4.14.171-stable review
Subsystems (3)
Greg Kroah-Hartman
maintainercover 4.14.171-stable review
This is the start of the stable review cycle for the 4.14.171 release. There are 173 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 Sat, 15 Feb 2020 15:16:41 +0000. 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/v4.x/stable-review/patch-4.14.171-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-4.14.y
and the diffstat can be found below.
thanks,
greg k-h
-------------
Pseudo-Shortlog of commits:
Greg Kroah-Hartman <[email protected]>
Linux 4.14.171-rc1
Nicolai Stange <[email protected]>
libertas: make lbs_ibss_join_existing() return error code on rates overflow
Nicolai Stange <[email protected]>
libertas: don't exit from lbs_ibss_join_existing() with RCU read lock held
Qing Xu <[email protected]>
mwifiex: Fix possible buffer overflows in mwifiex_cmd_append_vsie_tlv()
Qing Xu <[email protected]>
mwifiex: Fix possible buffer overflows in mwifiex_ret_wmm_get_status()
Mike Snitzer <[email protected]>
dm: fix potential for q->make_request_fn NULL pointer
Anand Lodnoor <[email protected]>
scsi: megaraid_sas: Do not initiate OCR if controller is not in ready state
Geert Uytterhoeven <[email protected]>
pinctrl: sh-pfc: r8a7778: Fix duplicate SDSELF_B and SD1_CLK_B
Gustavo A. R. Silva <[email protected]>
media: i2c: adv748x: Fix unsafe macros
Eric Biggers <[email protected]>
crypto: atmel-sha - fix error handling when setting hmac key
Eric Biggers <[email protected]>
crypto: artpec6 - return correct error code for failed setkey()
Gavin Shan <[email protected]>
KVM: arm/arm64: Fix young bit from mmu notifier
Suzuki K Poulose <[email protected]>
arm64: cpufeature: Fix the type of no FP/SIMD capability
Olof Johansson <[email protected]>
ARM: 8949/1: mm: mark free_memmap as __init
Eric Auger <[email protected]>
KVM: arm/arm64: vgic-its: Fix restoration of unmapped collections
Alexey Kardashevskiy <[email protected]>
powerpc/pseries: Allow not having ibm, hypertas-functions::hcall-multi-tce for DDW
Tyrel Datwyler <[email protected]>
powerpc/pseries/vio: Fix iommu_table use-after-free refcount warning
Zhengyuan Liu <[email protected]>
tools/power/acpi: fix compilation error
Alexandre Belloni <[email protected]>
ARM: dts: at91: sama5d3: define clock rate range for tcb1
Alexandre Belloni <[email protected]>
ARM: dts: at91: sama5d3: fix maximum peripheral clock rates
Mika Westerberg <[email protected]>
platform/x86: intel_mid_powerbtn: Take a copy of ddata
Jose Abreu <[email protected]>
ARC: [plat-axs10x]: Add missing multicast filter number to GMAC node
Andy Shevchenko <[email protected]>
rtc: cmos: Stop using shared IRQ
Paul Kocialkowski <[email protected]>
rtc: hym8563: Return -EINVAL if the time is known to be invalid
Shubhrajyoti Datta <[email protected]>
serial: uartps: Add a timeout to the tx empty wait
Robert Milkowski <[email protected]>
NFSv4: try lease recovery on NFS4ERR_EXPIRED
Trond Myklebust <[email protected]>
NFS/pnfs: Fix pnfs_generic_prepare_to_resend_writes()
Geert Uytterhoeven <[email protected]>
nfs: NFS_SWAP should depend on SWAP
Logan Gunthorpe <[email protected]>
PCI: Don't disable bridge BARs when assigning bus resources
Logan Gunthorpe <[email protected]>
PCI/switchtec: Fix vep_vector_number ioread width
Bean Huo <[email protected]>
scsi: ufs: Fix ufshcd_probe_hba() reture value in case ufshcd_scsi_add_wlus() fails
Håkon Bugge <[email protected]>
RDMA/netlink: Do not always generate an ACK for some netlink operations
Sunil Muthuswamy <[email protected]>
hv_sock: Remove the accept port restriction
Ranjani Sridharan <[email protected]>
ASoC: pcm: update FE/BE trigger order based on the command
David Howells <[email protected]>
rxrpc: Fix service call disconnection
Song Liu <[email protected]>
perf/core: Fix mlock accounting in perf_mmap()
Konstantin Khlebnikov <[email protected]>
clocksource: Prevent double add_timer_on() for watchdog_timer
Ronnie Sahlberg <[email protected]>
cifs: fail i/o on soft mounts if sessionsetup errors out
Christian Borntraeger <[email protected]>
KVM: s390: do not clobber registers during guest reset/store status
Sean Christopherson <[email protected]>
KVM: Play nice with read-only memslots when querying host page size
Sean Christopherson <[email protected]>
KVM: Use vcpu-specific gva->hva translation when querying host page size
Miaohe Lin <[email protected]>
KVM: nVMX: vmread should not set rflags to specify success in case of #PF
Sean Christopherson <[email protected]>
KVM: VMX: Add non-canonical check on writes to RTIT address MSRs
Sean Christopherson <[email protected]>
KVM: x86/mmu: Apply max PA check for MMIO sptes to 32-bit KVM
Sean Christopherson <[email protected]>
KVM: x86: Fix potential put_fpu() w/o load_fpu() on MPX platform
Josef Bacik <[email protected]>
btrfs: flush write bio if we loop in extent_write_cache_pages
Marios Pomonis <[email protected]>
KVM: x86: Protect pmu_intel.c from Spectre-v1/L1TF attacks
Wayne Lin <[email protected]>
drm/dp_mst: Remove VCPI while disabling topology mgr
Claudiu Beznea <[email protected]>
drm: atmel-hlcdc: enable clock before configuring timing engine
Filipe Manana <[email protected]>
Btrfs: fix race between adding and putting tree mod seq elements and nodes
David Sterba <[email protected]>
btrfs: remove trivial locking wrappers of tree mod log
Josef Bacik <[email protected]>
btrfs: free block groups after free'ing fs trees
Anand Jain <[email protected]>
btrfs: use bool argument in free_root_pointers()
Filipe Manana <[email protected]>
Btrfs: fix missing hole after hole punching and fsync when using NO_HOLES
Filipe Manana <[email protected]>
Btrfs: fix assertion failure on fsync with NO_HOLES enabled
Qu Wenruo <[email protected]>
btrfs: Get rid of the confusing btrfs_file_extent_inline_len
Eric Biggers <[email protected]>
ext4: fix deadlock allocating crypto bounce page from mempool
Eric Dumazet <[email protected]>
bonding/alb: properly access headers in bond_alb_xmit()
Harini Katakam <[email protected]>
net: macb: Limit maximum GEM TX length in TSO
Harini Katakam <[email protected]>
net: macb: Remove unnecessary alignment check for TSO
Florian Fainelli <[email protected]>
net: systemport: Avoid RBUF stuck in Wake-on-LAN mode
Cong Wang <[email protected]>
net_sched: fix a resource leak in tcindex_set_parms()
Florian Fainelli <[email protected]>
net: dsa: bcm_sf2: Only 7278 supports 2Gb/sec IMP port
Andreas Kemnade <[email protected]>
mfd: rn5t618: Mark ADC control register volatile
Marco Felsch <[email protected]>
mfd: da9062: Fix watchdog compatible string
Dan Carpenter <[email protected]>
ubi: Fix an error pointer dereference in error handling code
Sascha Hauer <[email protected]>
ubi: fastmap: Fix inverted logic in seen selfcheck
Arnd Bergmann <[email protected]>
nfsd: fix jiffies/time_t mixup in LRU list
Arnd Bergmann <[email protected]>
nfsd: fix delay timer on 32-bit architectures
Yishai Hadas <[email protected]>
IB/core: Fix ODP get user pages flow
Prabhath Sajeepa <[email protected]>
IB/mlx5: Fix outstanding_pi index for GSI qps
Nathan Chancellor <[email protected]>
net: tulip: Adjust indentation in {dmfe, uli526x}_init_module
Nathan Chancellor <[email protected]>
net: smc911x: Adjust indentation in smc911x_phy_configure
Nathan Chancellor <[email protected]>
ppp: Adjust indentation into ppp_async_input
Nathan Chancellor <[email protected]>
NFC: pn544: Adjust indentation in pn544_hci_check_presence
Nathan Chancellor <[email protected]>
powerpc/44x: Adjust indentation in ibm4xx_denali_fixup_memsize
Nathan Chancellor <[email protected]>
ext2: Adjust indentation in ext2_fill_super
Nathan Chancellor <[email protected]>
phy: qualcomm: Adjust indentation in read_poll_timeout
Asutosh Das <[email protected]>
scsi: ufs: Recheck bkops level if bkops is disabled
Nathan Chancellor <[email protected]>
scsi: qla4xxx: Adjust indentation in qla4xxx_mem_free
Nathan Chancellor <[email protected]>
scsi: csiostor: Adjust indentation in csio_device_reset
Bart Van Assche <[email protected]>
scsi: qla2xxx: Fix the endianness of the qla82xx_get_fw_size() return type
Stephen Warren <[email protected]>
clk: tegra: Mark fuse clock as critical
Sean Christopherson <[email protected]>
KVM: x86: Free wbinvd_dirty_mask if vCPU creation fails
Sean Christopherson <[email protected]>
KVM: PPC: Book3S PR: Free shared page if mmu initialization fails
Sean Christopherson <[email protected]>
KVM: PPC: Book3S HV: Uninit vCPU if vcore creation fails
Marios Pomonis <[email protected]>
KVM: x86: Protect MSR-based index computations in fixed_msr_to_seg_unit() from Spectre-v1/L1TF attacks
Marios Pomonis <[email protected]>
KVM: x86: Protect x86_decode_insn from Spectre-v1/L1TF attacks
Marios Pomonis <[email protected]>
KVM: x86: Protect MSR-based index computations from Spectre-v1/L1TF attacks in x86.c
Marios Pomonis <[email protected]>
KVM: x86: Protect ioapic_read_indirect() from Spectre-v1/L1TF attacks
Marios Pomonis <[email protected]>
KVM: x86: Protect MSR-based index computations in pmu.h from Spectre-v1/L1TF attacks
Marios Pomonis <[email protected]>
KVM: x86: Protect ioapic_write_indirect() from Spectre-v1/L1TF attacks
Marios Pomonis <[email protected]>
KVM: x86: Protect kvm_hv_msr_[get|set]_crash_data() from Spectre-v1/L1TF attacks
Marios Pomonis <[email protected]>
KVM: x86: Protect kvm_lapic_reg_write() from Spectre-v1/L1TF attacks
Marios Pomonis <[email protected]>
KVM: x86: Protect DR-based index computations from Spectre-v1/L1TF attacks
Marios Pomonis <[email protected]>
KVM: x86: Refactor prefix decoding to prevent Spectre-v1/L1TF attacks
Marios Pomonis <[email protected]>
KVM: x86: Refactor picdev_write() to prevent Spectre-v1/L1TF attacks
Juergen Gross <[email protected]>
xen/balloon: Support xend-based toolstack take two
Gavin Shan <[email protected]>
tools/kvm_stat: Fix kvm_exit filter name
Roberto Bergantinos Corpas <[email protected]>
sunrpc: expiry_time should be seconds not timeval
Brian Norris <[email protected]>
mwifiex: fix unbalanced locking in mwifiex_process_country_ie()
Luca Coelho <[email protected]>
iwlwifi: don't throw error when trying to remove IGTK
Stephen Warren <[email protected]>
ARM: tegra: Enable PLLP bypass during Tegra124 LP1
Josef Bacik <[email protected]>
btrfs: set trans->drity in btrfs_commit_transaction
Trond Myklebust <[email protected]>
NFS: Directory page cache pages need to be locked when read
Trond Myklebust <[email protected]>
NFS: Fix memory leaks and corruption in readdir
Arun Easi <[email protected]>
scsi: qla2xxx: Fix unbound NVME response length
Chuhong Yuan <[email protected]>
crypto: picoxcell - adjust the position of tasklet_init and fix missed tasklet_kill
Herbert Xu <[email protected]>
crypto: api - Fix race condition in crypto_spawn_alg
Tudor Ambarus <[email protected]>
crypto: atmel-aes - Fix counter overflow in CTR mode
Herbert Xu <[email protected]>
crypto: pcrypt - Do not clear MAY_SLEEP flag in original request
Ard Biesheuvel <[email protected]>
crypto: ccp - set max RSA modulus size for v3 platform devices as well
Toke Høiland-Jørgensen <[email protected]>
samples/bpf: Don't try to remove user's homedir on clean
Steven Rostedt (VMware) <[email protected]>
ftrace: Protect ftrace_graph_hash with ftrace_sync
Steven Rostedt (VMware) <[email protected]>
ftrace: Add comment to why rcu_dereference_sched() is open coded
Amol Grover <[email protected]>
tracing: Annotate ftrace_graph_notrace_hash pointer with __rcu
Amol Grover <[email protected]>
tracing: Annotate ftrace_graph_hash pointer with __rcu
Herbert Xu <[email protected]>
padata: Remove broken queue flushing
Milan Broz <[email protected]>
dm crypt: fix benbi IV constructor crash if used in authenticated mode
Joe Thornber <[email protected]>
dm space map common: fix to ensure new block isn't already in use
Dmitry Fomichev <[email protected]>
dm zoned: support zone sizes smaller than 128MiB
Michael Ellerman <[email protected]>
of: Add OF_DMA_DEFAULT_COHERENT & select it on powerpc
Rafael J. Wysocki <[email protected]>
PM: core: Fix handling of devices deleted during system-wide resume
Chengguang Xu <[email protected]>
f2fs: code cleanup for f2fs_statfs_project()
Chengguang Xu <[email protected]>
f2fs: fix miscounted block limit in f2fs_statfs_project()
Chengguang Xu <[email protected]>
f2fs: choose hardlimit when softlimit is larger than hardlimit in f2fs_statfs_project()
Sven Van Asbroeck <[email protected]>
power: supply: ltc2941-battery-gauge: fix use-after-free
Quinn Tran <[email protected]>
scsi: qla2xxx: Fix mtcp dump collection failure
Herbert Xu <[email protected]>
crypto: api - Check spawn->alg under lock in crypto_drop_spawn
Tianyu Lan <[email protected]>
hv_balloon: Balloon up according to request page number
Eric Biggers <[email protected]>
ubifs: don't trigger assertion on invalid no-key filename
Michał Mirosław <[email protected]>
mmc: sdhci-of-at91: fix memleak on clk_get failure
Yurii Monakov <[email protected]>
PCI: keystone: Fix link training retries initiation
Zhihao Cheng <[email protected]>
ubifs: Fix deadlock in concurrent bulk-read and writepage
Eric Biggers <[email protected]>
ubifs: Fix FS_IOC_SETFLAGS unexpectedly clearing encrypt flag
Hou Tao <[email protected]>
ubifs: Reject unsupported ioctl flags explicitly
Stephen Boyd <[email protected]>
alarmtimer: Unregister wakeup source when module get fails
Hans de Goede <[email protected]>
ACPI: video: Do not export a non working backlight interface on MSI MS-7721 boards
Linus Walleij <[email protected]>
mmc: spi: Toggle SPI polarity, do not hardcode it
Pingfan Liu <[email protected]>
powerpc/pseries: Advance pfn if section is not present in lmb_is_removable()
Sukadev Bhattiprolu <[email protected]>
powerpc/xmon: don't access ASDR in VMs
Gerald Schaefer <[email protected]>
s390/mm: fix dynamic pagetable upgrade for hugetlbfs
Alexander Lobakin <[email protected]>
MIPS: fix indentation of the 'RELOCS' message
Christoffer Dall <[email protected]>
KVM: arm64: Only sign-extend MMIO up to register width
Mika Westerberg <[email protected]>
platform/x86: intel_scu_ipc: Fix interrupt support
Kevin Hao <[email protected]>
irqdomain: Fix a memory leak in irq_domain_push_irq()
Gustavo A. R. Silva <[email protected]>
lib/test_kasan.c: fix memory leak in kmalloc_oob_krealloc_more()
Helen Koike <[email protected]>
media: v4l2-rect.h: fix v4l2_rect_map_inside() top/left adjustments
John Hubbard <[email protected]>
media/v4l2-core: set pages dirty upon releasing DMA buffers
Takashi Iwai <[email protected]>
ALSA: dummy: Fix PCM format loop in proc output
Bryan O'Donoghue <[email protected]>
usb: gadget: f_ecm: Use atomic_t to track in-flight request
Bryan O'Donoghue <[email protected]>
usb: gadget: f_ncm: Use atomic_t to track in-flight request
Roger Quadros <[email protected]>
usb: gadget: legacy: set max_speed to super-speed
Navid Emamdoost <[email protected]>
brcmfmac: Fix memory leak in brcmf_usbdev_qinit
Mathieu Desnoyers <[email protected]>
tracing: Fix sched switch start/stop refcount racy updates
Oliver Neukum <[email protected]>
mfd: dln2: More sanity checking for endpoints
Will Deacon <[email protected]>
media: uvcvideo: Avoid cyclic entity chains due to malformed USB descriptors
David Howells <[email protected]>
rxrpc: Fix NULL pointer deref due to call->conn being cleared on disconnect
David Howells <[email protected]>
rxrpc: Fix insufficient receive notification generation
Eric Dumazet <[email protected]>
tcp: clear tp->segs_{in|out} in tcp_disconnect()Eric Dumazet <[email protected]>
tcp: clear tp->data_segs{in|out} in tcp_disconnect()Eric Dumazet <[email protected]>
tcp: clear tp->delivered in tcp_disconnect()
Eric Dumazet <[email protected]>
tcp: clear tp->total_retrans in tcp_disconnect()
Michael Chan <[email protected]>
bnxt_en: Fix TC queue mapping.
