strcmp(type, "s64") == 0)
 		return val;
 
 	if (strcmp(type, "s8") == 0)
Signature · 1 line
2.20.1
author maintainer
#54 lore raw

PATCH 54/137 perf top: Always sample time to satisfy needs of use of ordered queuing

In reply to Greg Kroah-Hartman

[ Upstream commit 1e6db2ee86e6a4399fc0ae5689e55e0fd1c43caf ]

Bastian reported broken 'perf top -p PID' command, it won't display any data.

The problem is that for -p option we monitor single thread, so we don't enable time in samples, because it's not needed.

However since commit 16c66bc167cc we use ordered queues to stash data plus later commits added logic for dropping samples in case there's big load and we don't keep up. All this needs timestamp for sample. Enabling it unconditionally for perf top.

---
 tools/perf/builtin-top.c | 1 +
 1 file changed, 1 insertion(+)
tools/perf/builtin-top.c +1 -0
index 616408251e258..63750a711123f 100644
@@ -1393,6 +1393,7 @@ int cmd_top(int argc, const char **argv)
 			 * */
 			.overwrite	= 0,
 			.sample_time	= true,
+			.sample_time_set = true,
 		},
 		.max_stack	     = sysctl__max_stack(),
 		.annotation_opts     = annotation__default_options,
Signature · 1 line
2.20.1
author maintainer
#55 lore raw

PATCH 55/137 ipmi: ipmi_si_hardcode.c: init si_type array to fix a crash

In reply to Greg Kroah-Hartman

[ Upstream commit a885bcfd152f97b25005298ab2d6b741aed9b49c ]

The intended behavior of function ipmi_hardcode_init_one() is to default to kcs interface when no type argument is presented when initializing ipmi with hard coded addresses.

However, the array of char pointers allocated on the stack by function ipmi_hardcode_init() was not inited to zeroes, so it contained stack debris.

Consequently, passing the cruft stored in this array to function ipmi_hardcode_init_one() caused a crash when it was unable to detect that the char * being passed was nonsense and tried to access the address specified by the bogus pointer.

The fix is simply to initialize the si_type array to zeroes, so if there were no type argument given to at the command line, function ipmi_hardcode_init_one() could properly default to the kcs interface.

Message-Id: <[email protected]>

---
 drivers/char/ipmi/ipmi_si_hardcode.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/char/ipmi/ipmi_si_hardcode.c +2 -0
index 1e5783961b0dc..ab7180c46d8dd 100644
@@ -201,6 +201,8 @@ void __init ipmi_hardcode_init(void)
 	char *str;
 	char *si_type[SI_MAX_PARMS];
 
+	memset(si_type, 0, sizeof(si_type));
+
 	/* Parse out the si_type string into its components. */
 	str = si_type_str;
 	if (*str != '\0') {
Signature · 1 line
2.20.1
author maintainer
#56 lore raw

PATCH 56/137 ocelot: Dont sleep in atomic context (irqs_disabled())

In reply to Greg Kroah-Hartman

[ Upstream commit a8fd48b50deaa20808bbf0f6685f6f1acba6a64c ]

Preemption disabled at:

 [<ffff000008cabd54>] dev_set_rx_mode+0x1c/0x38
 Call trace:
 [<ffff00000808a5c0>] dump_backtrace+0x0/0x3d0
 [<ffff00000808a9a4>] show_stack+0x14/0x20
 [<ffff000008e6c0c0>] dump_stack+0xac/0xe4
 [<ffff0000080fe76c>] ___might_sleep+0x164/0x238
 [<ffff0000080fe890>] __might_sleep+0x50/0x88
 [<ffff0000082261e4>] kmem_cache_alloc+0x17c/0x1d0
 [<ffff000000ea0ae8>] ocelot_set_rx_mode+0x108/0x188 [mscc_ocelot_common]
 [<ffff000008cabcf0>] __dev_set_rx_mode+0x58/0xa0
 [<ffff000008cabd5c>] dev_set_rx_mode+0x24/0x38
Fixes a556c76adc05 ("net: mscc: Add initial Ocelot switch support")
---
 drivers/net/ethernet/mscc/ocelot.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/mscc/ocelot.c +1 -1
index 215a45374d7b0..0ef95abde6bb0 100644
@@ -613,7 +613,7 @@ static int ocelot_mact_mc_add(struct ocelot_port *port,
 			      struct netdev_hw_addr *hw_addr)
 {
 	struct ocelot *ocelot = port->ocelot;
-	struct netdev_hw_addr *ha = kzalloc(sizeof(*ha), GFP_KERNEL);
+	struct netdev_hw_addr *ha = kzalloc(sizeof(*ha), GFP_ATOMIC);
 
 	if (!ha)
 		return -ENOMEM;
Signature · 1 line
2.20.1
author maintainer
#57 lore raw

PATCH 57/137 perf tools: Fix map reference counting

In reply to Greg Kroah-Hartman

[ Upstream commit b9abbdfa88024d52c8084d8f46ea4f161606c692 ]

By calling maps__insert() we assume to get 2 references on the map, which we relese within maps__remove call.

However if there's already same map name, we currently don't bump the reference and can crash, like:

Program received signal SIGABRT, Aborted. 0x00007ffff75e60f5 in raise () from /lib64/libc.so.6

(gdb) bt

  #0  0x00007ffff75e60f5 in raise () from /lib64/libc.so.6
  #1  0x00007ffff75d0895 in abort () from /lib64/libc.so.6
  #2  0x00007ffff75d0769 in __assert_fail_base.cold () from /lib64/libc.so.6
  #3  0x00007ffff75de596 in __assert_fail () from /lib64/libc.so.6
  #4  0x00000000004fc006 in refcount_sub_and_test (i=1, r=0x1224e88) at tools/include/linux/refcount.h:131
  #5  refcount_dec_and_test (r=0x1224e88) at tools/include/linux/refcount.h:148
  #6  map__put (map=0x1224df0) at util/map.c:299
  #7  0x00000000004fdb95 in __maps__remove (map=0x1224df0, maps=0xb17d80) at util/map.c:953
  #8  maps__remove (maps=0xb17d80, map=0x1224df0) at util/map.c:959
  #9  0x00000000004f7d8a in map_groups__remove (map=<optimized out>, mg=<optimized out>) at util/map_groups.h:65

#10 machine__process_ksymbol_unregister (sample=<optimized out>, event=0x7ffff7279670, machine=<optimized out>) at util/machine.c:728 #11 machine__process_ksymbol (machine=<optimized out>, event=0x7ffff7279670, sample=<optimized out>) at util/machine.c:741
#12 0x00000000004fffbb in perf_session__deliver_event (session=0xb11390, event=0x7ffff7279670, tool=0x7fffffffc7b0, file_offset=13936) at util/session.c:1362 #13 0x00000000005039bb in do_flush (show_progress=false, oe=0xb17e80) at util/ordered-events.c:243
#14 __ordered_events__flush (oe=0xb17e80, how=OE_FLUSH__ROUND, timestamp=<optimized out>) at util/ordered-events.c:322
#15 0x00000000005005e4 in perf_session__process_user_event (session=session@entry=0xb11390, event=event@entry=0x7ffff72a4af8,
...

Add the map to the list and getting the reference event if we find the map with same name.

Cc Adrian Hunter <[email protected]>
Cc Alexander Shishkin <[email protected]>
Cc Alexei Starovoitov <[email protected]>
Cc Andi Kleen <[email protected]>
Cc Daniel Borkmann <[email protected]>
Cc Eric Saint-Etienne <[email protected]>
Cc Namhyung Kim <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Song Liu <[email protected]>
Fixes 1e6285699b30 ("perf symbols: Fix slowness due to -ffunction-section")
---
 tools/perf/util/map.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)
tools/perf/util/map.c +1 -3
index 2b37f56f05493..e33f20d16c8d6 100644
@@ -904,10 +904,8 @@ static void __maps__insert_name(struct maps *maps, struct map *map)
 		rc = strcmp(m->dso->short_name, map->dso->short_name);
 		if (rc < 0)
 			p = &(*p)->rb_left;
-		else if (rc  > 0)
-			p = &(*p)->rb_right;
 		else
-			return;
+			p = &(*p)->rb_right;
 	}
 	rb_link_node(&map->rb_node_name, parent, p);
 	rb_insert_color(&map->rb_node_name, &maps->names);
Signature · 1 line
2.20.1
author maintainer
#58 lore raw

PATCH 58/137 scsi: aic7xxx: fix EISA support

In reply to Greg Kroah-Hartman

[ Upstream commit 144ec97493af34efdb77c5aba146e9c7de8d0a06 ]

Instead of relying on the now removed NULL argument to pci_alloc_consistent, switch to the generic DMA API, and store the struct device so that we can pass it.

Fixes 4167b2ad5182 ("PCI: Remove NULL device handling from PCI DMA API")
---
 drivers/scsi/aic7xxx/aic7770_osm.c     |  1 +
 drivers/scsi/aic7xxx/aic7xxx.h         |  1 +
 drivers/scsi/aic7xxx/aic7xxx_osm.c     | 10 ++++------
 drivers/scsi/aic7xxx/aic7xxx_osm_pci.c |  1 +
 4 files changed, 7 insertions(+), 6 deletions(-)
drivers/scsi/aic7xxx/aic7770_osm.c +1 -0
index 3d401d02c0195..bdd177e3d7622 100644
@@ -91,6 +91,7 @@ aic7770_probe(struct device *dev)
 	ahc = ahc_alloc(&aic7xxx_driver_template, name);
 	if (ahc == NULL)
 		return (ENOMEM);
+	ahc->dev = dev;
 	error = aic7770_config(ahc, aic7770_ident_table + edev->id.driver_data,
 			       eisaBase);
 	if (error != 0) {
drivers/scsi/aic7xxx/aic7xxx.h +1 -0
index 5614921b4041a..88b90f9806c99 100644
@@ -943,6 +943,7 @@ struct ahc_softc {
 	 * Platform specific device information.
 	 */
 	ahc_dev_softc_t		  dev_softc;
+	struct device		  *dev;
 
 	/*
 	 * Bus specific device information.
drivers/scsi/aic7xxx/aic7xxx_osm.c +4 -6
index 3c9c17450bb39..d5c4a0d237062 100644
@@ -860,8 +860,8 @@ int
 ahc_dmamem_alloc(struct ahc_softc *ahc, bus_dma_tag_t dmat, void** vaddr,
 		 int flags, bus_dmamap_t *mapp)
 {
-	*vaddr = pci_alloc_consistent(ahc->dev_softc,
-				      dmat->maxsize, mapp);
+	/* XXX: check if we really need the GFP_ATOMIC and unwind this mess! */
+	*vaddr = dma_alloc_coherent(ahc->dev, dmat->maxsize, mapp, GFP_ATOMIC);
 	if (*vaddr == NULL)
 		return ENOMEM;
 	return 0;
@@ -871,8 +871,7 @@ void
 ahc_dmamem_free(struct ahc_softc *ahc, bus_dma_tag_t dmat,
 		void* vaddr, bus_dmamap_t map)
 {
-	pci_free_consistent(ahc->dev_softc, dmat->maxsize,
-			    vaddr, map);
+	dma_free_coherent(ahc->dev, dmat->maxsize, vaddr, map);
 }
 
 int
@@ -1123,8 +1122,7 @@ ahc_linux_register_host(struct ahc_softc *ahc, struct scsi_host_template *templa
 
 	host->transportt = ahc_linux_transport_template;
 
-	retval = scsi_add_host(host,
-			(ahc->dev_softc ? &ahc->dev_softc->dev : NULL));
+	retval = scsi_add_host(host, ahc->dev);
 	if (retval) {
 		printk(KERN_WARNING "aic7xxx: scsi_add_host failed\n");
 		scsi_host_put(host);
drivers/scsi/aic7xxx/aic7xxx_osm_pci.c +1 -0
index 0fc14dac7070c..717d8d1082ce1 100644
@@ -250,6 +250,7 @@ ahc_linux_pci_dev_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 		}
 	}
 	ahc->dev_softc = pci;
+	ahc->dev = &pci->dev;
 	error = ahc_pci_config(ahc, entry);
 	if (error != 0) {
 		ahc_free(ahc);
Signature · 1 line
2.20.1
author maintainer
#59 lore raw

PATCH 59/137 slab: store tagged freelist for off-slab slabmgmt

In reply to Greg Kroah-Hartman

[ Upstream commit 1a62b18d51e5c5ecc0345c85bb9fef870ab721ed ]

Commit 51dedad06b5f ("kasan, slab: make freelist stored without tags") calls kasan_reset_tag() for off-slab slab management object leading to freelist being stored non-tagged.

However, cache_grow_begin() calls alloc_slabmgmt() which calls kmem_cache_alloc_node() assigns a tag for the address and stores it in the shadow address. As the result, it causes endless errors below during boot due to drain_freelist() -> slab_destroy() -> kasan_slab_free() which compares already untagged freelist against the stored tag in the shadow address.

Since off-slab slab management object freelist is such a special case, just store it tagged. Non-off-slab management object freelist is still stored untagged which has not been assigned a tag and should not cause any other troubles with this inconsistency.

  BUG: KASAN: double-free or invalid-free in slab_destroy+0x84/0x88
  Pointer tag: [ff], memory tag: [99]

  CPU: 0 PID: 1376 Comm: kworker/0:4 Tainted: G        W 5.1.0-rc3+ #8
  Hardware name: HPE Apollo 70             /C01_APACHE_MB         , BIOS L50_5.13_1.0.6 07/10/2018
  Workqueue: cgroup_destroy css_killed_work_fn
  Call trace:
   print_address_description+0x74/0x2a4
   kasan_report_invalid_free+0x80/0xc0
   __kasan_slab_free+0x204/0x208
   kasan_slab_free+0xc/0x18
   kmem_cache_free+0xe4/0x254
   slab_destroy+0x84/0x88
   drain_freelist+0xd0/0x104
   __kmem_cache_shrink+0x1ac/0x224
   __kmemcg_cache_deactivate+0x1c/0x28
   memcg_deactivate_kmem_caches+0xa0/0xe8
   memcg_offline_kmem+0x8c/0x3d4
   mem_cgroup_css_offline+0x24c/0x290
   css_killed_work_fn+0x154/0x618
   process_one_work+0x9cc/0x183c
   worker_thread+0x9b0/0xe38
   kthread+0x374/0x390
   ret_from_fork+0x10/0x18

  Allocated by task 1625:
   __kasan_kmalloc+0x168/0x240
   kasan_slab_alloc+0x18/0x20
   kmem_cache_alloc_node+0x1f8/0x3a0
   cache_grow_begin+0x4fc/0xa24
   cache_alloc_refill+0x2f8/0x3e8
   kmem_cache_alloc+0x1bc/0x3bc
   sock_alloc_inode+0x58/0x334
   alloc_inode+0xb8/0x164
   new_inode_pseudo+0x20/0xec
   sock_alloc+0x74/0x284
   __sock_create+0xb0/0x58c
   sock_create+0x98/0xb8
   __sys_socket+0x60/0x138
   __arm64_sys_socket+0xa4/0x110
   el0_svc_handler+0x2c0/0x47c
   el0_svc+0x8/0xc

  Freed by task 1625:
   __kasan_slab_free+0x114/0x208
   kasan_slab_free+0xc/0x18
   kfree+0x1a8/0x1e0
   single_release+0x7c/0x9c
   close_pdeo+0x13c/0x43c
   proc_reg_release+0xec/0x108
   __fput+0x2f8/0x784
   ____fput+0x1c/0x28
   task_work_run+0xc0/0x1b0
   do_notify_resume+0xb44/0x1278
   work_pending+0x8/0x10

  The buggy address belongs to the object at ffff809681b89e00
   which belongs to the cache kmalloc-128 of size 128
  The buggy address is located 0 bytes inside of
   128-byte region [ffff809681b89e00, ffff809681b89e80)
  The buggy address belongs to the page:
  page:ffff7fe025a06e00 count:1 mapcount:0 mapping:01ff80082000fb00
  index:0xffff809681b8fe04
  flags: 0x17ffffffc000200(slab)
  raw: 017ffffffc000200 ffff7fe025a06d08 ffff7fe022ef7b88 01ff80082000fb00
  raw: ffff809681b8fe04 ffff809681b80000 00000001000000e0 0000000000000000
  page dumped because: kasan: bad access detected
  page allocated via order 0, migratetype Unmovable, gfp_mask
  0x2420c0(__GFP_IO|__GFP_FS|__GFP_NOWARN|__GFP_COMP|__GFP_THISNODE)
   prep_new_page+0x4e0/0x5e0
   get_page_from_freelist+0x4ce8/0x50d4
   __alloc_pages_nodemask+0x738/0x38b8
   cache_grow_begin+0xd8/0xa24
   ____cache_alloc_node+0x14c/0x268
   __kmalloc+0x1c8/0x3fc
   ftrace_free_mem+0x408/0x1284
   ftrace_free_init_mem+0x20/0x28
   kernel_init+0x24/0x548
   ret_from_fork+0x10/0x18

  Memory state around the buggy address:
   ffff809681b89c00: fe fe fe fe fe fe fe fe fe fe fe fe fe fe fe fe
   ffff809681b89d00: fe fe fe fe fe fe fe fe fe fe fe fe fe fe fe fe
  >ffff809681b89e00: 99 99 99 99 99 99 99 99 fe fe fe fe fe fe fe fe
                     ^
   ffff809681b89f00: 43 43 43 43 43 fe fe fe fe fe fe fe fe fe fe fe
   ffff809681b8a000: 6d fe fe fe fe fe fe fe fe fe fe fe fe fe fe fe
Fixes 51dedad06b5f ("kasan, slab: make freelist stored without tags")
Cc Christoph Lameter <[email protected]>
Cc Pekka Enberg <[email protected]>
Cc David Rientjes <[email protected]>
Cc Joonsoo Kim <[email protected]>
Cc Andrey Ryabinin <[email protected]>
Cc Alexander Potapenko <[email protected]>
Cc Dmitry Vyukov <[email protected]>
---
 mm/slab.c | 1 -
 1 file changed, 1 deletion(-)
mm/slab.c +0 -1
index 188c4b65255dc..f4bbc53008f3b 100644
@@ -2371,7 +2371,6 @@ static void *alloc_slabmgmt(struct kmem_cache *cachep,
 		/* Slab management obj is off-slab. */
 		freelist = kmem_cache_alloc_node(cachep->freelist_cache,
 					      local_flags, nodeid);
-		freelist = kasan_reset_tag(freelist);
 		if (!freelist)
 			return NULL;
 	} else {
Signature · 1 line
2.20.1
author maintainer
#60 lore raw

PATCH 60/137 mm/hotplug: treat CMA pages as unmovable

In reply to Greg Kroah-Hartman

[ Upstream commit 1a9f219157b22d0ffb340a9c5f431afd02cd2cf3 ]

has_unmovable_pages() is used by allocating CMA and gigantic pages as well as the memory hotplug. The later doesn't know how to offline CMA pool properly now, but if an unused (free) CMA page is encountered, then has_unmovable_pages() happily considers it as a free memory and propagates this up the call chain. Memory offlining code then frees the page without a proper CMA tear down which leads to an accounting issues. Moreover if the same memory range is onlined again then the memory never gets back to the CMA pool.

State after memory offline:

 # grep cma /proc/vmstat
 nr_free_cma 205824
 # cat /sys/kernel/debug/cma/cma-kvm_cma/count
 209920

Also, kmemleak still think those memory address are reserved below but have already been used by the buddy allocator after onlining. This patch fixes the situation by treating CMA pageblocks as unmovable except when has_unmovable_pages() is called as part of CMA allocation.

Offlined Pages 4096
kmemleak: Cannot insert 0xc000201f7d040008 into the object search tree (overlaps existing)

  Call Trace:
    dump_stack+0xb0/0xf4 (unreliable)
    create_object+0x344/0x380
    __kmalloc_node+0x3ec/0x860
    kvmalloc_node+0x58/0x110
    seq_read+0x41c/0x620
    __vfs_read+0x3c/0x70
    vfs_read+0xbc/0x1a0
    ksys_read+0x7c/0x140
    system_call+0x5c/0x70
  kmemleak: Kernel memory leak detector disabled
  kmemleak: Object 0xc000201cc8000000 (size 13757317120):
  kmemleak:   comm "swapper/0", pid 0, jiffies 4294937297
  kmemleak:   min_count = -1
  kmemleak:   count = 0
  kmemleak:   flags = 0x5
  kmemleak:   checksum = 0
  kmemleak:   backtrace:
       cma_declare_contiguous+0x2a4/0x3b0
       kvm_cma_reserve+0x11c/0x134
       setup_arch+0x300/0x3f8
       start_kernel+0x9c/0x6e8
       start_here_common+0x1c/0x4b0
  kmemleak: Automatic memory scanning thread ended
---
 mm/page_alloc.c | 30 ++++++++++++++++++------------
 1 file changed, 18 insertions(+), 12 deletions(-)
mm/page_alloc.c +18 -12
index 318ef6ccdb3b5..eedb57f9b40b5 100644
@@ -7945,7 +7945,10 @@ void *__init alloc_large_system_hash(const char *tablename,
 bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
 			 int migratetype, int flags)
 {
-	unsigned long pfn, iter, found;
+	unsigned long found;
+	unsigned long iter = 0;
+	unsigned long pfn = page_to_pfn(page);
+	const char *reason = "unmovable page";
 
 	/*
 	 * TODO we could make this much more efficient by not checking every
@@ -7955,17 +7958,20 @@ bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
 	 * can still lead to having bootmem allocations in zone_movable.
 	 */
 
-	/*
-	 * CMA allocations (alloc_contig_range) really need to mark isolate
-	 * CMA pageblocks even when they are not movable in fact so consider
-	 * them movable here.
-	 */
-	if (is_migrate_cma(migratetype) &&
-			is_migrate_cma(get_pageblock_migratetype(page)))
-		return false;
+	if (is_migrate_cma_page(page)) {
+		/*
+		 * CMA allocations (alloc_contig_range) really need to mark
+		 * isolate CMA pageblocks even when they are not movable in fact
+		 * so consider them movable here.
+		 */
+		if (is_migrate_cma(migratetype))
+			return false;
+
+		reason = "CMA page";
+		goto unmovable;
+	}
 
-	pfn = page_to_pfn(page);
-	for (found = 0, iter = 0; iter < pageblock_nr_pages; iter++) {
+	for (found = 0; iter < pageblock_nr_pages; iter++) {
 		unsigned long check = pfn + iter;
 
 		if (!pfn_valid_within(check))
@@ -8045,7 +8051,7 @@ bool has_unmovable_pages(struct zone *zone, struct page *page, int count,
 unmovable:
 	WARN_ON_ONCE(zone_idx(zone) == ZONE_MOVABLE);
 	if (flags & REPORT_FAILURE)
-		dump_page(pfn_to_page(pfn+iter), "unmovable page");
+		dump_page(pfn_to_page(pfn + iter), reason);
 	return true;
 }
 
Signature · 1 line
2.20.1
author maintainer
#61 lore raw

PATCH 61/137 mm: fix inactive list balancing between NUMA nodes and cgroups

In reply to Greg Kroah-Hartman

[ Upstream commit 3b991208b897f52507168374033771a984b947b1 ]

During !CONFIG_CGROUP reclaim, we expand the inactive list size if it's thrashing on the node that is about to be reclaimed. But when cgroups are enabled, we suddenly ignore the node scope and use the cgroup scope only. The result is that pressure bleeds between NUMA nodes depending on whether cgroups are merely compiled into Linux. This behavioral difference is unexpected and undesirable.

When the refault adaptivity of the inactive list was first introduced, there were no statistics at the lruvec level - the intersection of node and memcg - so it was better than nothing.

But now that we have that infrastructure, use lruvec_page_state() to make the list balancing decision always NUMA aware.

[[email protected]: fix bisection hole]
Fixes 2a2e48854d70 ("mm: vmscan: fix IO/refault regression in cache workingset transition")
Cc Roman Gushchin <[email protected]>
Cc Michal Hocko <[email protected]>
---
 mm/vmscan.c | 29 +++++++++--------------------
 1 file changed, 9 insertions(+), 20 deletions(-)
mm/vmscan.c +9 -20
index e979705bbf325..022afabac3f69 100644
@@ -2199,7 +2199,6 @@ static void shrink_active_list(unsigned long nr_to_scan,
  *   10TB     320        32GB
  */
 static bool inactive_list_is_low(struct lruvec *lruvec, bool file,
-				 struct mem_cgroup *memcg,
 				 struct scan_control *sc, bool actual_reclaim)
 {
 	enum lru_list active_lru = file * LRU_FILE + LRU_ACTIVE;
@@ -2220,16 +2219,12 @@ static bool inactive_list_is_low(struct lruvec *lruvec, bool file,
 	inactive = lruvec_lru_size(lruvec, inactive_lru, sc->reclaim_idx);
 	active = lruvec_lru_size(lruvec, active_lru, sc->reclaim_idx);
 
-	if (memcg)
-		refaults = memcg_page_state(memcg, WORKINGSET_ACTIVATE);
-	else
-		refaults = node_page_state(pgdat, WORKINGSET_ACTIVATE);
-
 	/*
 	 * When refaults are being observed, it means a new workingset
 	 * is being established. Disable active list protection to get
 	 * rid of the stale workingset quickly.
 	 */
+	refaults = lruvec_page_state(lruvec, WORKINGSET_ACTIVATE);
 	if (file && actual_reclaim && lruvec->refaults != refaults) {
 		inactive_ratio = 0;
 	} else {
@@ -2250,12 +2245,10 @@ static bool inactive_list_is_low(struct lruvec *lruvec, bool file,
 }
 
 static unsigned long shrink_list(enum lru_list lru, unsigned long nr_to_scan,
-				 struct lruvec *lruvec, struct mem_cgroup *memcg,
-				 struct scan_control *sc)
+				 struct lruvec *lruvec, struct scan_control *sc)
 {
 	if (is_active_lru(lru)) {
-		if (inactive_list_is_low(lruvec, is_file_lru(lru),
-					 memcg, sc, true))
+		if (inactive_list_is_low(lruvec, is_file_lru(lru), sc, true))
 			shrink_active_list(nr_to_scan, lruvec, sc, lru);
 		return 0;
 	}
@@ -2355,7 +2348,7 @@ static void get_scan_count(struct lruvec *lruvec, struct mem_cgroup *memcg,
 			 * anonymous pages on the LRU in eligible zones.
 			 * Otherwise, the small LRU gets thrashed.
 			 */
-			if (!inactive_list_is_low(lruvec, false, memcg, sc, false) &&
+			if (!inactive_list_is_low(lruvec, false, sc, false) &&
 			    lruvec_lru_size(lruvec, LRU_INACTIVE_ANON, sc->reclaim_idx)
 					>> sc->priority) {
 				scan_balance = SCAN_ANON;
@@ -2373,7 +2366,7 @@ static void get_scan_count(struct lruvec *lruvec, struct mem_cgroup *memcg,
 	 * lruvec even if it has plenty of old anonymous pages unless the
 	 * system is under heavy pressure.
 	 */
-	if (!inactive_list_is_low(lruvec, true, memcg, sc, false) &&
+	if (!inactive_list_is_low(lruvec, true, sc, false) &&
 	    lruvec_lru_size(lruvec, LRU_INACTIVE_FILE, sc->reclaim_idx) >> sc->priority) {
 		scan_balance = SCAN_FILE;
 		goto out;
@@ -2526,7 +2519,7 @@ static void shrink_node_memcg(struct pglist_data *pgdat, struct mem_cgroup *memc
 				nr[lru] -= nr_to_scan;
 
 				nr_reclaimed += shrink_list(lru, nr_to_scan,
-							    lruvec, memcg, sc);
+							    lruvec, sc);
 			}
 		}
 
@@ -2593,7 +2586,7 @@ static void shrink_node_memcg(struct pglist_data *pgdat, struct mem_cgroup *memc
 	 * Even if we did not try to evict anon pages at all, we want to
 	 * rebalance the anon lru active/inactive ratio.
 	 */
-	if (inactive_list_is_low(lruvec, false, memcg, sc, true))
+	if (inactive_list_is_low(lruvec, false, sc, true))
 		shrink_active_list(SWAP_CLUSTER_MAX, lruvec,
 				   sc, LRU_ACTIVE_ANON);
 }
@@ -2993,12 +2986,8 @@ static void snapshot_refaults(struct mem_cgroup *root_memcg, pg_data_t *pgdat)
 		unsigned long refaults;
 		struct lruvec *lruvec;
 
-		if (memcg)
-			refaults = memcg_page_state(memcg, WORKINGSET_ACTIVATE);
-		else
-			refaults = node_page_state(pgdat, WORKINGSET_ACTIVATE);
-
 		lruvec = mem_cgroup_lruvec(pgdat, memcg);
+		refaults = lruvec_page_state(lruvec, WORKINGSET_ACTIVATE);
 		lruvec->refaults = refaults;
 	} while ((memcg = mem_cgroup_iter(root_memcg, memcg, NULL)));
 }
@@ -3363,7 +3352,7 @@ static void age_active_anon(struct pglist_data *pgdat,
 	do {
 		struct lruvec *lruvec = mem_cgroup_lruvec(pgdat, memcg);
 
-		if (inactive_list_is_low(lruvec, false, memcg, sc, true))
+		if (inactive_list_is_low(lruvec, false, sc, true))
 			shrink_active_list(SWAP_CLUSTER_MAX, lruvec,
 					   sc, LRU_ACTIVE_ANON);
 
Signature · 1 line
2.20.1
author maintainer
#62 lore raw

PATCH 62/137 init: initialize jump labels before command line option parsing

In reply to Greg Kroah-Hartman

[ Upstream commit 6041186a32585fc7a1d0f6cfe2f138b05fdc3c82 ]

When a module option, or core kernel argument, toggles a static-key it requires jump labels to be initialized early. While x86, PowerPC, and ARM64 arrange for jump_label_init() to be called before parse_args(), ARM does not.

Kernel command line: rdinit=/sbin/init page_alloc.shuffle=1 panic=-1 console=ttyAMA0,115200 page_alloc.shuffle=1

  ------------[ cut here ]------------
  WARNING: CPU: 0 PID: 0 at ./include/linux/jump_label.h:303
  page_alloc_shuffle+0x12c/0x1ac
  static_key_enable(): static key 'page_alloc_shuffle_key+0x0/0x4' used
  before call to jump_label_init()
  Modules linked in:
  CPU: 0 PID: 0 Comm: swapper Not tainted
  5.1.0-rc4-next-20190410-00003-g3367c36ce744 #1
  Hardware name: ARM Integrator/CP (Device Tree)
  [<c0011c68>] (unwind_backtrace) from [<c000ec48>] (show_stack+0x10/0x18)
  [<c000ec48>] (show_stack) from [<c07e9710>] (dump_stack+0x18/0x24)
  [<c07e9710>] (dump_stack) from [<c001bb1c>] (__warn+0xe0/0x108)
  [<c001bb1c>] (__warn) from [<c001bb88>] (warn_slowpath_fmt+0x44/0x6c)
  [<c001bb88>] (warn_slowpath_fmt) from [<c0b0c4a8>]
  (page_alloc_shuffle+0x12c/0x1ac)
  [<c0b0c4a8>] (page_alloc_shuffle) from [<c0b0c550>] (shuffle_store+0x28/0x48)
  [<c0b0c550>] (shuffle_store) from [<c003e6a0>] (parse_args+0x1f4/0x350)
  [<c003e6a0>] (parse_args) from [<c0ac3c00>] (start_kernel+0x1c0/0x488)

Move the fallback call to jump_label_init() to occur before parse_args().

The redundant calls to jump_label_init() in other archs are left intact in case they have static key toggling use cases that are even earlier than option parsing.

---
 init/main.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
init/main.c +2 -2
index c86a1c8f19f40..7ae8245452650 100644
@@ -574,6 +574,8 @@ asmlinkage __visible void __init start_kernel(void)
 	page_alloc_init();
 
 	pr_notice("Kernel command line: %s\n", boot_command_line);
+	/* parameters may set static keys */
+	jump_label_init();
 	parse_early_param();
 	after_dashes = parse_args("Booting kernel",
 				  static_command_line, __start___param,
@@ -583,8 +585,6 @@ asmlinkage __visible void __init start_kernel(void)
 		parse_args("Setting init args", after_dashes, NULL, 0, -1, -1,
 			   NULL, set_init_arg);
 
-	jump_label_init();
-
 	/*
 	 * These use large bootmem allocations and must precede
 	 * kmem_cache_init()
Signature · 1 line
2.20.1
author maintainer
#63 lore raw

PATCH 63/137 drm: bridge: dw-hdmi: Fix overflow workaround for Rockchip SoCs

In reply to Greg Kroah-Hartman

[ Upstream commit d15d9fd02575ecfada92d42f655940c4f10af842 ]

The Rockchip RK3288 SoC (v2.00a) and RK3328/RK3399 SoCs (v2.11a) have also been identified as needing this workaround with a single iteration.

---
 drivers/gpu/drm/bridge/synopsys/dw-hdmi.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/gpu/drm/bridge/synopsys/dw-hdmi.c +4 -0
index 64c3cf0275182..14223c0ee7843 100644
@@ -1655,6 +1655,8 @@ static void dw_hdmi_clear_overflow(struct dw_hdmi *hdmi)
 	 * iteration for others.
 	 * The Amlogic Meson GX SoCs (v2.01a) have been identified as needing
 	 * the workaround with a single iteration.
+	 * The Rockchip RK3288 SoC (v2.00a) and RK3328/RK3399 SoCs (v2.11a) have
+	 * been identified as needing the workaround with a single iteration.
 	 */
 
 	switch (hdmi->version) {
@@ -1663,7 +1665,9 @@ static void dw_hdmi_clear_overflow(struct dw_hdmi *hdmi)
 		break;
 	case 0x131a:
 	case 0x132a:
+	case 0x200a:
 	case 0x201a:
+	case 0x211a:
 	case 0x212a:
 		count = 1;
 		break;
Signature · 1 line
2.20.1
author maintainer
#64 lore raw

PATCH 64/137 selftests: netfilter: check icmp pkttoobig errors are set as related

In reply to Greg Kroah-Hartman

[ Upstream commit becf2319f320cae43e20cf179cc51a355a0deb5f ]

When an icmp error such as pkttoobig is received, conntrack checks if the "inner" header (header of packet that did not fit link mtu) is matches an existing connection, and, if so, sets that packet as being related to the conntrack entry it found.

It was recently reported that this "related" setting also works if the inner header is from another, different connection (i.e., artificial/forged icmp error).

Add a test, followup patch will add additional "inner dst matches outer dst in reverse direction" check before setting related state.

---
 tools/testing/selftests/netfilter/Makefile    |   2 +-
 .../netfilter/conntrack_icmp_related.sh       | 283 ++++++++++++++++++
 2 files changed, 284 insertions(+), 1 deletion(-)
 create mode 100755 tools/testing/selftests/netfilter/conntrack_icmp_related.sh
tools/testing/selftests/netfilter/Makefile +1 -1
index c9ff2b47bd1ca..a37cb1192c6a6 100644
@@ -1,6 +1,6 @@
 # SPDX-License-Identifier: GPL-2.0
 # Makefile for netfilter selftests
 
-TEST_PROGS := nft_trans_stress.sh nft_nat.sh
+TEST_PROGS := nft_trans_stress.sh nft_nat.sh conntrack_icmp_related.sh
 
 include ../lib.mk
tools/testing/selftests/netfilter/conntrack_icmp_related.sh new file +283 -0

283 lines not on this page · raw message

Signature · 1 line
2.20.1
author maintainer
#65 lore raw

PATCH 65/137 ipvs: do not schedule icmp errors from tunnels

In reply to Greg Kroah-Hartman

[ Upstream commit 0261ea1bd1eb0da5c0792a9119b8655cf33c80a3 ]

We can receive ICMP errors from client or from
tunneling real server. While the former can be
scheduled to real server, the latter should
not be scheduled, they are decapsulated only when
existing connection is found.

Fixes 6044eeffafbe ("ipvs: attempt to schedule icmp packets")
---
 net/netfilter/ipvs/ip_vs_core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/netfilter/ipvs/ip_vs_core.c +1 -1
index 235205c93e14b..df112b27246a3 100644
@@ -1647,7 +1647,7 @@ ip_vs_in_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, int *related,
 	if (!cp) {
 		int v;
 
-		if (!sysctl_schedule_icmp(ipvs))
+		if (ipip || !sysctl_schedule_icmp(ipvs))
 			return NF_ACCEPT;
 
 		if (!ip_vs_try_to_schedule(ipvs, AF_INET, skb, pd, &v, &cp, &ciph))
Signature · 1 line
2.20.1
author maintainer
#66 lore raw

PATCH 66/137 netfilter: ctnetlink: dont use conntrack/expect object addresses as id

In reply to Greg Kroah-Hartman

[ Upstream commit 3c79107631db1f7fd32cf3f7368e4672004a3010 ]

else, we leak the addresses to userspace via ctnetlink events and dumps.

Compute an ID on demand based on the immutable parts of nf_conn struct.

Another advantage compared to using an address is that there is no immediate re-use of the same ID in case the conntrack entry is freed and reallocated again immediately.

Fixes 3583240249ef ("[NETFILTER]: nf_conntrack_expect: kill unique ID")
Fixes 7f85f914721f ("[NETFILTER]: nf_conntrack: kill unique ID")
---
 include/net/netfilter/nf_conntrack.h |  2 ++
 net/netfilter/nf_conntrack_core.c    | 35 ++++++++++++++++++++++++++++
 net/netfilter/nf_conntrack_netlink.c | 34 +++++++++++++++++++++++----
 3 files changed, 66 insertions(+), 5 deletions(-)
include/net/netfilter/nf_conntrack.h +2 -0
index 249d0a5b12b82..63fd47e924b92 100644
@@ -318,6 +318,8 @@ struct nf_conn *nf_ct_tmpl_alloc(struct net *net,
 				 gfp_t flags);
 void nf_ct_tmpl_free(struct nf_conn *tmpl);
 
+u32 nf_ct_get_id(const struct nf_conn *ct);
+
 static inline void
 nf_ct_set(struct sk_buff *skb, struct nf_conn *ct, enum ip_conntrack_info info)
 {
net/netfilter/nf_conntrack_core.c +35 -0
index 9dd4c2048a2ba..1982faf21ebb5 100644
@@ -25,6 +25,7 @@
 #include <linux/slab.h>
 #include <linux/random.h>
 #include <linux/jhash.h>
+#include <linux/siphash.h>
 #include <linux/err.h>
 #include <linux/percpu.h>
 #include <linux/moduleparam.h>
@@ -424,6 +425,40 @@ nf_ct_invert_tuple(struct nf_conntrack_tuple *inverse,
 }
 EXPORT_SYMBOL_GPL(nf_ct_invert_tuple);
 
+/* Generate a almost-unique pseudo-id for a given conntrack.
+ *
+ * intentionally doesn't re-use any of the seeds used for hash
+ * table location, we assume id gets exposed to userspace.
+ *
+ * Following nf_conn items do not change throughout lifetime
+ * of the nf_conn after it has been committed to main hash table:
+ *
+ * 1. nf_conn address
+ * 2. nf_conn->ext address
+ * 3. nf_conn->master address (normally NULL)
+ * 4. tuple
+ * 5. the associated net namespace
+ */
+u32 nf_ct_get_id(const struct nf_conn *ct)
+{
+	static __read_mostly siphash_key_t ct_id_seed;
+	unsigned long a, b, c, d;
+
+	net_get_random_once(&ct_id_seed, sizeof(ct_id_seed));
+
+	a = (unsigned long)ct;
+	b = (unsigned long)ct->master ^ net_hash_mix(nf_ct_net(ct));
+	c = (unsigned long)ct->ext;
+	d = (unsigned long)siphash(&ct->tuplehash, sizeof(ct->tuplehash),
+				   &ct_id_seed);
+#ifdef CONFIG_64BIT
+	return siphash_4u64((u64)a, (u64)b, (u64)c, (u64)d, &ct_id_seed);
+#else
+	return siphash_4u32((u32)a, (u32)b, (u32)c, (u32)d, &ct_id_seed);
+#endif
+}
+EXPORT_SYMBOL_GPL(nf_ct_get_id);
+
 static void
 clean_from_lists(struct nf_conn *ct)
 {
net/netfilter/nf_conntrack_netlink.c +29 -5
index 1213beb5a7146..36619ad8ab8c2 100644
@@ -29,6 +29,7 @@
 #include <linux/spinlock.h>
 #include <linux/interrupt.h>
 #include <linux/slab.h>
+#include <linux/siphash.h>
 
 #include <linux/netfilter.h>
 #include <net/netlink.h>
@@ -485,7 +486,9 @@ static int ctnetlink_dump_ct_synproxy(struct sk_buff *skb, struct nf_conn *ct)
 
 static int ctnetlink_dump_id(struct sk_buff *skb, const struct nf_conn *ct)
 {
-	if (nla_put_be32(skb, CTA_ID, htonl((unsigned long)ct)))
+	__be32 id = (__force __be32)nf_ct_get_id(ct);
+
+	if (nla_put_be32(skb, CTA_ID, id))
 		goto nla_put_failure;
 	return 0;
 
@@ -1286,8 +1289,9 @@ static int ctnetlink_del_conntrack(struct net *net, struct sock *ctnl,
 	}
 
 	if (cda[CTA_ID]) {
-		u_int32_t id = ntohl(nla_get_be32(cda[CTA_ID]));
-		if (id != (u32)(unsigned long)ct) {
+		__be32 id = nla_get_be32(cda[CTA_ID]);
+
+		if (id != (__force __be32)nf_ct_get_id(ct)) {
 			nf_ct_put(ct);
 			return -ENOENT;
 		}
@@ -2694,6 +2698,25 @@ static int ctnetlink_exp_dump_mask(struct sk_buff *skb,
 
 static const union nf_inet_addr any_addr;
 
+static __be32 nf_expect_get_id(const struct nf_conntrack_expect *exp)
+{
+	static __read_mostly siphash_key_t exp_id_seed;
+	unsigned long a, b, c, d;
+
+	net_get_random_once(&exp_id_seed, sizeof(exp_id_seed));
+
+	a = (unsigned long)exp;
+	b = (unsigned long)exp->helper;
+	c = (unsigned long)exp->master;
+	d = (unsigned long)siphash(&exp->tuple, sizeof(exp->tuple), &exp_id_seed);
+
+#ifdef CONFIG_64BIT
+	return (__force __be32)siphash_4u64((u64)a, (u64)b, (u64)c, (u64)d, &exp_id_seed);
+#else
+	return (__force __be32)siphash_4u32((u32)a, (u32)b, (u32)c, (u32)d, &exp_id_seed);
+#endif
+}
+
 static int
 ctnetlink_exp_dump_expect(struct sk_buff *skb,
 			  const struct nf_conntrack_expect *exp)
@@ -2741,7 +2764,7 @@ ctnetlink_exp_dump_expect(struct sk_buff *skb,
 	}
 #endif
 	if (nla_put_be32(skb, CTA_EXPECT_TIMEOUT, htonl(timeout)) ||
-	    nla_put_be32(skb, CTA_EXPECT_ID, htonl((unsigned long)exp)) ||
+	    nla_put_be32(skb, CTA_EXPECT_ID, nf_expect_get_id(exp)) ||
 	    nla_put_be32(skb, CTA_EXPECT_FLAGS, htonl(exp->flags)) ||
 	    nla_put_be32(skb, CTA_EXPECT_CLASS, htonl(exp->class)))
 		goto nla_put_failure;
@@ -3046,7 +3069,8 @@ static int ctnetlink_get_expect(struct net *net, struct sock *ctnl,
 
 	if (cda[CTA_EXPECT_ID]) {
 		__be32 id = nla_get_be32(cda[CTA_EXPECT_ID]);
-		if (ntohl(id) != (u32)(unsigned long)exp) {
+
+		if (id != nf_expect_get_id(exp)) {
 			nf_ct_expect_put(exp);
 			return -ENOENT;
 		}
Signature · 1 line
2.20.1
author maintainer
#67 lore raw

PATCH 67/137 netfilter: nf_tables: prevent shift wrap in nft_chain_parse_hook()

In reply to Greg Kroah-Hartman

[ Upstream commit 33d1c018179d0a30c39cc5f1682b77867282694b ]

I believe that "hook->num" can be up to UINT_MAX. Shifting more than 31 bits would is undefined in C but in practice it would lead to shift wrapping. That would lead to an array overflow in nf_tables_addchain():

	ops->hook       = hook.type->hooks[ops->hooknum];
Fixes fe19c04ca137 ("netfilter: nf_tables: remove nhooks field from struct nft_af_info")
---
 net/netfilter/nf_tables_api.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/netfilter/nf_tables_api.c +1 -1
index e2aac80f9b7b1..25c2b98b9a960 100644
@@ -1502,7 +1502,7 @@ static int nft_chain_parse_hook(struct net *net,
 		if (IS_ERR(type))
 			return PTR_ERR(type);
 	}
-	if (!(type->hook_mask & (1 << hook->num)))
+	if (hook->num > NF_MAX_HOOKS || !(type->hook_mask & (1 << hook->num)))
 		return -EOPNOTSUPP;
 
 	if (type->type == NFT_CHAIN_T_NAT &&
Signature · 1 line
2.20.1
author maintainer
#68 lore raw

PATCH 68/137 netfilter: nat: fix icmp id randomization

In reply to Greg Kroah-Hartman

[ Upstream commit 5bdac418f33f60b07a34e01e722889140ee8fac9 ]

Sven Auhagen reported that a 2nd ping request will fail if 'fully-random' mode is used.

Reason is that if no proto information is given, min/max are both 0, so we set the icmp id to 0 instead of chosing a random value between 0 and 65535.

Update test case as well to catch this, without fix this yields:
[..]
ERROR: cannot ping ns1 from ns2 with ip masquerade fully-random (attempt 2)
ERROR: cannot ping ns1 from ns2 with ipv6 masquerade fully-random (attempt 2)

... becaus 2nd ping clashes with existing 'id 0' icmp conntrack and gets dropped.

Fixes 203f2e78200c27e ("netfilter: nat: remove l4proto->unique_tuple")
---
 net/netfilter/nf_nat_core.c                  | 11 ++++--
 tools/testing/selftests/netfilter/nft_nat.sh | 36 +++++++++++++++-----
 2 files changed, 35 insertions(+), 12 deletions(-)
net/netfilter/nf_nat_core.c +8 -3
index d159e9e7835b4..ade527565127b 100644
@@ -358,9 +358,14 @@ static void nf_nat_l4proto_unique_tuple(struct nf_conntrack_tuple *tuple,
 	case IPPROTO_ICMPV6:
 		/* id is same for either direction... */
 		keyptr = &tuple->src.u.icmp.id;
-		min = range->min_proto.icmp.id;
-		range_size = ntohs(range->max_proto.icmp.id) -
-			     ntohs(range->min_proto.icmp.id) + 1;
+		if (!(range->flags & NF_NAT_RANGE_PROTO_SPECIFIED)) {
+			min = 0;
+			range_size = 65536;
+		} else {
+			min = ntohs(range->min_proto.icmp.id);
+			range_size = ntohs(range->max_proto.icmp.id) -
+				     ntohs(range->min_proto.icmp.id) + 1;
+		}
 		goto find_free_id;
 #if IS_ENABLED(CONFIG_NF_CT_PROTO_GRE)
 	case IPPROTO_GRE:
tools/testing/selftests/netfilter/nft_nat.sh +27 -9
index 8ec76681605cc..3194007cf8d1b 100755
@@ -321,6 +321,7 @@ EOF
 
 test_masquerade6()
 {
+	local natflags=$1
 	local lret=0
 
 	ip netns exec ns0 sysctl net.ipv6.conf.all.forwarding=1 > /dev/null
@@ -354,13 +355,13 @@ ip netns exec ns0 nft -f - <<EOF
 table ip6 nat {
 	chain postrouting {
 		type nat hook postrouting priority 0; policy accept;
-		meta oif veth0 masquerade
+		meta oif veth0 masquerade $natflags
 	}
 }
 EOF
 	ip netns exec ns2 ping -q -c 1 dead:1::99 > /dev/null # ping ns2->ns1
 	if [ $? -ne 0 ] ; then
-		echo "ERROR: cannot ping ns1 from ns2 with active ipv6 masquerading"
+		echo "ERROR: cannot ping ns1 from ns2 with active ipv6 masquerade $natflags"
 		lret=1
 	fi
 
@@ -397,19 +398,26 @@ EOF
 		fi
 	done
 
+	ip netns exec ns2 ping -q -c 1 dead:1::99 > /dev/null # ping ns2->ns1
+	if [ $? -ne 0 ] ; then
+		echo "ERROR: cannot ping ns1 from ns2 with active ipv6 masquerade $natflags (attempt 2)"
+		lret=1
+	fi
+
 	ip netns exec ns0 nft flush chain ip6 nat postrouting
 	if [ $? -ne 0 ]; then
 		echo "ERROR: Could not flush ip6 nat postrouting" 1>&2
 		lret=1
 	fi
 
-	test $lret -eq 0 && echo "PASS: IPv6 masquerade for ns2"
+	test $lret -eq 0 && echo "PASS: IPv6 masquerade $natflags for ns2"
 
 	return $lret
 }
 
 test_masquerade()
 {
+	local natflags=$1
 	local lret=0
 
 	ip netns exec ns0 sysctl net.ipv4.conf.veth0.forwarding=1 > /dev/null
@@ -417,7 +425,7 @@ test_masquerade()
 
 	ip netns exec ns2 ping -q -c 1 10.0.1.99 > /dev/null # ping ns2->ns1
 	if [ $? -ne 0 ] ; then
-		echo "ERROR: canot ping ns1 from ns2"
+		echo "ERROR: cannot ping ns1 from ns2 $natflags"
 		lret=1
 	fi
 
@@ -443,13 +451,13 @@ ip netns exec ns0 nft -f - <<EOF
 table ip nat {
 	chain postrouting {
 		type nat hook postrouting priority 0; policy accept;
-		meta oif veth0 masquerade
+		meta oif veth0 masquerade $natflags
 	}
 }
 EOF
 	ip netns exec ns2 ping -q -c 1 10.0.1.99 > /dev/null # ping ns2->ns1
 	if [ $? -ne 0 ] ; then
-		echo "ERROR: cannot ping ns1 from ns2 with active ip masquerading"
+		echo "ERROR: cannot ping ns1 from ns2 with active ip masquere $natflags"
 		lret=1
 	fi
 
@@ -485,13 +493,19 @@ EOF
 		fi
 	done
 
+	ip netns exec ns2 ping -q -c 1 10.0.1.99 > /dev/null # ping ns2->ns1
+	if [ $? -ne 0 ] ; then
+		echo "ERROR: cannot ping ns1 from ns2 with active ip masquerade $natflags (attempt 2)"
+		lret=1
+	fi
+
 	ip netns exec ns0 nft flush chain ip nat postrouting
 	if [ $? -ne 0 ]; then
 		echo "ERROR: Could not flush nat postrouting" 1>&2
 		lret=1
 	fi
 
-	test $lret -eq 0 && echo "PASS: IP masquerade for ns2"
+	test $lret -eq 0 && echo "PASS: IP masquerade $natflags for ns2"
 
 	return $lret
 }
@@ -750,8 +764,12 @@ test_local_dnat
 test_local_dnat6
 
 reset_counters
-test_masquerade
-test_masquerade6
+test_masquerade ""
+test_masquerade6 ""
+
+reset_counters
+test_masquerade "fully-random"
+test_masquerade6 "fully-random"
 
 reset_counters
 test_redirect
Signature · 1 line
2.20.1
author maintainer
#69 lore raw

PATCH 69/137 MIPS: perf: ath79: Fix perfcount IRQ assignment

In reply to Greg Kroah-Hartman

[ Upstream commit a1e8783db8e0d58891681bc1e6d9ada66eae8e20 ]

Currently it's not possible to use perf on ath79 due to genirq flags mismatch happening on static virtual IRQ 13 which is used for performance counters hardware IRQ 5.

On TP-Link Archer C7v5:

CPU0

  2:          0      MIPS   2  ath9k
  4:        318      MIPS   4  19000000.eth
  7:      55034      MIPS   7  timer
  8:       1236      MISC   3  ttyS0
 12:          0      INTC   1  ehci_hcd:usb1
 13:          0  gpio-ath79   2  keys
 14:          0  gpio-ath79   5  keys
 15:         31  AR724X PCI    1  ath10k_pci
 $ perf top
 genirq: Flags mismatch irq 13. 00014c83 (mips_perf_pmu) vs. 00002003 (keys)

On TP-Link Archer C7v4:

CPU0

  4:          0      MIPS   4  19000000.eth
  5:       7135      MIPS   5  1a000000.eth
  7:      98379      MIPS   7  timer
  8:         30      MISC   3  ttyS0
 12:      90028      INTC   0  ath9k
 13:       5520      INTC   1  ehci_hcd:usb1
 14:       4623      INTC   2  ehci_hcd:usb2
 15:      32844  AR724X PCI    1  ath10k_pci
 16:          0  gpio-ath79  16  keys
 23:          0  gpio-ath79  23  keys
 $ perf top
 genirq: Flags mismatch irq 13. 00014c80 (mips_perf_pmu) vs. 00000080 (ehci_hcd:usb1)

This problem is happening, because currently statically assigned virtual IRQ 13 for performance counters is not claimed during the initialization of MIPS PMU during the bootup, so the IRQ subsystem doesn't know, that this interrupt isn't available for further use.

So this patch fixes the issue by simply booking hardware IRQ 5 for MIPS PMU.

---
 arch/mips/ath79/setup.c          |  6 ------
 drivers/irqchip/irq-ath79-misc.c | 11 +++++++++++
 2 files changed, 11 insertions(+), 6 deletions(-)
arch/mips/ath79/setup.c +0 -6
index 9728abcb18fac..c04ae685003f7 100644
@@ -211,12 +211,6 @@ const char *get_system_type(void)
 	return ath79_sys_type;
 }
 
-int get_c0_perfcount_int(void)
-{
-	return ATH79_MISC_IRQ(5);
-}
-EXPORT_SYMBOL_GPL(get_c0_perfcount_int);
-
 unsigned int get_c0_compare_int(void)
 {
 	return CP0_LEGACY_COMPARE_IRQ;
drivers/irqchip/irq-ath79-misc.c +11 -0
index aa72907846360..0390603170b40 100644
@@ -22,6 +22,15 @@
 #define AR71XX_RESET_REG_MISC_INT_ENABLE	4
 
 #define ATH79_MISC_IRQ_COUNT			32
+#define ATH79_MISC_PERF_IRQ			5
+
+static int ath79_perfcount_irq;
+
+int get_c0_perfcount_int(void)
+{
+	return ath79_perfcount_irq;
+}
+EXPORT_SYMBOL_GPL(get_c0_perfcount_int);
 
 static void ath79_misc_irq_handler(struct irq_desc *desc)
 {
@@ -113,6 +122,8 @@ static void __init ath79_misc_intc_domain_init(
 {
 	void __iomem *base = domain->host_data;
 
+	ath79_perfcount_irq = irq_create_mapping(domain, ATH79_MISC_PERF_IRQ);
+
 	/* Disable and clear all interrupts */
 	__raw_writel(0, base + AR71XX_RESET_REG_MISC_INT_ENABLE);
 	__raw_writel(0, base + AR71XX_RESET_REG_MISC_INT_STATUS);
Signature · 1 line
2.20.1
author maintainer
#70 lore raw

PATCH 70/137 IB/mlx5: Fix scatter to CQE in DCT QP creation

In reply to Greg Kroah-Hartman

[ Upstream commit 7249c8ea227a582c14f63e9e8853eb7369122f10 ]

When scatter to CQE is enabled on a DCT QP it corrupts the mailbox command since it tried to treat it as as QP create mailbox command instead of a DCT create command.