Cong Wang <[email protected]>
net_sched: fix an OOB access in cls_tcindex
Eric Dumazet <[email protected]>
net: hsr: fix possible NULL deref in hsr_handle_frame()
Ridge Kennedy <[email protected]>
l2tp: Allow duplicate session creation with UDP
Taehee Yoo <[email protected]>
gtp: use __GFP_NOWARN to avoid memalloc warning
Eric Dumazet <[email protected]>
cls_rsvp: fix rsvp_policy
Arnd Bergmann <[email protected]>
sparc32: fix struct ipc64_perm type definition
Luca Coelho <[email protected]>
iwlwifi: mvm: fix NVM check for 3168 devices
Pawan Gupta <[email protected]>
x86/cpu: Update cached HLE state on write to TSX_CTRL_CPUID_CLEAR
Johan Hovold <[email protected]>
media: iguanair: fix endpoint sanity check
YueHaibing <[email protected]>
kernel/module: Fix memleak in module_add_modinfo_attrs()
-------------
Diffstat:
Makefile | 4 +- arch/arc/boot/dts/axs10x_mb.dtsi | 1 + arch/arm/boot/dts/sama5d3.dtsi | 28 +- arch/arm/boot/dts/sama5d3_can.dtsi | 4 +- arch/arm/boot/dts/sama5d3_tcb1.dtsi | 1 + arch/arm/boot/dts/sama5d3_uart.dtsi | 4 +- arch/arm/include/asm/kvm_emulate.h | 5 + arch/arm/include/asm/kvm_mmio.h | 2 + arch/arm/mach-tegra/sleep-tegra30.S | 11 + arch/arm/mm/init.c | 2 +- arch/arm64/include/asm/kvm_emulate.h | 5 + arch/arm64/include/asm/kvm_mmio.h | 6 +- arch/arm64/kernel/cpufeature.c | 2 +- arch/mips/Makefile.postlink | 2 +- arch/powerpc/Kconfig | 1 + arch/powerpc/boot/4xx.c | 2 +- arch/powerpc/kvm/book3s_hv.c | 4 +- arch/powerpc/kvm/book3s_pr.c | 4 +- arch/powerpc/platforms/pseries/hotplug-memory.c | 4 +- arch/powerpc/platforms/pseries/iommu.c | 43 +- arch/powerpc/platforms/pseries/vio.c | 2 + arch/powerpc/xmon/xmon.c | 9 +- arch/s390/include/asm/page.h | 2 + arch/s390/kvm/kvm-s390.c | 6 +- arch/s390/mm/hugetlbpage.c | 100 +- arch/sparc/include/uapi/asm/ipcbuf.h | 22 +- arch/x86/kernel/cpu/tsx.c | 13 +- arch/x86/kvm/emulate.c | 27 +- arch/x86/kvm/hyperv.c | 10 +- arch/x86/kvm/i8259.c | 6 +- arch/x86/kvm/ioapic.c | 15 +- arch/x86/kvm/lapic.c | 13 +- arch/x86/kvm/mmu.c | 6 +- arch/x86/kvm/mtrr.c | 8 +- arch/x86/kvm/pmu.h | 18 +- arch/x86/kvm/pmu_intel.c | 24 +- arch/x86/kvm/vmx.c | 4 +- arch/x86/kvm/vmx/vmx.c | 8033 ++++++++++++++++++++ arch/x86/kvm/x86.c | 27 +- crypto/algapi.c | 22 +- crypto/api.c | 3 +- crypto/internal.h | 1 - crypto/pcrypt.c | 1 - drivers/acpi/video_detect.c | 13 + drivers/base/power/main.c | 42 +- drivers/clk/tegra/clk-tegra-periph.c | 6 +- drivers/crypto/atmel-aes.c | 37 +- drivers/crypto/atmel-sha.c | 7 +- drivers/crypto/axis/artpec6_crypto.c | 2 +- drivers/crypto/ccp/ccp-dev-v3.c | 1 + drivers/crypto/picoxcell_crypto.c | 15 +- drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_crtc.c | 8 +- drivers/gpu/drm/drm_dp_mst_topology.c | 12 + drivers/hv/hv_balloon.c | 13 +- drivers/infiniband/core/addr.c | 2 +- drivers/infiniband/core/sa_query.c | 4 +- drivers/infiniband/core/umem_odp.c | 2 +- drivers/infiniband/hw/mlx5/gsi.c | 3 +- drivers/md/dm-crypt.c | 10 +- drivers/md/dm-zoned-metadata.c | 23 +- drivers/md/dm.c | 9 +- drivers/md/persistent-data/dm-space-map-common.c | 27 + drivers/md/persistent-data/dm-space-map-common.h | 2 + drivers/md/persistent-data/dm-space-map-disk.c | 6 +- drivers/md/persistent-data/dm-space-map-metadata.c | 5 +- drivers/media/i2c/adv748x/adv748x.h | 8 +- drivers/media/rc/iguanair.c | 2 +- drivers/media/usb/uvc/uvc_driver.c | 12 + drivers/media/v4l2-core/videobuf-dma-sg.c | 5 +- drivers/mfd/da9062-core.c | 2 +- drivers/mfd/dln2.c | 13 +- drivers/mfd/rn5t618.c | 1 + drivers/mmc/host/mmc_spi.c | 11 +- drivers/mmc/host/sdhci-of-at91.c | 9 +- drivers/mtd/ubi/fastmap.c | 23 +- drivers/net/bonding/bond_alb.c | 44 +- drivers/net/dsa/bcm_sf2.c | 4 +- drivers/net/ethernet/broadcom/bcmsysport.c | 3 + drivers/net/ethernet/broadcom/bnxt/bnxt.c | 2 +- drivers/net/ethernet/cadence/macb_main.c | 14 +- drivers/net/ethernet/dec/tulip/dmfe.c | 7 +- drivers/net/ethernet/dec/tulip/uli526x.c | 4 +- drivers/net/ethernet/smsc/smc911x.c | 2 +- drivers/net/gtp.c | 6 +- drivers/net/ppp/ppp_async.c | 18 +- .../net/wireless/broadcom/brcm80211/brcmfmac/usb.c | 1 + drivers/net/wireless/intel/iwlwifi/mvm/nvm.c | 2 +- drivers/net/wireless/intel/iwlwifi/mvm/sta.c | 10 +- drivers/net/wireless/marvell/libertas/cfg.c | 2 + drivers/net/wireless/marvell/mwifiex/scan.c | 7 + drivers/net/wireless/marvell/mwifiex/sta_ioctl.c | 1 + drivers/net/wireless/marvell/mwifiex/wmm.c | 4 + drivers/nfc/pn544/pn544.c | 2 +- drivers/of/Kconfig | 4 + drivers/of/address.c | 6 +- drivers/pci/dwc/pci-keystone-dw.c | 2 +- drivers/pci/setup-bus.c | 20 +- drivers/pci/switch/switchtec.c | 2 +- drivers/phy/qualcomm/phy-qcom-apq8064-sata.c | 2 +- drivers/pinctrl/sh-pfc/pfc-r8a7778.c | 4 +- drivers/platform/x86/intel_mid_powerbtn.c | 5 +- drivers/platform/x86/intel_scu_ipc.c | 21 +- drivers/power/supply/ltc2941-battery-gauge.c | 2 +- drivers/rtc/rtc-cmos.c | 2 +- drivers/rtc/rtc-hym8563.c | 2 +- drivers/scsi/csiostor/csio_scsi.c | 2 +- drivers/scsi/megaraid/megaraid_sas_base.c | 3 +- drivers/scsi/megaraid/megaraid_sas_fusion.c | 3 +- drivers/scsi/megaraid/megaraid_sas_fusion.h | 1 + drivers/scsi/qla2xxx/qla_dbg.c | 6 - drivers/scsi/qla2xxx/qla_dbg.h | 6 + drivers/scsi/qla2xxx/qla_isr.c | 12 + drivers/scsi/qla2xxx/qla_mbx.c | 3 +- drivers/scsi/qla2xxx/qla_nx.c | 7 +- drivers/scsi/qla4xxx/ql4_os.c | 2 +- drivers/scsi/ufs/ufshcd.c | 6 +- drivers/tty/serial/xilinx_uartps.c | 14 +- drivers/usb/gadget/function/f_ecm.c | 16 +- drivers/usb/gadget/function/f_ncm.c | 17 +- drivers/usb/gadget/legacy/cdc2.c | 2 +- drivers/usb/gadget/legacy/g_ffs.c | 2 +- drivers/usb/gadget/legacy/multi.c | 2 +- drivers/usb/gadget/legacy/ncm.c | 2 +- drivers/xen/xen-balloon.c | 2 +- fs/btrfs/ctree.c | 64 +- fs/btrfs/ctree.h | 32 +- fs/btrfs/delayed-ref.c | 8 +- fs/btrfs/disk-io.c | 22 +- fs/btrfs/extent_io.c | 8 + fs/btrfs/file-item.c | 2 +- fs/btrfs/file.c | 3 +- fs/btrfs/inode.c | 12 +- fs/btrfs/print-tree.c | 4 +- fs/btrfs/send.c | 17 +- fs/btrfs/tests/btrfs-tests.c | 1 - fs/btrfs/transaction.c | 8 + fs/btrfs/tree-log.c | 397 +- fs/cifs/smb2pdu.c | 10 +- fs/ext2/super.c | 6 +- fs/ext4/page-io.c | 19 +- fs/f2fs/super.c | 14 +- fs/nfs/Kconfig | 2 +- fs/nfs/dir.c | 47 +- fs/nfs/direct.c | 4 +- fs/nfs/nfs3xdr.c | 5 +- fs/nfs/nfs4proc.c | 5 + fs/nfs/nfs4xdr.c | 5 +- fs/nfs/pnfs_nfs.c | 7 +- fs/nfs/write.c | 4 +- fs/nfsd/nfs4layouts.c | 2 +- fs/nfsd/nfs4state.c | 2 +- fs/nfsd/state.h | 2 +- fs/ubifs/dir.c | 2 + fs/ubifs/file.c | 4 +- fs/ubifs/ioctl.c | 11 +- include/linux/kvm_host.h | 2 +- include/media/v4l2-rect.h | 8 +- include/trace/events/btrfs.h | 2 +- kernel/events/core.c | 10 +- kernel/irq/irqdomain.c | 1 + kernel/module.c | 2 + kernel/padata.c | 46 +- kernel/time/alarmtimer.c | 8 +- kernel/time/clocksource.c | 11 +- kernel/trace/ftrace.c | 15 +- kernel/trace/trace.h | 29 +- kernel/trace/trace_sched_switch.c | 4 +- lib/test_kasan.c | 1 + net/hsr/hsr_slave.c | 2 + net/ipv4/tcp.c | 6 + net/l2tp/l2tp_core.c | 7 +- net/rxrpc/ar-internal.h | 1 + net/rxrpc/call_object.c | 4 +- net/rxrpc/conn_client.c | 3 +- net/rxrpc/conn_object.c | 3 +- net/rxrpc/input.c | 3 +- net/rxrpc/output.c | 26 +- net/sched/cls_rsvp.h | 6 +- net/sched/cls_tcindex.c | 43 +- net/sunrpc/auth_gss/svcauth_gss.c | 4 + net/vmw_vsock/hyperv_transport.c | 68 +- samples/bpf/Makefile | 2 +- sound/drivers/dummy.c | 2 +- sound/soc/soc-pcm.c | 95 +- tools/kvm/kvm_stat/kvm_stat | 8 +- tools/power/acpi/Makefile.config | 2 +- virt/kvm/arm/mmio.c | 6 + virt/kvm/arm/mmu.c | 3 +- virt/kvm/arm/vgic/vgic-its.c | 3 +- virt/kvm/kvm_main.c | 4 +- 190 files changed, 9334 insertions(+), 990 deletions(-)
Greg Kroah-Hartman
maintainerPATCH 1/173 kernel/module: Fix memleak in module_add_modinfo_attrs()
From: YueHaibing <[email protected]>
[ Upstream commit f6d061d617124abbd55396a3bc37b9bf7d33233c ]
In module_add_modinfo_attrs() if sysfs_create_file() fails on the first iteration of the loop (so i = 0), we forget to free the modinfo_attrs.
--- kernel/module.c | 2 ++ 1 file changed, 2 insertions(+)
kernel/module.c +2 -0
@@ -1730,6 +1730,8 @@ static int module_add_modinfo_attrs(struct module *mod) error_out:if (i >0 )module_remove_modinfo_attrs (mod, --i); +else +kfree (mod->modinfo_attrs);return error; }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 2/173 media: iguanair: fix endpoint sanity check
In reply to Greg Kroah-HartmanFrom: Johan Hovold <[email protected]>
[ Upstream commit 1b257870a78b0a9ce98fdfb052c58542022ffb5b ]
Make sure to use the current alternate setting, which need not be the first one by index, when verifying the endpoint descriptors and initialising the URBs.
Failing to do so could cause the driver to misbehave or trigger a WARN() in usb_submit_urb() that kernels with panic_on_warn set would choke on.
--- drivers/media/rc/iguanair.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/rc/iguanair.c +1 -1
@@ -427,7 +427,7 @@ static int iguanair_probe(struct usb_interface *intf,int ret, pipein, pipeout;struct usb_host_interface *idesc; - idesc = intf->altsetting; + idesc = intf->cur_altsetting;if (idesc->desc.bNumEndpoints <2 )return -ENODEV ;
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 3/173 x86/cpu: Update cached HLE state on write to TSX_CTRL_CPUID_CLEAR
In reply to Greg Kroah-HartmanFrom: Pawan Gupta <[email protected]>
[ Upstream commit 5efc6fa9044c3356d6046c6e1da6d02572dbed6b ]
/proc/cpuinfo currently reports Hardware Lock Elision (HLE) feature to be present on boot cpu even if it was disabled during the bootup. This is because cpuinfo_x86->x86_capability HLE bit is not updated after TSX state is changed via the new MSR IA32_TSX_CTRL.
Update the cached HLE bit also since it is expected to change after an update to CPUID_CLEAR bit in MSR IA32_TSX_CTRL.
--- arch/x86/kernel/cpu/tsx.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-)
arch/x86/kernel/cpu/tsx.c +7 -6
@@ -115,11 +115,12 @@ void __init tsx_init(void)tsx_disable ();/* -* tsx_disable() will change the state of the -* RTM CPUID bit. Clear it here since it is now -* expected to be not set. +* tsx_disable() will change the state of the RTM and HLE CPUID +* bits. Clear them here since they are now expected to be not +* set. */ setup_clear_cpu_cap (X86_FEATURE_RTM ); +setup_clear_cpu_cap (X86_FEATURE_HLE ); }else if (tsx_ctrl_state ==TSX_CTRL_ENABLE ) {/*
@@ -131,10 +132,10 @@ void __init tsx_init(void)tsx_enable ();/* -* tsx_enable() will change the state of the -* RTM CPUID bit. Force it here since it is now -* expected to be set. +* tsx_enable() will change the state of the RTM and HLE CPUID +* bits. Force them here since they are now expected to be set. */ setup_force_cpu_cap (X86_FEATURE_RTM ); +setup_force_cpu_cap (X86_FEATURE_HLE ); } }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 4/173 iwlwifi: mvm: fix NVM check for 3168 devices
In reply to Greg Kroah-HartmanFrom: Luca Coelho <[email protected]>
[ Upstream commit b3f20e098293892388d6a0491d6bbb2efb46fbff ]
We had a check on !NVM_EXT and then a check for NVM_SDP in the else block of this if. The else block, obviously, could only be reached if using NVM_EXT, so it would never be NVM_SDP.
Fix that by checking whether the nvm_type is IWL_NVM instead of checking for !IWL_NVM_EXT to solve this issue.
--- drivers/net/wireless/intel/iwlwifi/mvm/nvm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/wireless/intel/iwlwifi/mvm/nvm.c +1 -1
@@ -298,7 +298,7 @@ iwl_parse_nvm_sections(struct iwl_mvm *mvm)int regulatory_type;/* Checking for required sections */ -if (mvm->trans->cfg->nvm_type !=IWL_NVM_EXT ) { +if (mvm->trans->cfg->nvm_type ==IWL_NVM ) {if (!mvm->nvm_sections[NVM_SECTION_TYPE_SW ].data || !mvm->nvm_sections[mvm->cfg->nvm_hw_section_num].data) {IWL_ERR (mvm,"Can't parse empty OTP/NVM sections\n" );
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 5/173 sparc32: fix struct ipc64_perm type definition
In reply to Greg Kroah-HartmanFrom: Arnd Bergmann <[email protected]>
[ Upstream commit 34ca70ef7d3a9fa7e89151597db5e37ae1d429b4 ]
As discussed in the strace issue tracker, it appears that the sparc32 sysvipc support has been broken for the past 11 years. It was however working in compat mode, which is how it must have escaped most of the regular testing.
The problem is that a cleanup patch inadvertently changed the uid/gid fields in struct ipc64_perm from 32-bit types to 16-bit types in uapi headers.
Both glibc and uclibc-ng still use the original types, so they should work fine with compat mode, but not natively. Change the definitions to use __kernel_uid32_t and __kernel_gid32_t again.
--- arch/sparc/include/uapi/asm/ipcbuf.h | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-)
arch/sparc/include/uapi/asm/ipcbuf.h +11 -11
@@ -15,19 +15,19 @@struct ipc64_perm { -__kernel_key_t key; -__kernel_uid_t uid; -__kernel_gid_t gid; -__kernel_uid_t cuid; -__kernel_gid_t cgid; +__kernel_key_t key; +__kernel_uid32_t uid; +__kernel_gid32_t gid; +__kernel_uid32_t cuid; +__kernel_gid32_t cgid;#ifndef __arch64__ -unsigned short __pad0; +unsigned short __pad0;#endif -__kernel_mode_t mode; -unsigned short __pad1; -unsigned short seq; -unsigned long long __unused1; -unsigned long long __unused2; +__kernel_mode_t mode; +unsigned short __pad1; +unsigned short seq; +unsigned long long __unused1; +unsigned long long __unused2; };#endif /* __SPARC_IPCBUF_H */
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 6/173 cls_rsvp: fix rsvp_policy
In reply to Greg Kroah-HartmanFrom: Eric Dumazet <[email protected]>
[ Upstream commit cb3c0e6bdf64d0d124e94ce43cbe4ccbb9b37f51 ]
NLA_BINARY can be confusing, since .len value represents the max size of the blob.
cls_rsvp really wants user space to provide long enough data for TCA_RSVP_DST and TCA_RSVP_SRC attributes.
BUG: KMSAN: uninit-value in rsvp_get net/sched/cls_rsvp.h:258 [inline] BUG: KMSAN: uninit-value in gen_handle net/sched/cls_rsvp.h:402 [inline] BUG: KMSAN: uninit-value in rsvp_change+0x1ae9/0x4220 net/sched/cls_rsvp.h:572 CPU: 1 PID: 13228 Comm: syz-executor.1 Not tainted 5.5.0-rc5-syzkaller #0 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+0x1c9/0x220 lib/dump_stack.c:118 kmsan_report+0xf7/0x1e0 mm/kmsan/kmsan_report.c:118 __msan_warning+0x58/0xa0 mm/kmsan/kmsan_instr.c:215 rsvp_get net/sched/cls_rsvp.h:258 [inline] gen_handle net/sched/cls_rsvp.h:402 [inline] rsvp_change+0x1ae9/0x4220 net/sched/cls_rsvp.h:572 tc_new_tfilter+0x31fe/0x5010 net/sched/cls_api.c:2104 rtnetlink_rcv_msg+0xcb7/0x1570 net/core/rtnetlink.c:5415 netlink_rcv_skb+0x451/0x650 net/netlink/af_netlink.c:2477 rtnetlink_rcv+0x50/0x60 net/core/rtnetlink.c:5442 netlink_unicast_kernel net/netlink/af_netlink.c:1302 [inline] netlink_unicast+0xf9e/0x1100 net/netlink/af_netlink.c:1328 netlink_sendmsg+0x1248/0x14d0 net/netlink/af_netlink.c:1917 sock_sendmsg_nosec net/socket.c:639 [inline] sock_sendmsg net/socket.c:659 [inline] ____sys_sendmsg+0x12b6/0x1350 net/socket.c:2330 ___sys_sendmsg net/socket.c:2384 [inline] __sys_sendmsg+0x451/0x5f0 net/socket.c:2417 __do_sys_sendmsg net/socket.c:2426 [inline] __se_sys_sendmsg+0x97/0xb0 net/socket.c:2424 __x64_sys_sendmsg+0x4a/0x70 net/socket.c:2424 do_syscall_64+0xb8/0x160 arch/x86/entry/common.c:296 entry_SYSCALL_64_after_hwframe+0x44/0xa9 RIP: 0033:0x45b349 Code: ad b6 fb ff c3 66 2e 0f 1f 84 00 00 00 00 00 66 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 7b b6 fb ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007f269d43dc78 EFLAGS: 00000246 ORIG_RAX: 000000000000002e RAX: ffffffffffffffda RBX: 00007f269d43e6d4 RCX: 000000000045b349 RDX: 0000000000000000 RSI: 00000000200001c0 RDI: 0000000000000003 RBP: 000000000075bfc8 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 00000000ffffffff R13: 00000000000009c2 R14: 00000000004cb338 R15: 000000000075bfd4
Uninit was created at:
kmsan_save_stack_with_flags mm/kmsan/kmsan.c:144 [inline] kmsan_internal_poison_shadow+0x66/0xd0 mm/kmsan/kmsan.c:127 kmsan_slab_alloc+0x8a/0xe0 mm/kmsan/kmsan_hooks.c:82 slab_alloc_node mm/slub.c:2774 [inline] __kmalloc_node_track_caller+0xb40/0x1200 mm/slub.c:4382 __kmalloc_reserve net/core/skbuff.c:141 [inline] __alloc_skb+0x2fd/0xac0 net/core/skbuff.c:209 alloc_skb include/linux/skbuff.h:1049 [inline] netlink_alloc_large_skb net/netlink/af_netlink.c:1174 [inline] netlink_sendmsg+0x7d3/0x14d0 net/netlink/af_netlink.c:1892 sock_sendmsg_nosec net/socket.c:639 [inline] sock_sendmsg net/socket.c:659 [inline] ____sys_sendmsg+0x12b6/0x1350 net/socket.c:2330 ___sys_sendmsg net/socket.c:2384 [inline] __sys_sendmsg+0x451/0x5f0 net/socket.c:2417 __do_sys_sendmsg net/socket.c:2426 [inline] __se_sys_sendmsg+0x97/0xb0 net/socket.c:2424 __x64_sys_sendmsg+0x4a/0x70 net/socket.c:2424 do_syscall_64+0xb8/0x160 arch/x86/entry/common.c:296 entry_SYSCALL_64_after_hwframe+0x44/0xa9
--- net/sched/cls_rsvp.h | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-)
net/sched/cls_rsvp.h +2 -4
@@ -475,10 +475,8 @@ static u32 gen_tunnel(struct rsvp_head *static const struct nla_policy rsvp_policy[TCA_RSVP_MAX +1 ] = { [TCA_RSVP_CLASSID ] = { .type =NLA_U32 }, - [TCA_RSVP_DST ] = { .type =NLA_BINARY , - .len =RSVP_DST_LEN *sizeof (u32) }, - [TCA_RSVP_SRC ] = { .type =NLA_BINARY , - .len =RSVP_DST_LEN *sizeof (u32) }, + [TCA_RSVP_DST ] = { .len =RSVP_DST_LEN *sizeof (u32) }, + [TCA_RSVP_SRC ] = { .len =RSVP_DST_LEN *sizeof (u32) }, [TCA_RSVP_PINFO ] = { .len =sizeof (struct tc_rsvp_pinfo ) }, };
Greg Kroah-Hartman
maintainerPATCH 7/173 gtp: use __GFP_NOWARN to avoid memalloc warning
In reply to Greg Kroah-HartmanFrom: Taehee Yoo <[email protected]>
[ Upstream commit bd5cd35b782abf5437fbd01dfaee12437d20e832 ]
gtp hashtable size is received by user-space.
So, this hashtable size could be too large. If so, kmalloc will internally print a warning message.
This warning message is actually not necessary for the gtp module. So, this patch adds __GFP_NOWARN to avoid this message.