The corrupted mailbox command causes userspace to malfunction as the device doesn't create the QP as expected.

A new mlx5 capability is exposed to user-space which ensures that it will not enable the feature on DCT without this fix in the kernel.

Fixes 5d6ff1babe78 ("IB/mlx5: Support scatter to CQE for DC transport type")
---
 drivers/infiniband/hw/mlx5/main.c |  2 ++
 drivers/infiniband/hw/mlx5/qp.c   | 11 +++++++----
 include/uapi/rdma/mlx5-abi.h      |  1 +
 3 files changed, 10 insertions(+), 4 deletions(-)
drivers/infiniband/hw/mlx5/main.c +2 -0
index 497181f5ba091..c6bdd0d16c4b6 100644
@@ -1025,6 +1025,8 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
 		if (MLX5_CAP_GEN(mdev, qp_packet_based))
 			resp.flags |=
 				MLX5_IB_QUERY_DEV_RESP_PACKET_BASED_CREDIT_MODE;
+
+		resp.flags |= MLX5_IB_QUERY_DEV_RESP_FLAGS_SCAT2CQE_DCT;
 	}
 
 	if (field_avail(typeof(resp), sw_parsing_caps,
drivers/infiniband/hw/mlx5/qp.c +7 -4
index 7db778d96ef5c..afc88e6e172e7 100644
@@ -1724,13 +1724,16 @@ static void configure_responder_scat_cqe(struct ib_qp_init_attr *init_attr,
 
 	rcqe_sz = mlx5_ib_get_cqe_size(init_attr->recv_cq);
 
-	if (rcqe_sz == 128) {
-		MLX5_SET(qpc, qpc, cs_res, MLX5_RES_SCAT_DATA64_CQE);
+	if (init_attr->qp_type == MLX5_IB_QPT_DCT) {
+		if (rcqe_sz == 128)
+			MLX5_SET(dctc, qpc, cs_res, MLX5_RES_SCAT_DATA64_CQE);
+
 		return;
 	}
 
-	if (init_attr->qp_type != MLX5_IB_QPT_DCT)
-		MLX5_SET(qpc, qpc, cs_res, MLX5_RES_SCAT_DATA32_CQE);
+	MLX5_SET(qpc, qpc, cs_res,
+		 rcqe_sz == 128 ? MLX5_RES_SCAT_DATA64_CQE :
+				  MLX5_RES_SCAT_DATA32_CQE);
 }
 
 static void configure_requester_scat_cqe(struct mlx5_ib_dev *dev,
include/uapi/rdma/mlx5-abi.h +1 -0
index 87b3198f4b5d7..f4d4010b7e3e5 100644
@@ -238,6 +238,7 @@ enum mlx5_ib_query_dev_resp_flags {
 	MLX5_IB_QUERY_DEV_RESP_FLAGS_CQE_128B_COMP = 1 << 0,
 	MLX5_IB_QUERY_DEV_RESP_FLAGS_CQE_128B_PAD  = 1 << 1,
 	MLX5_IB_QUERY_DEV_RESP_PACKET_BASED_CREDIT_MODE = 1 << 2,
+	MLX5_IB_QUERY_DEV_RESP_FLAGS_SCAT2CQE_DCT = 1 << 3,
 };
 
 enum mlx5_ib_tunnel_offloads {
Signature · 1 line
2.20.1
author maintainer
#71 lore raw

PATCH 71/137 s390: ctcm: fix ctcm_new_device error return code

In reply to Greg Kroah-Hartman

[ Upstream commit 27b141fc234a3670d21bd742c35d7205d03cbb3a ]

clang points out that the return code from this function is undefined for one of the error paths:

../drivers/s390/net/ctcm_main.c:1595:7: warning: variable 'result' is used uninitialized whenever 'if' condition is true

      [-Wsometimes-uninitialized]
                if (priv->channel[direction] == NULL) {
                    ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../drivers/s390/net/ctcm_main.c:1638:9: note: uninitialized use occurs here
        return result;
               ^~~~~~
../drivers/s390/net/ctcm_main.c:1595:3: note: remove the 'if' if its condition is always false
                if (priv->channel[direction] == NULL) {
                ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../drivers/s390/net/ctcm_main.c:1539:12: note: initialize the variable 'result' to silence this warning
        int result;
                  ^

Make it return -ENODEV here, as in the related failure cases. gcc has a known bug in underreporting some of these warnings when it has already eliminated the assignment of the return code based on some earlier optimization step.

---
 drivers/s390/net/ctcm_main.c | 1 +
 1 file changed, 1 insertion(+)
drivers/s390/net/ctcm_main.c +1 -0
index 7617d21cb2960..f63c5c871d3dd 100644
@@ -1595,6 +1595,7 @@ static int ctcm_new_device(struct ccwgroup_device *cgdev)
 		if (priv->channel[direction] == NULL) {
 			if (direction == CTCM_WRITE)
 				channel_free(priv->channel[CTCM_READ]);
+			result = -ENODEV;
 			goto out_dev;
 		}
 		priv->channel[direction]->netdev = dev;
Signature · 1 line
2.20.1
author maintainer
#72 lore raw

PATCH 72/137 drm/sun4i: Set device driver data at bind time for use in unbind

In reply to Greg Kroah-Hartman

[ Upstream commit 02b92adbe33e6dbd15dc6e32540b22f47c4ff0a2 ]

Our sun4i_drv_unbind gets the drm device using dev_get_drvdata. However, that driver data is never set in sun4i_drv_bind.

Set it there to avoid getting a NULL pointer at unbind time.

Fixes 9026e0d122ac ("drm: Add Allwinner A10 Display Engine support")
---
 drivers/gpu/drm/sun4i/sun4i_drv.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/gpu/drm/sun4i/sun4i_drv.c +2 -0
index 9e4c375ccc96f..c6b65a9699794 100644
@@ -85,6 +85,8 @@ static int sun4i_drv_bind(struct device *dev)
 		ret = -ENOMEM;
 		goto free_drm;
 	}
+
+	dev_set_drvdata(dev, drm);
 	drm->dev_private = drv;
 	INIT_LIST_HEAD(&drv->frontend_list);
 	INIT_LIST_HEAD(&drv->engine_list);
Signature · 1 line
2.20.1
author maintainer
#73 lore raw

PATCH 73/137 drm/sun4i: Fix component unbinding and component master deletion

In reply to Greg Kroah-Hartman

[ Upstream commit f5a9ed867c83875546c9aadd4ed8e785e9adcc3c ]

For our component-backed driver to be properly removed, we need to delete the component master in sun4i_drv_remove and make sure to call component_unbind_all in the master's unbind so that all components are unbound when the master is.

Fixes 9026e0d122ac ("drm: Add Allwinner A10 Display Engine support")
---
 drivers/gpu/drm/sun4i/sun4i_drv.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/gpu/drm/sun4i/sun4i_drv.c +4 -0
index c6b65a9699794..9a5713fa03b25 100644
@@ -148,6 +148,8 @@ static void sun4i_drv_unbind(struct device *dev)
 	drm_mode_config_cleanup(drm);
 	of_reserved_mem_device_release(dev);
 	drm_dev_put(drm);
+
+	component_unbind_all(dev, NULL);
 }
 
 static const struct component_master_ops sun4i_drv_master_ops = {
@@ -395,6 +397,8 @@ static int sun4i_drv_probe(struct platform_device *pdev)
 
 static int sun4i_drv_remove(struct platform_device *pdev)
 {
+	component_master_del(&pdev->dev, &sun4i_drv_master_ops);
+
 	return 0;
 }
 
Signature · 1 line
2.20.1
author maintainer
#74 lore raw

PATCH 74/137 of_net: Fix residues after of_get_nvmem_mac_address removal

In reply to Greg Kroah-Hartman

[ Upstream commit 36ad7022536e0c65f8baeeaa5efde11dec44808a ]

I've discovered following discrepancy in the bindings/net/ethernet.txt documentation, where it states following:

- nvmem-cells: phandle, reference to an nvmem node for the MAC address;
- nvmem-cell-names: string, should be "mac-address" if nvmem is to be..

which is actually misleading and confusing. There are only two ethernet drivers in the tree, cadence/macb and davinci which supports this properties.

This nvmem-cell* properties were introduced in commit 9217e566bdee ("of_net: Implement of_get_nvmem_mac_address helper"), but commit afa64a72b862 ("of: net: kill of_get_nvmem_mac_address()") forget to properly clean up this parts.

So this patch fixes the documentation by moving the nvmem-cell* properties at the appropriate places. While at it, I've removed unused include as well.

Cc Bartosz Golaszewski <[email protected]>
Fixes afa64a72b862 ("of: net: kill of_get_nvmem_mac_address()")
---
 Documentation/devicetree/bindings/net/davinci_emac.txt | 2 ++
 Documentation/devicetree/bindings/net/ethernet.txt     | 2 --
 Documentation/devicetree/bindings/net/macb.txt         | 4 ++++
 drivers/of/of_net.c                                    | 1 -
 4 files changed, 6 insertions(+), 3 deletions(-)
Documentation/devicetree/bindings/net/davinci_emac.txt +2 -0
index 24c5cdaba8d27..ca83dcc84fb8e 100644
@@ -20,6 +20,8 @@ Required properties:
 Optional properties:
 - phy-handle: See ethernet.txt file in the same directory.
               If absent, davinci_emac driver defaults to 100/FULL.
+- nvmem-cells: phandle, reference to an nvmem node for the MAC address
+- nvmem-cell-names: string, should be "mac-address" if nvmem is to be used
 - ti,davinci-rmii-en: 1 byte, 1 means use RMII
 - ti,davinci-no-bd-ram: boolean, does EMAC have BD RAM?
 
Documentation/devicetree/bindings/net/ethernet.txt +0 -2
index cfc376bc977aa..2974e63ba311a 100644
@@ -10,8 +10,6 @@ Documentation/devicetree/bindings/phy/phy-bindings.txt.
   the boot program; should be used in cases where the MAC address assigned to
   the device by the boot program is different from the "local-mac-address"
   property;
-- nvmem-cells: phandle, reference to an nvmem node for the MAC address;
-- nvmem-cell-names: string, should be "mac-address" if nvmem is to be used;
 - max-speed: number, specifies maximum speed in Mbit/s supported by the device;
 - max-frame-size: number, maximum transfer unit (IEEE defined MTU), rather than
   the maximum frame size (there's contradiction in the Devicetree
Documentation/devicetree/bindings/net/macb.txt +4 -0
index 3e17ac1d5d58c..1a914116f4c2c 100644
@@ -26,6 +26,10 @@ Required properties:
 	Optional elements: 'tsu_clk'
 - clocks: Phandles to input clocks.
 
+Optional properties:
+- nvmem-cells: phandle, reference to an nvmem node for the MAC address
+- nvmem-cell-names: string, should be "mac-address" if nvmem is to be used
+
 Optional properties for PHY child node:
 - reset-gpios : Should specify the gpio for phy reset
 - magic-packet : If present, indicates that the hardware supports waking
drivers/of/of_net.c +0 -1
index 810ab0fbcccbf..d820f3edd4311 100644
@@ -7,7 +7,6 @@
  */
 #include <linux/etherdevice.h>
 #include <linux/kernel.h>
-#include <linux/nvmem-consumer.h>
 #include <linux/of_net.h>
 #include <linux/phy.h>
 #include <linux/export.h>
Signature · 1 line
2.20.1
author maintainer
#75 lore raw

PATCH 75/137 selftests/net: correct the return value for run_netsocktests

In reply to Greg Kroah-Hartman

[ Upstream commit 30c04d796b693e22405c38e9b78e9a364e4c77e6 ]

The run_netsocktests will be marked as passed regardless the actual test result from the ./socket:

    selftests: net: run_netsocktests
    ========================================
    --------------------
    running socket test
    --------------------
    [FAIL]
    ok 1..6 selftests: net: run_netsocktests [PASS]

This is because the test script itself has been successfully executed. Fix this by exit 1 when the test failed.

---
 tools/testing/selftests/net/run_netsocktests | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/testing/selftests/net/run_netsocktests +1 -1
index b093f39c298c3..14e41faf2c574 100755
@@ -7,7 +7,7 @@ echo "--------------------"
 ./socket
 if [ $? -ne 0 ]; then
 	echo "[FAIL]"
+	exit 1
 else
 	echo "[PASS]"
 fi
-
Signature · 1 line
2.20.1
author maintainer
#76 lore raw

PATCH 76/137 selftests/net: correct the return value for run_afpackettests

In reply to Greg Kroah-Hartman

[ Upstream commit 8c03557c3f25271e62e39154af66ebdd1b59c9ca ]

The run_afpackettests will be marked as passed regardless the return value of those sub-tests in the script:

    --------------------
    running psock_tpacket test
    --------------------
    [FAIL]
    selftests: run_afpackettests [PASS]

Fix this by changing the return value for each tests.

---
 tools/testing/selftests/net/run_afpackettests | 5 +++++
 1 file changed, 5 insertions(+)
tools/testing/selftests/net/run_afpackettests +5 -0
index 2dc95fda7ef76..ea5938ec009a5 100755
@@ -6,12 +6,14 @@ if [ $(id -u) != 0 ]; then
 	exit 0
 fi
 
+ret=0
 echo "--------------------"
 echo "running psock_fanout test"
 echo "--------------------"
 ./in_netns.sh ./psock_fanout
 if [ $? -ne 0 ]; then
 	echo "[FAIL]"
+	ret=1
 else
 	echo "[PASS]"
 fi
@@ -22,6 +24,7 @@ echo "--------------------"
 ./in_netns.sh ./psock_tpacket
 if [ $? -ne 0 ]; then
 	echo "[FAIL]"
+	ret=1
 else
 	echo "[PASS]"
 fi
@@ -32,6 +35,8 @@ echo "--------------------"
 ./in_netns.sh ./txring_overwrite
 if [ $? -ne 0 ]; then
 	echo "[FAIL]"
+	ret=1
 else
 	echo "[PASS]"
 fi
+exit $ret
Signature · 1 line
2.20.1
author maintainer
#77 lore raw

PATCH 77/137 netfilter: never get/set skb->tstamp

In reply to Greg Kroah-Hartman

[ Upstream commit 916f6efae62305796e012e7c3a7884a267cbacbf ]

setting net.netfilter.nf_conntrack_timestamp=1 breaks xmit with fq scheduler. skb->tstamp might be "refreshed" using ktime_get_real(), but fq expects CLOCK_MONOTONIC.

This patch removes all places in netfilter that check/set skb->tstamp:

1. To fix the bogus "start" time seen with conntrack timestamping for outgoing packets, never use skb->tstamp and always use current time.
2. In nfqueue and nflog, only use skb->tstamp for incoming packets, as determined by current hook (prerouting, input, forward).
3. xt_time has to use system clock as well rather than skb->tstamp. We could still use skb->tstamp for prerouting/input/foward, but I see no advantage to make this conditional.

Fixes fb420d5d91c1 ("tcp/fq: move back to CLOCK_MONOTONIC")
Cc Eric Dumazet <[email protected]>
---
 net/netfilter/nf_conntrack_core.c |  7 ++-----
 net/netfilter/nfnetlink_log.c     |  2 +-
 net/netfilter/nfnetlink_queue.c   |  2 +-
 net/netfilter/xt_time.c           | 23 ++++++++++++++---------
 4 files changed, 18 insertions(+), 16 deletions(-)
net/netfilter/nf_conntrack_core.c +2 -5
index 1982faf21ebb5..d7ac2f82bb6d8 100644
@@ -983,12 +983,9 @@ __nf_conntrack_confirm(struct sk_buff *skb)
 
 	/* set conntrack timestamp, if enabled. */
 	tstamp = nf_conn_tstamp_find(ct);
-	if (tstamp) {
-		if (skb->tstamp == 0)
-			__net_timestamp(skb);
+	if (tstamp)
+		tstamp->start = ktime_get_real_ns();
 
-		tstamp->start = ktime_to_ns(skb->tstamp);
-	}
 	/* Since the lookup is lockless, hash insertion must be done after
 	 * starting the timer and setting the CONFIRMED bit. The RCU barriers
 	 * guarantee that no other CPU can find the conntrack before the above
net/netfilter/nfnetlink_log.c +1 -1
index b1f9c5303f026..0b3347570265c 100644
@@ -540,7 +540,7 @@ __build_packet_message(struct nfnl_log_net *log,
 			goto nla_put_failure;
 	}
 
-	if (skb->tstamp) {
+	if (hooknum <= NF_INET_FORWARD && skb->tstamp) {
 		struct nfulnl_msg_packet_timestamp ts;
 		struct timespec64 kts = ktime_to_timespec64(skb->tstamp);
 		ts.sec = cpu_to_be64(kts.tv_sec);
net/netfilter/nfnetlink_queue.c +1 -1
index 0dcc3592d053f..e057b2961d313 100644
@@ -582,7 +582,7 @@ nfqnl_build_packet_message(struct net *net, struct nfqnl_instance *queue,
 	if (nfqnl_put_bridge(entry, skb) < 0)
 		goto nla_put_failure;
 
-	if (entskb->tstamp) {
+	if (entry->state.hook <= NF_INET_FORWARD && entskb->tstamp) {
 		struct nfqnl_msg_packet_timestamp ts;
 		struct timespec64 kts = ktime_to_timespec64(entskb->tstamp);
 
net/netfilter/xt_time.c +14 -9
index c13bcd0ab4913..8dbb4d48f2ed5 100644
@@ -163,19 +163,24 @@ time_mt(const struct sk_buff *skb, struct xt_action_param *par)
 	s64 stamp;
 
 	/*
-	 * We cannot use get_seconds() instead of __net_timestamp() here.
+	 * We need real time here, but we can neither use skb->tstamp
+	 * nor __net_timestamp().
+	 *
+	 * skb->tstamp and skb->skb_mstamp_ns overlap, however, they
+	 * use different clock types (real vs monotonic).
+	 *
 	 * Suppose you have two rules:
-	 * 	1. match before 13:00
-	 * 	2. match after 13:00
+	 *	1. match before 13:00
+	 *	2. match after 13:00
+	 *
 	 * If you match against processing time (get_seconds) it
 	 * may happen that the same packet matches both rules if
-	 * it arrived at the right moment before 13:00.
+	 * it arrived at the right moment before 13:00, so it would be
+	 * better to check skb->tstamp and set it via __net_timestamp()
+	 * if needed.  This however breaks outgoing packets tx timestamp,
+	 * and causes them to get delayed forever by fq packet scheduler.
 	 */
-	if (skb->tstamp == 0)
-		__net_timestamp((struct sk_buff *)skb);
-
-	stamp = ktime_to_ns(skb->tstamp);
-	stamp = div_s64(stamp, NSEC_PER_SEC);
+	stamp = get_seconds();
 
 	if (info->flags & XT_TIME_LOCAL_TZ)
 		/* Adjust for local timezone */
Signature · 1 line
2.20.1
author maintainer
#78 lore raw

PATCH 78/137 netfilter: fix nf_l4proto_log_invalid to log invalid packets

In reply to Greg Kroah-Hartman

[ Upstream commit d48668052b2603b6262459625c86108c493588dd ]

It doesn't log a packet if sysctl_log_invalid isn't equal to protonum OR sysctl_log_invalid isn't equal to IPPROTO_RAW. This sentence is always true. I believe we need to replace OR to AND.

Cc Florian Westphal <[email protected]>
Fixes c4f3db1595827 ("netfilter: conntrack: add and use nf_l4proto_log_invalid")
---
 net/netfilter/nf_conntrack_proto.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/netfilter/nf_conntrack_proto.c +1 -1
index 859f5d07a9159..78361e462e802 100644
@@ -86,7 +86,7 @@ void nf_l4proto_log_invalid(const struct sk_buff *skb,
 	struct va_format vaf;
 	va_list args;
 
-	if (net->ct.sysctl_log_invalid != protonum ||
+	if (net->ct.sysctl_log_invalid != protonum &&
 	    net->ct.sysctl_log_invalid != IPPROTO_RAW)
 		return;
 
Signature · 1 line
2.20.1
author maintainer
#79 lore raw

PATCH 79/137 dmaengine: bcm2835: Avoid GFP_KERNEL in device_prep_slave_sg

In reply to Greg Kroah-Hartman

[ Upstream commit f147384774a7b24dda4783a3dcd61af272757ea8 ]

The commit af19b7ce76ba ("mmc: bcm2835: Avoid possible races on data requests") introduces a possible circular locking dependency, which is triggered by swapping to the sdhost interface.

So instead of reintroduce the race condition again, we could also avoid this situation by using GFP_NOWAIT for the allocation of the DMA buffer descriptors.

---
 drivers/dma/bcm2835-dma.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/dma/bcm2835-dma.c +1 -1
index ae10f5614f953..bf51192036378 100644
@@ -674,7 +674,7 @@ static struct dma_async_tx_descriptor *bcm2835_dma_prep_slave_sg(
 	d = bcm2835_dma_create_cb_chain(chan, direction, false,
 					info, extra,
 					frames, src, dst, 0, 0,
-					GFP_KERNEL);
+					GFP_NOWAIT);
 	if (!d)
 		return NULL;
 
Signature · 1 line
2.20.1
author maintainer
#80 lore raw

PATCH 80/137 arm64/module: ftrace: deal with place relative nature of PLTs

In reply to Greg Kroah-Hartman

[ Upstream commit 4e69ecf4da1ee0b2ac735e1f1bb13935acd5a38d ]

Another bodge for the ftrace PLT code: plt_entries_equal() now takes the place relative nature of the ADRP/ADD based PLT entries into account, which means that a struct trampoline instance on the stack is no longer equal to the same set of opcodes in the module struct, given that they don't point to the same place in memory anymore.

Work around this by using memcmp() in the ftrace PLT handling code.

---
 arch/arm64/kernel/ftrace.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)
arch/arm64/kernel/ftrace.c +7 -2
index 07b2981201820..65a51331088eb 100644
@@ -103,10 +103,15 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 		 * to be revisited if support for multiple ftrace entry points
 		 * is added in the future, but for now, the pr_err() below
 		 * deals with a theoretical issue only.
+		 *
+		 * Note that PLTs are place relative, and plt_entries_equal()
+		 * checks whether they point to the same target. Here, we need
+		 * to check if the actual opcodes are in fact identical,
+		 * regardless of the offset in memory so use memcmp() instead.
 		 */
 		trampoline = get_plt_entry(addr, mod->arch.ftrace_trampoline);
-		if (!plt_entries_equal(mod->arch.ftrace_trampoline,
-				       &trampoline)) {
+		if (memcmp(mod->arch.ftrace_trampoline, &trampoline,
+			   sizeof(trampoline))) {
 			if (plt_entry_is_initialized(mod->arch.ftrace_trampoline)) {
 				pr_err("ftrace: far branches to multiple entry points unsupported inside a single module\n");
 				return -EINVAL;
Signature · 1 line
2.20.1
author maintainer
#81 lore raw

PATCH 81/137 gpu: ipu-v3: dp: fix CSC handling

In reply to Greg Kroah-Hartman

[ Upstream commit d4fad0a426c6e26f48c9a7cdd21a7fe9c198d645 ]

Initialize the flow input colorspaces to unknown and reset to that value when the channel gets disabled. This avoids the state getting mixed up with a previous mode.

Also keep the CSC settings for the background flow intact when disabling the foreground flow.

---
 drivers/gpu/ipu-v3/ipu-dp.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)
drivers/gpu/ipu-v3/ipu-dp.c +9 -3
index 9b2b3fa479c46..5e44ff1f20851 100644
@@ -195,7 +195,8 @@ int ipu_dp_setup_channel(struct ipu_dp *dp,
 		ipu_dp_csc_init(flow, flow->foreground.in_cs, flow->out_cs,
 				DP_COM_CONF_CSC_DEF_BOTH);
 	} else {
-		if (flow->foreground.in_cs == flow->out_cs)
+		if (flow->foreground.in_cs == IPUV3_COLORSPACE_UNKNOWN ||
+		    flow->foreground.in_cs == flow->out_cs)
 			/*
 			 * foreground identical to output, apply color
 			 * conversion on background
@@ -261,6 +262,8 @@ void ipu_dp_disable_channel(struct ipu_dp *dp, bool sync)
 	struct ipu_dp_priv *priv = flow->priv;
 	u32 reg, csc;
 
+	dp->in_cs = IPUV3_COLORSPACE_UNKNOWN;
+
 	if (!dp->foreground)
 		return;
 
@@ -268,8 +271,9 @@ void ipu_dp_disable_channel(struct ipu_dp *dp, bool sync)
 
 	reg = readl(flow->base + DP_COM_CONF);
 	csc = reg & DP_COM_CONF_CSC_DEF_MASK;
-	if (csc == DP_COM_CONF_CSC_DEF_FG)
-		reg &= ~DP_COM_CONF_CSC_DEF_MASK;
+	reg &= ~DP_COM_CONF_CSC_DEF_MASK;
+	if (csc == DP_COM_CONF_CSC_DEF_BOTH || csc == DP_COM_CONF_CSC_DEF_BG)
+		reg |= DP_COM_CONF_CSC_DEF_BG;
 
 	reg &= ~DP_COM_CONF_FG_EN;
 	writel(reg, flow->base + DP_COM_CONF);
@@ -347,6 +351,8 @@ int ipu_dp_init(struct ipu_soc *ipu, struct device *dev, unsigned long base)
 	mutex_init(&priv->mutex);
 
 	for (i = 0; i < IPUV3_NUM_FLOWS; i++) {
+		priv->flow[i].background.in_cs = IPUV3_COLORSPACE_UNKNOWN;
+		priv->flow[i].foreground.in_cs = IPUV3_COLORSPACE_UNKNOWN;
 		priv->flow[i].foreground.foreground = true;
 		priv->flow[i].base = priv->base + ipu_dp_flow_base[i];
 		priv->flow[i].priv = priv;
Signature · 1 line
2.20.1
author maintainer
#82 lore raw

PATCH 82/137 drm/imx: dont skip DP channel disable for background plane

In reply to Greg Kroah-Hartman

[ Upstream commit 7bcde275eb1d0ac8793c77c7e666a886eb16633d ]

In order to make sure that the plane color space gets reset correctly.

---
 drivers/gpu/drm/imx/ipuv3-crtc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/imx/ipuv3-crtc.c +1 -1
index 058b53c0aa7ec..1bb3e598cb843 100644
@@ -70,7 +70,7 @@ static void ipu_crtc_disable_planes(struct ipu_crtc *ipu_crtc,
 	if (disable_partial)
 		ipu_plane_disable(ipu_crtc->plane[1], true);
 	if (disable_full)
-		ipu_plane_disable(ipu_crtc->plane[0], false);
+		ipu_plane_disable(ipu_crtc->plane[0], true);
 }
 
 static void ipu_crtc_atomic_disable(struct drm_crtc *crtc,
Signature · 1 line
2.20.1
author maintainer
#83 lore raw

PATCH 83/137 ARM: fix function graph tracer and unwinder dependencies

In reply to Greg Kroah-Hartman

[ Upstream commit 503621628b32782a07b2318e4112bd4372aa3401 ]

Naresh Kamboju recently reported that the function-graph tracer crashes on ARM. The function-graph tracer assumes that the kernel is built with frame pointers.

We explicitly disabled the function-graph tracer when building Thumb2, since the Thumb2 ABI doesn't have frame pointers.

We recently changed the way the unwinder method was selected, which seems to have made it more likely that we can end up with the function- graph tracer enabled but without the kernel built with frame pointers.

Fix up the function graph tracer dependencies so the option is not available when we have no possibility of having frame pointers, and adjust the dependencies on the unwinder option to hide the non-frame pointer unwinder options if the function-graph tracer is enabled.

---
 arch/arm/Kconfig       | 2 +-
 arch/arm/Kconfig.debug | 6 +++---
 2 files changed, 4 insertions(+), 4 deletions(-)
arch/arm/Kconfig +1 -1
index e5d56d9b712c2..3b353af9c48dc 100644
@@ -69,7 +69,7 @@ config ARM
 	select HAVE_EFFICIENT_UNALIGNED_ACCESS if (CPU_V6 || CPU_V6K || CPU_V7) && MMU
 	select HAVE_EXIT_THREAD
 	select HAVE_FTRACE_MCOUNT_RECORD if !XIP_KERNEL
-	select HAVE_FUNCTION_GRAPH_TRACER if !THUMB2_KERNEL
+	select HAVE_FUNCTION_GRAPH_TRACER if !THUMB2_KERNEL && !CC_IS_CLANG
 	select HAVE_FUNCTION_TRACER if !XIP_KERNEL
 	select HAVE_GCC_PLUGINS
 	select HAVE_GENERIC_DMA_COHERENT
arch/arm/Kconfig.debug +3 -3
index 6d6e0330930b5..e388af4594a6e 100644
@@ -47,8 +47,8 @@ config DEBUG_WX
 
 choice
 	prompt "Choose kernel unwinder"
-	default UNWINDER_ARM if AEABI && !FUNCTION_GRAPH_TRACER
-	default UNWINDER_FRAME_POINTER if !AEABI || FUNCTION_GRAPH_TRACER
+	default UNWINDER_ARM if AEABI
+	default UNWINDER_FRAME_POINTER if !AEABI
 	help
 	  This determines which method will be used for unwinding kernel stack
 	  traces for panics, oopses, bugs, warnings, perf, /proc/<pid>/stack,
@@ -65,7 +65,7 @@ config UNWINDER_FRAME_POINTER
 
 config UNWINDER_ARM
 	bool "ARM EABI stack unwinder"
-	depends on AEABI
+	depends on AEABI && !FUNCTION_GRAPH_TRACER
 	select ARM_UNWIND
 	help
 	  This option enables stack unwinding support in the kernel
Signature · 1 line
2.20.1
author maintainer
#84 lore raw

PATCH 84/137 ARM: 8856/1: NOMMU: Fix CCR register faulty initialization when MPU is disabled

In reply to Greg Kroah-Hartman

[ Upstream commit c3143967807adb1357c36b68a7563fc0c4e1f615 ]

When CONFIG_ARM_MPU is not defined, the base address of v7M SCB register is not initialized with correct value. This prevents enabling I/D caches when the L1 cache poilcy is applied in kernel.

Fixes 3c24121039c9da14692eb48f6e39565b28c0f3cf ("ARM: 8756/1: NOMMU: Postpone MPU activation till __after_proc_init")
---
 arch/arm/kernel/head-nommu.S | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/kernel/head-nommu.S +1 -1
index ec29de2500764..cab89479d15ef 100644
@@ -133,9 +133,9 @@ __secondary_data:
  */
 	.text
 __after_proc_init:
-#ifdef CONFIG_ARM_MPU
 M_CLASS(movw	r12, #:lower16:BASEADDR_V7M_SCB)
 M_CLASS(movt	r12, #:upper16:BASEADDR_V7M_SCB)
+#ifdef CONFIG_ARM_MPU
 M_CLASS(ldr	r3, [r12, 0x50])
 AR_CLASS(mrc	p15, 0, r3, c0, c1, 4)          @ Read ID_MMFR0
 	and	r3, r3, #(MMFR0_PMSA)           @ PMSA field
Signature · 1 line
2.20.1
author maintainer
#85 lore raw

PATCH 85/137 spi: Micrel eth switch: declare missing of table

In reply to Greg Kroah-Hartman

[ Upstream commit 2f23a2a768bee7ad2ff1e9527c3f7e279e794a46 ]

Add missing <of_device_id> table for SPI driver relying on SPI device match since compatible is in a DT binding or in a DTS.

Before this patch:
modinfo drivers/net/phy/spi_ks8995.ko | grep alias

alias:          spi:ksz8795
alias:          spi:ksz8864
alias:          spi:ks8995

After this patch:
modinfo drivers/net/phy/spi_ks8995.ko | grep alias

alias:          spi:ksz8795
alias:          spi:ksz8864
alias:          spi:ks8995
alias:          of:N*T*Cmicrel,ksz8795C*
alias:          of:N*T*Cmicrel,ksz8795
alias:          of:N*T*Cmicrel,ksz8864C*
alias:          of:N*T*Cmicrel,ksz8864
alias:          of:N*T*Cmicrel,ks8995C*
alias:          of:N*T*Cmicrel,ks8995
---
 drivers/net/phy/spi_ks8995.c | 9 +++++++++
 1 file changed, 9 insertions(+)
drivers/net/phy/spi_ks8995.c +9 -0
index f17b3441779bf..d8ea4147dfe78 100644
@@ -162,6 +162,14 @@ static const struct spi_device_id ks8995_id[] = {
 };
 MODULE_DEVICE_TABLE(spi, ks8995_id);
 
+static const struct of_device_id ks8895_spi_of_match[] = {
+        { .compatible = "micrel,ks8995" },
+        { .compatible = "micrel,ksz8864" },
+        { .compatible = "micrel,ksz8795" },
+        { },
+ };
+MODULE_DEVICE_TABLE(of, ks8895_spi_of_match);
+
 static inline u8 get_chip_id(u8 val)
 {
 	return (val >> ID1_CHIPID_S) & ID1_CHIPID_M;
@@ -529,6 +537,7 @@ static int ks8995_remove(struct spi_device *spi)
 static struct spi_driver ks8995_driver = {
 	.driver = {
 		.name	    = "spi-ks8995",
+		.of_match_table = of_match_ptr(ks8895_spi_of_match),
 	},
 	.probe	  = ks8995_probe,
 	.remove	  = ks8995_remove,
Signature · 1 line
2.20.1
author maintainer
#86 lore raw

PATCH 86/137 spi: ST ST95HF NFC: declare missing of table

In reply to Greg Kroah-Hartman

[ Upstream commit d04830531d0c4a99c897a44038e5da3d23331d2f ]

Add missing <of_device_id> table for SPI driver relying on SPI device match since compatible is in a DT binding or in a DTS.

Before this patch:
modinfo drivers/nfc/st95hf/st95hf.ko | grep alias
alias: spi:st95hf

After this patch:
modinfo drivers/nfc/st95hf/st95hf.ko | grep alias

alias:          spi:st95hf
alias:          of:N*T*Cst,st95hfC*
alias:          of:N*T*Cst,st95hf
---
 drivers/nfc/st95hf/core.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/nfc/st95hf/core.c +7 -0
index 2b26f762fbc3b..01acb6e533655 100644
@@ -1074,6 +1074,12 @@ static const struct spi_device_id st95hf_id[] = {
 };
 MODULE_DEVICE_TABLE(spi, st95hf_id);
 
+static const struct of_device_id st95hf_spi_of_match[] = {
+        { .compatible = "st,st95hf" },
+        { },
+};
+MODULE_DEVICE_TABLE(of, st95hf_spi_of_match);
+
 static int st95hf_probe(struct spi_device *nfc_spi_dev)
 {
 	int ret;
@@ -1260,6 +1266,7 @@ static struct spi_driver st95hf_driver = {
 	.driver = {
 		.name = "st95hf",
 		.owner = THIS_MODULE,
+		.of_match_table = of_match_ptr(st95hf_spi_of_match),
 	},
 	.id_table = st95hf_id,
 	.probe = st95hf_probe,
Signature · 1 line
2.20.1
author maintainer
#87 lore raw

PATCH 87/137 ceph: handle the case where a dentry has been renamed on outstanding req

In reply to Greg Kroah-Hartman

[ Upstream commit 4b8222870032715f9d995f3eb7c7acd8379a275d ]

It's possible for us to issue a lookup to revalidate a dentry concurrently with a rename. If done in the right order, then we could end up processing dentry info in the reply that no longer reflects the state of the dentry.

If req->r_dentry->d_name differs from the one in the trace, then just ignore the trace in the reply. We only need to do this however if the parent's i_rwsem is not held.

---
 fs/ceph/inode.c | 16 +++++++++++++++-
 1 file changed, 15 insertions(+), 1 deletion(-)
fs/ceph/inode.c +15 -1
index f7f9e305aaf87..fd3db2e112d6e 100644
@@ -1152,6 +1152,19 @@ static int splice_dentry(struct dentry **pdn, struct inode *in)
 	return 0;
 }
 
+static int d_name_cmp(struct dentry *dentry, const char *name, size_t len)
+{
+	int ret;
+
+	/* take d_lock to ensure dentry->d_name stability */
+	spin_lock(&dentry->d_lock);
+	ret = dentry->d_name.len - len;
+	if (!ret)
+		ret = memcmp(dentry->d_name.name, name, len);
+	spin_unlock(&dentry->d_lock);
+	return ret;
+}
+
 /*
  * Incorporate results into the local cache.  This is either just
  * one inode, or a directory, dentry, and possibly linked-to inode (e.g.,
@@ -1401,7 +1414,8 @@ int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req)
 		err = splice_dentry(&req->r_dentry, in);
 		if (err < 0)
 			goto done;
-	} else if (rinfo->head->is_dentry) {
+	} else if (rinfo->head->is_dentry &&
+		   !d_name_cmp(req->r_dentry, rinfo->dname, rinfo->dname_len)) {
 		struct ceph_vino *ptvino = NULL;
 
 		if ((le32_to_cpu(rinfo->diri.in->cap.caps) & CEPH_CAP_FILE_SHARED) ||
Signature · 1 line
2.20.1
author maintainer
#88 lore raw

PATCH 88/137 Revert "drm/virtio: drop prime import/export callbacks"

In reply to Greg Kroah-Hartman

[ Upstream commit a0cecc23cfcbf2626497a8c8770856dd56b67917 ]

This patch does more harm than good, as it breaks both Xwayland and gnome-shell with X11.

Xwayland requires DRI3 & DRI3 requires PRIME.

X11 crash for obscure double-free reason which are hard to debug (starting X11 by hand doesn't trigger the crash).

I don't see an apparent problem implementing those stub prime functions, they may return an error at run-time, and it seems to be handled fine by GNOME at least.

This reverts commit b318e3ff7ca065d6b107e424c85a63d7a6798a69. [airlied:
This broke userspace for virtio-gpus, and regressed things from DRI3 to DRI2.

This brings back the original problem, but it's better than regressions.]

Fixes b318e3ff7ca065d6b107e424c85a63d7a6798a ("drm/virtio: drop prime import/export callbacks")
---
 drivers/gpu/drm/virtio/virtgpu_drv.c   |  4 ++++
 drivers/gpu/drm/virtio/virtgpu_drv.h   |  4 ++++
 drivers/gpu/drm/virtio/virtgpu_prime.c | 12 ++++++++++++
 3 files changed, 20 insertions(+)
drivers/gpu/drm/virtio/virtgpu_drv.c +4 -0
index 2d1aaca491050..f7f32a885af79 100644
@@ -127,10 +127,14 @@ static struct drm_driver driver = {
 #if defined(CONFIG_DEBUG_FS)
 	.debugfs_init = virtio_gpu_debugfs_init,
 #endif
+	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
+	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
 	.gem_prime_export = drm_gem_prime_export,
 	.gem_prime_import = drm_gem_prime_import,
 	.gem_prime_pin = virtgpu_gem_prime_pin,
 	.gem_prime_unpin = virtgpu_gem_prime_unpin,
+	.gem_prime_get_sg_table = virtgpu_gem_prime_get_sg_table,
+	.gem_prime_import_sg_table = virtgpu_gem_prime_import_sg_table,
 	.gem_prime_vmap = virtgpu_gem_prime_vmap,
 	.gem_prime_vunmap = virtgpu_gem_prime_vunmap,
 	.gem_prime_mmap = virtgpu_gem_prime_mmap,
drivers/gpu/drm/virtio/virtgpu_drv.h +4 -0
index 0c15000f926eb..1deb41d42ea4d 100644
@@ -372,6 +372,10 @@ int virtio_gpu_object_wait(struct virtio_gpu_object *bo, bool no_wait);
 /* virtgpu_prime.c */
 int virtgpu_gem_prime_pin(struct drm_gem_object *obj);
 void virtgpu_gem_prime_unpin(struct drm_gem_object *obj);
+struct sg_table *virtgpu_gem_prime_get_sg_table(struct drm_gem_object *obj);
+struct drm_gem_object *virtgpu_gem_prime_import_sg_table(
+	struct drm_device *dev, struct dma_buf_attachment *attach,
+	struct sg_table *sgt);
 void *virtgpu_gem_prime_vmap(struct drm_gem_object *obj);
 void virtgpu_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr);
 int virtgpu_gem_prime_mmap(struct drm_gem_object *obj,
drivers/gpu/drm/virtio/virtgpu_prime.c +12 -0
index c59ec34c80a5d..eb51a78e11991 100644
@@ -39,6 +39,18 @@ void virtgpu_gem_prime_unpin(struct drm_gem_object *obj)
 	WARN_ONCE(1, "not implemented");
 }
 
+struct sg_table *virtgpu_gem_prime_get_sg_table(struct drm_gem_object *obj)
+{
+	return ERR_PTR(-ENODEV);
+}
+
+struct drm_gem_object *virtgpu_gem_prime_import_sg_table(
+	struct drm_device *dev, struct dma_buf_attachment *attach,
+	struct sg_table *table)
+{
+	return ERR_PTR(-ENODEV);
+}
+
 void *virtgpu_gem_prime_vmap(struct drm_gem_object *obj)
 {
 	struct virtio_gpu_object *bo = gem_to_virtio_gpu_obj(obj);
Signature · 1 line
2.20.1
author maintainer
#89 lore raw

PATCH 89/137 drm/sun4i: Unbind components before releasing DRM and memory

In reply to Greg Kroah-Hartman

[ Upstream commit e02bc29b2cfa7806830d6da8b2322cddd67e8dfe ]

Our components may still be using the DRM device driver (if only to access our driver's private data), so make sure to unbind them before the final drm_dev_put.

Also release our reserved memory after component unbind instead of before to match reverse creation order.

Fixes f5a9ed867c83 ("drm/sun4i: Fix component unbinding and component master deletion")
---
 drivers/gpu/drm/sun4i/sun4i_drv.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/gpu/drm/sun4i/sun4i_drv.c +3 -2
index 9a5713fa03b25..f8bf5bbec2df3 100644
@@ -146,10 +146,11 @@ static void sun4i_drv_unbind(struct device *dev)
 	drm_dev_unregister(drm);
 	drm_kms_helper_poll_fini(drm);
 	drm_mode_config_cleanup(drm);
-	of_reserved_mem_device_release(dev);
-	drm_dev_put(drm);
 
 	component_unbind_all(dev, NULL);
+	of_reserved_mem_device_release(dev);
+
+	drm_dev_put(drm);
 }
 
 static const struct component_master_ops sun4i_drv_master_ops = {
Signature · 1 line
2.20.1
author maintainer
#90 lore raw

PATCH 90/137 Input: snvs_pwrkey - make it depend on ARCH_MXC

In reply to Greg Kroah-Hartman

[ Upstream commit f06eba72274788db6a43012a05a99915c0283aef ]

The SNVS power key is not only used on i.MX6SX and i.MX7D, it is also used by i.MX6UL and NXP's latest ARMv8 based i.MX8M series SOC. So update the config dependency to use ARCH_MXC, and add the COMPILE_TEST too.

---
 drivers/input/keyboard/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/input/keyboard/Kconfig +1 -1
index a878351f16439..52d7f55fca329 100644
@@ -420,7 +420,7 @@ config KEYBOARD_MPR121
 
 config KEYBOARD_SNVS_PWRKEY
 	tristate "IMX SNVS Power Key Driver"
-	depends on SOC_IMX6SX || SOC_IMX7D
+	depends on ARCH_MXC || COMPILE_TEST
 	depends on OF
 	help
 	  This is the snvs powerkey driver for the Freescale i.MX application
Signature · 1 line
2.20.1
author maintainer
#91 lore raw

PATCH 91/137 Input: synaptics-rmi4 - fix possible double free

In reply to Greg Kroah-Hartman

[ Upstream commit bce1a78423961fce676ac65540a31b6ffd179e6d ]

The RMI4 function structure has been released in rmi_register_function if error occurs. However, it will be released again in the function rmi_create_function, which may result in a double-free bug.

---
 drivers/input/rmi4/rmi_driver.c | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)
drivers/input/rmi4/rmi_driver.c +1 -5
index fc3ab93b7aea4..7fb358f961957 100644
@@ -860,7 +860,7 @@ static int rmi_create_function(struct rmi_device *rmi_dev,
 
 	error = rmi_register_function(fn);
 	if (error)
-		goto err_put_fn;
+		return error;
 
 	if (pdt->function_number == 0x01)
 		data->f01_container = fn;
@@ -870,10 +870,6 @@ static int rmi_create_function(struct rmi_device *rmi_dev,
 	list_add_tail(&fn->node, &data->function_list);
 
 	return RMI_SCAN_CONTINUE;
-
-err_put_fn:
-	put_device(&fn->dev);
-	return error;
 }
 
 void rmi_enable_irq(struct rmi_device *rmi_dev, bool clear_wake)
Signature · 1 line
2.20.1
author maintainer
#92 lore raw

PATCH 92/137 net: vrf: Fix operation not supported when set vrf mac

In reply to Greg Kroah-Hartman

[ Upstream commit 6819e3f6d83a24777813b0d031ebe0861694db5a ]

Vrf device is not able to change mac address now because lack of ndo_set_mac_address. Complete this in case some apps need to do this.

---
 drivers/net/vrf.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/net/vrf.c +2 -0
index cd15c32b2e436..9ee4d7402ca23 100644
@@ -875,6 +875,7 @@ static const struct net_device_ops vrf_netdev_ops = {
 	.ndo_init		= vrf_dev_init,
 	.ndo_uninit		= vrf_dev_uninit,
 	.ndo_start_xmit		= vrf_xmit,
+	.ndo_set_mac_address	= eth_mac_addr,
 	.ndo_get_stats64	= vrf_get_stats64,
 	.ndo_add_slave		= vrf_add_slave,
 	.ndo_del_slave		= vrf_del_slave,
@@ -1274,6 +1275,7 @@ static void vrf_setup(struct net_device *dev)
 	/* default to no qdisc; user can add if desired */
 	dev->priv_flags |= IFF_NO_QUEUE;
 	dev->priv_flags |= IFF_NO_RX_HANDLER;
+	dev->priv_flags |= IFF_LIVE_ADDR_CHANGE;
 
 	/* VRF devices do not care about MTU, but if the MTU is set
 	 * too low then the ipv4 and ipv6 protocols are disabled
Signature · 1 line
2.20.1
author maintainer
#93 lore raw

PATCH 93/137 gpio: Fix gpiochip_add_data_with_key() error path

In reply to Greg Kroah-Hartman

[ Upstream commit 357798909164bf423eac6a78ff7da7e98d2d7f7f ]

The err_remove_chip block is too coarse, and may perform cleanup that must not be done. E.g. if of_gpiochip_add() fails, of_gpiochip_remove() is still called, causing:

    OF: ERROR: Bad of_node_put() on /soc/gpio@e6050000
    CPU: 1 PID: 20 Comm: kworker/1:1 Not tainted 5.1.0-rc2-koelsch+ #407
    Hardware name: Generic R-Car Gen2 (Flattened Device Tree)
    Workqueue: events deferred_probe_work_func
    [<c020ec74>] (unwind_backtrace) from [<c020ae58>] (show_stack+0x10/0x14)
    [<c020ae58>] (show_stack) from [<c07c1224>] (dump_stack+0x7c/0x9c)
    [<c07c1224>] (dump_stack) from [<c07c5a80>] (kobject_put+0x94/0xbc)
    [<c07c5a80>] (kobject_put) from [<c0470420>] (gpiochip_add_data_with_key+0x8d8/0xa3c)
    [<c0470420>] (gpiochip_add_data_with_key) from [<c0473738>] (gpio_rcar_probe+0x1d4/0x314)
    [<c0473738>] (gpio_rcar_probe) from [<c052fca8>] (platform_drv_probe+0x48/0x94)

and later, if a GPIO consumer tries to use a GPIO from a failed controller:

    WARNING: CPU: 0 PID: 1 at lib/refcount.c:156 kobject_get+0x38/0x4c
    refcount_t: increment on 0; use-after-free.
    Modules linked in:
    CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.1.0-rc2-koelsch+ #407
    Hardware name: Generic R-Car Gen2 (Flattened Device Tree)
    [<c020ec74>] (unwind_backtrace) from [<c020ae58>] (show_stack+0x10/0x14)
    [<c020ae58>] (show_stack) from [<c07c1224>] (dump_stack+0x7c/0x9c)
    [<c07c1224>] (dump_stack) from [<c0221580>] (__warn+0xd0/0xec)
    [<c0221580>] (__warn) from [<c02215e0>] (warn_slowpath_fmt+0x44/0x6c)
    [<c02215e0>] (warn_slowpath_fmt) from [<c07c58fc>] (kobject_get+0x38/0x4c)
    [<c07c58fc>] (kobject_get) from [<c068b3ec>] (of_node_get+0x14/0x1c)
    [<c068b3ec>] (of_node_get) from [<c0686f24>] (of_find_node_by_phandle+0xc0/0xf0)
    [<c0686f24>] (of_find_node_by_phandle) from [<c0686fbc>] (of_phandle_iterator_next+0x68/0x154)
    [<c0686fbc>] (of_phandle_iterator_next) from [<c0687fe4>] (__of_parse_phandle_with_args+0x40/0xd0)
    [<c0687fe4>] (__of_parse_phandle_with_args) from [<c0688204>] (of_parse_phandle_with_args_map+0x100/0x3ac)
    [<c0688204>] (of_parse_phandle_with_args_map) from [<c0471240>] (of_get_named_gpiod_flags+0x38/0x380)
    [<c0471240>] (of_get_named_gpiod_flags) from [<c046f864>] (gpiod_get_from_of_node+0x24/0xd8)
    [<c046f864>] (gpiod_get_from_of_node) from [<c0470aa4>] (devm_fwnode_get_index_gpiod_from_child+0xa0/0x144)
    [<c0470aa4>] (devm_fwnode_get_index_gpiod_from_child) from [<c05f425c>] (gpio_keys_probe+0x418/0x7bc)
    [<c05f425c>] (gpio_keys_probe) from [<c052fca8>] (platform_drv_probe+0x48/0x94)

Fix this by splitting the cleanup block, and adding a missing call to gpiochip_irqchip_remove().

Fixes 28355f81969962cf ("gpio: defer probe if pinctrl cannot be found")
---
 drivers/gpio/gpiolib.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)
drivers/gpio/gpiolib.c +8 -4
index d1adfdf50fb30..34fbf879411f6 100644
@@ -1379,7 +1379,7 @@ int gpiochip_add_data_with_key(struct gpio_chip *chip, void *data,
 
 	status = gpiochip_add_irqchip(chip, lock_key, request_key);
 	if (status)
-		goto err_remove_chip;
+		goto err_free_gpiochip_mask;
 
 	status = of_gpiochip_add(chip);
 	if (status)
@@ -1387,7 +1387,7 @@ int gpiochip_add_data_with_key(struct gpio_chip *chip, void *data,
 
 	status = gpiochip_init_valid_mask(chip);
 	if (status)
-		goto err_remove_chip;
+		goto err_remove_of_chip;
 
 	for (i = 0; i < chip->ngpio; i++) {
 		struct gpio_desc *desc = &gdev->descs[i];
@@ -1415,14 +1415,18 @@ int gpiochip_add_data_with_key(struct gpio_chip *chip, void *data,
 	if (gpiolib_initialized) {
 		status = gpiochip_setup_dev(gdev);
 		if (status)
-			goto err_remove_chip;
+			goto err_remove_acpi_chip;
 	}
 	return 0;
 
-err_remove_chip:
+err_remove_acpi_chip:
 	acpi_gpiochip_remove(chip);
+err_remove_of_chip:
 	gpiochip_free_hogs(chip);
 	of_gpiochip_remove(chip);
+err_remove_chip:
+	gpiochip_irqchip_remove(chip);
+err_free_gpiochip_mask:
 	gpiochip_free_valid_mask(chip);
 err_remove_irqchip_mask:
 	gpiochip_irqchip_free_valid_mask(chip);
Signature · 1 line
2.20.1
author maintainer
#94 lore raw

PATCH 94/137 RDMA/hns: Bugfix for mapping user db

In reply to Greg Kroah-Hartman

[ Upstream commit 2557fabd6e29f349bfa0ac13f38ac98aa5eafc74 ]

When the maximum send wr delivered by the user is zero, the qp does not have a sq.

When allocating the sq db buffer to store the user sq pi pointer and map it to the kernel mode, max_send_wr is used as the trigger condition, while the kernel does not consider the max_send_wr trigger condition when mapmping db. It will cause sq record doorbell map fail and create qp fail.

The failed print information as follows:

hns3 0000:7d:00.1: Send cmd: tail - 418, opcode - 0x8504, flag - 0x0011, retval - 0x0000 hns3 0000:7d:00.1: Send cmd: 0xe59dc000 0x00000000 0x00000000 0x00000000 0x00000116 0x0000ffff hns3 0000:7d:00.1: sq record doorbell map failed!
hns3 0000:7d:00.1: Create RC QP failed

Fixes 0425e3e6e0c7 ("RDMA/hns: Support flush cqe for hip08 in kernel space")
---
 drivers/infiniband/hw/hns/hns_roce_qp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/infiniband/hw/hns/hns_roce_qp.c +1 -1
index 54031c5b53fa9..89dd2380fc812 100644
@@ -517,7 +517,7 @@ static int hns_roce_set_kernel_sq_size(struct hns_roce_dev *hr_dev,
 
 static int hns_roce_qp_has_sq(struct ib_qp_init_attr *attr)
 {
-	if (attr->qp_type == IB_QPT_XRC_TGT)
+	if (attr->qp_type == IB_QPT_XRC_TGT || !attr->cap.max_send_wr)
 		return 0;
 
 	return 1;
Signature · 1 line
2.20.1
author maintainer
#95 lore raw

PATCH 95/137 mm/memory_hotplug.c: drop memory device reference after find_memory_block()

In reply to Greg Kroah-Hartman

[ Upstream commit 89c02e69fc5245f8a2f34b58b42d43a737af1a5e ]

Right now we are using find_memory_block() to get the node id for the pfn range to online. We are missing to drop a reference to the memory block device. While the device still gets unregistered via device_unregister(), resulting in no user visible problem, the device is never released via device_release(), resulting in a memory leak. Fix that by properly using a put_device().

---
 mm/memory_hotplug.c | 1 +
 1 file changed, 1 insertion(+)
mm/memory_hotplug.c +1 -0
index 11593a03c051f..7493f50ee8800 100644
@@ -858,6 +858,7 @@ int __ref online_pages(unsigned long pfn, unsigned long nr_pages, int online_typ
 	 */
 	mem = find_memory_block(__pfn_to_section(pfn));
 	nid = mem->nid;
+	put_device(&mem->dev);
 
 	/* associate pfn range with the zone */
 	zone = move_pfn_range(online_type, nid, pfn, nr_pages);
Signature · 1 line
2.20.1
author maintainer
#96 lore raw

PATCH 96/137 mm/page_alloc.c: avoid potential NULL pointer dereference

In reply to Greg Kroah-Hartman

[ Upstream commit 8139ad043d632c0e9e12d760068a7a8e91659aa1 ]

ac.preferred_zoneref->zone passed to alloc_flags_nofragment() can be NULL. 'zone' pointer unconditionally derefernced in alloc_flags_nofragment(). Bail out on NULL zone to avoid potential crash. Currently we don't see any crashes only because alloc_flags_nofragment() has another bug which allows compiler to optimize away all accesses to 'zone'.