Splat looks like:
[ 2171.200049][ T1860] WARNING: CPU: 1 PID: 1860 at mm/page_alloc.c:4713 __alloc_pages_nodemask+0x2f3/0x740 [ 2171.238885][ T1860] Modules linked in: gtp veth openvswitch nsh nf_conncount nf_nat nf_conntrack nf_defrag_ipv] [ 2171.262680][ T1860] CPU: 1 PID: 1860 Comm: gtp-link Not tainted 5.5.0+ #321 [ 2171.263567][ T1860] Hardware name: innotek GmbH VirtualBox/VirtualBox, BIOS VirtualBox 12/01/2006 [ 2171.264681][ T1860] RIP: 0010:__alloc_pages_nodemask+0x2f3/0x740 [ 2171.265332][ T1860] Code: 64 fe ff ff 65 48 8b 04 25 c0 0f 02 00 48 05 f0 12 00 00 41 be 01 00 00 00 49 89 47 0 [ 2171.267301][ T1860] RSP: 0018:ffff8880b51af1f0 EFLAGS: 00010246 [ 2171.268320][ T1860] RAX: ffffed1016a35e43 RBX: 0000000000000000 RCX: 0000000000000000 [ 2171.269517][ T1860] RDX: 0000000000000000 RSI: 000000000000000b RDI: 0000000000000000 [ 2171.270305][ T1860] RBP: 0000000000040cc0 R08: ffffed1018893109 R09: dffffc0000000000 [ 2171.275973][ T1860] R10: 0000000000000001 R11: ffffed1018893108 R12: 1ffff11016a35e43 [ 2171.291039][ T1860] R13: 000000000000000b R14: 000000000000000b R15: 00000000000f4240 [ 2171.292328][ T1860] FS: 00007f53cbc83740(0000) GS:ffff8880da000000(0000) knlGS:0000000000000000 [ 2171.293409][ T1860] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 2171.294586][ T1860] CR2: 000055f540014508 CR3: 00000000b49f2004 CR4: 00000000000606e0 [ 2171.295424][ T1860] Call Trace: [ 2171.295756][ T1860] ? mark_held_locks+0xa5/0xe0 [ 2171.296659][ T1860] ? __alloc_pages_slowpath+0x21b0/0x21b0 [ 2171.298283][ T1860] ? gtp_encap_enable_socket+0x13e/0x400 [gtp] [ 2171.298962][ T1860] ? alloc_pages_current+0xc1/0x1a0 [ 2171.299475][ T1860] kmalloc_order+0x22/0x80 [ 2171.299936][ T1860] kmalloc_order_trace+0x1d/0x140 [ 2171.300437][ T1860] __kmalloc+0x302/0x3a0 [ 2171.300896][ T1860] gtp_newlink+0x293/0xba0 [gtp]
[ ... ]
--- drivers/net/gtp.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/net/gtp.c +4 -2
@@ -771,11 +771,13 @@ static int gtp_hashtable_new(struct gtp_ {int i; - gtp->addr_hash =kmalloc (sizeof (struct hlist_head ) * hsize,GFP_KERNEL ); + gtp->addr_hash =kmalloc (sizeof (struct hlist_head ) * hsize, +GFP_KERNEL | __GFP_NOWARN);if (gtp->addr_hash ==NULL )return -ENOMEM ; - gtp->tid_hash =kmalloc (sizeof (struct hlist_head ) * hsize,GFP_KERNEL ); + gtp->tid_hash =kmalloc (sizeof (struct hlist_head ) * hsize, +GFP_KERNEL | __GFP_NOWARN);if (gtp->tid_hash ==NULL ) goto err1;
Greg Kroah-Hartman
maintainerPATCH 8/173 l2tp: Allow duplicate session creation with UDP
In reply to Greg Kroah-HartmanFrom: Ridge Kennedy <[email protected]>
[ Upstream commit 0d0d9a388a858e271bb70e71e99e7fe2a6fd6f64 ]
In the past it was possible to create multiple L2TPv3 sessions with the same session id as long as the sessions belonged to different tunnels. The resulting sessions had issues when used with IP encapsulated tunnels, but worked fine with UDP encapsulated ones. Some applications began to rely on this behaviour to avoid having to negotiate unique session ids.
Some time ago a change was made to require session ids to be unique across all tunnels, breaking the applications making use of this "feature".
This change relaxes the duplicate session id check to allow duplicates if both of the colliding sessions belong to UDP encapsulated tunnels.
--- net/l2tp/l2tp_core.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
net/l2tp/l2tp_core.c +6 -1
@@ -358,8 +358,13 @@ static int l2tp_session_add_to_tunnel(stspin_lock_bh (&pn->l2tp_session_hlist_lock); +/* IP encap expects session IDs to be globally unique, while +* UDP encap doesn't. +*/ hlist_for_each_entry (session_walk, g_head, global_hlist) -if (session_walk->session_id == session->session_id) { +if (session_walk->session_id == session->session_id && + (session_walk->tunnel->encap ==L2TP_ENCAPTYPE_IP || + tunnel->encap ==L2TP_ENCAPTYPE_IP )) { err = -EEXIST ; goto err_tlock_pnlock; }
Greg Kroah-Hartman
maintainerPATCH 9/173 net: hsr: fix possible NULL deref in hsr_handle_frame()
In reply to Greg Kroah-HartmanFrom: Eric Dumazet <[email protected]>
[ Upstream commit 2b5b8251bc9fe2f9118411f037862ee17cf81e97 ]
hsr_port_get_rcu() can return NULL, so we need to be careful.
general protection fault, probably for non-canonical address 0xdffffc0000000006: 0000 [#1] PREEMPT SMP KASAN KASAN: null-ptr-deref in range [0x0000000000000030-0x0000000000000037] CPU: 1 PID: 10249 Comm: syz-executor.5 Not tainted 5.5.0-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 RIP: 0010:__read_once_size include/linux/compiler.h:199 [inline] RIP: 0010:hsr_addr_is_self+0x86/0x330 net/hsr/hsr_framereg.c:44 Code: 04 00 f3 f3 f3 65 48 8b 04 25 28 00 00 00 48 89 45 d0 31 c0 e8 6b ff 94 f9 4c 89 f2 48 b8 00 00 00 00 00 fc ff df 48 c1 ea 03 <80> 3c 02 00 0f 85 75 02 00 00 48 8b 43 30 49 39 c6 49 89 47 c0 0f RSP: 0018:ffffc90000da8a90 EFLAGS: 00010206 RAX: dffffc0000000000 RBX: 0000000000000000 RCX: ffffffff87e0cc33 RDX: 0000000000000006 RSI: ffffffff87e035d5 RDI: 0000000000000000 RBP: ffffc90000da8b20 R08: ffff88808e7de040 R09: ffffed1015d2707c R10: ffffed1015d2707b R11: ffff8880ae9383db R12: ffff8880a689bc5e R13: 1ffff920001b5153 R14: 0000000000000030 R15: ffffc90000da8af8 FS: 00007fd7a42be700(0000) GS:ffff8880ae900000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000001b32338000 CR3: 00000000a928c000 CR4: 00000000001406e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: <IRQ> hsr_handle_frame+0x1c5/0x630 net/hsr/hsr_slave.c:31 __netif_receive_skb_core+0xfbc/0x30b0 net/core/dev.c:5099 __netif_receive_skb_one_core+0xa8/0x1a0 net/core/dev.c:5196 __netif_receive_skb+0x2c/0x1d0 net/core/dev.c:5312 process_backlog+0x206/0x750 net/core/dev.c:6144 napi_poll net/core/dev.c:6582 [inline] net_rx_action+0x508/0x1120 net/core/dev.c:6650 __do_softirq+0x262/0x98c kernel/softirq.c:292 do_softirq_own_stack+0x2a/0x40 arch/x86/entry/entry_64.S:1082 </IRQ>
--- net/hsr/hsr_slave.c | 2 ++ 1 file changed, 2 insertions(+)
net/hsr/hsr_slave.c +2 -0
@@ -32,6 +32,8 @@ static rx_handler_result_t hsr_handle_frrcu_read_lock ();/* hsr->node_db, hsr->ports */ port =hsr_port_get_rcu (skb->dev); +if (!port) + goto finish_pass;if (hsr_addr_is_self (port->hsr,eth_hdr (skb)->h_source)) {/* Directly kill frames sent by ourselves */
Greg Kroah-Hartman
maintainerPATCH 10/173 net_sched: fix an OOB access in cls_tcindex
In reply to Greg Kroah-HartmanFrom: Cong Wang <[email protected]>
[ Upstream commit 599be01ee567b61f4471ee8078870847d0a11e8e ]
As Eric noticed, tcindex_alloc_perfect_hash() uses cp->hash to compute the size of memory allocation, but cp->hash is set again after the allocation, this caused an out-of-bound access.
So we have to move all cp->hash initialization and computation before the memory allocation. Move cp->mask and cp->shift together as cp->hash may need them for computation too.
--- net/sched/cls_tcindex.c | 40 ++++++++++++++++++++-------------------- 1 file changed, 20 insertions(+), 20 deletions(-)
net/sched/cls_tcindex.c +20 -20
@@ -351,12 +351,31 @@ tcindex_set_parms(struct net *net, struc cp->fall_through = p->fall_through; cp->tp = tp; +if (tb[TCA_TCINDEX_HASH ]) + cp->hash =nla_get_u32 (tb[TCA_TCINDEX_HASH ]); + +if (tb[TCA_TCINDEX_MASK ]) + cp->mask =nla_get_u16 (tb[TCA_TCINDEX_MASK ]); + +if (tb[TCA_TCINDEX_SHIFT ]) + cp->shift =nla_get_u32 (tb[TCA_TCINDEX_SHIFT ]); + +if (!cp->hash) { +/* Hash not specified, use perfect hash if the upper limit +* of the hashing index is below the threshold. +*/ +if ((cp->mask >> cp->shift) <PERFECT_HASH_THRESHOLD ) + cp->hash = (cp->mask >> cp->shift) +1 ; +else + cp->hash =DEFAULT_HASH_SIZE ; + } +if (p->perfect) {int i;if (tcindex_alloc_perfect_hash (cp) <0 ) goto errout; -for (i =0 ; i < cp->hash; i++) +for (i =0 ; i <min (cp->hash, p->hash); i++) cp->perfect[i].res = p->perfect[i].res; balloc =1 ; }
@@ -368,15 +387,6 @@ tcindex_set_parms(struct net *net, strucif (old_r) cr = r->res; -if (tb[TCA_TCINDEX_HASH ]) - cp->hash =nla_get_u32 (tb[TCA_TCINDEX_HASH ]); - -if (tb[TCA_TCINDEX_MASK ]) - cp->mask =nla_get_u16 (tb[TCA_TCINDEX_MASK ]); - -if (tb[TCA_TCINDEX_SHIFT ]) - cp->shift =nla_get_u32 (tb[TCA_TCINDEX_SHIFT ]); - err = -EBUSY ;/* Hash already allocated, make sure that we still meet the
@@ -394,16 +404,6 @@ tcindex_set_parms(struct net *net, strucif (tb[TCA_TCINDEX_FALL_THROUGH ]) cp->fall_through =nla_get_u32 (tb[TCA_TCINDEX_FALL_THROUGH ]); -if (!cp->hash) { -/* Hash not specified, use perfect hash if the upper limit -* of the hashing index is below the threshold. -*/ -if ((cp->mask >> cp->shift) <PERFECT_HASH_THRESHOLD ) - cp->hash = (cp->mask >> cp->shift) +1 ; -else - cp->hash =DEFAULT_HASH_SIZE ; - } -if (!cp->perfect && !cp->h) cp->alloc_hash = cp->hash;
Greg Kroah-Hartman
maintainerPATCH 11/173 bnxt_en: Fix TC queue mapping.
In reply to Greg Kroah-HartmanFrom: Michael Chan <[email protected]>
[ Upstream commit 18e4960c18f484ac288f41b43d0e6c4c88e6ea78 ]
The driver currently only calls netdev_set_tc_queue when the number of TCs is greater than 1. Instead, the comparison should be greater than or equal to 1. Even with 1 TC, we need to set the queue mapping.
This bug can cause warnings when the number of TCs is changed back to 1.
--- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/broadcom/bnxt/bnxt.c +1 -1
@@ -5375,7 +5375,7 @@ static void bnxt_setup_msix(struct bnxtint tcs, i; tcs =netdev_get_num_tc (dev); -if (tcs >1 ) { +if (tcs) {int i, off, count;for (i =0 ; i < tcs; i++) {
Greg Kroah-Hartman
maintainerPATCH 12/173 tcp: clear tp->total_retrans in tcp_disconnect()
In reply to Greg Kroah-HartmanFrom: Eric Dumazet <[email protected]>
[ Upstream commit c13c48c00a6bc1febc73902505bdec0967bd7095 ]
total_retrans needs to be cleared in tcp_disconnect().
tcp_disconnect() is rarely used, but it is worth fixing it.
--- net/ipv4/tcp.c | 1 + 1 file changed, 1 insertion(+)
net/ipv4/tcp.c +1 -0
@@ -2366,6 +2366,7 @@ int tcp_disconnect(struct sock *sk, inttcp_set_ca_state (sk, TCP_CA_Open); tp->is_sack_reneg =0 ;tcp_clear_retrans (tp); + tp->total_retrans =0 ;inet_csk_delack_init (sk);/* Initialize rcv_mss to TCP_MIN_MSS to avoid division by 0 * issue in __tcp_select_window()
Greg Kroah-Hartman
maintainerPATCH 13/173 tcp: clear tp->delivered in tcp_disconnect()
In reply to Greg Kroah-HartmanFrom: Eric Dumazet <[email protected]>
[ Upstream commit 2fbdd56251b5c62f96589f39eded277260de7267 ]
tp->delivered needs to be cleared in tcp_disconnect().
tcp_disconnect() is rarely used, but it is worth fixing it.
--- net/ipv4/tcp.c | 1 + 1 file changed, 1 insertion(+)
net/ipv4/tcp.c +1 -0
@@ -2363,6 +2363,7 @@ int tcp_disconnect(struct sock *sk, int tp->snd_ssthresh =TCP_INFINITE_SSTHRESH ; tp->snd_cwnd_cnt =0 ; tp->window_clamp =0 ; + tp->delivered =0 ;tcp_set_ca_state (sk, TCP_CA_Open); tp->is_sack_reneg =0 ;tcp_clear_retrans (tp);
Greg Kroah-Hartman
maintainerPATCH 14/173 tcp: clear tp->data_segs{in|out} in tcp_disconnect()
In reply to Greg Kroah-HartmanFrom: Eric Dumazet <[email protected]>
[ Upstream commit db7ffee6f3eb3683cdcaeddecc0a630a14546fe3 ]
tp->data_segs_in and tp->data_segs_out need to be cleared in tcp_disconnect().
tcp_disconnect() is rarely used, but it is worth fixing it.
--- net/ipv4/tcp.c | 2 ++ 1 file changed, 2 insertions(+)
net/ipv4/tcp.c +2 -0
@@ -2381,6 +2381,8 @@ int tcp_disconnect(struct sock *sk, inttcp_saved_syn_free (tp); tp->bytes_acked =0 ; tp->bytes_received =0 ; + tp->data_segs_in =0 ; + tp->data_segs_out =0 ;/* Clean up fastopen related fields */ tcp_free_fastopen_req (tp);
Greg Kroah-Hartman
maintainerPATCH 15/173 tcp: clear tp->segs_{in|out} in tcp_disconnect()
In reply to Greg Kroah-HartmanFrom: Eric Dumazet <[email protected]>
[ Upstream commit 784f8344de750a41344f4bbbebb8507a730fc99c ]
tp->segs_in and tp->segs_out need to be cleared in tcp_disconnect().
tcp_disconnect() is rarely used, but it is worth fixing it.
--- net/ipv4/tcp.c | 2 ++ 1 file changed, 2 insertions(+)
net/ipv4/tcp.c +2 -0
@@ -2379,6 +2379,8 @@ int tcp_disconnect(struct sock *sk, intdst_release (sk->sk_rx_dst); sk->sk_rx_dst =NULL ;tcp_saved_syn_free (tp); + tp->segs_in =0 ; + tp->segs_out =0 ; tp->bytes_acked =0 ; tp->bytes_received =0 ; tp->data_segs_in =0 ;
Greg Kroah-Hartman
maintainerPATCH 16/173 rxrpc: Fix insufficient receive notification generation
In reply to Greg Kroah-HartmanFrom: David Howells <[email protected]>
[ Upstream commit f71dbf2fb28489a79bde0dca1c8adfb9cdb20a6b ]
In rxrpc_input_data(), rxrpc_notify_socket() is called if the base sequence number of the packet is immediately following the hard-ack point at the end of the function. However, this isn't sufficient, since the recvmsg side may have been advancing the window and then overrun the position in which we're adding - at which point rx_hard_ack >= seq0 and no notification is generated.
Fix this by always generating a notification at the end of the input function.
Without this, a long call may stall, possibly indefinitely.
--- net/rxrpc/input.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
net/rxrpc/input.c +1 -2
@@ -585,8 +585,7 @@ ack: immediate_ack, true, rxrpc_propose_ack_input_data); -if (sp->hdr.seq ==READ_ONCE (call->rx_hard_ack) +1 ) -rxrpc_notify_socket (call); +rxrpc_notify_socket (call);_leave (" [queued]" ); }
Greg Kroah-Hartman
maintainerPATCH 17/173 rxrpc: Fix NULL pointer deref due to call->conn being cleared on disconnect
In reply to Greg Kroah-HartmanFrom: David Howells <[email protected]>
[ Upstream commit 5273a191dca65a675dc0bcf3909e59c6933e2831 ]
When a call is disconnected, the connection pointer from the call is cleared to make sure it isn't used again and to prevent further attempted transmission for the call. Unfortunately, there might be a daemon trying to use it at the same time to transmit a packet.
Fix this by keeping call->conn set, but setting a flag on the call to indicate disconnection instead.
Remove also the bits in the transmission functions where the conn pointer is checked and a ref taken under spinlock as this is now redundant.
--- net/rxrpc/ar-internal.h | 1 + net/rxrpc/call_object.c | 4 ++-- net/rxrpc/conn_client.c | 3 +-- net/rxrpc/conn_object.c | 4 ++-- net/rxrpc/output.c | 26 +++++++++----------------- 5 files changed, 15 insertions(+), 23 deletions(-)
net/rxrpc/ar-internal.h +1 -0
@@ -451,6 +451,7 @@ enum rxrpc_call_flag {RXRPC_CALL_SEND_PING ,/* A ping will need to be sent */ RXRPC_CALL_PINGING ,/* Ping in process */ RXRPC_CALL_RETRANS_TIMEOUT ,/* Retransmission due to timeout occurred */ + RXRPC_CALL_DISCONNECTED, /* The call has been disconnected */ };/*
net/rxrpc/call_object.c +2 -2
@@ -505,7 +505,7 @@ void rxrpc_release_call(struct rxrpc_soc_debug ("RELEASE CALL %p (%d CONN %p)" , call, call->debug_id, conn); -if (conn) +if (conn && !test_bit (RXRPC_CALL_DISCONNECTED , &call->flags))rxrpc_disconnect_call (call);for (i =0 ; i <RXRPC_RXTX_BUFF_SIZE ; i++) {
@@ -639,6 +639,7 @@ static void rxrpc_rcu_destroy_call(struc { struct rxrpc_call *call = container_of(rcu, struct rxrpc_call, rcu); +rxrpc_put_connection (call->conn);rxrpc_put_peer (call->peer);kfree (call->rxtx_buffer);kfree (call->rxtx_annotations);
@@ -660,7 +661,6 @@ void rxrpc_cleanup_call(struct rxrpc_cal ASSERTCMP(call->state, ==, RXRPC_CALL_COMPLETE);ASSERT (test_bit (RXRPC_CALL_RELEASED , &call->flags)); - ASSERTCMP(call->conn, ==, NULL);/* Clean up the Rx/Tx buffer */ for (i =0 ; i <RXRPC_RXTX_BUFF_SIZE ; i++)
net/rxrpc/conn_client.c +1 -2
@@ -762,9 +762,9 @@ void rxrpc_disconnect_client_call(structstruct rxrpc_net *rxnet =rxrpc_net (sock_net (&call->socket->sk));trace_rxrpc_client (conn, channel, rxrpc_client_chan_disconnect); - call->conn =NULL ;spin_lock (&conn->channel_lock); +set_bit (RXRPC_CALL_DISCONNECTED , &call->flags);/* Calls that have never actually been assigned a channel can simply be * discarded. If the conn didn't get used either, it will follow
@@ -863,7 +863,6 @@ out:spin_unlock (&rxnet->client_conn_cache_lock); out_2:spin_unlock (&conn->channel_lock); -rxrpc_put_connection (conn);_leave ("" );return ;
net/rxrpc/conn_object.c +2 -2
@@ -163,6 +163,8 @@ void __rxrpc_disconnect_call(struct rxrp_enter ("%d,%x" , conn->debug_id, call->cid); +set_bit (RXRPC_CALL_DISCONNECTED , &call->flags); +if (rcu_access_pointer (chan->call) == call) {/* Save the result of the call so that we can repeat it if necessary * through the channel, whilst disposing of the actual call record.
@@ -207,9 +209,7 @@ void rxrpc_disconnect_call(struct rxrpc___rxrpc_disconnect_call (conn, call);spin_unlock (&conn->channel_lock); - call->conn =NULL ; conn->idle_timestamp = jiffies; -rxrpc_put_connection (conn); }/*
net/rxrpc/output.c +9 -17
@@ -96,7 +96,7 @@ static size_t rxrpc_fill_out_ack(struct*/ int rxrpc_send_ack_packet (struct rxrpc_call *call,bool ping) { -struct rxrpc_connection *conn =NULL ; +struct rxrpc_connection *conn;struct rxrpc_ack_buffer *pkt;struct msghdr msg;struct kvec iov[2 ];
@@ -106,18 +106,14 @@ int rxrpc_send_ack_packet(struct rxrpc_cint ret;u8 reason; -spin_lock_bh (&call->lock); -if (call->conn) - conn =rxrpc_get_connection_maybe (call->conn); -spin_unlock_bh (&call->lock); -if (!conn) +if (test_bit (RXRPC_CALL_DISCONNECTED , &call->flags))return -ECONNRESET ; pkt =kzalloc (sizeof (*pkt),GFP_KERNEL ); -if (!pkt) { -rxrpc_put_connection (conn); +if (!pkt)return -ENOMEM ; - } + + conn = call->conn; msg.msg_name = &call->peer->srx.transport; msg.msg_namelen = call->peer->srx.transport_len;
@@ -204,7 +200,6 @@ int rxrpc_send_ack_packet(struct rxrpc_c } out: -rxrpc_put_connection (conn);kfree (pkt);return ret; }
@@ -214,20 +209,18 @@ out:*/ int rxrpc_send_abort_packet (struct rxrpc_call *call) { -struct rxrpc_connection *conn =NULL ; +struct rxrpc_connection *conn;struct rxrpc_abort_buffer pkt;struct msghdr msg;struct kvec iov[1 ];rxrpc_serial_t serial;int ret; -spin_lock_bh (&call->lock); -if (call->conn) - conn =rxrpc_get_connection_maybe (call->conn); -spin_unlock_bh (&call->lock); -if (!conn) +if (test_bit (RXRPC_CALL_DISCONNECTED , &call->flags))return -ECONNRESET ; + conn = call->conn; + msg.msg_name = &call->peer->srx.transport; msg.msg_namelen = call->peer->srx.transport_len; msg.msg_control =NULL ;
@@ -255,7 +248,6 @@ int rxrpc_send_abort_packet(struct rxrpc ret =kernel_sendmsg (conn->params.local->socket, &msg, iov,1 ,sizeof (pkt)); -rxrpc_put_connection (conn);return ret; }
Greg Kroah-Hartman
maintainerPATCH 18/173 media: uvcvideo: Avoid cyclic entity chains due to malformed USB descriptors
In reply to Greg Kroah-HartmanFrom: Will Deacon <[email protected]>
commit 68035c80e129c4cfec659aac4180354530b26527 upstream.
Way back in 2017, fuzzing the 4.14-rc2 USB stack with syzkaller kicked up the following WARNING from the UVC chain scanning code:
| list_add double add: new=ffff880069084010, prev=ffff880069084010, | next=ffff880067d22298. | ------------[ cut here ]------------ | WARNING: CPU: 1 PID: 1846 at lib/list_debug.c:31 __list_add_valid+0xbd/0xf0 | Modules linked in: | CPU: 1 PID: 1846 Comm: kworker/1:2 Not tainted | 4.14.0-rc2-42613-g1488251d1a98 #238 | Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011 | Workqueue: usb_hub_wq hub_event | task: ffff88006b01ca40 task.stack: ffff880064358000 | RIP: 0010:__list_add_valid+0xbd/0xf0 lib/list_debug.c:29 | RSP: 0018:ffff88006435ddd0 EFLAGS: 00010286 | RAX: 0000000000000058 RBX: ffff880067d22298 RCX: 0000000000000000 | RDX: 0000000000000058 RSI: ffffffff85a58800 RDI: ffffed000c86bbac | RBP: ffff88006435dde8 R08: 1ffff1000c86ba52 R09: 0000000000000000 | R10: 0000000000000002 R11: 0000000000000000 R12: ffff880069084010 | R13: ffff880067d22298 R14: ffff880069084010 R15: ffff880067d222a0 | FS: 0000000000000000(0000) GS:ffff88006c900000(0000) knlGS:0000000000000000 | CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 | CR2: 0000000020004ff2 CR3: 000000006b447000 CR4: 00000000000006e0 | Call Trace: | __list_add ./include/linux/list.h:59 | list_add_tail+0x8c/0x1b0 ./include/linux/list.h:92 | uvc_scan_chain_forward.isra.8+0x373/0x416 | drivers/media/usb/uvc/uvc_driver.c:1471 | uvc_scan_chain drivers/media/usb/uvc/uvc_driver.c:1585 | uvc_scan_device drivers/media/usb/uvc/uvc_driver.c:1769 | uvc_probe+0x77f2/0x8f00 drivers/media/usb/uvc/uvc_driver.c:2104
Looking into the output from usbmon, the interesting part is the following data packet:
ffff880069c63e00 30710169 C Ci:1:002:0 0 143 = 09028f00 01030080 00090403 00000e01 00000924 03000103 7c003328 010204db
If we drop the lead configuration and interface descriptors, we're left with an output terminal descriptor describing a generic display:
/* Output terminal descriptor */
buf[0] 09
buf[1] 24
buf[2] 03 /* UVC_VC_OUTPUT_TERMINAL */
buf[3] 00 /* ID */
buf[4] 01 /* type == 0x0301 (UVC_OTT_DISPLAY) */
buf[5] 03
buf[6] 7c
buf[7] 00 /* source ID refers to self! */
buf[8] 33
The problem with this descriptor is that it is self-referential: the source ID of 0 matches itself! This causes the 'struct uvc_entity' representing the display to be added to its chain list twice during 'uvc_scan_chain()': once via 'uvc_scan_chain_entity()' when it is processed directly from the 'dev->entities' list and then again immediately afterwards when trying to follow the source ID in 'uvc_scan_chain_forward()'
Add a check before adding an entity to a chain list to ensure that the entity is not already part of a chain.
--- drivers/media/usb/uvc/uvc_driver.c | 12 ++++++++++++ 1 file changed, 12 insertions(+)
drivers/media/usb/uvc/uvc_driver.c +12 -0
@@ -1446,6 +1446,11 @@ static int uvc_scan_chain_forward(structbreak ;if (forward == prev)continue ; +if (forward->chain.next || forward->chain.prev) { +uvc_trace (UVC_TRACE_DESCR ,"Found reference to " +"entity %d already in chain.\n" , forward->id); +return -EINVAL ; + }switch (UVC_ENTITY_TYPE (forward)) {case UVC_VC_EXTENSION_UNIT :
@@ -1527,6 +1532,13 @@ static int uvc_scan_chain_backward(strucreturn -1 ; } +if (term->chain.next || term->chain.prev) { +uvc_trace (UVC_TRACE_DESCR ,"Found reference to " +"entity %d already in chain.\n" , + term->id); +return -EINVAL ; + } +if (uvc_trace_param &UVC_TRACE_PROBE )printk (KERN_CONT " %d" , term->id);
Greg Kroah-Hartman
maintainerPATCH 19/173 mfd: dln2: More sanity checking for endpoints
In reply to Greg Kroah-HartmanFrom: Oliver Neukum <[email protected]>
commit 2b8bd606b1e60ca28c765f69c1eedd7d2a2e9dca upstream.
It is not enough to check for the number of endpoints. The types must also be correct.
--- drivers/mfd/dln2.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-)
drivers/mfd/dln2.c +11 -2
@@ -729,6 +729,8 @@ static int dln2_probe(struct usb_interfa const struct usb_device_id *usb_id) {struct usb_host_interface *hostif = interface->cur_altsetting; +struct usb_endpoint_descriptor *epin; +struct usb_endpoint_descriptor *epout;struct device *dev = &interface->dev;struct dln2_dev *dln2;int ret;
@@ -738,12 +740,19 @@ static int dln2_probe(struct usb_interfa hostif->desc.bNumEndpoints < 2)return -ENODEV ; + epin = &hostif->endpoint[0 ].desc; + epout = &hostif->endpoint[1 ].desc; +if (!usb_endpoint_is_bulk_out (epout)) +return -ENODEV ; +if (!usb_endpoint_is_bulk_in (epin)) +return -ENODEV ; + dln2 =kzalloc (sizeof (*dln2),GFP_KERNEL );if (!dln2)return -ENOMEM ; - dln2->ep_out = hostif->endpoint[0 ].desc.bEndpointAddress; - dln2->ep_in = hostif->endpoint[1 ].desc.bEndpointAddress; + dln2->ep_out = epout->bEndpointAddress; + dln2->ep_in = epin->bEndpointAddress; dln2->usb_dev =usb_get_dev (interface_to_usbdev (interface)); dln2->interface = interface;usb_set_intfdata (interface, dln2);
Greg Kroah-Hartman
maintainerPATCH 20/173 tracing: Fix sched switch start/stop refcount racy updates
In reply to Greg Kroah-HartmanFrom: Mathieu Desnoyers <[email protected]>
commit 64ae572bc7d0060429e40e1c8d803ce5eb31a0d6 upstream.
Reading the sched_cmdline_ref and sched_tgid_ref initial state within tracing_start_sched_switch without holding the sched_register_mutex is racy against concurrent updates, which can lead to tracepoint probes being registered more than once (and thus trigger warnings within tracepoint.c).
--- kernel/trace/trace_sched_switch.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
kernel/trace/trace_sched_switch.c +3 -1
@@ -89,8 +89,10 @@ static void tracing_sched_unregister(voistatic void tracing_start_sched_switch (int ops) { -bool sched_register = (!sched_cmdline_ref && !sched_tgid_ref); +bool sched_register; +mutex_lock (&sched_register_mutex); + sched_register = (!sched_cmdline_ref && !sched_tgid_ref);switch (ops) {case RECORD_CMDLINE :
Greg Kroah-Hartman
maintainerPATCH 21/173 brcmfmac: Fix memory leak in brcmf_usbdev_qinit
In reply to Greg Kroah-HartmanFrom: Navid Emamdoost <[email protected]>
commit 4282dc057d750c6a7dd92953564b15c26b54c22c upstream.
In the implementation of brcmf_usbdev_qinit() the allocated memory for reqs is leaking if usb_alloc_urb() fails. Release reqs in the error handling path.
--- drivers/net/wireless/broadcom/brcm80211/brcmfmac/usb.c | 1 + 1 file changed, 1 insertion(+)
drivers/net/wireless/broadcom/brcm80211/brcmfmac/usb.c +1 -0
@@ -441,6 +441,7 @@ fail:usb_free_urb (req->urb);list_del (q->next); } +kfree (reqs);return NULL ; }
Greg Kroah-Hartman
maintainerPATCH 22/173 usb: gadget: legacy: set max_speed to super-speed
In reply to Greg Kroah-HartmanFrom: Roger Quadros <[email protected]>
commit 463f67aec2837f981b0a0ce8617721ff59685c00 upstream.
These interfaces do support super-speed so let's not limit maximum speed to high-speed.
--- drivers/usb/gadget/legacy/cdc2.c | 2 +- drivers/usb/gadget/legacy/g_ffs.c | 2 +- drivers/usb/gadget/legacy/multi.c | 2 +- drivers/usb/gadget/legacy/ncm.c | 2 +- 4 files changed, 4 insertions(+), 4 deletions(-)
drivers/usb/gadget/legacy/cdc2.c +1 -1
@@ -229,7 +229,7 @@ static struct usb_composite_driver cdc_d .name = "g_cdc", .dev = &device_desc, .strings = dev_strings, - .max_speed = USB_SPEED_HIGH, + .max_speed = USB_SPEED_SUPER, .bind = cdc_bind, .unbind = cdc_unbind, };
drivers/usb/gadget/legacy/g_ffs.c +1 -1
@@ -153,7 +153,7 @@ static struct usb_composite_driver gfs_d .name = DRIVER_NAME, .dev = &gfs_dev_desc, .strings = gfs_dev_strings, - .max_speed = USB_SPEED_HIGH, + .max_speed = USB_SPEED_SUPER, .bind = gfs_bind, .unbind = gfs_unbind, };
drivers/usb/gadget/legacy/multi.c +1 -1
@@ -486,7 +486,7 @@ static struct usb_composite_driver multi .name = "g_multi", .dev = &device_desc, .strings = dev_strings, - .max_speed = USB_SPEED_HIGH, + .max_speed = USB_SPEED_SUPER, .bind = multi_bind, .unbind = multi_unbind, .needs_serial = 1,
drivers/usb/gadget/legacy/ncm.c +1 -1
@@ -203,7 +203,7 @@ static struct usb_composite_driver ncm_d .name = "g_ncm", .dev = &device_desc, .strings = dev_strings, - .max_speed = USB_SPEED_HIGH, + .max_speed = USB_SPEED_SUPER, .bind = gncm_bind, .unbind = gncm_unbind, };
Greg Kroah-Hartman
maintainerPATCH 23/173 usb: gadget: f_ncm: Use atomic_t to track in-flight request
In reply to Greg Kroah-HartmanFrom: Bryan O'Donoghue <[email protected]>
commit 5b24c28cfe136597dc3913e1c00b119307a20c7e upstream.
Currently ncm->notify_req is used to flag when a request is in-flight. ncm->notify_req is set to NULL and when a request completes it is subsequently reset.
This is fundamentally buggy in that the unbind logic of the NCM driver will unconditionally free ncm->notify_req leading to a NULL pointer dereference.
--- drivers/usb/gadget/function/f_ncm.c | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-)
drivers/usb/gadget/function/f_ncm.c +13 -4
@@ -58,6 +58,7 @@ struct f_ncm {struct usb_ep *notify;struct usb_request *notify_req;u8 notify_state; +atomic_t notify_count;bool is_open;const struct ndp_parser_opts *parser_opts;
@@ -553,7 +554,7 @@ static void ncm_do_notify(struct f_ncm *int status;/* notification already in flight? */ -if (!req) +if (atomic_read (&ncm->notify_count))return ; event = req->buf;
@@ -593,7 +594,8 @@ static void ncm_do_notify(struct f_ncm * event->bmRequestType =0xA1 ; event->wIndex =cpu_to_le16 (ncm->ctrl_id); - ncm->notify_req =NULL ; +atomic_inc (&ncm->notify_count); +/* * In double buffering if there is a space in FIFO, * completion callback can be called right after the call,
@@ -603,7 +605,7 @@ static void ncm_do_notify(struct f_ncm * status =usb_ep_queue (ncm->notify, req,GFP_ATOMIC );spin_lock (&ncm->lock);if (status <0 ) { - ncm->notify_req = req; +atomic_dec (&ncm->notify_count);DBG (cdev,"notify --> %d\n" , status); } }
@@ -638,17 +640,19 @@ static void ncm_notify_complete(struct ucase 0 :VDBG (cdev,"Notification %02x sent\n" , event->bNotificationType); +atomic_dec (&ncm->notify_count);break ;case -ECONNRESET :case -ESHUTDOWN : +atomic_set (&ncm->notify_count,0 ); ncm->notify_state =NCM_NOTIFY_NONE ;break ;default :DBG (cdev,"event %02x --> %d\n" , event->bNotificationType, req->status); +atomic_dec (&ncm->notify_count);break ; } - ncm->notify_req = req;ncm_do_notify (ncm);spin_unlock (&ncm->lock); }
@@ -1632,6 +1636,11 @@ static void ncm_unbind(struct usb_config ncm_string_defs[0 ].id =0 ;usb_free_all_descriptors (f); +if (atomic_read (&ncm->notify_count)) { +usb_ep_dequeue (ncm->notify, ncm->notify_req); +atomic_set (&ncm->notify_count,0 ); + } +kfree (ncm->notify_req->buf);usb_ep_free_request (ncm->notify, ncm->notify_req); }
Greg Kroah-Hartman
maintainerPATCH 24/173 usb: gadget: f_ecm: Use atomic_t to track in-flight request
In reply to Greg Kroah-HartmanFrom: Bryan O'Donoghue <[email protected]>
commit d710562e01c48d59be3f60d58b7a85958b39aeda upstream.
Currently ecm->notify_req is used to flag when a request is in-flight. ecm->notify_req is set to NULL and when a request completes it is subsequently reset.
This is fundamentally buggy in that the unbind logic of the ECM driver will unconditionally free ecm->notify_req leading to a NULL pointer dereference.
--- drivers/usb/gadget/function/f_ecm.c | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-)
drivers/usb/gadget/function/f_ecm.c +12 -4
@@ -56,6 +56,7 @@ struct f_ecm {struct usb_ep *notify;struct usb_request *notify_req;u8 notify_state; +atomic_t notify_count;bool is_open;/* FIXME is_open needs some irq-ish locking
@@ -384,7 +385,7 @@ static void ecm_do_notify(struct f_ecm *int status;/* notification already in flight? */ -if (!req) +if (atomic_read (&ecm->notify_count))return ; event = req->buf;
@@ -424,10 +425,10 @@ static void ecm_do_notify(struct f_ecm * event->bmRequestType =0xA1 ; event->wIndex =cpu_to_le16 (ecm->ctrl_id); - ecm->notify_req =NULL ; +atomic_inc (&ecm->notify_count); status =usb_ep_queue (ecm->notify, req,GFP_ATOMIC );if (status <0 ) { - ecm->notify_req = req; +atomic_dec (&ecm->notify_count);DBG (cdev,"notify --> %d\n" , status); } }
@@ -452,17 +453,19 @@ static void ecm_notify_complete(struct uswitch (req->status) {case 0 :/* no fault */ +atomic_dec (&ecm->notify_count);break ;case -ECONNRESET :case -ESHUTDOWN : +atomic_set (&ecm->notify_count,0 ); ecm->notify_state =ECM_NOTIFY_NONE ;break ;default :DBG (cdev,"event %02x --> %d\n" , event->bNotificationType, req->status); +atomic_dec (&ecm->notify_count);break ; } - ecm->notify_req = req;ecm_do_notify (ecm); }
@@ -909,6 +912,11 @@ static void ecm_unbind(struct usb_configusb_free_all_descriptors (f); +if (atomic_read (&ecm->notify_count)) { +usb_ep_dequeue (ecm->notify, ecm->notify_req); +atomic_set (&ecm->notify_count,0 ); + } +kfree (ecm->notify_req->buf);usb_ep_free_request (ecm->notify, ecm->notify_req); }
Greg Kroah-Hartman
maintainerPATCH 25/173 ALSA: dummy: Fix PCM format loop in proc output
In reply to Greg Kroah-HartmanFrom: Takashi Iwai <[email protected]>
commit 2acf25f13ebe8beb40e97a1bbe76f36277c64f1e upstream.
The loop termination for iterating over all formats should contain SNDRV_PCM_FORMAT_LAST, not less than it.
--- sound/drivers/dummy.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
sound/drivers/dummy.c +1 -1
@@ -933,7 +933,7 @@ static void print_formats(struct snd_dum {int i; -for (i =0 ; i <SNDRV_PCM_FORMAT_LAST ; i++) { +for (i =0 ; i <=SNDRV_PCM_FORMAT_LAST ; i++) {if (dummy->pcm_hw.formats & (1ULL << i))snd_iprintf (buffer," %s" ,snd_pcm_format_name (i)); }
Greg Kroah-Hartman
maintainerPATCH 26/173 media/v4l2-core: set pages dirty upon releasing DMA buffers
In reply to Greg Kroah-HartmanFrom: John Hubbard <[email protected]>
commit 3c7470b6f68434acae459482ab920d1e3fabd1c7 upstream.
After DMA is complete, and the device and CPU caches are synchronized, it's still required to mark the CPU pages as dirty, if the data was coming from the device. However, this driver was just issuing a bare put_page() call, without any set_page_dirty*() call.
Fix the problem, by calling set_page_dirty_lock() if the CPU pages were potentially receiving data from the device.
--- drivers/media/v4l2-core/videobuf-dma-sg.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/media/v4l2-core/videobuf-dma-sg.c +4 -1
@@ -352,8 +352,11 @@ int videobuf_dma_free(struct videobuf_dmBUG_ON (dma->sglen);if (dma->pages) { -for (i =0 ; i < dma->nr_pages; i++) +for (i =0 ; i < dma->nr_pages; i++) { +if (dma->direction ==DMA_FROM_DEVICE ) +set_page_dirty_lock (dma->pages[i]);put_page (dma->pages[i]); + }kfree (dma->pages); dma->pages =NULL ; }
Greg Kroah-Hartman
maintainerPATCH 27/173 media: v4l2-rect.h: fix v4l2_rect_map_inside() top/left adjustments
In reply to Greg Kroah-HartmanFrom: Helen Koike <[email protected]>
commit f51e50db4c20d46930b33be3f208851265694f3e upstream.
boundary->width and boundary->height are sizes relative to boundary->left and boundary->top coordinates, but they were not being taken into consideration to adjust r->left and r->top, leading to the following error:
Consider the follow as initial values for boundary and r:
struct v4l2_rect boundary = {
.left = 100,
.top = 100,
.width = 800,
.height = 600,
}struct v4l2_rect r = {
.left = 0,
.top = 0,
.width = 1920,
.height = 960,
}calling v4l2_rect_map_inside(&r, &boundary) was modifying r to:
r = {
.left = 0,
.top = 0,
.width = 800,
.height = 600,
}Which is wrongly outside the boundary rectangle, because:
v4l2_rect_set_max_size(r, boundary); // r->width = 800, r->height = 600 ... if (r->left + r->width > boundary->width) // true r->left = boundary->width - r->width; // r->left = 800 - 800 if (r->top + r->height > boundary->height) // true r->top = boundary->height - r->height; // r->height = 600 - 600
Fix this by considering top/left coordinates from boundary.