Fixes 6bb154504f8b ("mm, page_alloc: spread allocations across zones before introducing fragmentation")
---
 mm/page_alloc.c | 3 +++
 1 file changed, 3 insertions(+)
mm/page_alloc.c +3 -0
index eedb57f9b40b5..d59be95ba45cf 100644
@@ -3385,6 +3385,9 @@ alloc_flags_nofragment(struct zone *zone, gfp_t gfp_mask)
 		alloc_flags |= ALLOC_KSWAPD;
 
 #ifdef CONFIG_ZONE_DMA32
+	if (!zone)
+		return alloc_flags;
+
 	if (zone_idx(zone) != ZONE_NORMAL)
 		goto out;
 
Signature · 1 line
2.20.1
author maintainer
#97 lore raw

PATCH 97/137 bpf: only test gso type on gso packets

In reply to Greg Kroah-Hartman

[ Upstream commit 4c3024debf62de4c6ac6d3cb4c0063be21d4f652 ]

BPF can adjust gso only for tcp bytestreams. Fail on other gso types.

But only on gso packets. It does not touch this field if !gso_size.

Fixes b90efd225874 ("bpf: only adjust gso_size on bytestream protocols")
---
 include/linux/skbuff.h | 4 ++--
 net/core/filter.c      | 8 ++++----
 2 files changed, 6 insertions(+), 6 deletions(-)
include/linux/skbuff.h +2 -2
index bdb9563c64a01..b8679dcba96f8 100644
@@ -4212,10 +4212,10 @@ static inline bool skb_is_gso_sctp(const struct sk_buff *skb)
 	return skb_shinfo(skb)->gso_type & SKB_GSO_SCTP;
 }
 
+/* Note: Should be called only if skb_is_gso(skb) is true */
 static inline bool skb_is_gso_tcp(const struct sk_buff *skb)
 {
-	return skb_is_gso(skb) &&
-	       skb_shinfo(skb)->gso_type & (SKB_GSO_TCPV4 | SKB_GSO_TCPV6);
+	return skb_shinfo(skb)->gso_type & (SKB_GSO_TCPV4 | SKB_GSO_TCPV6);
 }
 
 static inline void skb_gso_reset(struct sk_buff *skb)
net/core/filter.c +4 -4
index f7d0004fc1609..ff07996515f2d 100644
@@ -2789,7 +2789,7 @@ static int bpf_skb_proto_4_to_6(struct sk_buff *skb)
 	u32 off = skb_mac_header_len(skb);
 	int ret;
 
-	if (!skb_is_gso_tcp(skb))
+	if (skb_is_gso(skb) && !skb_is_gso_tcp(skb))
 		return -ENOTSUPP;
 
 	ret = skb_cow(skb, len_diff);
@@ -2830,7 +2830,7 @@ static int bpf_skb_proto_6_to_4(struct sk_buff *skb)
 	u32 off = skb_mac_header_len(skb);
 	int ret;
 
-	if (!skb_is_gso_tcp(skb))
+	if (skb_is_gso(skb) && !skb_is_gso_tcp(skb))
 		return -ENOTSUPP;
 
 	ret = skb_unclone(skb, GFP_ATOMIC);
@@ -2955,7 +2955,7 @@ static int bpf_skb_net_grow(struct sk_buff *skb, u32 len_diff)
 	u32 off = skb_mac_header_len(skb) + bpf_skb_net_base_len(skb);
 	int ret;
 
-	if (!skb_is_gso_tcp(skb))
+	if (skb_is_gso(skb) && !skb_is_gso_tcp(skb))
 		return -ENOTSUPP;
 
 	ret = skb_cow(skb, len_diff);
@@ -2984,7 +2984,7 @@ static int bpf_skb_net_shrink(struct sk_buff *skb, u32 len_diff)
 	u32 off = skb_mac_header_len(skb) + bpf_skb_net_base_len(skb);
 	int ret;
 
-	if (!skb_is_gso_tcp(skb))
+	if (skb_is_gso(skb) && !skb_is_gso_tcp(skb))
 		return -ENOTSUPP;
 
 	ret = skb_unclone(skb, GFP_ATOMIC);
Signature · 1 line
2.20.1
author maintainer
#98 lore raw

PATCH 98/137 net: sched: fix cleanup NULL pointer exception in act_mirr

In reply to Greg Kroah-Hartman

[ Upstream commit 064c5d6881e897077639e04973de26440ee205e6 ]

A new mirred action is created by the tcf_mirred_init function. This contains a list head struct which is inserted into a global list on successful creation of a new action. However, after a creation, it is still possible to error out and call the tcf_idr_release function. This, in turn, calls the act_mirr cleanup function via __tcf_idr_release and __tcf_action_put. This cleanup function tries to delete the list entry which is as yet uninitialised, leading to a NULL pointer exception.

Fix this by initialising the list entry on creation of a new action.

Bug report:

BUG: unable to handle kernel NULL pointer dereference at 0000000000000008
PGD 8000000840c73067 P4D 8000000840c73067 PUD 858dcc067 PMD 0
Oops: 0002 [#1] SMP PTI
CPU: 32 PID: 5636 Comm: handler194 Tainted: G           OE     5.0.0+ #186
Hardware name: Dell Inc. PowerEdge R730/0599V5, BIOS 1.3.6 06/03/2015
RIP: 0010:tcf_mirred_release+0x42/0xa7 [act_mirred]
Code: f0 90 39 c0 e8 52 04 57 c8 48 c7 c7 b8 80 39 c0 e8 94 fa d4 c7 48 8b 93 d0 00 00 00 48 8b 83 d8 00 00 00 48 c7 c7 f0 90 39 c0 <48> 89 42 08 48 89 10 48 b8 00 01 00 00 00 00 ad de 48 89 83 d0 00
RSP: 0018:ffffac4aa059f688 EFLAGS: 00010282
RAX: 0000000000000000 RBX: ffff9dcd1b214d00 RCX: 0000000000000000
RDX: 0000000000000000 RSI: ffff9dcd1fa165f8 RDI: ffffffffc03990f0
RBP: ffff9dccf9c7af80 R08: 0000000000000a3b R09: 0000000000000000
R10: ffff9dccfa11f420 R11: 0000000000000000 R12: 0000000000000001
R13: ffff9dcd16b433c0 R14: ffff9dcd1b214d80 R15: 0000000000000000
FS:  00007f441bfff700(0000) GS:ffff9dcd1fa00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000008 CR3: 0000000839e64004 CR4: 00000000001606e0
Call Trace:
tcf_action_cleanup+0x59/0xca
__tcf_action_put+0x54/0x6b
__tcf_idr_release.cold.33+0x9/0x12
tcf_mirred_init.cold.20+0x22e/0x3b0 [act_mirred]
tcf_action_init_1+0x3d0/0x4c0
tcf_action_init+0x9c/0x130
tcf_exts_validate+0xab/0xc0
fl_change+0x1ca/0x982 [cls_flower]
tc_new_tfilter+0x647/0x8d0
? load_balance+0x14b/0x9e0
rtnetlink_rcv_msg+0xe3/0x370
? __switch_to_asm+0x40/0x70
? __switch_to_asm+0x34/0x70
? _cond_resched+0x15/0x30
? __kmalloc_node_track_caller+0x1d4/0x2b0
? rtnl_calcit.isra.31+0xf0/0xf0
netlink_rcv_skb+0x49/0x110
netlink_unicast+0x16f/0x210
netlink_sendmsg+0x1df/0x390
sock_sendmsg+0x36/0x40
___sys_sendmsg+0x27b/0x2c0
? futex_wake+0x80/0x140
? do_futex+0x2b9/0xac0
? ep_scan_ready_list.constprop.22+0x1f2/0x210
? ep_poll+0x7a/0x430
__sys_sendmsg+0x47/0x80
do_syscall_64+0x55/0x100
entry_SYSCALL_64_after_hwframe+0x44/0xa9
Fixes 4e232818bd32 ("net: sched: act_mirred: remove dependency on rtnl lock")
---
 net/sched/act_mirred.c | 3 +++
 1 file changed, 3 insertions(+)
net/sched/act_mirred.c +3 -0
index c8cf4d10c4355..971dc03304f42 100644
@@ -159,6 +159,9 @@ static int tcf_mirred_init(struct net *net, struct nlattr *nla,
 	}
 	m = to_mirred(*a);
 
+	if (ret == ACT_P_CREATED)
+		INIT_LIST_HEAD(&m->tcfm_list);
+
 	spin_lock_bh(&m->tcf_lock);
 	m->tcf_action = parm->action;
 	m->tcfm_eaction = parm->eaction;
Signature · 1 line
2.20.1
author maintainer
#99 lore raw

PATCH 99/137 net: mvpp2: fix validate for PPv2.1

In reply to Greg Kroah-Hartman

[ Upstream commit 8b318f30ab4ef9bbc1241e6f8c1db366dbd347f2 ]

The Phylink validate function is the Marvell PPv2 driver makes a check on the GoP id. This is valid an has to be done when using PPv2.2 engines but makes no sense when using PPv2.1. The check done when using an RGMII interface makes sure the GoP id is not 0, but this breaks PPv2.1. Fixes it.

Fixes 0fb628f0f250 ("net: mvpp2: fix phylink handling of invalid PHY modes")
---
 drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c +1 -1
index 931beac3359d1..70031e2b22944 100644
@@ -4370,7 +4370,7 @@ static void mvpp2_phylink_validate(struct net_device *dev,
 	case PHY_INTERFACE_MODE_RGMII_ID:
 	case PHY_INTERFACE_MODE_RGMII_RXID:
 	case PHY_INTERFACE_MODE_RGMII_TXID:
-		if (port->gop_id == 0)
+		if (port->priv->hw_version == MVPP22 && port->gop_id == 0)
 			goto empty_set;
 		break;
 	default:
Signature · 1 line
2.20.1
author maintainer
#100 lore raw

PATCH 100/137 drm/rockchip: fix for mailbox read validation.

In reply to Greg Kroah-Hartman

[ Upstream commit e4056bbb6719fe713bfc4030ac78e8e97ddf7574 ]

This is basically the same fix as in
commit fa68d4f8476b ("drm/rockchip: fix for mailbox read size") but for cdn_dp_mailbox_validate_receive function.

See patchwork.kernel.org/patch/10671981/ for details.

---
 drivers/gpu/drm/rockchip/cdn-dp-reg.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/rockchip/cdn-dp-reg.c +1 -1
index 5a485489a1e23..6c8b14fb1d2f3 100644
@@ -113,7 +113,7 @@ static int cdp_dp_mailbox_write(struct cdn_dp_device *dp, u8 val)
 
 static int cdn_dp_mailbox_validate_receive(struct cdn_dp_device *dp,
 					   u8 module_id, u8 opcode,
-					   u8 req_size)
+					   u16 req_size)
 {
 	u32 mbox_size, i;
 	u8 header[4];
Signature · 1 line
2.20.1
author maintainer
#101 lore raw

PATCH 101/137 cw1200: fix missing unlock on error in cw1200_hw_scan()

In reply to Greg Kroah-Hartman

From: Wei Yongjun <[email protected]>

commit 51c8d24101c79ffce3e79137e2cee5dfeb956dd7 upstream.

Add the missing unlock before return from function cw1200_hw_scan() in the error handling case.

Fixes 4f68ef64cd7f ("cw1200: Fix concurrency use-after-free bugs in cw1200_hw_scan()")
---
 drivers/net/wireless/st/cw1200/scan.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/net/wireless/st/cw1200/scan.c +4 -1
@@ -84,8 +84,11 @@ int cw1200_hw_scan(struct ieee80211_hw *
 
 	frame.skb = ieee80211_probereq_get(hw, priv->vif->addr, NULL, 0,
 		req->ie_len);
-	if (!frame.skb)
+	if (!frame.skb) {
+		mutex_unlock(&priv->conf_mutex);
+		up(&priv->scan.lock);
 		return -ENOMEM;
+	}
 
 	if (req->ie_len)
 		skb_put_data(frame.skb, req->ie, req->ie_len);
author maintainer
#102 lore raw

PATCH 102/137 mwl8k: Fix rate_idx underflow

In reply to Greg Kroah-Hartman

From: Petr Štetiar <[email protected]>

commit 6b583201fa219b7b1b6aebd8966c8fd9357ef9f4 upstream.

It was reported on OpenWrt bug tracking system[1], that several users are affected by the endless reboot of their routers if they configure 5GHz interface with channel 44 or 48.

The reboot loop is caused by the following excessive number of WARN_ON messages:

 WARNING: CPU: 0 PID: 0 at backports-4.19.23-1/net/mac80211/rx.c:4516
                             ieee80211_rx_napi+0x1fc/0xa54 [mac80211]

as the messages are being correctly emitted by the following guard:

case RX_ENC_LEGACY:

      if (WARN_ON(status->rate_idx >= sband->n_bitrates))

as the rate_idx is in this case erroneously set to 251 (0xfb). This fix simply converts previously used magic number to proper constant and guards against substraction which is leading to the currently observed underflow.

1. https://bugs.openwrt.org/index.php?do=details&task_id=2218

Fixes 854783444bab ("mwl8k: properly set receive status rate index on 5 GHz receive")
---
 drivers/net/wireless/marvell/mwl8k.c |   13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)
drivers/net/wireless/marvell/mwl8k.c +9 -4
@@ -441,6 +441,9 @@ static const struct ieee80211_rate mwl8k
 #define MWL8K_CMD_UPDATE_STADB		0x1123
 #define MWL8K_CMD_BASTREAM		0x1125
 
+#define MWL8K_LEGACY_5G_RATE_OFFSET \
+	(ARRAY_SIZE(mwl8k_rates_24) - ARRAY_SIZE(mwl8k_rates_50))
+
 static const char *mwl8k_cmd_name(__le16 cmd, char *buf, int bufsize)
 {
 	u16 command = le16_to_cpu(cmd);
@@ -1016,8 +1019,9 @@ mwl8k_rxd_ap_process(void *_rxd, struct
 
 	if (rxd->channel > 14) {
 		status->band = NL80211_BAND_5GHZ;
-		if (!(status->encoding == RX_ENC_HT))
-			status->rate_idx -= 5;
+		if (!(status->encoding == RX_ENC_HT) &&
+		    status->rate_idx >= MWL8K_LEGACY_5G_RATE_OFFSET)
+			status->rate_idx -= MWL8K_LEGACY_5G_RATE_OFFSET;
 	} else {
 		status->band = NL80211_BAND_2GHZ;
 	}
@@ -1124,8 +1128,9 @@ mwl8k_rxd_sta_process(void *_rxd, struct
 
 	if (rxd->channel > 14) {
 		status->band = NL80211_BAND_5GHZ;
-		if (!(status->encoding == RX_ENC_HT))
-			status->rate_idx -= 5;
+		if (!(status->encoding == RX_ENC_HT) &&
+		    status->rate_idx >= MWL8K_LEGACY_5G_RATE_OFFSET)
+			status->rate_idx -= MWL8K_LEGACY_5G_RATE_OFFSET;
 	} else {
 		status->band = NL80211_BAND_2GHZ;
 	}
author maintainer
#103 lore raw

PATCH 103/137 rtlwifi: rtl8723ae: Fix missing break in switch statement

In reply to Greg Kroah-Hartman

From: Gustavo A. R. Silva <[email protected]>

commit 84242b82d81c54e009a2aaa74d3d9eff70babf56 upstream.

Add missing break statement in order to prevent the code from falling through to case 0x1025, and erroneously setting rtlhal->oem_id to RT_CID_819X_ACER when rtlefuse->eeprom_svid is equal to 0x10EC and none of the cases in switch (rtlefuse->eeprom_smid) match.

This bug was found thanks to the ongoing efforts to enable -Wimplicit-fallthrough.

Fixes 238ad2ddf34b ("rtlwifi: rtl8723ae: Clean up the hardware info routine")
---
 drivers/net/wireless/realtek/rtlwifi/rtl8723ae/hw.c |    1 +
 1 file changed, 1 insertion(+)
drivers/net/wireless/realtek/rtlwifi/rtl8723ae/hw.c +1 -0
@@ -1697,6 +1697,7 @@ static void _rtl8723e_read_adapter_info(
 					rtlhal->oem_id = RT_CID_819X_LENOVO;
 					break;
 				}
+				break;
 			case 0x1025:
 				rtlhal->oem_id = RT_CID_819X_ACER;
 				break;
author maintainer
#104 lore raw

PATCH 104/137 Dont jump to compute_result state from check_result state

In reply to Greg Kroah-Hartman

From: Nigel Croxon <[email protected]>

commit 4f4fd7c5798bbdd5a03a60f6269cf1177fbd11ef upstream.

Changing state from check_state_check_result to check_state_compute_result not only is unsafe but also doesn't appear to serve a valid purpose. A raid6 check should only be pushing out extra writes if doing repair and a mis-match occurs. The stripe dev management will already try and do repair writes for failing sectors.

This patch makes the raid6 check_state_check_result handling work more like raid5's. If somehow too many failures for a check, just quit the check operation for the stripe. When any checks pass, don't try and use check_state_compute_result for a purpose it isn't needed for and is unsafe for. Just mark the stripe as in sync for passing its parity checks and let the stripe dev read/write code and the bad blocks list do their job handling I/O errors.

Repro steps from Xiao:

These are the steps to reproduce this problem:
1. redefined OPT_MEDIUM_ERR_ADDR to 12000 in scsi_debug.c
2. insmod scsi_debug.ko dev_size_mb=11000 max_luns=1 num_tgts=1
3. mdadm --create /dev/md127 --level=6 --raid-devices=5 /dev/sde1 /dev/sde2 /dev/sde3 /dev/sde5 /dev/sde6 sde is the disk created by scsi_debug
4. echo "2" >/sys/module/scsi_debug/parameters/opts
5. raid-check

It panic:

[ 4854.730899] md: data-check of RAID array md127
[ 4854.857455] sd 5:0:0:0: [sdr] tag#80 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
[ 4854.859246] sd 5:0:0:0: [sdr] tag#80 Sense Key : Medium Error [current]
[ 4854.860694] sd 5:0:0:0: [sdr] tag#80 Add. Sense: Unrecovered read error
[ 4854.862207] sd 5:0:0:0: [sdr] tag#80 CDB: Read(10) 28 00 00 00 2d 88 00 04 00 00
[ 4854.864196] print_req_error: critical medium error, dev sdr, sector 11656 flags 0
[ 4854.867409] sd 5:0:0:0: [sdr] tag#100 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
[ 4854.869469] sd 5:0:0:0: [sdr] tag#100 Sense Key : Medium Error [current]
[ 4854.871206] sd 5:0:0:0: [sdr] tag#100 Add. Sense: Unrecovered read error
[ 4854.872858] sd 5:0:0:0: [sdr] tag#100 CDB: Read(10) 28 00 00 00 2e e0 00 00 08 00
[ 4854.874587] print_req_error: critical medium error, dev sdr, sector 12000 flags 4000
[ 4854.876456] sd 5:0:0:0: [sdr] tag#101 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
[ 4854.878552] sd 5:0:0:0: [sdr] tag#101 Sense Key : Medium Error [current]
[ 4854.880278] sd 5:0:0:0: [sdr] tag#101 Add. Sense: Unrecovered read error
[ 4854.881846] sd 5:0:0:0: [sdr] tag#101 CDB: Read(10) 28 00 00 00 2e e8 00 00 08 00
[ 4854.883691] print_req_error: critical medium error, dev sdr, sector 12008 flags 4000
[ 4854.893927] sd 5:0:0:0: [sdr] tag#166 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
[ 4854.896002] sd 5:0:0:0: [sdr] tag#166 Sense Key : Medium Error [current]
[ 4854.897561] sd 5:0:0:0: [sdr] tag#166 Add. Sense: Unrecovered read error
[ 4854.899110] sd 5:0:0:0: [sdr] tag#166 CDB: Read(10) 28 00 00 00 2e e0 00 00 10 00
[ 4854.900989] print_req_error: critical medium error, dev sdr, sector 12000 flags 0
[ 4854.902757] md/raid:md127: read error NOT corrected!! (sector 9952 on sdr1).
[ 4854.904375] md/raid:md127: read error NOT corrected!! (sector 9960 on sdr1).
[ 4854.906201] ------------[ cut here ]------------
[ 4854.907341] kernel BUG at drivers/md/raid5.c:4190!

raid5.c:4190 above is this BUG_ON:

    handle_parity_checks6()
        ...
        BUG_ON(s->uptodate < disks - 1); /* We don't need Q to recover */

Cc: <[email protected]> # v3.16+
OriginalAuthor: David Jeffery <[email protected]>

---
 drivers/md/raid5.c |   19 ++++---------------
 1 file changed, 4 insertions(+), 15 deletions(-)
drivers/md/raid5.c +4 -15
@@ -4233,26 +4233,15 @@ static void handle_parity_checks6(struct
 	case check_state_check_result:
 		sh->check_state = check_state_idle;
 
+		if (s->failed > 1)
+			break;
 		/* handle a successful check operation, if parity is correct
 		 * we are done.  Otherwise update the mismatch count and repair
 		 * parity if !MD_RECOVERY_CHECK
 		 */
 		if (sh->ops.zero_sum_result == 0) {
-			/* both parities are correct */
-			if (!s->failed)
-				set_bit(STRIPE_INSYNC, &sh->state);
-			else {
-				/* in contrast to the raid5 case we can validate
-				 * parity, but still have a failure to write
-				 * back
-				 */
-				sh->check_state = check_state_compute_result;
-				/* Returning at this point means that we may go
-				 * off and bring p and/or q uptodate again so
-				 * we make sure to check zero_sum_result again
-				 * to verify if p or q need writeback
-				 */
-			}
+			/* Any parity checked was correct */
+			set_bit(STRIPE_INSYNC, &sh->state);
 		} else {
 			atomic64_add(STRIPE_SECTORS, &conf->mddev->resync_mismatches);
 			if (test_bit(MD_RECOVERY_CHECK, &conf->mddev->recovery)) {
author maintainer
#105 lore raw

PATCH 105/137 bonding: fix arp_validate toggling in active-backup mode

In reply to Greg Kroah-Hartman

From: Jarod Wilson <[email protected]>

[ Upstream commit a9b8a2b39ce65df45687cf9ef648885c2a99fe75 ]

There's currently a problem with toggling arp_validate on and off with an active-backup bond. At the moment, you can start up a bond, like so:

modprobe bonding mode=1 arp_interval=100 arp_validate=0 arp_ip_targets=192.168.1.1 ip link set bond0 down
echo "ens4f0" > /sys/class/net/bond0/bonding/slaves
echo "ens4f1" > /sys/class/net/bond0/bonding/slaves
ip link set bond0 up
ip addr add 192.168.1.2/24 dev bond0

Pings to 192.168.1.1 work just fine. Now turn on arp_validate:

echo 1 > /sys/class/net/bond0/bonding/arp_validate

Pings to 192.168.1.1 continue to work just fine. Now when you go to turn arp_validate off again, the link falls flat on it's face:

echo 0 > /sys/class/net/bond0/bonding/arp_validate dmesg
...

[133191.911987] bond0: Setting arp_validate to none (0)
[133194.257793] bond0: bond_should_notify_peers: slave ens4f0
[133194.258031] bond0: link status definitely down for interface ens4f0, disabling it
[133194.259000] bond0: making interface ens4f1 the new active one
[133197.330130] bond0: link status definitely down for interface ens4f1, disabling it
[133197.331191] bond0: now running without any active interface!

The problem lies in bond_options.c, where passing in arp_validate=0 results in bond->recv_probe getting set to NULL. This flies directly in the face of commit 3fe68df97c7f, which says we need to set recv_probe = bond_arp_recv, even if we're not using arp_validate. Said commit fixed this in bond_option_arp_interval_set, but missed that we can get to that same state in bond_option_arp_validate_set as well.

One solution would be to universally set recv_probe = bond_arp_recv here as well, but I don't think bond_option_arp_validate_set has any business touching recv_probe at all, and that should be left to the arp_interval code, so we can just make things much tidier here.

Fixes 3fe68df97c7f ("bonding: always set recv_probe to bond_arp_rcv in arp monitor")
CC Jay Vosburgh <[email protected]>
CC Veaceslav Falico <[email protected]>
CC Andy Gospodarek <[email protected]>
CC "David S. Miller" <[email protected]>
---
 drivers/net/bonding/bond_options.c |    7 -------
 1 file changed, 7 deletions(-)
drivers/net/bonding/bond_options.c +0 -7
@@ -1098,13 +1098,6 @@ static int bond_option_arp_validate_set(
 {
 	netdev_dbg(bond->dev, "Setting arp_validate to %s (%llu)\n",
 		   newval->string, newval->value);
-
-	if (bond->dev->flags & IFF_UP) {
-		if (!newval->value)
-			bond->recv_probe = NULL;
-		else if (bond->params.arp_interval)
-			bond->recv_probe = bond_arp_rcv;
-	}
 	bond->params.arp_validate = newval->value;
 
 	return 0;
author maintainer
#106 lore raw

PATCH 106/137 bridge: Fix error path for kobject_init_and_add()

In reply to Greg Kroah-Hartman

From: "Tobin C. Harding" <[email protected]>

[ Upstream commit bdfad5aec1392b93495b77b864d58d7f101dc1c1 ]

Currently error return from kobject_init_and_add() is not followed by a call to kobject_put(). This means there is a memory leak. We currently set p to NULL so that kfree() may be called on it as a noop, the code is arguably clearer if we move the kfree() up closer to where it is called (instead of after goto jump).

Remove a goto label 'err1' and jump to call to kobject_put() in error return from kobject_init_and_add() fixing the memory leak. Re-name goto label 'put_back' to 'err1' now that we don't use err1, following current nomenclature (err1, err2 ...). Move call to kfree out of the error code at bottom of function up to closer to where memory was allocated. Add comment to clarify call to kfree().

---
 net/bridge/br_if.c |   13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)
net/bridge/br_if.c +6 -7
@@ -602,13 +602,15 @@ int br_add_if(struct net_bridge *br, str
 	call_netdevice_notifiers(NETDEV_JOIN, dev);
 
 	err = dev_set_allmulti(dev, 1);
-	if (err)
-		goto put_back;
+	if (err) {
+		kfree(p);	/* kobject not yet init'd, manually free */
+		goto err1;
+	}
 
 	err = kobject_init_and_add(&p->kobj, &brport_ktype, &(dev->dev.kobj),
 				   SYSFS_BRIDGE_PORT_ATTR);
 	if (err)
-		goto err1;
+		goto err2;
 
 	err = br_sysfs_addif(p);
 	if (err)
@@ -700,12 +702,9 @@ err3:
 	sysfs_remove_link(br->ifobj, p->dev->name);
 err2:
 	kobject_put(&p->kobj);
-	p = NULL; /* kobject_put frees */
-err1:
 	dev_set_allmulti(dev, -1);
-put_back:
+err1:
 	dev_put(dev);
-	kfree(p);
 	return err;
 }
 
author maintainer
#107 lore raw

PATCH 107/137 dpaa_eth: fix SG frame cleanup

In reply to Greg Kroah-Hartman

From: Laurentiu Tudor <[email protected]>

[ Upstream commit 17170e6570c082717c142733d9a638bcd20551f8 ]

Fix issue with the entry indexing in the sg frame cleanup code being off-by-1. This problem showed up when doing some basic iperf tests and manifested in traffic coming to a halt.