--- include/media/v4l2-rect.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-)
include/media/v4l2-rect.h +4 -4
@@ -75,10 +75,10 @@ static inline void v4l2_rect_map_inside( r->left = boundary->left;if (r->top < boundary->top) r->top = boundary->top; -if (r->left + r->width > boundary->width) - r->left = boundary->width - r->width; -if (r->top + r->height > boundary->height) - r->top = boundary->height - r->height; +if (r->left + r->width > boundary->left + boundary->width) + r->left = boundary->left + boundary->width - r->width; +if (r->top + r->height > boundary->top + boundary->height) + r->top = boundary->top + boundary->height - r->height; }/**
Greg Kroah-Hartman
maintainerPATCH 28/173 lib/test_kasan.c: fix memory leak in kmalloc_oob_krealloc_more()
In reply to Greg Kroah-HartmanFrom: Gustavo A. R. Silva <[email protected]>
commit 3e21d9a501bf99aee2e5835d7f34d8c823f115b5 upstream.
In case memory resources for _ptr2_ were allocated, release them before return.
Notice that in case _ptr1_ happens to be NULL, krealloc() behaves exactly like kmalloc().
Addresses-Coverity-ID: 1490594 ("Resource leak")
--- lib/test_kasan.c | 1 + 1 file changed, 1 insertion(+)
lib/test_kasan.c +1 -0
@@ -126,6 +126,7 @@ static noinline void __init kmalloc_oob_if (!ptr1 || !ptr2) {pr_err ("Allocation failed\n" );kfree (ptr1); +kfree (ptr2);return ; }
Greg Kroah-Hartman
maintainerPATCH 29/173 irqdomain: Fix a memory leak in irq_domain_push_irq()
In reply to Greg Kroah-HartmanFrom: Kevin Hao <[email protected]>
commit 0f394daef89b38d58c91118a2b08b8a1b316703b upstream.
Fix a memory leak reported by kmemleak:
unreferenced object 0xffff000bc6f50e80 (size 128):
comm "kworker/23:2", pid 201, jiffies 4294894947 (age 942.132s) hex dump (first 32 bytes):
00 00 00 00 41 00 00 00 86 c0 03 00 00 00 00 00 ....A...........
00 a0 b2 c6 0b 00 ff ff 40 51 fd 10 00 80 ff ff ........@Q......
backtrace:
[<00000000e62d2240>] kmem_cache_alloc_trace+0x1a4/0x320
[<00000000279143c9>] irq_domain_push_irq+0x7c/0x188
[<00000000d9f4c154>] thunderx_gpio_probe+0x3ac/0x438
[<00000000fd09ec22>] pci_device_probe+0xe4/0x198
[<00000000d43eca75>] really_probe+0xdc/0x320
[<00000000d3ebab09>] driver_probe_device+0x5c/0xf0
[<000000005b3ecaa0>] __device_attach_driver+0x88/0xc0
[<000000004e5915f5>] bus_for_each_drv+0x7c/0xc8
[<0000000079d4db41>] __device_attach+0xe4/0x140
[<00000000883bbda9>] device_initial_probe+0x18/0x20
[<000000003be59ef6>] bus_probe_device+0x98/0xa0
[<0000000039b03d3f>] deferred_probe_work_func+0x74/0xa8
[<00000000870934ce>] process_one_work+0x1c8/0x470
[<00000000e3cce570>] worker_thread+0x1f8/0x428
[<000000005d64975e>] kthread+0xfc/0x128
[<00000000f0eaa764>] ret_from_fork+0x10/0x18--- kernel/irq/irqdomain.c | 1 + 1 file changed, 1 insertion(+)
kernel/irq/irqdomain.c +1 -0
@@ -1538,6 +1538,7 @@ int irq_domain_push_irq(struct irq_domaiif (rv) {/* Restore the original irq_data. */ *root_irq_data = *child_irq_data; +kfree (child_irq_data); goto error; }
Greg Kroah-Hartman
maintainerPATCH 30/173 platform/x86: intel_scu_ipc: Fix interrupt support
In reply to Greg Kroah-HartmanFrom: Mika Westerberg <[email protected]>
commit e48b72a568bbd641c91dad354138d3c17d03ee6f upstream.
Currently the driver has disabled interrupt support for Tangier but actually interrupt works just fine if the command is not written twice in a row. Also we need to ack the interrupt in the handler.
--- drivers/platform/x86/intel_scu_ipc.c | 21 ++++++++------------- 1 file changed, 8 insertions(+), 13 deletions(-)
drivers/platform/x86/intel_scu_ipc.c +8 -13
@@ -69,26 +69,22 @@struct intel_scu_ipc_pdata_t {u32 i2c_base;u32 i2c_len; -u8 irq_mode; };static const struct intel_scu_ipc_pdata_t intel_scu_ipc_lincroft_pdata = { .i2c_base =0xff12b000 , .i2c_len =0x10 , - .irq_mode =0 , };/* Penwell and Cloverview */ static const struct intel_scu_ipc_pdata_t intel_scu_ipc_penwell_pdata = { .i2c_base =0xff12b000 , .i2c_len =0x10 , - .irq_mode =1 , };static const struct intel_scu_ipc_pdata_t intel_scu_ipc_tangier_pdata = { .i2c_base =0xff00d000 , .i2c_len =0x10 , - .irq_mode =0 , };struct intel_scu_ipc_dev {
@@ -101,6 +97,9 @@ struct intel_scu_ipc_dev {static struct intel_scu_ipc_dev ipcdev;/* Only one for now */ +#define IPC_STATUS 0x04 +#define IPC_STATUS_IRQ BIT(2) +/* * IPC Read Buffer (Read Only): * 16 byte buffer for receiving data from SCU, if IPC command
@@ -122,11 +121,8 @@ static DEFINE_MUTEX(ipclock); /* lock us*/ static inline void ipc_command (struct intel_scu_ipc_dev *scu,u32 cmd) { -if (scu->irq_mode) { -reinit_completion (&scu->cmd_complete); -writel (cmd |IPC_IOC , scu->ipc_base); - } -writel (cmd, scu->ipc_base); +reinit_completion (&scu->cmd_complete); +writel (cmd |IPC_IOC , scu->ipc_base); }/*
@@ -612,9 +608,10 @@ EXPORT_SYMBOL(intel_scu_ipc_i2c_cntrl);static irqreturn_t ioc (int irq,void *dev_id) {struct intel_scu_ipc_dev *scu = dev_id; +int status =ipc_read_status (scu); -if (scu->irq_mode) -complete (&scu->cmd_complete); +writel (status |IPC_STATUS_IRQ , scu->ipc_base +IPC_STATUS ); +complete (&scu->cmd_complete);return IRQ_HANDLED ; }
@@ -640,8 +637,6 @@ static int ipc_probe(struct pci_dev *pdeif (!pdata)return -ENODEV ; - scu->irq_mode = pdata->irq_mode; - err =pcim_enable_device (pdev);if (err)return err;
Greg Kroah-Hartman
maintainerPATCH 31/173 KVM: arm64: Only sign-extend MMIO up to register width
In reply to Greg Kroah-HartmanFrom: Christoffer Dall <[email protected]>
commit b6ae256afd32f96bec0117175b329d0dd617655e upstream.
On AArch64 you can do a sign-extended load to either a 32-bit or 64-bit register, and we should only sign extend the register up to the width of the register as specified in the operation (by using the 32-bit Wn or 64-bit Xn register specifier).
As it turns out, the architecture provides this decoding information in the SF ("Sixty-Four" -- how cute...) bit.
Let's take advantage of this with the usual 32-bit/64-bit header file dance and do the right thing on AArch64 hosts.
--- arch/arm/include/asm/kvm_emulate.h | 5 +++++ arch/arm/include/asm/kvm_mmio.h | 2 ++ arch/arm64/include/asm/kvm_emulate.h | 5 +++++ arch/arm64/include/asm/kvm_mmio.h | 6 ++---- virt/kvm/arm/mmio.c | 6 ++++++ 5 files changed, 20 insertions(+), 4 deletions(-)
arch/arm/include/asm/kvm_emulate.h +5 -0
@@ -144,6 +144,11 @@ static inline bool kvm_vcpu_dabt_issext(return kvm_vcpu_get_hsr (vcpu) &HSR_SSE ; } +static inline bool kvm_vcpu_dabt_issf (const struct kvm_vcpu *vcpu) +{ +return false; +} +static inline int kvm_vcpu_dabt_get_rd (struct kvm_vcpu *vcpu) {return (kvm_vcpu_get_hsr (vcpu) &HSR_SRT_MASK ) >>HSR_SRT_SHIFT ;
arch/arm/include/asm/kvm_mmio.h +2 -0
@@ -26,6 +26,8 @@struct kvm_decode {unsigned long rt;bool sign_extend; +/* Not used on 32-bit arm */ +bool sixty_four; };void kvm_mmio_write_buf (void *buf,unsigned int len,unsigned long data);
arch/arm64/include/asm/kvm_emulate.h +5 -0
@@ -188,6 +188,11 @@ static inline bool kvm_vcpu_dabt_issext(return !!(kvm_vcpu_get_hsr (vcpu) & ESR_ELx_SSE); } +static inline bool kvm_vcpu_dabt_issf (const struct kvm_vcpu *vcpu) +{ +return !!(kvm_vcpu_get_hsr (vcpu) & ESR_ELx_SF); +} +static inline int kvm_vcpu_dabt_get_rd (const struct kvm_vcpu *vcpu) {return (kvm_vcpu_get_hsr (vcpu) & ESR_ELx_SRT_MASK) >> ESR_ELx_SRT_SHIFT;
arch/arm64/include/asm/kvm_mmio.h +2 -4
@@ -21,13 +21,11 @@#include <linux/kvm_host.h> #include <asm/kvm_arm.h> -/* -* This is annoying. The mmio code requires this, even if we don't -* need any decoding. To be fixed. -*/ struct kvm_decode {unsigned long rt;bool sign_extend; +/* Witdth of the register accessed by the faulting instruction is 64-bits */ +bool sixty_four; };void kvm_mmio_write_buf (void *buf,unsigned int len,unsigned long data);
virt/kvm/arm/mmio.c +6 -0
@@ -117,6 +117,9 @@ int kvm_handle_mmio_return(struct kvm_vc data = (data ^ mask) - mask; } +if (!vcpu->arch.mmio_decode.sixty_four) + data = data &0xffffffff ; +trace_kvm_mmio (KVM_TRACE_MMIO_READ , len, run->mmio.phys_addr, &data); data =vcpu_data_host_to_guest (vcpu, data, len);
@@ -137,6 +140,7 @@ static int decode_hsr(struct kvm_vcpu *vunsigned long rt;int access_size;bool sign_extend; +bool sixty_four;if (kvm_vcpu_dabt_iss1tw (vcpu)) {/* page table accesses IO mem: tell guest to fix its TTBR */
@@ -150,11 +154,13 @@ static int decode_hsr(struct kvm_vcpu *v *is_write =kvm_vcpu_dabt_iswrite (vcpu); sign_extend =kvm_vcpu_dabt_issext (vcpu); + sixty_four =kvm_vcpu_dabt_issf (vcpu); rt =kvm_vcpu_dabt_get_rd (vcpu); *len = access_size; vcpu->arch.mmio_decode.sign_extend = sign_extend; vcpu->arch.mmio_decode.rt = rt; + vcpu->arch.mmio_decode.sixty_four = sixty_four;return 0 ; }
Greg Kroah-Hartman
maintainerPATCH 32/173 MIPS: fix indentation of the RELOCS message
In reply to Greg Kroah-HartmanFrom: Alexander Lobakin <[email protected]>
commit a53998802e178451701d59d38e36f551422977ba upstream.
quiet_cmd_relocs lacks a whitespace which results in:
LD vmlinux SORTEX vmlinux SYSMAP System.map
RELOCS vmlinux
Building modules, stage 2.
MODPOST 64 modules
After this patch:
LD vmlinux SORTEX vmlinux SYSMAP System.map RELOCS vmlinux
Building modules, stage 2.
MODPOST 64 modules
Typo is present in kernel tree since the introduction of relocatable kernel support in commit e818fac595ab ("MIPS: Generate relocation table when CONFIG_RELOCATABLE"), but the relocation scripts were moved to Makefile.postlink later with commit 44079d3509ae ("MIPS: Use Makefile.postlink to insert relocations into vmlinux").
--- arch/mips/Makefile.postlink | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/mips/Makefile.postlink +1 -1
@@ -12,7 +12,7 @@ __archpost: include scripts/Kbuild.include CMD_RELOCS = arch/mips/boot/tools/relocs -quiet_cmd_relocs = RELOCS $@ +quiet_cmd_relocs = RELOCS $@ cmd_relocs = $(CMD_RELOCS) $@ # `@true` prevents complaint when there is nothing to be done
Greg Kroah-Hartman
maintainerPATCH 33/173 s390/mm: fix dynamic pagetable upgrade for hugetlbfs
In reply to Greg Kroah-HartmanFrom: Gerald Schaefer <[email protected]>
commit 5f490a520bcb393389a4d44bec90afcb332eb112 upstream.
Commit ee71d16d22bb ("s390/mm: make TASK_SIZE independent from the number of page table levels") changed the logic of TASK_SIZE and also removed the arch_mmap_check() implementation for s390. This combination has a subtle effect on how get_unmapped_area() for hugetlbfs pages works. It is now possible that a user process establishes a hugetlbfs mapping at an address above 4 TB, without triggering a dynamic pagetable upgrade from 3 to 4 levels.
This is because hugetlbfs mappings will not use mm->get_unmapped_area, but rather file->f_op->get_unmapped_area, which currently is the generic implementation of hugetlb_get_unmapped_area() that does not know about s390 dynamic pagetable upgrades, but with the new definition of TASK_SIZE, it will now allow mappings above 4 TB.
Subsequent access to such a mapped address above 4 TB will result in a page fault loop, because the CPU cannot translate such a large address with 3 pagetable levels. The fault handler will try to map in a hugepage at the address, but due to the folded pagetable logic it will end up with creating entries in the 3 level pagetable, possibly overwriting existing mappings, and then it all repeats when the access is retried.
Apart from the page fault loop, this can have various nasty effects, e.g. kernel panic from one of the BUG_ON() checks in memory management code, or even data loss if an existing mapping gets overwritten.
Fix this by implementing HAVE_ARCH_HUGETLB_UNMAPPED_AREA support for s390, providing an s390 version for hugetlb_get_unmapped_area() with pagetable upgrade support similar to arch_get_unmapped_area(), which will then be used instead of the generic version.
--- arch/s390/include/asm/page.h | 2 arch/s390/mm/hugetlbpage.c | 100 ++++++++++++++++++++++++++++++++++++++++++- 2 files changed, 101 insertions(+), 1 deletion(-)
arch/s390/include/asm/page.h +2 -0
@@ -33,6 +33,8 @@#define ARCH_HAS_PREPARE_HUGEPAGE #define ARCH_HAS_HUGEPAGE_CLEAR_FLUSH +#define HAVE_ARCH_HUGETLB_UNMAPPED_AREA +#include <asm/setup.h> #ifndef __ASSEMBLY__
arch/s390/mm/hugetlbpage.c +99 -1
115 lines not on this page · raw message
Greg Kroah-Hartman
maintainerPATCH 34/173 powerpc/xmon: dont access ASDR in VMs
In reply to Greg Kroah-HartmanFrom: Sukadev Bhattiprolu <[email protected]>
commit c2a20711fc181e7f22ee5c16c28cb9578af84729 upstream.
ASDR is HV-privileged and must only be accessed in HV-mode. Fixes a Program Check (0x700) when xmon in a VM dumps SPRs.
--- arch/powerpc/xmon/xmon.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-)
arch/powerpc/xmon/xmon.c +4 -5
@@ -1830,15 +1830,14 @@ static void dump_300_sprs(void)printf ("pidr = %.16lx tidr = %.16lx\n" ,mfspr (SPRN_PID ),mfspr (SPRN_TIDR )); -printf ("asdr = %.16lx psscr = %.16lx\n" , -mfspr (SPRN_ASDR ), hv ?mfspr (SPRN_PSSCR ) - :mfspr (SPRN_PSSCR_PR )); +printf ("psscr = %.16lx\n" , + hv ?mfspr (SPRN_PSSCR ) :mfspr (SPRN_PSSCR_PR ));if (!hv)return ; -printf ("ptcr = %.16lx\n" , -mfspr (SPRN_PTCR )); +printf ("ptcr = %.16lx asdr = %.16lx\n" , +mfspr (SPRN_PTCR ),mfspr (SPRN_ASDR ));#endif }
Greg Kroah-Hartman
maintainerPATCH 35/173 powerpc/pseries: Advance pfn if section is not present in lmb_is_removable()
In reply to Greg Kroah-HartmanFrom: Pingfan Liu <[email protected]>
commit fbee6ba2dca30d302efe6bddb3a886f5e964a257 upstream.
In lmb_is_removable(), if a section is not present, it should continue to test the rest of the sections in the block. But the current code fails to do so.
--- arch/powerpc/platforms/pseries/hotplug-memory.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
arch/powerpc/platforms/pseries/hotplug-memory.c +3 -1
@@ -452,8 +452,10 @@ static bool lmb_is_removable(struct of_dfor (i =0 ; i < scns_per_block; i++) { pfn =PFN_DOWN (phys_addr); -if (!pfn_present (pfn)) +if (!pfn_present (pfn)) { + phys_addr +=MIN_MEMORY_BLOCK_SIZE ;continue ; + } rc &=is_mem_section_removable (pfn,PAGES_PER_SECTION ); phys_addr +=MIN_MEMORY_BLOCK_SIZE ;
Greg Kroah-Hartman
maintainerPATCH 36/173 mmc: spi: Toggle SPI polarity, do not hardcode it
In reply to Greg Kroah-HartmanFrom: Linus Walleij <[email protected]>
commit af3ed119329cf9690598c5a562d95dfd128e91d6 upstream.
The code in mmc_spi_initsequence() tries to send a burst with high chipselect and for this reason hardcodes the device into SPI_CS_HIGH.
This is not good because the SPI_CS_HIGH flag indicates logical "asserted" CS not always the physical level. In some cases the signal is inverted in the GPIO library and in that case SPI_CS_HIGH is already set, and enforcing SPI_CS_HIGH again will actually drive it low.
Instead of hard-coding this, toggle the polarity so if the default is LOW it goes high to assert chipselect but if it is already high then toggle it low instead.
--- drivers/mmc/host/mmc_spi.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-)
drivers/mmc/host/mmc_spi.c +8 -3
@@ -1154,17 +1154,22 @@ static void mmc_spi_initsequence(struct* SPI protocol. Another is that when chipselect is released while * the card returns BUSY status, the clock must issue several cycles * with chipselect high before the card will stop driving its output. +* +* SPI_CS_HIGH means "asserted" here. In some cases like when using +* GPIOs for chip select, SPI_CS_HIGH is set but this will be logically +* inverted by gpiolib, so if we want to ascertain to drive it high +* we should toggle the default with an XOR as we do here. */ - host->spi->mode |=SPI_CS_HIGH ; + host->spi->mode ^=SPI_CS_HIGH ;if (spi_setup (host->spi) !=0 ) {/* Just warn; most cards work without it. */ dev_warn (&host->spi->dev,"can't change chip-select polarity\n" ); - host->spi->mode &= ~SPI_CS_HIGH ; + host->spi->mode ^=SPI_CS_HIGH ; }else {mmc_spi_readbytes (host,18 ); - host->spi->mode &= ~SPI_CS_HIGH ; + host->spi->mode ^=SPI_CS_HIGH ;if (spi_setup (host->spi) !=0 ) {/* Wot, we can't get the same setup we had before? */ dev_err(&host->spi->dev,
Greg Kroah-Hartman
maintainerPATCH 37/173 ACPI: video: Do not export a non working backlight interface on MSI MS-7721 boards
In reply to Greg Kroah-HartmanFrom: Hans de Goede <[email protected]>
commit d21a91629f4b8e794fc4c0e0c17c85cedf1d806c upstream.
Despite our heuristics to not wrongly export a non working ACPI backlight interface on desktop machines, we still end up exporting one on desktops using a motherboard from the MSI MS-7721 series.
I've looked at improving the heuristics, but in this case a quirk seems to be the only way to solve this.
While at it also add a comment to separate the video_detect_force_none entries in the video_detect_dmi_table from other type of entries, as we already do for the other entry types.