---
 drivers/net/ethernet/freescale/dpaa/dpaa_eth.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/freescale/dpaa/dpaa_eth.c +1 -1
@@ -1648,7 +1648,7 @@ static struct sk_buff *dpaa_cleanup_tx_f
 				 qm_sg_entry_get_len(&sgt[0]), dma_dir);
 
 		/* remaining pages were mapped with skb_frag_dma_map() */
-		for (i = 1; i < nr_frags; i++) {
+		for (i = 1; i <= nr_frags; i++) {
 			WARN_ON(qm_sg_entry_is_ext(&sgt[i]));
 
 			dma_unmap_page(dev, qm_sg_addr(&sgt[i]),
author maintainer
#108 lore raw

PATCH 108/137 fib_rules: return 0 directly if an exactly same rule exists when NLM_F_EXCL not supplied

In reply to Greg Kroah-Hartman

From: Hangbin Liu <[email protected]>

[ Upstream commit e9919a24d3022f72bcadc407e73a6ef17093a849 ]

With commit 153380ec4b9 ("fib_rules: Added NLM_F_EXCL support to fib_nl_newrule") we now able to check if a rule already exists. But this only works with iproute2. For other tools like libnl, NetworkManager, it still could add duplicate rules with only NLM_F_CREATE flag, like

[localhost ~ ]# ip rule

0:      from all lookup local
32766:  from all lookup main
32767:  from all lookup default

100000: from 192.168.7.5 lookup 5
100000: from 192.168.7.5 lookup 5

As it doesn't make sense to create two duplicate rules, let's just return 0 if the rule exists.

Fixes 153380ec4b9 ("fib_rules: Added NLM_F_EXCL support to fib_nl_newrule")
---
 net/core/fib_rules.c |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
net/core/fib_rules.c +3 -3
@@ -756,9 +756,9 @@ int fib_nl_newrule(struct sk_buff *skb,
 	if (err)
 		goto errout;
 
-	if ((nlh->nlmsg_flags & NLM_F_EXCL) &&
-	    rule_exists(ops, frh, tb, rule)) {
-		err = -EEXIST;
+	if (rule_exists(ops, frh, tb, rule)) {
+		if (nlh->nlmsg_flags & NLM_F_EXCL)
+			err = -EEXIST;
 		goto errout_free;
 	}
 
author maintainer
#109 lore raw

PATCH 109/137 ipv4: Fix raw socket lookup for local traffic

In reply to Greg Kroah-Hartman

From: David Ahern <[email protected]>

[ Upstream commit 19e4e768064a87b073a4b4c138b55db70e0cfb9f ]

inet_iif should be used for the raw socket lookup. inet_iif considers rt_iif which handles the case of local traffic.

As it stands, ping to a local address with the '-I <dev>' option fails ever since ping was changed to use SO_BINDTODEVICE instead of cmsg + IP_PKTINFO.

IPv6 works fine.

Fixes 1da177e4c3f4 ("Linux-2.6.12-rc2")
---
 net/ipv4/raw.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
net/ipv4/raw.c +2 -2
@@ -173,6 +173,7 @@ static int icmp_filter(const struct sock
 static int raw_v4_input(struct sk_buff *skb, const struct iphdr *iph, int hash)
 {
 	int sdif = inet_sdif(skb);
+	int dif = inet_iif(skb);
 	struct sock *sk;
 	struct hlist_head *head;
 	int delivered = 0;
@@ -185,8 +186,7 @@ static int raw_v4_input(struct sk_buff *
 
 	net = dev_net(skb->dev);
 	sk = __raw_v4_lookup(net, __sk_head(head), iph->protocol,
-			     iph->saddr, iph->daddr,
-			     skb->dev->ifindex, sdif);
+			     iph->saddr, iph->daddr, dif, sdif);
 
 	while (sk) {
 		delivered = 1;
author maintainer
#110 lore raw

PATCH 110/137 net: dsa: Fix error cleanup path in dsa_init_module

In reply to Greg Kroah-Hartman

From: YueHaibing <[email protected]>

[ Upstream commit 68be930249d051fd54d3d99156b3dcadcb2a1f9b ]

BUG: unable to handle kernel paging request at ffffffffa01c5430
PGD 3270067 P4D 3270067 PUD 3271063 PMD 230bc5067 PTE 0
Oops: 0000 [#1
CPU: 0 PID: 6159 Comm: modprobe Not tainted 5.1.0+ #33
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.9.3-0-ge2fc41e-prebuilt.qemu-project.org 04/01/2014
RIP: 0010:raw_notifier_chain_register+0x16/0x40
Code: 63 f8 66 90 e9 5d ff ff ff 90 90 90 90 90 90 90 90 90 90 90 55 48 8b 07 48 89 e5 48 85 c0 74 1c 8b 56 10 3b 50 10 7e 07 eb 12 <39> 50 10 7c 0d 48 8d 78 08 48 8b 40 08 48 85 c0 75 ee 48 89 46 08
RSP: 0018:ffffc90001c33c08 EFLAGS: 00010282
RAX: ffffffffa01c5420 RBX: ffffffffa01db420 RCX: 4fcef45928070a8b
RDX: 0000000000000000 RSI: ffffffffa01db420 RDI: ffffffffa01b0068
RBP: ffffc90001c33c08 R08: 000000003e0a33d0 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000094443661 R12: ffff88822c320700
R13: ffff88823109be80 R14: 0000000000000000 R15: ffffc90001c33e78
FS:  00007fab8bd08540(0000) GS:ffff888237a00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: ffffffffa01c5430 CR3: 00000002297ea000 CR4: 00000000000006f0
Call Trace:
 register_netdevice_notifier+0x43/0x250
 ? 0xffffffffa01e0000
 dsa_slave_register_notifier+0x13/0x70 [dsa_core
 ? 0xffffffffa01e0000
 dsa_init_module+0x2e/0x1000 [dsa_core
 do_one_initcall+0x6c/0x3cc
 ? do_init_module+0x22/0x1f1
 ? rcu_read_lock_sched_held+0x97/0xb0
 ? kmem_cache_alloc_trace+0x325/0x3b0
 do_init_module+0x5b/0x1f1
 load_module+0x1db1/0x2690
 ? m_show+0x1d0/0x1d0
 __do_sys_finit_module+0xc5/0xd0
 __x64_sys_finit_module+0x15/0x20
 do_syscall_64+0x6b/0x1d0
 entry_SYSCALL_64_after_hwframe+0x49/0xbe

Cleanup allocated resourses if there are errors,
otherwise it will trgger memleak.

Fixes c9eb3e0f8701 ("net: dsa: Add support for learning FDB through notification")
---
 net/dsa/dsa.c |   11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)
net/dsa/dsa.c +9 -2
@@ -342,15 +342,22 @@ static int __init dsa_init_module(void)
 
 	rc = dsa_slave_register_notifier();
 	if (rc)
-		return rc;
+		goto register_notifier_fail;
 
 	rc = dsa_legacy_register();
 	if (rc)
-		return rc;
+		goto legacy_register_fail;
 
 	dev_add_pack(&dsa_pack_type);
 
 	return 0;
+
+legacy_register_fail:
+	dsa_slave_unregister_notifier();
+register_notifier_fail:
+	destroy_workqueue(dsa_owq);
+
+	return rc;
 }
 module_init(dsa_init_module);
 
author maintainer
#111 lore raw

PATCH 111/137 net: ethernet: stmmac: dwmac-sun8i: enable support of unicast filtering

In reply to Greg Kroah-Hartman

From: Corentin Labbe <[email protected]>

[ Upstream commit d4c26eb6e721683a0f93e346ce55bc8dc3cbb175 ]

When adding more MAC addresses to a dwmac-sun8i interface, the device goes directly in promiscuous mode.
This is due to IFF_UNICAST_FLT missing flag.

So since the hardware support unicast filtering, let's add IFF_UNICAST_FLT.

Fixes 9f93ac8d4085 ("net-next: stmmac: Add dwmac-sun8i")
---
 drivers/net/ethernet/stmicro/stmmac/dwmac-sun8i.c |    2 ++
 1 file changed, 2 insertions(+)
drivers/net/ethernet/stmicro/stmmac/dwmac-sun8i.c +2 -0
@@ -1015,6 +1015,8 @@ static struct mac_device_info *sun8i_dwm
 	mac->mac = &sun8i_dwmac_ops;
 	mac->dma = &sun8i_dwmac_dma_ops;
 
+	priv->dev->priv_flags |= IFF_UNICAST_FLT;
+
 	/* The loopback bit seems to be re-set when link change
 	 * Simply mask it each time
 	 * Speed 10/100/1000 are set in BIT(2)/BIT(3)
author maintainer
#112 lore raw

PATCH 112/137 net: macb: Change interrupt and napi enable order in open

In reply to Greg Kroah-Hartman

From: Harini Katakam <[email protected]>

[ Upstream commit 0504453139ef5a593c9587e1e851febee859c7d8 ]

Current order in open:
-> Enable interrupts (macb_init_hw)
-> Enable NAPI
-> Start PHY

Sequence of RX handling:
-> RX interrupt occurs
-> Interrupt is cleared and interrupt bits disabled in handler -> NAPI is scheduled
-> In NAPI, RX budget is processed and RX interrupts are re-enabled

With the above, on QEMU or fixed link setups (where PHY state doesn't matter), there's a chance macb RX interrupt occurs before NAPI is enabled. This will result in NAPI being scheduled before it is enabled. Fix this macb open by changing the order.

Fixes ae1f2a56d273 ("net: macb: Added support for many RX queues")
---
 drivers/net/ethernet/cadence/macb_main.c |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/net/ethernet/cadence/macb_main.c +3 -3
@@ -2414,12 +2414,12 @@ static int macb_open(struct net_device *
 		return err;
 	}
 
-	bp->macbgem_ops.mog_init_rings(bp);
-	macb_init_hw(bp);
-
 	for (q = 0, queue = bp->queues; q < bp->num_queues; ++q, ++queue)
 		napi_enable(&queue->napi);
 
+	bp->macbgem_ops.mog_init_rings(bp);
+	macb_init_hw(bp);
+
 	/* schedule a link state check */
 	phy_start(dev->phydev);
 
author maintainer
#113 lore raw

PATCH 113/137 net: seeq: fix crash caused by not set dev.parent

In reply to Greg Kroah-Hartman

From: Thomas Bogendoerfer <[email protected]>

[ Upstream commit 5afcd14cfc7fed1bcc8abcee2cef82732772bfc2 ]

The old MIPS implementation of dma_cache_sync() didn't use the dev argument, but commit c9eb6172c328 ("dma-mapping: turn dma_cache_sync into a dma_map_ops method") changed that, so we now need to set dev.parent.

---
 drivers/net/ethernet/seeq/sgiseeq.c |    1 +
 1 file changed, 1 insertion(+)
drivers/net/ethernet/seeq/sgiseeq.c +1 -0
@@ -735,6 +735,7 @@ static int sgiseeq_probe(struct platform
 	}
 
 	platform_set_drvdata(pdev, dev);
+	SET_NETDEV_DEV(dev, &pdev->dev);
 	sp = netdev_priv(dev);
 
 	/* Make private data page aligned */
author maintainer
#114 lore raw

PATCH 114/137 net: ucc_geth - fix Oops when changing number of buffers in the ring

In reply to Greg Kroah-Hartman

From: Christophe Leroy <[email protected]>

[ Upstream commit ee0df19305d9fabd9479b785918966f6e25b733b ]

When changing the number of buffers in the RX ring while the interface is running, the following Oops is encountered due to the new number of buffers being taken into account immediately while their allocation is done when opening the device only.

[   69.882706] Unable to handle kernel paging request for data at address 0xf0000100
[   69.890172] Faulting instruction address: 0xc033e164
[   69.895122] Oops: Kernel access of bad area, sig: 11 [#1]
[   69.900494] BE PREEMPT CMPCPRO
[   69.907120] CPU: 0 PID: 0 Comm: swapper Not tainted 4.14.115-00006-g179ade8ce3-dirty #269
[   69.915956] task: c0684310 task.stack: c06da000
[   69.920470] NIP:  c033e164 LR: c02e44d0 CTR: c02e41fc
[   69.925504] REGS: dfff1e20 TRAP: 0300   Not tainted  (4.14.115-00006-g179ade8ce3-dirty)
[   69.934161] MSR:  00009032 <EE,ME,IR,DR,RI>  CR: 22004428  XER: 20000000
[   69.940869] DAR: f0000100 DSISR: 20000000
[   69.940869] GPR00: c0352d70 dfff1ed0 c0684310 f00000a4 00000040 dfff1f68 00000000 0000001f
[   69.940869] GPR08: df53f410 1cc00040 00000021 c0781640 42004424 100c82b6 f00000a4 df53f5b0
[   69.940869] GPR16: df53f6c0 c05daf84 00000040 00000000 00000040 c0782be4 00000000 00000001
[   69.940869] GPR24: 00000000 df53f400 000001b0 df53f410 df53f000 0000003f df708220 1cc00044
[   69.978348] NIP [c033e164] skb_put+0x0/0x5c
[   69.982528] LR [c02e44d0] ucc_geth_poll+0x2d4/0x3f8
[   69.987384] Call Trace:
[   69.989830] [dfff1ed0] [c02e4554] ucc_geth_poll+0x358/0x3f8 (unreliable)
[   69.996522] [dfff1f20] [c0352d70] net_rx_action+0x248/0x30c
[   70.002099] [dfff1f80] [c04e93e4] __do_softirq+0xfc/0x310
[   70.007492] [dfff1fe0] [c0021124] irq_exit+0xd0/0xd4
[   70.012458] [dfff1ff0] [c000e7e0] call_do_irq+0x24/0x3c
[   70.017683] [c06dbe80] [c0006bac] do_IRQ+0x64/0xc4
[   70.022474] [c06dbea0] [c001097c] ret_from_except+0x0/0x14
[   70.027964] --- interrupt: 501 at rcu_idle_exit+0x84/0x90
[   70.027964]     LR = rcu_idle_exit+0x74/0x90
[   70.037585] [c06dbf60] [20000000] 0x20000000 (unreliable)
[   70.042984] [c06dbf80] [c004bb0c] do_idle+0xb4/0x11c
[   70.047945] [c06dbfa0] [c004bd14] cpu_startup_entry+0x18/0x1c
[   70.053682] [c06dbfb0] [c05fb034] start_kernel+0x370/0x384
[   70.059153] [c06dbff0] [00003438] 0x3438
[   70.063062] Instruction dump:
[   70.066023] 38a00000 38800000 90010014 4bfff015 80010014 7c0803a6 3123ffff 7c691910
[   70.073767] 38210010 4e800020 38600000 4e800020 <80e3005c> 80c30098 3107ffff 7d083910
[   70.081690] ---[ end trace be7ccd9c1e1a9f12 ]---

This patch forbids the modification of the number of buffers in the ring while the interface is running.

Fixes ac421852b3a0 ("ucc_geth: add ethtool support")
---
 drivers/net/ethernet/freescale/ucc_geth_ethtool.c |    8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)
drivers/net/ethernet/freescale/ucc_geth_ethtool.c +3 -5
@@ -252,14 +252,12 @@ uec_set_ringparam(struct net_device *net
 		return -EINVAL;
 	}
 
+	if (netif_running(netdev))
+		return -EBUSY;
+
 	ug_info->bdRingLenRx[queue] = ring->rx_pending;
 	ug_info->bdRingLenTx[queue] = ring->tx_pending;
 
-	if (netif_running(netdev)) {
-		/* FIXME: restart automatically */
-		netdev_info(netdev, "Please re-open the interface\n");
-	}
-
 	return ret;
 }
 
author maintainer
#115 lore raw

PATCH 115/137 packet: Fix error path in packet_init

In reply to Greg Kroah-Hartman

From: YueHaibing <[email protected]>

[ Upstream commit 36096f2f4fa05f7678bc87397665491700bae757 ]

kernel BUG at lib/list_debug.c:47!
invalid opcode: 0000 [#1

CPU: 0 PID: 12914 Comm: rmmod Tainted: G        W         5.1.0+ #47
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.9.3-0-ge2fc41e-prebuilt.qemu-project.org 04/01/2014
RIP: 0010:__list_del_entry_valid+0x53/0x90
Code: 48 8b 32 48 39 fe 75 35 48 8b 50 08 48 39 f2 75 40 b8 01 00 00 00 5d c3 48
89 fe 48 89 c2 48 c7 c7 18 75 fe 82 e8 cb 34 78 ff <0f> 0b 48 89 fe 48 c7 c7 50 75 fe 82 e8 ba 34 78 ff 0f 0b 48 89 f2
RSP: 0018:ffffc90001c2fe40 EFLAGS: 00010286
RAX: 000000000000004e RBX: ffffffffa0184000 RCX: 0000000000000000
RDX: 0000000000000000 RSI: ffff888237a17788 RDI: 00000000ffffffff
RBP: ffffc90001c2fe40 R08: 0000000000000000 R09: 0000000000000000
R10: ffffc90001c2fe10 R11: 0000000000000000 R12: 0000000000000000
R13: ffffc90001c2fe50 R14: ffffffffa0184000 R15: 0000000000000000
FS:  00007f3d83634540(0000) GS:ffff888237a00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000555c350ea818 CR3: 0000000231677000 CR4: 00000000000006f0
Call Trace:
 unregister_pernet_operations+0x34/0x120
 unregister_pernet_subsys+0x1c/0x30
 packet_exit+0x1c/0x369 [af_packet
 __x64_sys_delete_module+0x156/0x260
 ? lockdep_hardirqs_on+0x133/0x1b0
 ? do_syscall_64+0x12/0x1f0
 do_syscall_64+0x6e/0x1f0
 entry_SYSCALL_64_after_hwframe+0x49/0xbe

When modprobe af_packet, register_pernet_subsys
fails and does a cleanup, ops->list is set to LIST_POISON1, but the module init is considered to success, then while rmmod it, BUG() is triggered in __list_del_entry_valid which is called from unregister_pernet_subsys. This patch fix error handing path in packet_init to avoid possilbe issue if some error occur.

---
 net/packet/af_packet.c |   25 ++++++++++++++++++++-----
 1 file changed, 20 insertions(+), 5 deletions(-)
net/packet/af_packet.c +20 -5
@@ -4604,14 +4604,29 @@ static void __exit packet_exit(void)
 
 static int __init packet_init(void)
 {
-	int rc = proto_register(&packet_proto, 0);
+	int rc;
 
-	if (rc != 0)
+	rc = proto_register(&packet_proto, 0);
+	if (rc)
 		goto out;
+	rc = sock_register(&packet_family_ops);
+	if (rc)
+		goto out_proto;
+	rc = register_pernet_subsys(&packet_net_ops);
+	if (rc)
+		goto out_sock;
+	rc = register_netdevice_notifier(&packet_netdev_notifier);
+	if (rc)
+		goto out_pernet;
 
-	sock_register(&packet_family_ops);
-	register_pernet_subsys(&packet_net_ops);
-	register_netdevice_notifier(&packet_netdev_notifier);
+	return 0;
+
+out_pernet:
+	unregister_pernet_subsys(&packet_net_ops);
+out_sock:
+	sock_unregister(PF_PACKET);
+out_proto:
+	proto_unregister(&packet_proto);
 out:
 	return rc;
 }
author maintainer
#116 lore raw

PATCH 116/137 selinux: do not report error on connect(AF_UNSPEC)

In reply to Greg Kroah-Hartman

From: Paolo Abeni <[email protected]>

[ Upstream commit c7e0d6cca86581092cbbf2cd868b3601495554cf ]

calling connect(AF_UNSPEC) on an already connected TCP socket is an established way to disconnect() such socket. After commit 68741a8adab9 ("selinux: Fix ltp test connect-syscall failure") it no longer works and, in the above scenario connect() fails with EAFNOSUPPORT.

Fix the above falling back to the generic/old code when the address family is not AF_INET{4,6}, but leave the SCTP code path untouched, as it has specific constraints.

Fixes 68741a8adab9 ("selinux: Fix ltp test connect-syscall failure")
---
 security/selinux/hooks.c |    8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
security/selinux/hooks.c +4 -4
@@ -4586,7 +4586,7 @@ static int selinux_socket_connect_helper
 		struct lsm_network_audit net = {0,};
 		struct sockaddr_in *addr4 = NULL;
 		struct sockaddr_in6 *addr6 = NULL;
-		unsigned short snum;
+		unsigned short snum = 0;
 		u32 sid, perm;
 
 		/* sctp_connectx(3) calls via selinux_sctp_bind_connect()
@@ -4609,12 +4609,12 @@ static int selinux_socket_connect_helper
 			break;
 		default:
 			/* Note that SCTP services expect -EINVAL, whereas
-			 * others expect -EAFNOSUPPORT.
+			 * others must handle this at the protocol level:
+			 * connect(AF_UNSPEC) on a connected socket is
+			 * a documented way disconnect the socket.
 			 */
 			if (sksec->sclass == SECCLASS_SCTP_SOCKET)
 				return -EINVAL;
-			else
-				return -EAFNOSUPPORT;
 		}
 
 		err = sel_netport_sid(sk->sk_protocol, snum, &sid);
author maintainer
#117 lore raw

PATCH 117/137 tipc: fix hanging clients using poll with EPOLLOUT flag

In reply to Greg Kroah-Hartman

From: Parthasarathy Bhuvaragan <[email protected]>

[ Upstream commit ff946833b70e0c7f93de9a3f5b329b5ae2287b38 ]

commit 517d7c79bdb398 ("tipc: fix hanging poll() for stream sockets") introduced a regression for clients using non-blocking sockets. After the commit, we send EPOLLOUT event to the client even in TIPC_CONNECTING state. This causes the subsequent send() to fail with ENOTCONN, as the socket is still not in TIPC_ESTABLISHED state.

In this commit, we:
- improve the fix for hanging poll() by replacing sk_data_ready() with sk_state_change() to wake up all clients.
- revert the faulty updates introduced by commit 517d7c79bdb398 ("tipc: fix hanging poll() for stream sockets").

Fixes 517d7c79bdb398 ("tipc: fix hanging poll() for stream sockets")
---
 net/tipc/socket.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
net/tipc/socket.c +2 -2
@@ -734,11 +734,11 @@ static __poll_t tipc_poll(struct file *f
 
 	switch (sk->sk_state) {
 	case TIPC_ESTABLISHED:
-	case TIPC_CONNECTING:
 		if (!tsk->cong_link_cnt && !tsk_conn_cong(tsk))
 			revents |= EPOLLOUT;
 		/* fall thru' */
 	case TIPC_LISTEN:
+	case TIPC_CONNECTING:
 		if (!skb_queue_empty(&sk->sk_receive_queue))
 			revents |= EPOLLIN | EPOLLRDNORM;
 		break;
@@ -2041,7 +2041,7 @@ static bool tipc_sk_filter_connect(struc
 			if (msg_data_sz(hdr))
 				return true;
 			/* Empty ACK-, - wake up sleeping connect() and drop */
-			sk->sk_data_ready(sk);
+			sk->sk_state_change(sk);
 			msg_set_dest_droppable(hdr, 1);
 			return false;
 		}
author maintainer
#118 lore raw

PATCH 118/137 vlan: disable SIOCSHWTSTAMP in container

In reply to Greg Kroah-Hartman

From: Hangbin Liu <[email protected]>

[ Upstream commit 873017af778439f2f8e3d87f28ddb1fcaf244a76 ]

With NET_ADMIN enabled in container, a normal user could be mapped to root and is able to change the real device's rx filter via ioctl on vlan, which would affect the other ptp process on host. Fix it by disabling SIOCSHWTSTAMP in container.

Fixes a6111d3c93d0 ("vlan: Pass SIOC[SG]HWTSTAMP ioctls to real device")
---
 net/8021q/vlan_dev.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
net/8021q/vlan_dev.c +3 -1
@@ -368,10 +368,12 @@ static int vlan_dev_ioctl(struct net_dev
 	ifrr.ifr_ifru = ifr->ifr_ifru;
 
 	switch (cmd) {
+	case SIOCSHWTSTAMP:
+		if (!net_eq(dev_net(dev), &init_net))
+			break;
 	case SIOCGMIIPHY:
 	case SIOCGMIIREG:
 	case SIOCSMIIREG:
-	case SIOCSHWTSTAMP:
 	case SIOCGHWTSTAMP:
 		if (netif_device_present(real_dev) && ops->ndo_do_ioctl)
 			err = ops->ndo_do_ioctl(real_dev, &ifrr, cmd);
author maintainer
#119 lore raw

PATCH 119/137 vrf: sit mtu should not be updated when vrf netdev is the link

In reply to Greg Kroah-Hartman

From: Stephen Suryaputra <[email protected]>

[ Upstream commit ff6ab32bd4e073976e4d8797b4d514a172cfe6cb ]

VRF netdev mtu isn't typically set and have an mtu of 65536. When the link of a tunnel is set, the tunnel mtu is changed from 1480 to the link mtu minus tunnel header. In the case of VRF netdev is the link, then the tunnel mtu becomes 65516. So, fix it by not setting the tunnel mtu in this case.

---
 net/ipv6/sit.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/ipv6/sit.c +1 -1
@@ -1084,7 +1084,7 @@ static void ipip6_tunnel_bind_dev(struct
 	if (!tdev && tunnel->parms.link)
 		tdev = __dev_get_by_index(tunnel->net, tunnel->parms.link);
 
-	if (tdev) {
+	if (tdev && !netif_is_l3_master(tdev)) {
 		int t_hlen = tunnel->hlen + sizeof(struct iphdr);
 
 		dev->hard_header_len = tdev->hard_header_len + sizeof(struct iphdr);
author maintainer
#120 lore raw

PATCH 120/137 aqc111: fix endianness issue in aqc111_change_mtu

In reply to Greg Kroah-Hartman

From: Oliver Neukum <[email protected]>

[ Upstream commit b8b277525e9df2fd2dc3d1f4fe01c6796bb107fc ]

If the MTU is large enough, the first write to the device is just repeated. On BE architectures, however, the first word of the command will be swapped a second time and garbage will be written. Avoid that.

---
 drivers/net/usb/aqc111.c |    2 ++
 1 file changed, 2 insertions(+)
drivers/net/usb/aqc111.c +2 -0
@@ -453,6 +453,8 @@ static int aqc111_change_mtu(struct net_
 		reg16 = 0x1420;
 	else if (dev->net->mtu <= 16334)
 		reg16 = 0x1A20;
+	else
+		return 0;
 
 	aqc111_write16_cmd(dev, AQ_ACCESS_MAC, SFR_PAUSE_WATERLVL_LOW,
 			   2, &reg16);
author maintainer
#121 lore raw

PATCH 121/137 aqc111: fix writing to the phy on BE

In reply to Greg Kroah-Hartman

From: Oliver Neukum <[email protected]>

[ Upstream commit 369b46e9fbcfa5136f2cb5f486c90e5f7fa92630 ]

When writing to the phy on BE architectures an internal data structure was directly given, leading to it being byte swapped in the wrong way for the CPU in 50% of all cases. A temporary buffer must be used.