--- drivers/acpi/video_detect.c | 13 +++++++++++++ 1 file changed, 13 insertions(+)
drivers/acpi/video_detect.c +13 -0
@@ -328,6 +328,11 @@ static const struct dmi_system_id video_ DMI_MATCH(DMI_PRODUCT_NAME, "Precision 7510"), }, }, + +/* +* Desktops which falsely report a backlight and which our heuristics +* for this do not catch. +*/ { .callback = video_detect_force_none, .ident = "Dell OptiPlex 9020M",
@@ -336,6 +341,14 @@ static const struct dmi_system_id video_ DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 9020M"), }, }, + { + .callback = video_detect_force_none, + .ident = "MSI MS-7721", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "MSI"), + DMI_MATCH(DMI_PRODUCT_NAME, "MS-7721"), + }, + }, { }, };
Greg Kroah-Hartman
maintainerPATCH 38/173 alarmtimer: Unregister wakeup source when module get fails
In reply to Greg Kroah-HartmanFrom: Stephen Boyd <[email protected]>
commit 6b6d188aae79a630957aefd88ff5c42af6553ee3 upstream.
The alarmtimer_rtc_add_device() function creates a wakeup source and then tries to grab a module reference. If that fails the function returns early with an error code, but fails to remove the wakeup source.
Cleanup this exit path so there is no dangling wakeup source, which is named 'alarmtime' left allocated which will conflict with another RTC device that may be registered later.
--- kernel/time/alarmtimer.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)
kernel/time/alarmtimer.c +5 -3
@@ -91,6 +91,7 @@ static int alarmtimer_rtc_add_device(strunsigned long flags;struct rtc_device *rtc =to_rtc_device (dev);struct wakeup_source *__ws; +int ret =0 ;if (rtcdev)return -EBUSY ;
@@ -105,8 +106,8 @@ static int alarmtimer_rtc_add_device(strspin_lock_irqsave (&rtcdev_lock, flags);if (!rtcdev) {if (!try_module_get (rtc->owner)) { -spin_unlock_irqrestore (&rtcdev_lock, flags); -return -1 ; + ret =-1 ; + goto unlock; } rtcdev = rtc;
@@ -115,11 +116,12 @@ static int alarmtimer_rtc_add_device(str ws = __ws; __ws =NULL ; } +unlock:spin_unlock_irqrestore (&rtcdev_lock, flags);wakeup_source_unregister (__ws); -return 0 ; +return ret; }static inline void alarmtimer_rtc_timer_init (void )
Greg Kroah-Hartman
maintainerPATCH 39/173 ubifs: Reject unsupported ioctl flags explicitly
In reply to Greg Kroah-HartmanFrom: Hou Tao <[email protected]>
commit 2fe8b2d5578d7d142982e3bf62e4c0caf8b8fe02 upstream.
Reject unsupported ioctl flags explicitly, so the following command on a regular ubifs file will fail:
chattr +d ubifs_file
And xfstests generic/424 will pass.
--- fs/ubifs/ioctl.c | 8 ++++++++ 1 file changed, 8 insertions(+)
fs/ubifs/ioctl.c +8 -0
@@ -28,6 +28,11 @@#include <linux/mount.h> #include "ubifs.h" +/* Need to be kept consistent with checked flags in ioctl2ubifs() */ +#define UBIFS_SUPPORTED_IOCTL_FLAGS \ + (FS_COMPR_FL | FS_SYNC_FL | FS_APPEND_FL | \ + FS_IMMUTABLE_FL | FS_DIRSYNC_FL) +/** * ubifs_set_inode_flags - set VFS inode flags. * @inode: VFS inode to set flags for
@@ -166,6 +171,9 @@ long ubifs_ioctl(struct file *file, unsiif (get_user (flags, (int __user *) arg))return -EFAULT ; +if (flags & ~UBIFS_SUPPORTED_IOCTL_FLAGS ) +return -EOPNOTSUPP ; +if (!S_ISDIR (inode->i_mode)) flags &= ~FS_DIRSYNC_FL ;
Greg Kroah-Hartman
maintainerPATCH 40/173 ubifs: Fix FS_IOC_SETFLAGS unexpectedly clearing encrypt flag
In reply to Greg Kroah-HartmanFrom: Eric Biggers <[email protected]>
commit 2b57067a7778484c10892fa191997bfda29fea13 upstream.
UBIFS's implementation of FS_IOC_SETFLAGS fails to preserve existing inode flags that aren't settable by FS_IOC_SETFLAGS, namely the encrypt flag. This causes the encrypt flag to be unexpectedly cleared.
Fix it by preserving existing unsettable flags, like ext4 and f2fs do.
Test case with kvm-xfstests shell:
FSTYP=ubifs KEYCTL_PROG=keyctl
. fs/ubifs/config
. ~/xfstests/common/encrypt
dev=$(__blkdev_to_ubi_volume /dev/vdc)
ubiupdatevol -t $dev
mount $dev /mnt -t ubifs
k=$(_generate_session_encryption_key)
mkdir /mnt/edir
xfs_io -c "set_encpolicy $k" /mnt/edir
echo contents > /mnt/edir/file
chattr +i /mnt/edir/file
chattr -i /mnt/edir/file
With the bug, the following errors occur on the last command:
[ 18.081559] fscrypt (ubifs, inode 67): Inconsistent encryption context (parent directory: 65)
chattr: Operation not permitted while reading flags on /mnt/edir/file--- fs/ubifs/ioctl.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
fs/ubifs/ioctl.c +2 -1
@@ -129,7 +129,8 @@ static int setflags(struct inode *inode, } } - ui->flags =ioctl2ubifs (flags); + ui->flags &= ~ioctl2ubifs (UBIFS_SUPPORTED_IOCTL_FLAGS ); + ui->flags |=ioctl2ubifs (flags);ubifs_set_inode_flags (inode); inode->i_ctime =current_time (inode); release = ui->dirty;
Greg Kroah-Hartman
maintainerPATCH 41/173 ubifs: Fix deadlock in concurrent bulk-read and writepage
In reply to Greg Kroah-HartmanFrom: Zhihao Cheng <[email protected]>
commit f5de5b83303e61b1f3fb09bd77ce3ac2d7a475f2 upstream.
In ubifs, concurrent execution of writepage and bulk read on the same file may cause ABBA deadlock, for example (Reproduce method see Link):
Process A(Bulk-read starts from page4) Process B(write page4 back)
vfs_read wb_workfn or fsync ... ... generic_file_buffered_read write_cache_pages
ubifs_readpage LOCK(page4)
ubifs_bulk_read ubifs_writepage
LOCK(ui->ui_mutex) ubifs_write_inode
ubifs_do_bulk_read LOCK(ui->ui_mutex)
find_or_create_page(alloc page4) ↑
LOCK(page4) <-- ABBA deadlock occurs!
In order to ensure the serialization execution of bulk read, we can't remove the big lock 'ui->ui_mutex' in ubifs_bulk_read(). Instead, we allow ubifs_do_bulk_read() to lock page failed by replacing find_or_create_page(FGP_LOCK) with pagecache_get_page(FGP_LOCK | FGP_NOWAIT).
--- fs/ubifs/file.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
fs/ubifs/file.c +3 -1
@@ -797,7 +797,9 @@ static int ubifs_do_bulk_read(struct ubiif (page_offset > end_index)break ; - page =find_or_create_page (mapping, page_offset, ra_gfp_mask); + page =pagecache_get_page (mapping, page_offset, +FGP_LOCK |FGP_ACCESSED |FGP_CREAT |FGP_NOWAIT , + ra_gfp_mask);if (!page)break ;if (!PageUptodate (page))
Greg Kroah-Hartman
maintainerPATCH 42/173 PCI: keystone: Fix link training retries initiation
In reply to Greg Kroah-HartmanFrom: Yurii Monakov <[email protected]>
[ Upstream commit 6df19872d881641e6394f93ef2938cffcbdae5bb ]
ks_pcie_stop_link() function does not clear LTSSM_EN_VAL bit so link training was not triggered more than once after startup. In configurations where link can be unstable during early boot, for example, under low temperature, it will never be established.
--- drivers/pci/dwc/pci-keystone-dw.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/pci/dwc/pci-keystone-dw.c +1 -1
@@ -510,7 +510,7 @@ void ks_dw_pcie_initiate_link_train(struct keystone_pcie *ks_pcie)/* Disable Link training */ val =ks_dw_app_readl (ks_pcie,CMD_STATUS ); val &= ~LTSSM_EN_VAL ; -ks_dw_app_writel (ks_pcie,CMD_STATUS ,LTSSM_EN_VAL | val); +ks_dw_app_writel (ks_pcie,CMD_STATUS , val);/* Initiate Link Training */ val =ks_dw_app_readl (ks_pcie,CMD_STATUS );
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 43/173 mmc: sdhci-of-at91: fix memleak on clk_get failure
In reply to Greg Kroah-HartmanFrom: Michał Mirosław <[email protected]>
[ Upstream commit a04184ce777b46e92c2b3c93c6dcb2754cb005e1 ]
sdhci_alloc_host() does its work not using managed infrastructure, so needs explicit free on error path. Add it where needed.
--- drivers/mmc/host/sdhci-of-at91.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-)
drivers/mmc/host/sdhci-of-at91.c +6 -3
@@ -331,19 +331,22 @@ static int sdhci_at91_probe(struct platform_device *pdev) priv->mainck =devm_clk_get (&pdev->dev,"baseclk" );if (IS_ERR (priv->mainck)) {dev_err (&pdev->dev,"failed to get baseclk\n" ); -return PTR_ERR (priv->mainck); + ret =PTR_ERR (priv->mainck); + goto sdhci_pltfm_free; } priv->hclock =devm_clk_get (&pdev->dev,"hclock" );if (IS_ERR (priv->hclock)) {dev_err (&pdev->dev,"failed to get hclock\n" ); -return PTR_ERR (priv->hclock); + ret =PTR_ERR (priv->hclock); + goto sdhci_pltfm_free; } priv->gck =devm_clk_get (&pdev->dev,"multclk" );if (IS_ERR (priv->gck)) {dev_err (&pdev->dev,"failed to get multclk\n" ); -return PTR_ERR (priv->gck); + ret =PTR_ERR (priv->gck); + goto sdhci_pltfm_free; } ret =sdhci_at91_set_clks_presets (&pdev->dev);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 44/173 ubifs: dont trigger assertion on invalid no-key filename
In reply to Greg Kroah-HartmanFrom: Eric Biggers <[email protected]>
[ Upstream commit f0d07a98a070bb5e443df19c3aa55693cbca9341 ]
If userspace provides an invalid fscrypt no-key filename which encodes a hash value with any of the UBIFS node type bits set (i.e. the high 3 bits), gracefully report ENOENT rather than triggering ubifs_assert().
Test case with kvm-xfstests shell:
. fs/ubifs/config
. ~/xfstests/common/encrypt dev=$(__blkdev_to_ubi_volume /dev/vdc) ubiupdatevol $dev -t
mount $dev /mnt -t ubifs
mkdir /mnt/edir
xfs_io -c set_encpolicy /mnt/edir
rm /mnt/edir/_,,,,,DAAAAAAAAAAAAAAAAAAAAAAAAAA
With the bug, the following assertion fails on the 'rm' command:
[ 19.066048] UBIFS error (ubi0:0 pid 379): ubifs_assert_failed: UBIFS assert failed: !(hash & ~UBIFS_S_KEY_HASH_MASK), in fs/ubifs/key.h:170
--- fs/ubifs/dir.c | 2 ++ 1 file changed, 2 insertions(+)
fs/ubifs/dir.c +2 -0
@@ -253,6 +253,8 @@ static struct dentry *ubifs_lookup(struct inode *dir, struct dentry *dentry,if (nm.hash) {ubifs_assert (fname_len (&nm) ==0 );ubifs_assert (fname_name (&nm) ==NULL ); +if (nm.hash & ~UBIFS_S_KEY_HASH_MASK ) + goto done;/* ENOENT */ dent_key_init_hash (c, &key, dir->i_ino, nm.hash); err =ubifs_tnc_lookup_dh (c, &key, dent, nm.minor_hash); }else {
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 45/173 hv_balloon: Balloon up according to request page number
In reply to Greg Kroah-HartmanFrom: Tianyu Lan <[email protected]>
commit d33c240d47dab4fd15123d9e73fc8810cbc6ed6a upstream.
Current code has assumption that balloon request memory size aligns with 2MB. But actually Hyper-V doesn't guarantee such alignment. When balloon driver receives non-aligned balloon request, it produces warning and balloon up more memory than requested in order to keep 2MB alignment. Remove the warning and balloon up memory according to actual requested memory size.
--- drivers/hv/hv_balloon.c | 13 +++---------- 1 file changed, 3 insertions(+), 10 deletions(-)
drivers/hv/hv_balloon.c +3 -10
@@ -1170,10 +1170,7 @@ static unsigned int alloc_balloon_pages(unsigned int i =0 ;struct page *pg; -if (num_pages < alloc_unit) -return 0 ; - -for (i =0 ; (i * alloc_unit) < num_pages; i++) { +for (i =0 ; i < num_pages / alloc_unit; i++) {if (bl_resp->hdr.size +sizeof (union dm_mem_page_range ) >PAGE_SIZE )return i * alloc_unit;
@@ -1207,7 +1204,7 @@ static unsigned int alloc_balloon_pages( } -return num_pages; +return i * alloc_unit; }static void balloon_up (struct work_struct *dummy)
@@ -1222,9 +1219,6 @@ static void balloon_up(struct work_struclong avail_pages;unsigned long floor; -/* The host balloons pages in 2M granularity. */ -WARN_ON_ONCE (num_pages %PAGES_IN_2M !=0 ); -/* * We will attempt 2M allocations. However, if we fail to * allocate 2M chunks, we will go back to 4k allocations.
@@ -1234,14 +1228,13 @@ static void balloon_up(struct work_struc avail_pages =si_mem_available (); floor =compute_balloon_floor (); -/* Refuse to balloon below the floor, keep the 2M granularity. */ +/* Refuse to balloon below the floor. */ if (avail_pages < num_pages || avail_pages - num_pages < floor) {pr_warn ("Balloon request will be partially fulfilled. %s\n" , avail_pages < num_pages ?"Not enough memory." :"Balloon floor reached." ); num_pages = avail_pages > floor ? (avail_pages - floor) :0 ; - num_pages -= num_pages %PAGES_IN_2M ; }while (!done) {
Greg Kroah-Hartman
maintainerPATCH 46/173 crypto: api - Check spawn->alg under lock in crypto_drop_spawn
In reply to Greg Kroah-HartmanFrom: Herbert Xu <[email protected]>
commit 7db3b61b6bba4310f454588c2ca6faf2958ad79f upstream.
We need to check whether spawn->alg is NULL under lock as otherwise the algorithm could be removed from under us after we have checked it and found it to be non-NULL. This could cause us to remove the spawn from a non-existent list.
--- crypto/algapi.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-)
crypto/algapi.c +2 -4
@@ -652,11 +652,9 @@ EXPORT_SYMBOL_GPL(crypto_grab_spawn);void crypto_drop_spawn (struct crypto_spawn *spawn) { -if (!spawn->alg) -return ; -down_write (&crypto_alg_sem); -list_del (&spawn->list); +if (spawn->alg) +list_del (&spawn->list);up_write (&crypto_alg_sem); }EXPORT_SYMBOL_GPL (crypto_drop_spawn);
Greg Kroah-Hartman
maintainerPATCH 47/173 scsi: qla2xxx: Fix mtcp dump collection failure
In reply to Greg Kroah-HartmanFrom: Quinn Tran <[email protected]>
commit 641e0efddcbde52461e017136acd3ce7f2ef0c14 upstream.
MTCP dump failed due to MB Reg 10 was picking garbage data from stack memory.
--- drivers/scsi/qla2xxx/qla_mbx.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/scsi/qla2xxx/qla_mbx.c +1 -2
@@ -5853,9 +5853,8 @@ qla2x00_dump_mctp_data(scsi_qla_host_t * mcp->mb[7 ] =LSW (MSD (req_dma)); mcp->mb[8 ] =MSW (addr);/* Setting RAM ID to valid */ - mcp->mb[10 ] |=BIT_7 ;/* For MCTP RAM ID is 0x40 */ - mcp->mb[10 ] |=0x40 ; + mcp->mb[10 ] =BIT_7 |0x40 ; mcp->out_mb |=MBX_10 |MBX_8 |MBX_7 |MBX_6 |MBX_5 |MBX_4 |MBX_3 |MBX_2 |MBX_1 |MBX_0 ;
Greg Kroah-Hartman
maintainerPATCH 48/173 power: supply: ltc2941-battery-gauge: fix use-after-free
In reply to Greg Kroah-HartmanFrom: Sven Van Asbroeck <[email protected]>
commit a60ec78d306c6548d4adbc7918b587a723c555cc upstream.
This driver's remove path calls cancel_delayed_work(). However, that function does not wait until the work function finishes. This could mean that the work function is still running after the driver's remove function has finished, which would result in a use-after-free.
Fix by calling cancel_delayed_work_sync(), which ensures that that the work is properly cancelled, no longer running, and unable to re-schedule itself.
This issue was detected with the help of Coccinelle.
--- drivers/power/supply/ltc2941-battery-gauge.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/power/supply/ltc2941-battery-gauge.c +1 -1
@@ -406,7 +406,7 @@ static int ltc294x_i2c_remove(struct i2c {struct ltc294x_info *info =i2c_get_clientdata (client); -cancel_delayed_work (&info->work); +cancel_delayed_work_sync (&info->work);power_supply_unregister (info->supply);return 0 ; }
Greg Kroah-Hartman
maintainerPATCH 49/173 f2fs: choose hardlimit when softlimit is larger than hardlimit in f2fs_statfs_project()
In reply to Greg Kroah-HartmanFrom: Chengguang Xu <[email protected]>
commit 909110c060f22e65756659ec6fa957ae75777e00 upstream.
Setting softlimit larger than hardlimit seems meaningless for disk quota but currently it is allowed. In this case, there may be a bit of comfusion for users when they run df comamnd to directory which has project quota.
For example, we set 20M softlimit and 10M hardlimit of block usage limit for project quota of test_dir(project id 123).
[root@hades f2fs]# repquota -P -a
--- fs/f2fs/super.c | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-)
fs/f2fs/super.c +14 -6
@@ -912,9 +912,13 @@ static int f2fs_statfs_project(struct sureturn PTR_ERR (dquot);spin_lock (&dq_data_lock); - limit = (dquot->dq_dqb.dqb_bsoftlimit ? - dquot->dq_dqb.dqb_bsoftlimit : - dquot->dq_dqb.dqb_bhardlimit) >> sb->s_blocksize_bits; + limit =0 ; +if (dquot->dq_dqb.dqb_bsoftlimit) + limit = dquot->dq_dqb.dqb_bsoftlimit; +if (dquot->dq_dqb.dqb_bhardlimit && + (!limit || dquot->dq_dqb.dqb_bhardlimit < limit)) + limit = dquot->dq_dqb.dqb_bhardlimit; +if (limit && buf->f_blocks > limit) { curblock = dquot->dq_dqb.dqb_curspace >> sb->s_blocksize_bits; buf->f_blocks = limit;
@@ -923,9 +927,13 @@ static int f2fs_statfs_project(struct su (buf->f_blocks - curblock) : 0; } - limit = dquot->dq_dqb.dqb_isoftlimit ? - dquot->dq_dqb.dqb_isoftlimit : - dquot->dq_dqb.dqb_ihardlimit; + limit =0 ; +if (dquot->dq_dqb.dqb_isoftlimit) + limit = dquot->dq_dqb.dqb_isoftlimit; +if (dquot->dq_dqb.dqb_ihardlimit && + (!limit || dquot->dq_dqb.dqb_ihardlimit < limit)) + limit = dquot->dq_dqb.dqb_ihardlimit; +if (limit && buf->f_files > limit) { buf->f_files = limit; buf->f_ffree =
Greg Kroah-Hartman
maintainerPATCH 50/173 f2fs: fix miscounted block limit in f2fs_statfs_project()
In reply to Greg Kroah-HartmanFrom: Chengguang Xu <[email protected]>
commit acdf2172172a511f97fa21ed0ee7609a6d3b3a07 upstream.
statfs calculates Total/Used/Avail disk space in block unit, so we should translate soft/hard prjquota limit to block unit as well.
Below testing result shows the block/inode numbers of Total/Used/Avail from df command are all correct afer applying this patch.
[root@localhost quota-tools]\# ./repquota -P /dev/sdb1
--- fs/f2fs/super.c | 2 ++ 1 file changed, 2 insertions(+)
fs/f2fs/super.c +2 -0
@@ -918,6 +918,8 @@ static int f2fs_statfs_project(struct suif (dquot->dq_dqb.dqb_bhardlimit && (!limit || dquot->dq_dqb.dqb_bhardlimit < limit)) limit = dquot->dq_dqb.dqb_bhardlimit; +if (limit) + limit >>= sb->s_blocksize_bits;if (limit && buf->f_blocks > limit) { curblock = dquot->dq_dqb.dqb_curspace >> sb->s_blocksize_bits;
Greg Kroah-Hartman
maintainerPATCH 51/173 f2fs: code cleanup for f2fs_statfs_project()
In reply to Greg Kroah-HartmanFrom: Chengguang Xu <[email protected]>
commit bf2cbd3c57159c2b639ee8797b52ab5af180bf83 upstream.
Calling min_not_zero() to simplify complicated prjquota limit comparison in f2fs_statfs_project().