---
 drivers/net/usb/aqc111.c |   23 +++++++++++++++++------
 1 file changed, 17 insertions(+), 6 deletions(-)
drivers/net/usb/aqc111.c +17 -6
@@ -320,6 +320,7 @@ static int aqc111_get_link_ksettings(str
 static void aqc111_set_phy_speed(struct usbnet *dev, u8 autoneg, u16 speed)
 {
 	struct aqc111_data *aqc111_data = dev->driver_priv;
+	u32 phy_on_the_wire;
 
 	aqc111_data->phy_cfg &= ~AQ_ADV_MASK;
 	aqc111_data->phy_cfg |= AQ_PAUSE;
@@ -361,7 +362,8 @@ static void aqc111_set_phy_speed(struct
 		}
 	}
 
-	aqc111_write32_cmd(dev, AQ_PHY_OPS, 0, 0, &aqc111_data->phy_cfg);
+	phy_on_the_wire = aqc111_data->phy_cfg;
+	aqc111_write32_cmd(dev, AQ_PHY_OPS, 0, 0, &phy_on_the_wire);
 }
 
 static int aqc111_set_link_ksettings(struct net_device *net,
@@ -755,6 +757,7 @@ static void aqc111_unbind(struct usbnet
 {
 	struct aqc111_data *aqc111_data = dev->driver_priv;
 	u16 reg16;
+	u32 phy_on_the_wire;
 
 	/* Force bz */
 	reg16 = SFR_PHYPWR_RSTCTL_BZ;
@@ -768,8 +771,9 @@ static void aqc111_unbind(struct usbnet
 	aqc111_data->phy_cfg &= ~AQ_ADV_MASK;
 	aqc111_data->phy_cfg |= AQ_LOW_POWER;
 	aqc111_data->phy_cfg &= ~AQ_PHY_POWER_EN;
+	phy_on_the_wire = aqc111_data->phy_cfg;
 	aqc111_write32_cmd_nopm(dev, AQ_PHY_OPS, 0, 0,
-				&aqc111_data->phy_cfg);
+				&phy_on_the_wire);
 
 	kfree(aqc111_data);
 }
@@ -992,6 +996,7 @@ static int aqc111_reset(struct usbnet *d
 {
 	struct aqc111_data *aqc111_data = dev->driver_priv;
 	u8 reg8 = 0;
+	u32 phy_on_the_wire;
 
 	dev->rx_urb_size = URB_SIZE;
 
@@ -1004,8 +1009,9 @@ static int aqc111_reset(struct usbnet *d
 
 	/* Power up ethernet PHY */
 	aqc111_data->phy_cfg = AQ_PHY_POWER_EN;
+	phy_on_the_wire = aqc111_data->phy_cfg;
 	aqc111_write32_cmd(dev, AQ_PHY_OPS, 0, 0,
-			   &aqc111_data->phy_cfg);
+			   &phy_on_the_wire);
 
 	/* Set the MAC address */
 	aqc111_write_cmd(dev, AQ_ACCESS_MAC, SFR_NODE_ID, ETH_ALEN,
@@ -1036,6 +1042,7 @@ static int aqc111_stop(struct usbnet *de
 {
 	struct aqc111_data *aqc111_data = dev->driver_priv;
 	u16 reg16 = 0;
+	u32 phy_on_the_wire;
 
 	aqc111_read16_cmd(dev, AQ_ACCESS_MAC, SFR_MEDIUM_STATUS_MODE,
 			  2, &reg16);
@@ -1047,8 +1054,9 @@ static int aqc111_stop(struct usbnet *de
 
 	/* Put PHY to low power*/
 	aqc111_data->phy_cfg |= AQ_LOW_POWER;
+	phy_on_the_wire = aqc111_data->phy_cfg;
 	aqc111_write32_cmd(dev, AQ_PHY_OPS, 0, 0,
-			   &aqc111_data->phy_cfg);
+			   &phy_on_the_wire);
 
 	netif_carrier_off(dev->net);
 
@@ -1324,6 +1332,7 @@ static int aqc111_suspend(struct usb_int
 	u16 temp_rx_ctrl = 0x00;
 	u16 reg16;
 	u8 reg8;
+	u32 phy_on_the_wire;
 
 	usbnet_suspend(intf, message);
 
@@ -1395,12 +1404,14 @@ static int aqc111_suspend(struct usb_int
 
 		aqc111_write_cmd(dev, AQ_WOL_CFG, 0, 0,
 				 WOL_CFG_SIZE, &wol_cfg);
+		phy_on_the_wire = aqc111_data->phy_cfg;
 		aqc111_write32_cmd(dev, AQ_PHY_OPS, 0, 0,
-				   &aqc111_data->phy_cfg);
+				   &phy_on_the_wire);
 	} else {
 		aqc111_data->phy_cfg |= AQ_LOW_POWER;
+		phy_on_the_wire = aqc111_data->phy_cfg;
 		aqc111_write32_cmd(dev, AQ_PHY_OPS, 0, 0,
-				   &aqc111_data->phy_cfg);
+				   &phy_on_the_wire);
 
 		/* Disable RX path */
 		aqc111_read16_cmd_nopm(dev, AQ_ACCESS_MAC,
author maintainer
#122 lore raw

PATCH 122/137 aqc111: fix double endianness swap on BE

In reply to Greg Kroah-Hartman

From: Oliver Neukum <[email protected]>

[ Upstream commit 2cf672709beb005f6e90cb4edbed6f2218ba953e ]

If you are using a function that does a swap in place, you cannot just reuse the buffer on the assumption that it has not been changed.

---
 drivers/net/usb/aqc111.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/net/usb/aqc111.c +4 -2
@@ -1428,7 +1428,7 @@ static int aqc111_resume(struct usb_inte
 {
 	struct usbnet *dev = usb_get_intfdata(intf);
 	struct aqc111_data *aqc111_data = dev->driver_priv;
-	u16 reg16;
+	u16 reg16, oldreg16;
 	u8 reg8;
 
 	netif_carrier_off(dev->net);
@@ -1444,9 +1444,11 @@ static int aqc111_resume(struct usb_inte
 	/* Configure RX control register => start operation */
 	reg16 = aqc111_data->rxctl;
 	reg16 &= ~SFR_RX_CTL_START;
+	/* needs to be saved in case endianness is swapped */
+	oldreg16 = reg16;
 	aqc111_write16_cmd_nopm(dev, AQ_ACCESS_MAC, SFR_RX_CTL, 2, &reg16);
 
-	reg16 |= SFR_RX_CTL_START;
+	reg16 = oldreg16 | SFR_RX_CTL_START;
 	aqc111_write16_cmd_nopm(dev, AQ_ACCESS_MAC, SFR_RX_CTL, 2, &reg16);
 
 	aqc111_set_phy_speed(dev, aqc111_data->autoneg,
author maintainer
#123 lore raw

PATCH 123/137 tuntap: fix dividing by zero in ebpf queue selection

In reply to Greg Kroah-Hartman

From: Jason Wang <[email protected]>

[ Upstream commit a35d310f03a692bf4798eb309a1950a06a150620 ]

We need check if tun->numqueues is zero (e.g for the persist device) before trying to use it for modular arithmetic.

Fixes 96f84061620c6("tun: add eBPF based queue selection method")
---
 drivers/net/tun.c |    7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/net/tun.c +6 -1
@@ -596,13 +596,18 @@ static u16 tun_automq_select_queue(struc
 static u16 tun_ebpf_select_queue(struct tun_struct *tun, struct sk_buff *skb)
 {
 	struct tun_prog *prog;
+	u32 numqueues;
 	u16 ret = 0;
 
+	numqueues = READ_ONCE(tun->numqueues);
+	if (!numqueues)
+		return 0;
+
 	prog = rcu_dereference(tun->steering_prog);
 	if (prog)
 		ret = bpf_prog_run_clear_cb(prog->prog, skb);
 
-	return ret % tun->numqueues;
+	return ret % numqueues;
 }
 
 static u16 tun_select_queue(struct net_device *dev, struct sk_buff *skb,
author maintainer
#124 lore raw

PATCH 124/137 tuntap: synchronize through tfiles array instead of tun->numqueues

In reply to Greg Kroah-Hartman

From: Jason Wang <[email protected]>

[ Upstream commit 9871a9e47a2646fe30ae7fd2e67668a8d30912f6 ]

When a queue(tfile) is detached through __tun_detach(), we move the last enabled tfile to the position where detached one sit but don't NULL out last position. We expect to synchronize the datapath through tun->numqueues. Unfortunately, this won't work since we're lacking sufficient mechanism to order or synchronize the access to tun->numqueues.

To fix this, NULL out the last position during detaching and check RCU protected tfile against NULL instead of checking tun->numqueues in datapath.

Cc YueHaibing <[email protected]>
Cc Cong Wang <[email protected]>
Cc weiyongjun (A) <[email protected]>
Cc Eric Dumazet <[email protected]>
Fixes c8d68e6be1c3b ("tuntap: multiqueue support")
---
 drivers/net/tun.c |    7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/net/tun.c +6 -1
@@ -705,6 +705,8 @@ static void __tun_detach(struct tun_file
 				   tun->tfiles[tun->numqueues - 1]);
 		ntfile = rtnl_dereference(tun->tfiles[index]);
 		ntfile->queue_index = index;
+		rcu_assign_pointer(tun->tfiles[tun->numqueues - 1],
+				   NULL);
 
 		--tun->numqueues;
 		if (clean) {
@@ -1087,7 +1089,7 @@ static netdev_tx_t tun_net_xmit(struct s
 	tfile = rcu_dereference(tun->tfiles[txq]);
 
 	/* Drop packet if interface is not attached */
-	if (txq >= tun->numqueues)
+	if (!tfile)
 		goto drop;
 
 	if (!rcu_dereference(tun->steering_prog))
@@ -1310,6 +1312,7 @@ static int tun_xdp_xmit(struct net_devic
 
 	rcu_read_lock();
 
+resample:
 	numqueues = READ_ONCE(tun->numqueues);
 	if (!numqueues) {
 		rcu_read_unlock();
@@ -1318,6 +1321,8 @@ static int tun_xdp_xmit(struct net_devic
 
 	tfile = rcu_dereference(tun->tfiles[smp_processor_id() %
 					    numqueues]);
+	if (unlikely(!tfile))
+		goto resample;
 
 	spin_lock(&tfile->tx_ring.producer_lock);
 	for (i = 0; i < n; i++) {
author maintainer
#125 lore raw

PATCH 125/137 net: phy: fix phy_validate_pause

In reply to Greg Kroah-Hartman

From: Heiner Kallweit <[email protected]>

[ Upstream commit b4010af981ac8cdf1f7f58eb6b131c482e5dee02 ]

We have valid scenarios where ETHTOOL_LINK_MODE_Pause_BIT doesn't need to be supported. Therefore extend the first check to check for rx_pause being set.

See also phy_set_asym_pause:
rx=0 and tx=1: advertise asym pause only
rx=0 and tx=0: stop advertising both pause modes

The fixed commit isn't wrong, it's just the one that introduced the linkmode bitmaps.

Fixes 3c1bcc8614db ("net: ethernet: Convert phydev advertize and supported from u32 to link mode")
---
 drivers/net/phy/phy_device.c |   11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)
drivers/net/phy/phy_device.c +7 -4
@@ -2083,11 +2083,14 @@ bool phy_validate_pause(struct phy_devic
 			struct ethtool_pauseparam *pp)
 {
 	if (!linkmode_test_bit(ETHTOOL_LINK_MODE_Pause_BIT,
-			       phydev->supported) ||
-	    (!linkmode_test_bit(ETHTOOL_LINK_MODE_Asym_Pause_BIT,
-				phydev->supported) &&
-	     pp->rx_pause != pp->tx_pause))
+			       phydev->supported) && pp->rx_pause)
 		return false;
+
+	if (!linkmode_test_bit(ETHTOOL_LINK_MODE_Asym_Pause_BIT,
+			       phydev->supported) &&
+	    pp->rx_pause != pp->tx_pause)
+		return false;
+
 	return true;
 }
 EXPORT_SYMBOL(phy_validate_pause);
author maintainer
#126 lore raw

PATCH 126/137 flow_dissector: disable preemption around BPF calls

In reply to Greg Kroah-Hartman

From: Eric Dumazet <[email protected]>

[ Upstream commit b1c17a9a353878602fd5bfe9103e4afe5e9a3f96 ]

Various things in eBPF really require us to disable preemption before running an eBPF program.

syzbot reported :

BUG: assuming atomic context at net/core/flow_dissector.c:737

in_atomic(): 0, irqs_disabled(): 0, pid: 24710, name: syz-executor.3
2 locks held by syz-executor.3/24710:
#0: 00000000e81a4bf1 (&tfile->napi_mutex){+.+.}, at: tun_get_user+0x168e/0x3ff0 drivers/net/tun.c:1850 #1: 00000000254afebd (rcu_read_lock){....}, at: __skb_flow_dissect+0x1e1/0x4bb0 net/core/flow_dissector.c:822

CPU: 1 PID: 24710 Comm: syz-executor.3 Not tainted 5.1.0+ #6
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
 __dump_stack lib/dump_stack.c:77 [inline]
 dump_stack+0x172/0x1f0 lib/dump_stack.c:113
 __cant_sleep kernel/sched/core.c:6165 [inline]
 __cant_sleep.cold+0xa3/0xbb kernel/sched/core.c:6142
 bpf_flow_dissect+0xfe/0x390 net/core/flow_dissector.c:737
 __skb_flow_dissect+0x362/0x4bb0 net/core/flow_dissector.c:853
 skb_flow_dissect_flow_keys_basic include/linux/skbuff.h:1322 [inline]
 skb_probe_transport_header include/linux/skbuff.h:2500 [inline]
 skb_probe_transport_header include/linux/skbuff.h:2493 [inline]
 tun_get_user+0x2cfe/0x3ff0 drivers/net/tun.c:1940
 tun_chr_write_iter+0xbd/0x156 drivers/net/tun.c:2037
 call_write_iter include/linux/fs.h:1872 [inline]
 do_iter_readv_writev+0x5fd/0x900 fs/read_write.c:693
 do_iter_write fs/read_write.c:970 [inline]
 do_iter_write+0x184/0x610 fs/read_write.c:951
 vfs_writev+0x1b3/0x2f0 fs/read_write.c:1015
 do_writev+0x15b/0x330 fs/read_write.c:1058
 __do_sys_writev fs/read_write.c:1131 [inline]
 __se_sys_writev fs/read_write.c:1128 [inline]
 __x64_sys_writev+0x75/0xb0 fs/read_write.c:1128
 do_syscall_64+0x103/0x670 arch/x86/entry/common.c:298
 entry_SYSCALL_64_after_hwframe+0x49/0xbe
Fixes d58e468b1112 ("flow_dissector: implements flow dissector BPF hook")
Cc Petar Penkov <[email protected]>
Cc Stanislav Fomichev <[email protected]>
---
 net/core/flow_dissector.c |    3 +++
 1 file changed, 3 insertions(+)
net/core/flow_dissector.c +3 -0
@@ -786,7 +786,10 @@ bool __skb_flow_dissect(const struct sk_
 		flow_keys.thoff = nhoff;
 
 		bpf_compute_data_pointers((struct sk_buff *)skb);
+
+		preempt_disable();
 		result = BPF_PROG_RUN(attached, skb);
+		preempt_enable();
 
 		/* Restore state */
 		memcpy(cb, &cb_saved, sizeof(cb_saved));
author maintainer
#127 lore raw

PATCH 127/137 isdn: bas_gigaset: use usb_fill_int_urb() properly

In reply to Greg Kroah-Hartman

From: Paul Bolle <[email protected]>

[ Upstream commit 4014dfae3ccaaf3ec19c9ae0691a3f14e7132eae ]

The switch to make bas_gigaset use usb_fill_int_urb() - instead of filling that urb "by hand" - missed the subtle ordering of the previous code.

See, before the switch urb->dev was set to a member somewhere deep in a complicated structure and then supplied to usb_rcvisocpipe() and usb_sndisocpipe(). After that switch urb->dev wasn't set to anything specific before being supplied to those two macros. This triggers a nasty oops:

    BUG: unable to handle kernel NULL pointer dereference at 00000000
    #PF error: [normal kernel read fault]
    *pde = 00000000
    Oops: 0000 [#1] SMP
    CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.1.0-0.rc4.1.local0.fc28.i686 #1
    Hardware name: IBM 2525FAG/2525FAG, BIOS 74ET64WW (2.09 ) 12/14/2006
    EIP: gigaset_init_bchannel+0x89/0x320 [bas_gigaset]
    Code: 75 07 83 8b 84 00 00 00 40 8d 47 74 c7 07 01 00 00 00 89 45 f0 8b 44 b7 68 85 c0 0f 84 6a 02 00 00 8b 48 28 8b 93 88 00 00 00 <8b> 09 8d 54 12 03 c1 e2 0f c1 e1 08 09 ca 8b 8b 8c 00 00 00 80 ca
    EAX: f05ec200 EBX: ed404200 ECX: 00000000 EDX: 00000000
    ESI: 00000000 EDI: f065a000 EBP: f30c9f40 ESP: f30c9f20
    DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068 EFLAGS: 00010086
    CR0: 80050033 CR2: 00000000 CR3: 0ddc7000 CR4: 000006d0
    Call Trace:
     <SOFTIRQ>
     ? gigaset_isdn_connD+0xf6/0x140 [gigaset]
     gigaset_handle_event+0x173e/0x1b90 [gigaset]
     tasklet_action_common.isra.16+0x4e/0xf0
     tasklet_action+0x1e/0x20
     __do_softirq+0xb2/0x293
     ? __irqentry_text_end+0x3/0x3
     call_on_stack+0x45/0x50
     </SOFTIRQ>
     ? irq_exit+0xb5/0xc0
     ? do_IRQ+0x78/0xd0
     ? acpi_idle_enter_s2idle+0x50/0x50
     ? common_interrupt+0xd4/0xdc
     ? acpi_idle_enter_s2idle+0x50/0x50
     ? sched_cpu_activate+0x1b/0xf0
     ? acpi_fan_resume.cold.7+0x9/0x18
     ? cpuidle_enter_state+0x152/0x4c0
     ? cpuidle_enter+0x14/0x20
     ? call_cpuidle+0x21/0x40
     ? do_idle+0x1c8/0x200
     ? cpu_startup_entry+0x25/0x30
     ? rest_init+0x88/0x8a
     ? arch_call_rest_init+0xd/0x19
     ? start_kernel+0x42f/0x448
     ? i386_start_kernel+0xac/0xb0
     ? startup_32_smp+0x164/0x168
    Modules linked in: ppp_generic slhc capi bas_gigaset gigaset kernelcapi nf_conntrack_netbios_ns nf_conntrack_broadcast xt_CT ip6t_rpfilter ip6t_REJECT nf_reject_ipv6 xt_conntrack ip_set nfnetlink ebtable_nat ebtable_broute bridge stp llc ip6table_nat ip6table_mangle ip6table_raw ip6table_security iptable_nat nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 libcrc32c iptable_mangle iptable_raw iptable_security ebtable_filter ebtables ip6table_filter ip6_tables sunrpc ipw2200 iTCO_wdt gpio_ich snd_intel8x0 libipw iTCO_vendor_support snd_ac97_codec lib80211 ppdev ac97_bus snd_seq cfg80211 snd_seq_device pcspkr thinkpad_acpi lpc_ich snd_pcm i2c_i801 snd_timer ledtrig_audio snd soundcore rfkill parport_pc parport pcc_cpufreq acpi_cpufreq i915 i2c_algo_bit drm_kms_helper syscopyarea sysfillrect sdhci_pci sysimgblt cqhci fb_sys_fops drm sdhci mmc_core tg3 ata_generic serio_raw yenta_socket pata_acpi video
    CR2: 0000000000000000
    ---[ end trace 1fe07487b9200c73 ]---
    EIP: gigaset_init_bchannel+0x89/0x320 [bas_gigaset]
    Code: 75 07 83 8b 84 00 00 00 40 8d 47 74 c7 07 01 00 00 00 89 45 f0 8b 44 b7 68 85 c0 0f 84 6a 02 00 00 8b 48 28 8b 93 88 00 00 00 <8b> 09 8d 54 12 03 c1 e2 0f c1 e1 08 09 ca 8b 8b 8c 00 00 00 80 ca
    EAX: f05ec200 EBX: ed404200 ECX: 00000000 EDX: 00000000
    ESI: 00000000 EDI: f065a000 EBP: f30c9f40 ESP: cddcb3bc
    DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068 EFLAGS: 00010086
    CR0: 80050033 CR2: 00000000 CR3: 0ddc7000 CR4: 000006d0
    Kernel panic - not syncing: Fatal exception in interrupt
    Kernel Offset: 0xcc00000 from 0xc0400000 (relocation range: 0xc0000000-0xf6ffdfff)
    ---[ end Kernel panic - not syncing: Fatal exception in interrupt ]---

No-one noticed because this Oops is apparently only triggered by setting up an ISDN data connection on a live ISDN line on a gigaset base (ie, the PBX that the gigaset driver support). Very few people do that running present day kernels.

Anyhow, a little code reorganization makes this problem go away, while avoiding the subtle ordering that was used in the past. So let's do that.

Fixes 78c696c19578 ("isdn: gigaset: use usb_fill_int_urb()")
---
 drivers/isdn/gigaset/bas-gigaset.c |    9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)
drivers/isdn/gigaset/bas-gigaset.c +5 -4
@@ -958,6 +958,7 @@ static void write_iso_callback(struct ur
  */
 static int starturbs(struct bc_state *bcs)
 {
+	struct usb_device *udev = bcs->cs->hw.bas->udev;
 	struct bas_bc_state *ubc = bcs->hw.bas;
 	struct urb *urb;
 	int j, k;
@@ -975,8 +976,8 @@ static int starturbs(struct bc_state *bc
 			rc = -EFAULT;
 			goto error;
 		}
-		usb_fill_int_urb(urb, bcs->cs->hw.bas->udev,
-				 usb_rcvisocpipe(urb->dev, 3 + 2 * bcs->channel),
+		usb_fill_int_urb(urb, udev,
+				 usb_rcvisocpipe(udev, 3 + 2 * bcs->channel),
 				 ubc->isoinbuf + k * BAS_INBUFSIZE,
 				 BAS_INBUFSIZE, read_iso_callback, bcs,
 				 BAS_FRAMETIME);
@@ -1006,8 +1007,8 @@ static int starturbs(struct bc_state *bc
 			rc = -EFAULT;
 			goto error;
 		}
-		usb_fill_int_urb(urb, bcs->cs->hw.bas->udev,
-				 usb_sndisocpipe(urb->dev, 4 + 2 * bcs->channel),
+		usb_fill_int_urb(urb, udev,
+				 usb_sndisocpipe(udev, 4 + 2 * bcs->channel),
 				 ubc->isooutbuf->data,
 				 sizeof(ubc->isooutbuf->data),
 				 write_iso_callback, &ubc->isoouturbs[k],
author maintainer
#128 lore raw

PATCH 128/137 drivers/virt/fsl_hypervisor.c: dereferencing error pointers in ioctl

In reply to Greg Kroah-Hartman

From: Dan Carpenter <[email protected]>

commit c8ea3663f7a8e6996d44500ee818c9330ac4fd88 upstream.

strndup_user() returns error pointers on error, and then in the error handling we pass the error pointers to kfree(). It will cause an Oops.

Fixes 6db7199407ca ("drivers/virt: introduce Freescale hypervisor management driver")
Cc Timur Tabi <[email protected]>
Cc Mihai Caraman <[email protected]>
Cc Kumar Gala <[email protected]>
---
 drivers/virt/fsl_hypervisor.c |   26 +++++++++++++-------------
 1 file changed, 13 insertions(+), 13 deletions(-)
drivers/virt/fsl_hypervisor.c +13 -13
@@ -331,8 +331,8 @@ static long ioctl_dtprop(struct fsl_hv_i
 	struct fsl_hv_ioctl_prop param;
 	char __user *upath, *upropname;
 	void __user *upropval;
-	char *path = NULL, *propname = NULL;
-	void *propval = NULL;
+	char *path, *propname;
+	void *propval;
 	int ret = 0;
 
 	/* Get the parameters from the user. */
@@ -344,32 +344,30 @@ static long ioctl_dtprop(struct fsl_hv_i
 	upropval = (void __user *)(uintptr_t)param.propval;
 
 	path = strndup_user(upath, FH_DTPROP_MAX_PATHLEN);
-	if (IS_ERR(path)) {
-		ret = PTR_ERR(path);
-		goto out;
-	}
+	if (IS_ERR(path))
+		return PTR_ERR(path);
 
 	propname = strndup_user(upropname, FH_DTPROP_MAX_PATHLEN);
 	if (IS_ERR(propname)) {
 		ret = PTR_ERR(propname);
-		goto out;
+		goto err_free_path;
 	}
 
 	if (param.proplen > FH_DTPROP_MAX_PROPLEN) {
 		ret = -EINVAL;
-		goto out;
+		goto err_free_propname;
 	}
 
 	propval = kmalloc(param.proplen, GFP_KERNEL);
 	if (!propval) {
 		ret = -ENOMEM;
-		goto out;
+		goto err_free_propname;
 	}
 
 	if (set) {
 		if (copy_from_user(propval, upropval, param.proplen)) {
 			ret = -EFAULT;
-			goto out;
+			goto err_free_propval;
 		}
 
 		param.ret = fh_partition_set_dtprop(param.handle,
@@ -388,7 +386,7 @@ static long ioctl_dtprop(struct fsl_hv_i
 			if (copy_to_user(upropval, propval, param.proplen) ||
 			    put_user(param.proplen, &p->proplen)) {
 				ret = -EFAULT;
-				goto out;
+				goto err_free_propval;
 			}
 		}
 	}
@@ -396,10 +394,12 @@ static long ioctl_dtprop(struct fsl_hv_i
 	if (put_user(param.ret, &p->ret))
 		ret = -EFAULT;
 
-out:
-	kfree(path);
+err_free_propval:
 	kfree(propval);
+err_free_propname:
 	kfree(propname);
+err_free_path:
+	kfree(path);
 
 	return ret;
 }
author maintainer
#129 lore raw

PATCH 129/137 drivers/virt/fsl_hypervisor.c: prevent integer overflow in ioctl

In reply to Greg Kroah-Hartman

From: Dan Carpenter <[email protected]>

commit 6a024330650e24556b8a18cc654ad00cfecf6c6c upstream.

The "param.count" value is a u64 thatcomes from the user. The code later in the function assumes that param.count is at least one and if it's not then it leads to an Oops when we dereference the ZERO_SIZE_PTR.

Also the addition can have an integer overflow which would lead us to allocate a smaller "pages" array than required. I can't immediately tell what the possible run times implications are, but it's safest to prevent the overflow.

Fixes 6db7199407ca ("drivers/virt: introduce Freescale hypervisor management driver")
Cc Timur Tabi <[email protected]>
Cc Mihai Caraman <[email protected]>
Cc Kumar Gala <[email protected]>
---
 drivers/virt/fsl_hypervisor.c |    3 +++
 1 file changed, 3 insertions(+)
drivers/virt/fsl_hypervisor.c +3 -0
@@ -215,6 +215,9 @@ static long ioctl_memcpy(struct fsl_hv_i
 	 * hypervisor.
 	 */
 	lb_offset = param.local_vaddr & (PAGE_SIZE - 1);
+	if (param.count == 0 ||
+	    param.count > U64_MAX - lb_offset - PAGE_SIZE + 1)
+		return -EINVAL;
 	num_pages = (param.count + lb_offset + PAGE_SIZE - 1) >> PAGE_SHIFT;
 
 	/* Allocate the buffers we need */
author maintainer
#130 lore raw

PATCH 130/137 powerpc/book3s/64: check for NULL pointer in pgd_alloc()

In reply to Greg Kroah-Hartman

From: Rick Lindsley <[email protected]>

commit f39356261c265a0689d7ee568132d516e8b6cecc upstream.