--- fs/f2fs/super.c | 16 ++++------------ 1 file changed, 4 insertions(+), 12 deletions(-)
fs/f2fs/super.c +4 -12
@@ -912,12 +912,8 @@ static int f2fs_statfs_project(struct sureturn PTR_ERR (dquot);spin_lock (&dq_data_lock); - limit =0 ; -if (dquot->dq_dqb.dqb_bsoftlimit) - limit = dquot->dq_dqb.dqb_bsoftlimit; -if (dquot->dq_dqb.dqb_bhardlimit && - (!limit || dquot->dq_dqb.dqb_bhardlimit < limit)) - limit = dquot->dq_dqb.dqb_bhardlimit; + limit =min_not_zero (dquot->dq_dqb.dqb_bsoftlimit, + dquot->dq_dqb.dqb_bhardlimit);if (limit) limit >>= sb->s_blocksize_bits;
@@ -929,12 +925,8 @@ static int f2fs_statfs_project(struct su (buf->f_blocks - curblock) : 0; } - limit =0 ; -if (dquot->dq_dqb.dqb_isoftlimit) - limit = dquot->dq_dqb.dqb_isoftlimit; -if (dquot->dq_dqb.dqb_ihardlimit && - (!limit || dquot->dq_dqb.dqb_ihardlimit < limit)) - limit = dquot->dq_dqb.dqb_ihardlimit; + limit =min_not_zero (dquot->dq_dqb.dqb_isoftlimit, + dquot->dq_dqb.dqb_ihardlimit);if (limit && buf->f_files > limit) { buf->f_files = limit;
Greg Kroah-Hartman
maintainerPATCH 52/173 PM: core: Fix handling of devices deleted during system-wide resume
In reply to Greg Kroah-HartmanFrom: Rafael J. Wysocki <[email protected]>
commit 0552e05fdfea191a2cf3a0abd33574b5ef9ca818 upstream.
If a device is deleted by one of its system-wide resume callbacks (for example, because it does not appear to be present or accessible any more) along with its children, the resume of the children may continue leading to use-after-free errors and other issues (potentially).
Namely, if the device's children are resumed asynchronously, their resume may have been scheduled already before the device's callback runs and so the device may be deleted while dpm_wait_for_superior() is being executed for them. The memory taken up by the parent device object may be freed then while dpm_wait() is waiting for the parent's resume callback to complete, which leads to a use-after-free. Moreover, the resume of the children is really not expected to continue after they have been unregistered, so it must be terminated right away in that case.
To address this problem, modify dpm_wait_for_superior() to check if the target device is still there in the system-wide PM list of devices and if so, to increment its parent's reference counter, both under dpm_list_mtx which prevents device_del() running for the child from dropping the parent's reference counter prematurely.
If the device is not present in the system-wide PM list of devices any more, the resume of it cannot continue, so check that again after dpm_wait() returns, which means that the parent's callback has been completed, and pass the result of that check to the caller of dpm_wait_for_superior() to allow it to abort the device's resume if it is not there any more.
--- drivers/base/power/main.c | 42 +++++++++++++++++++++++++++++++++++++----- 1 file changed, 37 insertions(+), 5 deletions(-)
drivers/base/power/main.c +37 -5
@@ -269,10 +269,38 @@ static void dpm_wait_for_suppliers(strucdevice_links_read_unlock (idx); } -static void dpm_wait_for_superior (struct device *dev,bool async) +static bool dpm_wait_for_superior (struct device *dev,bool async) { -dpm_wait (dev->parent, async); +struct device *parent; + +/* +* If the device is resumed asynchronously and the parent's callback +* deletes both the device and the parent itself, the parent object may +* be freed while this function is running, so avoid that by reference +* counting the parent once more unless the device has been deleted +* already (in which case return right away). +*/ +mutex_lock (&dpm_list_mtx); + +if (!device_pm_initialized (dev)) { +mutex_unlock (&dpm_list_mtx); +return false; + } + + parent =get_device (dev->parent); + +mutex_unlock (&dpm_list_mtx); + +dpm_wait (parent, async); +put_device (parent); +dpm_wait_for_suppliers (dev, async); + +/* +* If the parent's callback has deleted the device, attempting to resume +* it would be invalid, so avoid doing that then. +*/ +return device_pm_initialized (dev); }static void dpm_wait_for_consumers (struct device *dev,bool async)
@@ -551,7 +579,8 @@ static int device_resume_noirq(struct deif (!dev->power.is_noirq_suspended) goto Out; -dpm_wait_for_superior (dev, async); +if (!dpm_wait_for_superior (dev, async)) + goto Out;if (dev->pm_domain) { info ="noirq power domain " ;
@@ -691,7 +720,8 @@ static int device_resume_early(struct deif (!dev->power.is_late_suspended) goto Out; -dpm_wait_for_superior (dev, async); +if (!dpm_wait_for_superior (dev, async)) + goto Out;if (dev->pm_domain) { info ="early power domain " ;
@@ -823,7 +853,9 @@ static int device_resume(struct device * goto Complete; } -dpm_wait_for_superior (dev, async); +if (!dpm_wait_for_superior (dev, async)) + goto Complete; +dpm_watchdog_set (&wd, dev);device_lock (dev);
Greg Kroah-Hartman
maintainerPATCH 53/173 of: Add OF_DMA_DEFAULT_COHERENT & select it on powerpc
In reply to Greg Kroah-HartmanFrom: Michael Ellerman <[email protected]>
commit dabf6b36b83a18d57e3d4b9d50544ed040d86255 upstream.
There's an OF helper called of_dma_is_coherent(), which checks if a device has a "dma-coherent" property to see if the device is coherent for DMA.
But on some platforms devices are coherent by default, and on some platforms it's not possible to update existing device trees to add the "dma-coherent" property.
So add a Kconfig symbol to allow arch code to tell of_dma_is_coherent() that devices are coherent by default, regardless of the presence of the property.
Select that symbol on powerpc when NOT_COHERENT_CACHE is not set, ie. when the system has a coherent cache.
--- arch/powerpc/Kconfig | 1 + drivers/of/Kconfig | 4 ++++ drivers/of/address.c | 6 +++++- 3 files changed, 10 insertions(+), 1 deletion(-)
arch/powerpc/Kconfig +1 -0
@@ -225,6 +225,7 @@ config PPC select MODULES_USE_ELF_RELA select NO_BOOTMEM select OF + select OF_DMA_DEFAULT_COHERENT if !NOT_COHERENT_CACHE select OF_EARLY_FLATTREE select OF_RESERVED_MEM select OLD_SIGACTION if PPC32
drivers/of/Kconfig +4 -0
@@ -112,4 +112,8 @@ config OF_OVERLAY config OF_NUMA bool +config OF_DMA_DEFAULT_COHERENT + # arches should select this if DMA is coherent by default for OF devices + bool + endif # OF
drivers/of/address.c +5 -1
@@ -894,12 +894,16 @@ EXPORT_SYMBOL_GPL(of_dma_get_range);* @np: device node * * It returns true if "dma-coherent" property was found -* for this device in DT. +* for this device in the DT, or if DMA is coherent by +* default for OF devices on the current platform. */ bool of_dma_is_coherent (struct device_node *np) {struct device_node *node =of_node_get (np); +if (IS_ENABLED (CONFIG_OF_DMA_DEFAULT_COHERENT )) +return true; +while (node) {if (of_property_read_bool (node,"dma-coherent" )) {of_node_put (node);
Greg Kroah-Hartman
maintainerPATCH 54/173 dm zoned: support zone sizes smaller than 128MiB
In reply to Greg Kroah-HartmanFrom: Dmitry Fomichev <[email protected]>
commit b39962950339912978484cdac50069258545d753 upstream.
dm-zoned is observed to log failed kernel assertions and not work correctly when operating against a device with a zone size smaller than 128MiB (e.g. 32768 bits per 4K block). The reason is that the bitmap size per zone is calculated as zero with such a small zone size. Fix this problem and also make the code related to zone bitmap management be able to handle per zone bitmaps smaller than a single block.
A dm-zoned-tools patch is required to properly format dm-zoned devices with zone sizes smaller than 128MiB.
--- drivers/md/dm-zoned-metadata.c | 23 ++++++++++++++--------- 1 file changed, 14 insertions(+), 9 deletions(-)
drivers/md/dm-zoned-metadata.c +14 -9
@@ -132,6 +132,7 @@ struct dmz_metadata {sector_t zone_bitmap_size;unsigned int zone_nr_bitmap_blocks; +unsigned int zone_bits_per_mblk;unsigned int nr_bitmap_blocks;unsigned int nr_map_blocks;
@@ -1165,7 +1166,10 @@ static int dmz_init_zones(struct dmz_met/* Init */ zmd->zone_bitmap_size = dev->zone_nr_blocks >>3 ; - zmd->zone_nr_bitmap_blocks = zmd->zone_bitmap_size >>DMZ_BLOCK_SHIFT ; + zmd->zone_nr_bitmap_blocks = +max_t (sector_t,1 , zmd->zone_bitmap_size >>DMZ_BLOCK_SHIFT ); + zmd->zone_bits_per_mblk =min_t (sector_t, dev->zone_nr_blocks, +DMZ_BLOCK_SIZE_BITS );/* Allocate zone array */ zmd->zones =kcalloc (dev->nr_zones,sizeof (struct dm_zone ),GFP_KERNEL );
@@ -1982,7 +1986,7 @@ int dmz_copy_valid_blocks(struct dmz_metdmz_release_mblock (zmd, to_mblk);dmz_release_mblock (zmd, from_mblk); - chunk_block +=DMZ_BLOCK_SIZE_BITS ; + chunk_block += zmd->zone_bits_per_mblk; } to_zone->weight = from_zone->weight;
@@ -2043,7 +2047,7 @@ int dmz_validate_blocks(struct dmz_metad/* Set bits */ bit = chunk_block &DMZ_BLOCK_MASK_BITS ; - nr_bits =min (nr_blocks,DMZ_BLOCK_SIZE_BITS - bit); + nr_bits =min (nr_blocks, zmd->zone_bits_per_mblk - bit); count =dmz_set_bits ((unsigned long *)mblk->data, bit, nr_bits);if (count) {
@@ -2122,7 +2126,7 @@ int dmz_invalidate_blocks(struct dmz_met/* Clear bits */ bit = chunk_block &DMZ_BLOCK_MASK_BITS ; - nr_bits =min (nr_blocks,DMZ_BLOCK_SIZE_BITS - bit); + nr_bits =min (nr_blocks, zmd->zone_bits_per_mblk - bit); count =dmz_clear_bits ((unsigned long *)mblk->data, bit, nr_bits);
@@ -2182,6 +2186,7 @@ static int dmz_to_next_set_block(struct {struct dmz_mblock *mblk;unsigned int bit, set_bit, nr_bits; +unsigned int zone_bits = zmd->zone_bits_per_mblk;unsigned long *bitmap;int n =0 ;
@@ -2196,15 +2201,15 @@ static int dmz_to_next_set_block(struct/* Get offset */ bitmap = (unsigned long *) mblk->data; bit = chunk_block &DMZ_BLOCK_MASK_BITS ; - nr_bits =min (nr_blocks,DMZ_BLOCK_SIZE_BITS - bit); + nr_bits =min (nr_blocks, zone_bits - bit);if (set) - set_bit =find_next_bit (bitmap,DMZ_BLOCK_SIZE_BITS , bit); + set_bit =find_next_bit (bitmap, zone_bits, bit);else - set_bit =find_next_zero_bit (bitmap,DMZ_BLOCK_SIZE_BITS , bit); + set_bit =find_next_zero_bit (bitmap, zone_bits, bit);dmz_release_mblock (zmd, mblk); n += set_bit - bit; -if (set_bit <DMZ_BLOCK_SIZE_BITS ) +if (set_bit < zone_bits)break ; nr_blocks -= nr_bits;
@@ -2307,7 +2312,7 @@ static void dmz_get_zone_weight(struct d/* Count bits in this block */ bitmap = mblk->data; bit = chunk_block &DMZ_BLOCK_MASK_BITS ; - nr_bits =min (nr_blocks,DMZ_BLOCK_SIZE_BITS - bit); + nr_bits =min (nr_blocks, zmd->zone_bits_per_mblk - bit); n +=dmz_count_bits (bitmap, bit, nr_bits);dmz_release_mblock (zmd, mblk);
Greg Kroah-Hartman
maintainerPATCH 55/173 dm space map common: fix to ensure new block isnt already in use
In reply to Greg Kroah-HartmanFrom: Joe Thornber <[email protected]>
commit 4feaef830de7ffdd8352e1fe14ad3bf13c9688f8 upstream.
The space-maps track the reference counts for disk blocks allocated by both the thin-provisioning and cache targets. There are variants for tracking metadata blocks and data blocks.
Transactionality is implemented by never touching blocks from the previous transaction, so we can rollback in the event of a crash.
When allocating a new block we need to ensure the block is free (has reference count of 0) in both the current and previous transaction. Prior to this fix we were doing this by searching for a free block in the previous transaction, and relying on a 'begin' counter to track where the last allocation in the current transaction was. This 'begin' field was not being updated in all code paths (eg, increment of a data block reference count due to breaking sharing of a neighbour block in the same btree leaf).
This fix keeps the 'begin' field, but now it's just a hint to speed up the search. Instead the current transaction is searched for a free block, and then the old transaction is double checked to ensure it's free. Much simpler.
This fixes reports of sm_disk_new_block()'s BUG_ON() triggering when DM thin-provisioning's snapshots are heavily used.
--- drivers/md/persistent-data/dm-space-map-common.c | 27 +++++++++++++++++++++ drivers/md/persistent-data/dm-space-map-common.h | 2 + drivers/md/persistent-data/dm-space-map-disk.c | 6 +++- drivers/md/persistent-data/dm-space-map-metadata.c | 5 +++ 4 files changed, 37 insertions(+), 3 deletions(-)
drivers/md/persistent-data/dm-space-map-common.c +27 -0
@@ -382,6 +382,33 @@ int sm_ll_find_free_block(struct ll_diskreturn -ENOSPC ; } +int sm_ll_find_common_free_block (struct ll_disk *old_ll,struct ll_disk *new_ll, +dm_block_t begin,dm_block_t end,dm_block_t *b) +{ +int r; +uint32_t count; + +do { + r =sm_ll_find_free_block (new_ll, begin, new_ll->nr_blocks, b); +if (r) +break ; + +/* double check this block wasn't used in the old transaction */ +if (*b >= old_ll->nr_blocks) + count =0 ; +else { + r =sm_ll_lookup (old_ll, *b, &count); +if (r) +break ; + +if (count) + begin = *b +1 ; + } + }while (count); + +return r; +} +static int sm_ll_mutate (struct ll_disk *ll,dm_block_t b,int (*mutator)(void *context,uint32_t old,uint32_t *new),void *context,enum allocation_event *ev)
drivers/md/persistent-data/dm-space-map-common.h +2 -0
@@ -109,6 +109,8 @@ int sm_ll_lookup_bitmap(struct ll_disk *int sm_ll_lookup (struct ll_disk *ll,dm_block_t b,uint32_t *result);int sm_ll_find_free_block (struct ll_disk *ll,dm_block_t begin,dm_block_t end,dm_block_t *result); +int sm_ll_find_common_free_block (struct ll_disk *old_ll,struct ll_disk *new_ll, +dm_block_t begin,dm_block_t end,dm_block_t *result);int sm_ll_insert (struct ll_disk *ll,dm_block_t b,uint32_t ref_count,enum allocation_event *ev);int sm_ll_inc (struct ll_disk *ll,dm_block_t b,enum allocation_event *ev);int sm_ll_dec (struct ll_disk *ll,dm_block_t b,enum allocation_event *ev);
drivers/md/persistent-data/dm-space-map-disk.c +4 -2
@@ -167,8 +167,10 @@ static int sm_disk_new_block(struct dm_senum allocation_event ev; struct sm_disk *smd = container_of(sm, struct sm_disk, sm); -/* FIXME: we should loop round a couple of times */ - r =sm_ll_find_free_block (&smd->old_ll, smd->begin, smd->old_ll.nr_blocks, b); +/* +* Any block we allocate has to be free in both the old and current ll. +*/ + r =sm_ll_find_common_free_block (&smd->old_ll, &smd->ll, smd->begin, smd->ll.nr_blocks, b);if (r)return r;
drivers/md/persistent-data/dm-space-map-metadata.c +4 -1
@@ -447,7 +447,10 @@ static int sm_metadata_new_block_(structenum allocation_event ev; struct sm_metadata *smm = container_of(sm, struct sm_metadata, sm); - r =sm_ll_find_free_block (&smm->old_ll, smm->begin, smm->old_ll.nr_blocks, b); +/* +* Any block we allocate has to be free in both the old and current ll. +*/ + r =sm_ll_find_common_free_block (&smm->old_ll, &smm->ll, smm->begin, smm->ll.nr_blocks, b);if (r)return r;
Greg Kroah-Hartman
maintainerPATCH 56/173 dm crypt: fix benbi IV constructor crash if used in authenticated mode
In reply to Greg Kroah-HartmanFrom: Milan Broz <[email protected]>
commit 4ea9471fbd1addb25a4d269991dc724e200ca5b5 upstream.
If benbi IV is used in AEAD construction, for example:
cryptsetup luksFormat <device> --cipher twofish-xts-benbi --key-size 512 --integrity=hmac-sha256
the constructor uses wrong skcipher function and crashes:
BUG: kernel NULL pointer dereference, address: 00000014 ... EIP: crypt_iv_benbi_ctr+0x15/0x70 [dm_crypt] Call Trace: ? crypt_subkey_size+0x20/0x20 [dm_crypt] crypt_ctr+0x567/0xfc0 [dm_crypt] dm_table_add_target+0x15f/0x340 [dm_mod]
Fix this by properly using crypt_aead_blocksize() in this case.
--- drivers/md/dm-crypt.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)
drivers/md/dm-crypt.c +8 -2
@@ -485,8 +485,14 @@ static int crypt_iv_essiv_gen(struct crystatic int crypt_iv_benbi_ctr (struct crypt_config *cc,struct dm_target *ti,const char *opts) { -unsigned bs =crypto_skcipher_blocksize (any_tfm (cc)); -int log =ilog2 (bs); +unsigned bs; +int log; + +if (test_bit (CRYPT_MODE_INTEGRITY_AEAD , &cc->cipher_flags)) + bs =crypto_aead_blocksize (any_tfm_aead (cc)); +else + bs =crypto_skcipher_blocksize (any_tfm (cc)); + log =ilog2 (bs);/* we need to calculate how far we must shift the sector count * to get the cipher block count, we use this shift in _gen */
Greg Kroah-Hartman
maintainerPATCH 57/173 padata: Remove broken queue flushing
In reply to Greg Kroah-HartmanFrom: Herbert Xu <[email protected]>
[ Upstream commit 07928d9bfc81640bab36f5190e8725894d93b659 ]
The function padata_flush_queues is fundamentally broken because it cannot force padata users to complete the request that is underway. IOW padata has to passively wait for the completion of any outstanding work.
As it stands flushing is used in two places. Its use in padata_stop is simply unnecessary because nothing depends on the queues to be flushed afterwards.
The other use in padata_replace is more substantial as we depend on it to free the old pd structure. This patch instead uses the pd->refcnt to dynamically free the pd structure once all requests are complete.
--- kernel/padata.c | 46 ++++++++++++---------------------------------- 1 file changed, 12 insertions(+), 34 deletions(-)
kernel/padata.c +12 -34
@@ -34,6 +34,8 @@#define MAX_OBJ_NUM 1000 +static void padata_free_pd (struct parallel_data *pd); +static int padata_index_to_cpu (struct parallel_data *pd,int cpu_index) {int cpu, target_cpu;
@@ -292,6 +294,7 @@ static void padata_serial_worker(struct work_struct *serial_work)struct padata_serial_queue *squeue;struct parallel_data *pd;LIST_HEAD (local_list); +int cnt;local_bh_disable (); squeue = container_of(serial_work, struct padata_serial_queue, work);
@@ -301,6 +304,8 @@ static void padata_serial_worker(struct work_struct *serial_work)list_replace_init (&squeue->serial.list, &local_list);spin_unlock (&squeue->serial.lock); + cnt =0 ; +while (!list_empty (&local_list)) {struct padata_priv *padata;
@@ -310,9 +315,12 @@ static void padata_serial_worker(struct work_struct *serial_work)list_del_init (&padata->list); padata->serial (padata); -atomic_dec (&pd->refcnt); + cnt++; }local_bh_enable (); + +if (atomic_sub_and_test (cnt, &pd->refcnt)) +padata_free_pd (pd); }/**
@@ -435,8 +443,7 @@ static struct parallel_data *padata_alloc_pd(struct padata_instance *pinst,setup_timer (&pd->timer, padata_reorder_timer, (unsigned long )pd);atomic_set (&pd->seq_nr,-1 );atomic_set (&pd->reorder_objects,0 ); -atomic_set (&pd->refcnt,0 ); - pd->pinst = pinst; +atomic_set (&pd->refcnt,1 );spin_lock_init (&pd->lock);return pd;
@@ -460,31 +467,6 @@ static void padata_free_pd(struct parallel_data *pd)kfree (pd); } -/* Flush all objects out of the padata queues. */ -static void padata_flush_queues (struct parallel_data *pd) -{ -int cpu; -struct padata_parallel_queue *pqueue; -struct padata_serial_queue *squeue; - -for_each_cpu (cpu, pd->cpumask.pcpu) { - pqueue =per_cpu_ptr (pd->pqueue, cpu); -flush_work (&pqueue->work); - } - -del_timer_sync (&pd->timer); - -if (atomic_read (&pd->reorder_objects)) -padata_reorder (pd); - -for_each_cpu (cpu, pd->cpumask.cbcpu) { - squeue =per_cpu_ptr (pd->squeue, cpu); -flush_work (&squeue->work); - } - -BUG_ON (atomic_read (&pd->refcnt) !=0 ); -} -static void __padata_start (struct padata_instance *pinst) { pinst->flags |=PADATA_INIT ;
@@ -498,10 +480,6 @@ static void __padata_stop(struct padata_instance *pinst) pinst->flags &= ~PADATA_INIT ;synchronize_rcu (); - -get_online_cpus (); -padata_flush_queues (pinst->pd); -put_online_cpus (); }/* Replace the internal control structure with a new one. */
@@ -522,8 +500,8 @@ static void padata_replace(struct padata_instance *pinst,if (!cpumask_equal (pd_old->cpumask.cbcpu, pd_new->cpumask.cbcpu)) notification_mask |=PADATA_CPU_SERIAL ; -padata_flush_queues (pd_old); -padata_free_pd (pd_old); +if (atomic_dec_and_test (&pd_old->refcnt)) +padata_free_pd (pd_old);if (notification_mask) blocking_notifier_call_chain(&pinst->cpumask_change_notifier,
Signature · 1 line
2.20.1
PATCH v2 1/1 padata: Remove broken queue flushing
From: Herbert Xu <[email protected]>
[ Upstream commit 07928d9bfc81640bab36f5190e8725894d93b659 ]
The function padata_flush_queues is fundamentally broken because it cannot force padata users to complete the request that is underway. IOW padata has to passively wait for the completion of any outstanding work.