When the memset code was added to pgd_alloc(), it failed to consider that kmem_cache_alloc() can return NULL. It's uncommon, but not impossible under heavy memory contention. Example oops:

  Unable to handle kernel paging request for data at address 0x00000000
  Faulting instruction address: 0xc0000000000a4000
  Oops: Kernel access of bad area, sig: 11 [#1]
  LE SMP NR_CPUS=2048 NUMA pSeries
  CPU: 70 PID: 48471 Comm: entrypoint.sh Kdump: loaded Not tainted 4.14.0-115.6.1.el7a.ppc64le #1
  task: c000000334a00000 task.stack: c000000331c00000
  NIP:  c0000000000a4000 LR: c00000000012f43c CTR: 0000000000000020
  REGS: c000000331c039c0 TRAP: 0300   Not tainted  (4.14.0-115.6.1.el7a.ppc64le)
  MSR:  800000010280b033 <SF,VEC,VSX,EE,FP,ME,IR,DR,RI,LE,TM[E]>  CR: 44022840  XER: 20040000
  CFAR: c000000000008874 DAR: 0000000000000000 DSISR: 42000000 SOFTE: 1
  ...
  NIP [c0000000000a4000] memset+0x68/0x104
  LR [c00000000012f43c] mm_init+0x27c/0x2f0
  Call Trace:
    mm_init+0x260/0x2f0 (unreliable)
    copy_mm+0x11c/0x638
    copy_process.isra.28.part.29+0x6fc/0x1080
    _do_fork+0xdc/0x4c0
    ppc_clone+0x8/0xc
  Instruction dump:
  409e000c b0860000 38c60002 409d000c 90860000 38c60004 78a0d183 78a506a0
  7c0903a6 41820034 60000000 60420000 <f8860000> f8860008 f8860010 f8860018
Fixes fc5c2f4a55a2 ("powerpc/mm/hash64: Zero PGD pages on allocation")
Cc [email protected] # v4.16+
---
 arch/powerpc/include/asm/book3s/64/pgalloc.h |    3 +++
 1 file changed, 3 insertions(+)
arch/powerpc/include/asm/book3s/64/pgalloc.h +3 -0
@@ -81,6 +81,9 @@ static inline pgd_t *pgd_alloc(struct mm
 
 	pgd = kmem_cache_alloc(PGT_CACHE(PGD_INDEX_SIZE),
 			       pgtable_gfp_flags(mm, GFP_KERNEL));
+	if (unlikely(!pgd))
+		return pgd;
+
 	/*
 	 * Don't scan the PGD for pointers, it contains references to PUDs but
 	 * those references are not full pointers and so can't be recognised by
author maintainer
#131 lore raw

PATCH 131/137 powerpc/powernv/idle: Restore IAMR after idle

In reply to Greg Kroah-Hartman

From: Russell Currey <[email protected]>

commit a3f3072db6cad40895c585dce65e36aab997f042 upstream.

Without restoring the IAMR after idle, execution prevention on POWER9 with Radix MMU is overwritten and the kernel can freely execute userspace without faulting.

This is necessary when returning from any stop state that modifies user state, as well as hypervisor state.

To test how this fails without this patch, load the lkdtm driver and do the following:

  $ echo EXEC_USERSPACE > /sys/kernel/debug/provoke-crash/DIRECT

which won't fault, then boot the kernel with powersave=off, where it will fault. Applying this patch will fix this.

Fixes 3b10d0095a1e ("powerpc/mm/radix: Prevent kernel execution of user space")
Cc [email protected] # v4.10+
---
 arch/powerpc/kernel/idle_book3s.S |   20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)
arch/powerpc/kernel/idle_book3s.S +20 -0
@@ -170,6 +170,9 @@ core_idle_lock_held:
 	bne-	core_idle_lock_held
 	blr
 
+/* Reuse an unused pt_regs slot for IAMR */
+#define PNV_POWERSAVE_IAMR	_DAR
+
 /*
  * Pass requested state in r3:
  *	r3 - PNV_THREAD_NAP/SLEEP/WINKLE in POWER8
@@ -200,6 +203,12 @@ pnv_powersave_common:
 	/* Continue saving state */
 	SAVE_GPR(2, r1)
 	SAVE_NVGPRS(r1)
+
+BEGIN_FTR_SECTION
+	mfspr	r5, SPRN_IAMR
+	std	r5, PNV_POWERSAVE_IAMR(r1)
+END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S)
+
 	mfcr	r5
 	std	r5,_CCR(r1)
 	std	r1,PACAR1(r13)
@@ -924,6 +933,17 @@ BEGIN_FTR_SECTION
 END_FTR_SECTION_IFSET(CPU_FTR_HVMODE)
 	REST_NVGPRS(r1)
 	REST_GPR(2, r1)
+
+BEGIN_FTR_SECTION
+	/* IAMR was saved in pnv_powersave_common() */
+	ld	r5, PNV_POWERSAVE_IAMR(r1)
+	mtspr	SPRN_IAMR, r5
+	/*
+	 * We don't need an isync here because the upcoming mtmsrd is
+	 * execution synchronizing.
+	 */
+END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S)
+
 	ld	r4,PACAKMSR(r13)
 	ld	r5,_LINK(r1)
 	ld	r6,_CCR(r1)
author maintainer
#132 lore raw

PATCH 132/137 powerpc/booke64: set RI in default MSR

In reply to Greg Kroah-Hartman

From: Laurentiu Tudor <[email protected]>

commit 5266e58d6cd90ac85c187d673093ad9cb649e16d upstream.

Set RI in the default kernel's MSR so that the architected way of detecting unrecoverable machine check interrupts has a chance to work. This is inline with the MSR setup of the rest of booke powerpc architectures configured here.

---
 arch/powerpc/include/asm/reg_booke.h |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/powerpc/include/asm/reg_booke.h +1 -1
@@ -41,7 +41,7 @@
 #if defined(CONFIG_PPC_BOOK3E_64)
 #define MSR_64BIT	MSR_CM
 
-#define MSR_		(MSR_ME | MSR_CE)
+#define MSR_		(MSR_ME | MSR_RI | MSR_CE)
 #define MSR_KERNEL	(MSR_ | MSR_64BIT)
 #define MSR_USER32	(MSR_ | MSR_PR | MSR_EE)
 #define MSR_USER64	(MSR_USER32 | MSR_64BIT)
author maintainer
#133 lore raw

PATCH 133/137 virtio_ring: Fix potential mem leak in virtqueue_add_indirect_packed

In reply to Greg Kroah-Hartman

From: YueHaibing <[email protected]>

commit df0bfe7501e9319546ea380d39674a4179e059c3 upstream.

'desc' should be freed before leaving from err handing path.

Fixes 1ce9e6055fa0 ("virtio_ring: introduce packed ring support")

[email protected]

---
 drivers/virtio/virtio_ring.c |    1 +
 1 file changed, 1 insertion(+)
drivers/virtio/virtio_ring.c +1 -0
@@ -993,6 +993,7 @@ static int virtqueue_add_indirect_packed
 
 	if (unlikely(vq->vq.num_free < 1)) {
 		pr_debug("Can't add buf len 1 - avail = 0\n");
+		kfree(desc);
 		END_USE(vq);
 		return -ENOSPC;
 	}
author maintainer
#134 lore raw

PATCH 134/137 PCI: hv: Fix a memory leak in hv_eject_device_work()

In reply to Greg Kroah-Hartman

From: Dexuan Cui <[email protected]>

commit 05f151a73ec2b23ffbff706e5203e729a995cdc2 upstream.

When a device is created in new_pcichild_device(), hpdev->refs is set to 2 (i.e. the initial value of 1 plus the get_pcichild()).

When we hot remove the device from the host, in a Linux VM we first call hv_pci_eject_device(), which increases hpdev->refs by get_pcichild() and then schedules a work of hv_eject_device_work(), so hpdev->refs becomes 3 (let's ignore the paired get/put_pcichild() in other places). But in hv_eject_device_work(), currently we only call put_pcichild() twice, meaning the 'hpdev' struct can't be freed in put_pcichild().

Add one put_pcichild() to fix the memory leak.

The device can also be removed when we run "rmmod pci-hyperv". On this path (hv_pci_remove() -> hv_pci_bus_exit() -> hv_pci_devices_present()), hpdev->refs is 2, and we do correctly call put_pcichild() twice in pci_devices_present_work().

Fixes 4daace0d8ce8 ("PCI: hv: Add paravirtual PCI front-end for Microsoft Hyper-V VMs")
[[email protected]: commit log rework]
---
 drivers/pci/controller/pci-hyperv.c |    3 +++
 1 file changed, 3 insertions(+)
drivers/pci/controller/pci-hyperv.c +3 -0
@@ -1905,6 +1905,9 @@ static void hv_eject_device_work(struct
 			 sizeof(*ejct_pkt), (unsigned long)&ctxt.pkt,
 			 VM_PKT_DATA_INBAND, 0);
 
+	/* For the get_pcichild() in hv_pci_eject_device() */
+	put_pcichild(hpdev);
+	/* For the two refs got in new_pcichild_device() */
 	put_pcichild(hpdev);
 	put_pcichild(hpdev);
 	put_hvpcibus(hpdev->hbus);
author maintainer
#135 lore raw

PATCH 135/137 PCI: hv: Add hv_pci_remove_slots() when we unload the driver

In reply to Greg Kroah-Hartman

From: Dexuan Cui <[email protected]>

commit 15becc2b56c6eda3d9bf5ae993bafd5661c1fad1 upstream.

When we unload the pci-hyperv host controller driver, the host does not send us a PCI_EJECT message.

In this case we also need to make sure the sysfs PCI slot directory is removed, otherwise a command on a slot file eg:

"cat /sys/bus/pci/slots/2/address"

will trigger a

"BUG: unable to handle kernel paging request"

and, if we unload/reload the driver several times we would end up with stale slot entries in PCI slot directories in /sys/bus/pci/slots/

root@localhost:~# ls -rtl  /sys/bus/pci/slots/
total 0
drwxr-xr-x 2 root root 0 Feb  7 10:49 2
drwxr-xr-x 2 root root 0 Feb  7 10:49 2-1
drwxr-xr-x 2 root root 0 Feb  7 10:51 2-2

Add the missing code to remove the PCI slot and fix the current behaviour.

Fixes a15f2c08c708 ("PCI: hv: support reporting serial number as slot information")
[[email protected]: reformatted the log]
---
 drivers/pci/controller/pci-hyperv.c |   16 ++++++++++++++++
 1 file changed, 16 insertions(+)
drivers/pci/controller/pci-hyperv.c +16 -0
@@ -1491,6 +1491,21 @@ static void hv_pci_assign_slots(struct h
 	}
 }
 
+/*
+ * Remove entries in sysfs pci slot directory.
+ */
+static void hv_pci_remove_slots(struct hv_pcibus_device *hbus)
+{
+	struct hv_pci_dev *hpdev;
+
+	list_for_each_entry(hpdev, &hbus->children, list_entry) {
+		if (!hpdev->pci_slot)
+			continue;
+		pci_destroy_slot(hpdev->pci_slot);
+		hpdev->pci_slot = NULL;
+	}
+}
+
 /**
  * create_root_hv_pci_bus() - Expose a new root PCI bus
  * @hbus:	Root PCI bus, as understood by this driver
@@ -2685,6 +2700,7 @@ static int hv_pci_remove(struct hv_devic
 		pci_lock_rescan_remove();
 		pci_stop_root_bus(hbus->pci_bus);
 		pci_remove_root_bus(hbus->pci_bus);
+		hv_pci_remove_slots(hbus);
 		pci_unlock_rescan_remove();
 		hbus->state = hv_pcibus_removed;
 	}
author maintainer
#136 lore raw

PATCH 136/137 PCI: hv: Add pci_destroy_slot() in pci_devices_present_work(), if necessary

In reply to Greg Kroah-Hartman

From: Dexuan Cui <[email protected]>

commit 340d455699400f2c2c0f9b3f703ade3085cdb501 upstream.

When we hot-remove a device, usually the host sends us a PCI_EJECT message, and a PCI_BUS_RELATIONS message with bus_rel->device_count == 0.

When we execute the quick hot-add/hot-remove test, the host may not send us the PCI_EJECT message if the guest has not fully finished the initialization by sending the PCI_RESOURCES_ASSIGNED* message to the host, so it's potentially unsafe to only depend on the pci_destroy_slot() in hv_eject_device_work() because the code path

create_root_hv_pci_bus()
-> hv_pci_assign_slots()

is not called in this case. Note: in this case, the host still sends the guest a PCI_BUS_RELATIONS message with bus_rel->device_count == 0.

In the quick hot-add/hot-remove test, we can have such a race before the code path

pci_devices_present_work()
-> new_pcichild_device()

adds the new device into the hbus->children list, we may have already received the PCI_EJECT message, and since the tasklet handler

hv_pci_onchannelcallback()

may fail to find the "hpdev" by calling

get_pcichild_wslot(hbus, dev_message->wslot.slot)

hv_pci_eject_device() is not called; Later, by continuing execution

create_root_hv_pci_bus()
-> hv_pci_assign_slots()

creates the slot and the PCI_BUS_RELATIONS message with bus_rel->device_count == 0 removes the device from hbus->children, and we end up being unable to remove the slot in

hv_pci_remove()
-> hv_pci_remove_slots()

Remove the slot in pci_devices_present_work() when the device is removed to address this race.

pci_devices_present_work() and hv_eject_device_work() run in the singled-threaded hbus->wq, so there is not a double-remove issue for the slot.

We cannot offload hv_pci_eject_device() from hv_pci_onchannelcallback() to the workqueue, because we need the hv_pci_onchannelcallback() synchronously call hv_pci_eject_device() to poll the channel ringbuffer to work around the "hangs in hv_compose_msi_msg()" issue fixed in commit de0aa7b2f97d ("PCI: hv: Fix 2 hang issues in hv_compose_msi_msg()")

Fixes a15f2c08c708 ("PCI: hv: support reporting serial number as slot information")
[[email protected]: rewritten commit log]
---
 drivers/pci/controller/pci-hyperv.c |    4 ++++
 1 file changed, 4 insertions(+)
drivers/pci/controller/pci-hyperv.c +4 -0
@@ -1781,6 +1781,10 @@ static void pci_devices_present_work(str
 		hpdev = list_first_entry(&removed, struct hv_pci_dev,
 					 list_entry);
 		list_del(&hpdev->list_entry);
+
+		if (hpdev->pci_slot)
+			pci_destroy_slot(hpdev->pci_slot);
+
 		put_pcichild(hpdev);
 	}
 
author maintainer
#137 lore raw

PATCH 137/137 f2fs: Fix use of number of devices

In reply to Greg Kroah-Hartman

From: Damien Le Moal <[email protected]>

commit 0916878da355650d7e77104a7ac0fa1784eca852 upstream.

For a single device mount using a zoned block device, the zone information for the device is stored in the sbi->devs single entry array and sbi->s_ndevs is set to 1. This differs from a single device mount using a regular block device which does not allocate sbi->devs and sets sbi->s_ndevs to 0.

However, sbi->s_devs == 0 condition is used throughout the code to differentiate a single device mount from a multi-device mount where sbi->s_ndevs is always larger than 1. This results in problems with single zoned block device volumes as these are treated as multi-device mounts but do not have the start_blk and end_blk information set. One of the problem observed is skipping of zone discard issuing resulting in write commands being issued to full zones or unaligned to a zone write pointer.

Fix this problem by simply treating the cases sbi->s_ndevs == 0 (single regular block device mount) and sbi->s_ndevs == 1 (single zoned block device mount) in the same manner. This is done by introducing the helper function f2fs_is_multi_device() and using this helper in place of direct tests of sbi->s_ndevs value, improving code readability.

Fixes 7bb3a371d199 ("f2fs: Fix zoned block device support")
---
 fs/f2fs/data.c    |   17 +++++++++++------
 fs/f2fs/f2fs.h    |   13 ++++++++++++-
 fs/f2fs/file.c    |    2 +-
 fs/f2fs/gc.c      |    2 +-
 fs/f2fs/segment.c |   13 +++++++------
 5 files changed, 32 insertions(+), 15 deletions(-)
fs/f2fs/data.c +11 -6
@@ -218,12 +218,14 @@ struct block_device *f2fs_target_device(
 	struct block_device *bdev = sbi->sb->s_bdev;
 	int i;
 
-	for (i = 0; i < sbi->s_ndevs; i++) {
-		if (FDEV(i).start_blk <= blk_addr &&
-					FDEV(i).end_blk >= blk_addr) {
-			blk_addr -= FDEV(i).start_blk;
-			bdev = FDEV(i).bdev;
-			break;
+	if (f2fs_is_multi_device(sbi)) {
+		for (i = 0; i < sbi->s_ndevs; i++) {
+			if (FDEV(i).start_blk <= blk_addr &&
+			    FDEV(i).end_blk >= blk_addr) {
+				blk_addr -= FDEV(i).start_blk;
+				bdev = FDEV(i).bdev;
+				break;
+			}
 		}
 	}
 	if (bio) {
@@ -237,6 +239,9 @@ int f2fs_target_device_index(struct f2fs
 {
 	int i;
 
+	if (!f2fs_is_multi_device(sbi))
+		return 0;
+
 	for (i = 0; i < sbi->s_ndevs; i++)
 		if (FDEV(i).start_blk <= blkaddr && FDEV(i).end_blk >= blkaddr)
 			return i;
fs/f2fs/f2fs.h +12 -1
@@ -1364,6 +1364,17 @@ static inline bool time_to_inject(struct
 }
 #endif
 
+/*
+ * Test if the mounted volume is a multi-device volume.
+ *   - For a single regular disk volume, sbi->s_ndevs is 0.
+ *   - For a single zoned disk volume, sbi->s_ndevs is 1.
+ *   - For a multi-device volume, sbi->s_ndevs is always 2 or more.
+ */
+static inline bool f2fs_is_multi_device(struct f2fs_sb_info *sbi)
+{
+	return sbi->s_ndevs > 1;
+}
+
 /* For write statistics. Suppose sector size is 512 bytes,
  * and the return value is in kbytes. s is of struct f2fs_sb_info.
  */
@@ -3612,7 +3623,7 @@ static inline bool f2fs_force_buffered_i
 
 	if (f2fs_post_read_required(inode))
 		return true;
-	if (sbi->s_ndevs)
+	if (f2fs_is_multi_device(sbi))
 		return true;
 	/*
 	 * for blkzoned device, fallback direct IO to buffered IO, so
fs/f2fs/file.c +1 -1
@@ -2570,7 +2570,7 @@ static int f2fs_ioc_flush_device(struct
 							sizeof(range)))
 		return -EFAULT;
 
-	if (sbi->s_ndevs <= 1 || sbi->s_ndevs - 1 <= range.dev_num ||
+	if (!f2fs_is_multi_device(sbi) || sbi->s_ndevs - 1 <= range.dev_num ||
 			__is_large_section(sbi)) {
 		f2fs_msg(sbi->sb, KERN_WARNING,
 			"Can't flush %u in %d for segs_per_sec %u != 1\n",
fs/f2fs/gc.c +1 -1
@@ -1346,7 +1346,7 @@ void f2fs_build_gc_manager(struct f2fs_s
 	sbi->gc_pin_file_threshold = DEF_GC_FAILED_PINNED_FILES;
 
 	/* give warm/cold data area from slower device */
-	if (sbi->s_ndevs && !__is_large_section(sbi))
+	if (f2fs_is_multi_device(sbi) && !__is_large_section(sbi))
 		SIT_I(sbi)->last_victim[ALLOC_NEXT] =
 				GET_SEGNO(sbi, FDEV(0).end_blk) + 1;
 }
fs/f2fs/segment.c +7 -6
@@ -576,7 +576,7 @@ static int submit_flush_wait(struct f2fs
 	int ret = 0;
 	int i;
 
-	if (!sbi->s_ndevs)
+	if (!f2fs_is_multi_device(sbi))
 		return __submit_flush_wait(sbi, sbi->sb->s_bdev);
 
 	for (i = 0; i < sbi->s_ndevs; i++) {
@@ -644,7 +644,8 @@ int f2fs_issue_flush(struct f2fs_sb_info
 		return ret;
 	}
 
-	if (atomic_inc_return(&fcc->queued_flush) == 1 || sbi->s_ndevs > 1) {
+	if (atomic_inc_return(&fcc->queued_flush) == 1 ||
+	    f2fs_is_multi_device(sbi)) {
 		ret = submit_flush_wait(sbi, ino);
 		atomic_dec(&fcc->queued_flush);
 
@@ -750,7 +751,7 @@ int f2fs_flush_device_cache(struct f2fs_
 {
 	int ret = 0, i;
 
-	if (!sbi->s_ndevs)
+	if (!f2fs_is_multi_device(sbi))
 		return 0;
 
 	for (i = 1; i < sbi->s_ndevs; i++) {
@@ -1359,7 +1360,7 @@ static int __queue_discard_cmd(struct f2
 
 	trace_f2fs_queue_discard(bdev, blkstart, blklen);
 
-	if (sbi->s_ndevs) {
+	if (f2fs_is_multi_device(sbi)) {
 		int devi = f2fs_target_device_index(sbi, blkstart);
 
 		blkstart -= FDEV(devi).start_blk;
@@ -1714,7 +1715,7 @@ static int __f2fs_issue_discard_zone(str
 	block_t lblkstart = blkstart;
 	int devi = 0;
 
-	if (sbi->s_ndevs) {
+	if (f2fs_is_multi_device(sbi)) {
 		devi = f2fs_target_device_index(sbi, blkstart);
 		blkstart -= FDEV(devi).start_blk;
 	}
@@ -3071,7 +3072,7 @@ static void update_device_state(struct f
 	struct f2fs_sb_info *sbi = fio->sbi;
 	unsigned int devidx;
 
-	if (!sbi->s_ndevs)
+	if (!f2fs_is_multi_device(sbi))
 		return;
 
 	devidx = f2fs_target_device_index(sbi, fio->new_blkaddr);
#138 lore raw
In reply to Greg Kroah-Hartman
Quoted 10 lines · Greg Kroah-Hartman

This is the start of the stable review cycle for the 5.0.17 release. There are 137 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.

Responses should be made by Fri 17 May 2019 09:04:31 AM UTC. Anything received after that time might be too late.

The whole patch series can be found in one patch at:

Quoted 9 lines
        https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.0.17-rc1.gz

or in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-5.0.y
and the diffstat can be found below.

thanks,

greg k-h

Results from Linaro’s test farm.
No regressions on arm64, arm, x86_64, and i386.

Summary
------------------------------------------------------------------------

kernel: 5.0.17-rc1
git repo: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git
git branch: linux-5.0.y
git commit: 174f1e44d0165ce68f4e520718847304556717e3
git describe: v5.0.16-138-g174f1e44d016
Test details: https://qa-reports.linaro.org/lkft/linux-stable-rc-5.0-oe/build/v5.0.16-138-g174f1e44d016

No regressions (compared to build v5.0.16)

No fixes (compared to build v5.0.16)

Ran 24960 total tests in the following environments and test suites.

Environments
--------------
- dragonboard-410c
- hi6220-hikey
- i386
- juno-r2
- qemu_arm
- qemu_arm64
- qemu_i386
- qemu_x86_64
- x15
- x86

Test Suites
-----------
* build
* install-android-platform-tools-r2600
* kselftest
* libgpiod
* libhugetlbfs
* ltp-cap_bounds-tests
* ltp-commands-tests
* ltp-containers-tests
* ltp-cpuhotplug-tests
* ltp-cve-tests
* ltp-dio-tests
* ltp-fcntl-locktests-tests
* ltp-filecaps-tests
* ltp-fs_bind-tests
* ltp-fs_perms_simple-tests
* ltp-fsx-tests
* ltp-hugetlb-tests
* ltp-io-tests
* ltp-ipc-tests
* ltp-math-tests
* ltp-mm-tests
* ltp-nptl-tests
* ltp-pty-tests
* ltp-sched-tests
* ltp-securebits-tests
* ltp-syscalls-tests
* ltp-timers-tests
* perf
* spectre-meltdown-checker-test
* v4l2-compliance
* ltp-fs-tests
* ltp-open-posix-tests
* kvm-unit-tests
* kselftest-vsyscall-mode-native
* kselftest-vsyscall-mode-none
* ssuite

Signature · 2 lines
Linaro LKFT
https://lkft.linaro.org
#139 lore raw
In reply to Greg Kroah-Hartman
stable-rc/linux-5.0.y boot: 139 boots: 0 failed, 136 passed with 1 offline, 1 untried/unknown, 1 conflict (v5.0.16-138-g174f1e44d016)

stable-rc/linux-5.0.y boot: 139 boots: 0 failed, 136 passed with 1 offline, 1 untried/unknown, 1 conflict (v5.0.16-138-g174f1e44d016)

Full Boot Summary: https://kernelci.org/boot/all/job/stable-rc/branch/linux-5.0.y/kernel/v5.0.16-138-g174f1e44d016/
Full Build Summary: https://kernelci.org/build/stable-rc/branch/linux-5.0.y/kernel/v5.0.16-138-g174f1e44d016/

Tree: stable-rc
Branch: linux-5.0.y
Git Describe: v5.0.16-138-g174f1e44d016
Git Commit: 174f1e44d0165ce68f4e520718847304556717e3
Git URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git
Tested: 74 unique boards, 23 SoC families, 14 builds out of 208

Boot Regressions Detected:

arm:

    omap2plus_defconfig:
        gcc-8:
          omap4-panda:
              lab-baylibre: new failure (last pass: v5.0.16-105-gcedd923508e9)

Offline Platforms:

arm:

    multi_v7_defconfig:
        gcc-8
            stih410-b2120: 1 offline lab

Conflicting Boot Failure Detected: (These likely are not failures as other labs are reporting PASS. Needs review.)

arm:
omap2plus_defconfig:
omap4-panda:
lab-baylibre: FAIL (gcc-8)
lab-baylibre-seattle: PASS (gcc-8)

---
For more info write to <[email protected]>

#140 lore raw
In reply to Greg Kroah-Hartman
Quoted 8 lines · Greg Kroah-Hartman

This is the start of the stable review cycle for the 5.0.17 release. There are 137 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.

Responses should be made by Fri 17 May 2019 09:04:31 AM UTC. Anything received after that time might be too late.

Build results:

	total: 159 pass: 159 fail: 0
Qemu test results:
	total: 349 pass: 349 fail: 0

Guenter

maintainer
#141 lore raw
In reply to Greg Kroah-Hartman
Quoted 9 lines · Greg Kroah-Hartman

This is the start of the stable review cycle for the 5.0.17 release. There are 137 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.

Responses should be made by Fri 17 May 2019 09:04:31 AM UTC. Anything received after that time might be too late.

The whole patch series can be found in one patch at:

Quoted 8 lines
	https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.0.17-rc1.gz

or in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-5.0.y
and the diffstat can be found below.

thanks,

greg k-h

All tests are passing for Tegra ...

Test results for stable-v5.0:

    12 builds:	12 pass, 0 fail
    22 boots:	22 pass, 0 fail
    32 tests:	32 pass, 0 fail

Linux version: 5.0.17-rc1-g174f1e4
Boards tested: tegra124-jetson-tk1, tegra186-p2771-0000,

                tegra194-p2972-0000, tegra20-ventana,
                tegra210-p2371-2180, tegra30-cardhu-a04

Cheers
Jon

Signature · 1 line
nvpublic
#142 lore raw
In reply to Greg Kroah-Hartman
Quoted 9 lines · Greg Kroah-Hartman

This is the start of the stable review cycle for the 5.0.17 release. There are 137 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.

Responses should be made by Fri 17 May 2019 09:04:31 AM UTC. Anything received after that time might be too late.

The whole patch series can be found in one patch at:

Quoted 9 lines
	https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.0.17-rc1.gz

or in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-5.0.y
and the diffstat can be found below.

thanks,

greg k-h

Compiled and booted on my test system. No dmesg regressions.

thanks,
-- Shuah

#143 lore raw
In reply to Greg Kroah-Hartman
Quoted 9 lines · Greg Kroah-Hartman

This is the start of the stable review cycle for the 5.0.17 release. There are 137 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.

Responses should be made by Fri 17 May 2019 09:04:31 AM UTC. Anything received after that time might be too late.

The whole patch series can be found in one patch at:

Quoted 8 lines
	https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.0.17-rc1.gz

or in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-5.0.y
and the diffstat can be found below.

thanks,

greg k-h

Compiled and booted with no dmesg regressions on my system.

Cheers,
Kelsey