As it stands flushing is used in two places. Its use in padata_stop is simply unnecessary because nothing depends on the queues to be flushed afterwards.
The other use in padata_replace is more substantial as we depend on it to free the old pd structure. This patch instead uses the pd->refcnt to dynamically free the pd structure once all requests are complete.
--- kernel/padata.c | 45 ++++++++++++--------------------------------- 1 file changed, 12 insertions(+), 33 deletions(-)
kernel/padata.c +12 -33
@@ -34,6 +34,8 @@#define MAX_OBJ_NUM 1000 +static void padata_free_pd (struct parallel_data *pd); +static int padata_index_to_cpu (struct parallel_data *pd,int cpu_index) {int cpu, target_cpu;
@@ -292,6 +294,7 @@ static void padata_serial_worker(struct work_struct *serial_work)struct padata_serial_queue *squeue;struct parallel_data *pd;LIST_HEAD (local_list); +int cnt;local_bh_disable (); squeue = container_of(serial_work, struct padata_serial_queue, work);
@@ -301,6 +304,8 @@ static void padata_serial_worker(struct work_struct *serial_work)list_replace_init (&squeue->serial.list, &local_list);spin_unlock (&squeue->serial.lock); + cnt =0 ; +while (!list_empty (&local_list)) {struct padata_priv *padata;
@@ -310,9 +315,12 @@ static void padata_serial_worker(struct work_struct *serial_work)list_del_init (&padata->list); padata->serial (padata); -atomic_dec (&pd->refcnt); + cnt++; }local_bh_enable (); + +if (atomic_sub_and_test (cnt, &pd->refcnt)) +padata_free_pd (pd); }/**
@@ -435,7 +443,7 @@ static struct parallel_data *padata_alloc_pd(struct padata_instance *pinst,setup_timer (&pd->timer, padata_reorder_timer, (unsigned long )pd);atomic_set (&pd->seq_nr,-1 );atomic_set (&pd->reorder_objects,0 ); -atomic_set (&pd->refcnt,0 ); +atomic_set (&pd->refcnt,1 ); pd->pinst = pinst;spin_lock_init (&pd->lock);
@@ -460,31 +468,6 @@ static void padata_free_pd(struct parallel_data *pd)kfree (pd); } -/* Flush all objects out of the padata queues. */ -static void padata_flush_queues (struct parallel_data *pd) -{ -int cpu; -struct padata_parallel_queue *pqueue; -struct padata_serial_queue *squeue; - -for_each_cpu (cpu, pd->cpumask.pcpu) { - pqueue =per_cpu_ptr (pd->pqueue, cpu); -flush_work (&pqueue->work); - } - -del_timer_sync (&pd->timer); - -if (atomic_read (&pd->reorder_objects)) -padata_reorder (pd); - -for_each_cpu (cpu, pd->cpumask.cbcpu) { - squeue =per_cpu_ptr (pd->squeue, cpu); -flush_work (&squeue->work); - } - -BUG_ON (atomic_read (&pd->refcnt) !=0 ); -} -static void __padata_start (struct padata_instance *pinst) { pinst->flags |=PADATA_INIT ;
@@ -498,10 +481,6 @@ static void __padata_stop(struct padata_instance *pinst) pinst->flags &= ~PADATA_INIT ;synchronize_rcu (); - -get_online_cpus (); -padata_flush_queues (pinst->pd); -put_online_cpus (); }/* Replace the internal control structure with a new one. */
@@ -522,8 +501,8 @@ static void padata_replace(struct padata_instance *pinst,if (!cpumask_equal (pd_old->cpumask.cbcpu, pd_new->cpumask.cbcpu)) notification_mask |=PADATA_CPU_SERIAL ; -padata_flush_queues (pd_old); -padata_free_pd (pd_old); +if (atomic_dec_and_test (&pd_old->refcnt)) +padata_free_pd (pd_old);if (notification_mask) blocking_notifier_call_chain(&pinst->cpumask_change_notifier,
Signature · 1 line
2.25.0
Greg Kroah-Hartman
maintainerRe: [PATCH v2 4.14] padata: Remove broken queue flushing
Quoted 18 lines · Daniel Jordan
From: Herbert Xu <[email protected]>
[ Upstream commit
07928d9bfc81640bab36f5190e8725894d93b659]The function
padata_flush_queuesis fundamentally broken because it cannot force padata users to complete the request that is underway. IOW padata has to passively wait for the completion of any outstanding work.As it stands flushing is used in two places. Its use in
padata_stopis simply unnecessary because nothing depends on the queues to be flushed afterwards.The other use in
padata_replaceis more substantial as we depend on it to free the old pd structure. This patch instead uses thepd->refcntto dynamically free the pd structure once all requests are complete.
Quoted 7 lines
Fixes 2b73b07ab8a4 ("padata: Flush the padata queues actively")Cc <[email protected]>Signed-off-by Herbert Xu <[email protected]>Reviewed-by Daniel Jordan <[email protected]>Signed-off-by Herbert Xu <[email protected]>[dj: leave "pd->pinst = pinst" assignment in padata_alloc_pd()]Signed-off-by Daniel Jordan <[email protected]>
Thanks, all 3 backports now queued up.
greg k-h
Greg Kroah-Hartman
maintainerPATCH 58/173 tracing: Annotate ftrace_graph_hash pointer with __rcu
In reply to Greg Kroah-HartmanFrom: Amol Grover <[email protected]>
[ Upstream commit 24a9729f831462b1d9d61dc85ecc91c59037243f ]
Fix following instances of sparse error kernel/trace/ftrace.c:5664:29: error: incompatible types in comparison kernel/trace/ftrace.c:5785:21: error: incompatible types in comparison kernel/trace/ftrace.c:5864:36: error: incompatible types in comparison kernel/trace/ftrace.c:5866:25: error: incompatible types in comparison
Use rcu_dereference_protected to access the __rcu annotated pointer.
--- kernel/trace/ftrace.c | 2 +- kernel/trace/trace.h | 9 ++++++--- 2 files changed, 7 insertions(+), 4 deletions(-)
kernel/trace/ftrace.c +1 -1
@@ -5146,7 +5146,7 @@ static const struct file_operations ftrace_notrace_fops = {static DEFINE_MUTEX (graph_lock); -struct ftrace_hash *ftrace_graph_hash =EMPTY_HASH ; +struct ftrace_hash __rcu *ftrace_graph_hash =EMPTY_HASH ;struct ftrace_hash *ftrace_graph_notrace_hash =EMPTY_HASH ;enum graph_filter_type {
kernel/trace/trace.h +6 -3
@@ -868,22 +868,25 @@ extern void __trace_graph_return(struct trace_array *tr, unsigned long flags, int pc);#ifdef CONFIG_DYNAMIC_FTRACE -extern struct ftrace_hash *ftrace_graph_hash; +extern struct ftrace_hash __rcu *ftrace_graph_hash;extern struct ftrace_hash *ftrace_graph_notrace_hash;static inline int ftrace_graph_addr (struct ftrace_graph_ent *trace) {unsigned long addr = trace->func;int ret =0 ; +struct ftrace_hash *hash;preempt_disable_notrace (); -if (ftrace_hash_empty (ftrace_graph_hash)) { + hash =rcu_dereference_protected (ftrace_graph_hash, !preemptible ()); + +if (ftrace_hash_empty (hash)) { ret =1 ; goto out; } -if (ftrace_lookup_ip (ftrace_graph_hash, addr)) { +if (ftrace_lookup_ip (hash, addr)) {/* * This needs to be cleared on the return functions
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 59/173 tracing: Annotate ftrace_graph_notrace_hash pointer with __rcu
In reply to Greg Kroah-HartmanFrom: Amol Grover <[email protected]>
[ Upstream commit fd0e6852c407dd9aefc594f54ddcc21d84803d3b ]
Fix following instances of sparse error kernel/trace/ftrace.c:5667:29: error: incompatible types in comparison kernel/trace/ftrace.c:5813:21: error: incompatible types in comparison kernel/trace/ftrace.c:5868:36: error: incompatible types in comparison kernel/trace/ftrace.c:5870:25: error: incompatible types in comparison
Use rcu_dereference_protected to dereference the newly annotated pointer.
--- kernel/trace/ftrace.c | 2 +- kernel/trace/trace.h | 8 ++++++-- 2 files changed, 7 insertions(+), 3 deletions(-)
kernel/trace/ftrace.c +1 -1
@@ -5147,7 +5147,7 @@ static const struct file_operations ftrace_notrace_fops = {static DEFINE_MUTEX (graph_lock);struct ftrace_hash __rcu *ftrace_graph_hash =EMPTY_HASH ; -struct ftrace_hash *ftrace_graph_notrace_hash =EMPTY_HASH ; +struct ftrace_hash __rcu *ftrace_graph_notrace_hash =EMPTY_HASH ;enum graph_filter_type {GRAPH_FILTER_NOTRACE =0 ,
kernel/trace/trace.h +6 -2
@@ -869,7 +869,7 @@ extern void __trace_graph_return(struct trace_array *tr,#ifdef CONFIG_DYNAMIC_FTRACE extern struct ftrace_hash __rcu *ftrace_graph_hash; -extern struct ftrace_hash *ftrace_graph_notrace_hash; +extern struct ftrace_hash __rcu *ftrace_graph_notrace_hash;static inline int ftrace_graph_addr (struct ftrace_graph_ent *trace) {
@@ -922,10 +922,14 @@ static inline void ftrace_graph_addr_finish(struct ftrace_graph_ret *trace)static inline int ftrace_graph_notrace_addr (unsigned long addr) {int ret =0 ; +struct ftrace_hash *notrace_hash;preempt_disable_notrace (); -if (ftrace_lookup_ip (ftrace_graph_notrace_hash, addr)) + notrace_hash =rcu_dereference_protected (ftrace_graph_notrace_hash, + !preemptible ()); + +if (ftrace_lookup_ip (notrace_hash, addr)) ret =1 ;preempt_enable_notrace ();
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 60/173 ftrace: Add comment to why rcu_dereference_sched() is open coded
In reply to Greg Kroah-HartmanFrom: Steven Rostedt (VMware) <[email protected]>
[ Upstream commit 16052dd5bdfa16dbe18d8c1d4cde2ddab9d23177 ]
Because the function graph tracer can execute in sections where RCU is not "watching", the rcu_dereference_sched() for the has needs to be open coded. This is fine because the RCU "flavor" of the ftrace hash is protected by its own RCU handling (it does its own little synchronization on every CPU and does not rely on RCU sched).
--- kernel/trace/trace.h | 10 ++++++++++ 1 file changed, 10 insertions(+)
kernel/trace/trace.h +10 -0
@@ -879,6 +879,11 @@ static inline int ftrace_graph_addr(struct ftrace_graph_ent *trace)preempt_disable_notrace (); +/* +* Have to open code "rcu_dereference_sched()" because the +* function graph tracer can be called when RCU is not +* "watching". +*/ hash =rcu_dereference_protected (ftrace_graph_hash, !preemptible ());if (ftrace_hash_empty (hash)) {
@@ -926,6 +931,11 @@ static inline int ftrace_graph_notrace_addr(unsigned long addr)preempt_disable_notrace (); +/* +* Have to open code "rcu_dereference_sched()" because the +* function graph tracer can be called when RCU is not +* "watching". +*/ notrace_hash =rcu_dereference_protected (ftrace_graph_notrace_hash, !preemptible ());
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 61/173 ftrace: Protect ftrace_graph_hash with ftrace_sync
In reply to Greg Kroah-HartmanFrom: Steven Rostedt (VMware) <[email protected]>
[ Upstream commit 54a16ff6f2e50775145b210bcd94d62c3c2af117 ]
As function_graph tracer can run when RCU is not "watching", it can not be protected by synchronize_rcu() it requires running a task on each CPU before it can be freed. Calling schedule_on_each_cpu(ftrace_sync) needs to be used.
--- kernel/trace/ftrace.c | 11 +++++++++-- kernel/trace/trace.h | 2 ++ 2 files changed, 11 insertions(+), 2 deletions(-)
kernel/trace/ftrace.c +9 -2
@@ -5419,8 +5419,15 @@ ftrace_graph_release(struct inode *inode, struct file *file)mutex_unlock (&graph_lock); -/* Wait till all users are no longer using the old hash */ -synchronize_sched (); +/* +* We need to do a hard force of sched synchronization. +* This is because we use preempt_disable() to do RCU, but +* the function tracers can be called where RCU is not watching +* (like before user_exit()). We can not rely on the RCU +* infrastructure to do the synchronization, thus we must do it +* ourselves. +*/ +schedule_on_each_cpu (ftrace_sync);free_ftrace_hash (old_hash); }
kernel/trace/trace.h +2 -0
@@ -883,6 +883,7 @@ static inline int ftrace_graph_addr(struct ftrace_graph_ent *trace)* Have to open code "rcu_dereference_sched()" because the * function graph tracer can be called when RCU is not * "watching". +* Protected with schedule_on_each_cpu(ftrace_sync) */ hash =rcu_dereference_protected (ftrace_graph_hash, !preemptible ());
@@ -935,6 +936,7 @@ static inline int ftrace_graph_notrace_addr(unsigned long addr)* Have to open code "rcu_dereference_sched()" because the * function graph tracer can be called when RCU is not * "watching". +* Protected with schedule_on_each_cpu(ftrace_sync) */ notrace_hash =rcu_dereference_protected (ftrace_graph_notrace_hash, !preemptible ());
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 62/173 samples/bpf: Dont try to remove users homedir on clean
In reply to Greg Kroah-HartmanFrom: Toke Høiland-Jørgensen <[email protected]>
commit b2e5e93ae8af6a34bca536cdc4b453ab1e707b8b upstream.
The 'clean' rule in the samples/bpf Makefile tries to remove backup files (ending in ~). However, if no such files exist, it will instead try to remove the user's home directory. While the attempt is mostly harmless, it does lead to a somewhat scary warning like this:
rm: cannot remove '~': Is a directory
Fix this by using find instead of shell expansion to locate any actual backup files that need to be removed.
--- samples/bpf/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
samples/bpf/Makefile +1 -1
@@ -184,7 +184,7 @@ all: $(LIBBPF) clean: $(MAKE) -C ../../ M=$(CURDIR) clean - @rm -f *~ + @find $(CURDIR) -type f -name '*~' -delete $(LIBBPF): FORCE $(MAKE) -C $(dir $@) $(notdir $@)
Greg Kroah-Hartman
maintainerPATCH 63/173 crypto: ccp - set max RSA modulus size for v3 platform devices as well
In reply to Greg Kroah-HartmanFrom: Ard Biesheuvel <[email protected]>
commit 11548f5a5747813ff84bed6f2ea01100053b0d8d upstream.
AMD Seattle incorporates a non-PCI version of the v3 CCP crypto accelerator, and this version was left behind when the maximum RSA modulus size was parameterized in order to support v5 hardware which supports larger moduli than v3 hardware does. Due to this oversight, RSA acceleration no longer works at all on these systems.
Fix this by setting the .rsamax property to the appropriate value for v3 platform hardware.
--- drivers/crypto/ccp/ccp-dev-v3.c | 1 + 1 file changed, 1 insertion(+)
drivers/crypto/ccp/ccp-dev-v3.c +1 -0
@@ -590,6 +590,7 @@ const struct ccp_vdata ccpv3_platform = .setup = NULL, .perform = &ccp3_actions, .offset = 0, + .rsamax = CCP_RSA_MAX_WIDTH, };const struct ccp_vdata ccpv3 = {
Greg Kroah-Hartman
maintainerPATCH 64/173 crypto: pcrypt - Do not clear MAY_SLEEP flag in original request
In reply to Greg Kroah-HartmanFrom: Herbert Xu <[email protected]>
commit e8d998264bffade3cfe0536559f712ab9058d654 upstream.
We should not be modifying the original request's MAY_SLEEP flag upon completion. It makes no sense to do so anyway.
--- crypto/pcrypt.c | 1 - 1 file changed, 1 deletion(-)
crypto/pcrypt.c +0 -1
@@ -130,7 +130,6 @@ static void pcrypt_aead_done(struct crypstruct padata_priv *padata =pcrypt_request_padata (preq); padata->info = err; - req->base.flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP ;padata_do_serial (padata); }
Greg Kroah-Hartman
maintainerPATCH 65/173 crypto: atmel-aes - Fix counter overflow in CTR mode
In reply to Greg Kroah-HartmanFrom: Tudor Ambarus <[email protected]>
commit 781a08d9740afa73357f1a60d45d7c93d7cca2dd upstream.
32 bit counter is not supported by neither of our AES IPs, all implement a 16 bit block counter. Drop the 32 bit block counter logic.
--- drivers/crypto/atmel-aes.c | 37 ++++++++++++------------------------- 1 file changed, 12 insertions(+), 25 deletions(-)
drivers/crypto/atmel-aes.c +12 -25
@@ -91,7 +91,6 @@struct atmel_aes_caps {bool has_dualbuff;bool has_cfb64; -bool has_ctr32;bool has_gcm;bool has_xts;bool has_authenc;
@@ -990,8 +989,9 @@ static int atmel_aes_ctr_transfer(structstruct atmel_aes_ctr_ctx *ctx =atmel_aes_ctr_ctx_cast (dd->ctx);struct ablkcipher_request *req =ablkcipher_request_cast (dd->areq);struct scatterlist *src, *dst; -u32 ctr, blocks;size_t datalen; +u32 ctr; +u16 blocks, start, end;bool use_dma, fragmented = false;/* Check for transfer completion. */
@@ -1003,27 +1003,17 @@ static int atmel_aes_ctr_transfer(struct datalen = req->nbytes - ctx->offset; blocks =DIV_ROUND_UP (datalen,AES_BLOCK_SIZE ); ctr =be32_to_cpu (ctx->iv[3 ]); -if (dd->caps.has_ctr32) { -/* Check 32bit counter overflow. */ -u32 start = ctr; -u32 end = start + blocks -1 ; - -if (end < start) { - ctr |=0xffffffff ; - datalen =AES_BLOCK_SIZE * -start; - fragmented = true; - } - }else { -/* Check 16bit counter overflow. */ -u16 start = ctr &0xffff ; -u16 end = start + (u16 )blocks -1 ; - -if (blocks >>16 || end < start) { - ctr |=0xffff ; - datalen =AES_BLOCK_SIZE * (0x10000 -start); - fragmented = true; - } + +/* Check 16bit counter overflow. */ + start = ctr &0xffff ; + end = start + blocks -1 ; + +if (blocks >>16 || end < start) { + ctr |=0xffff ; + datalen =AES_BLOCK_SIZE * (0x10000 - start); + fragmented = true; } + use_dma = (datalen >=ATMEL_AES_DMA_THRESHOLD );/* Jump to offset. */
@@ -2536,7 +2526,6 @@ static void atmel_aes_get_cap(struct atm { dd->caps.has_dualbuff =0 ; dd->caps.has_cfb64 =0 ; - dd->caps.has_ctr32 =0 ; dd->caps.has_gcm =0 ; dd->caps.has_xts =0 ; dd->caps.has_authenc =0 ;
@@ -2547,7 +2536,6 @@ static void atmel_aes_get_cap(struct atmcase 0x500 : dd->caps.has_dualbuff =1 ; dd->caps.has_cfb64 =1 ; - dd->caps.has_ctr32 =1 ; dd->caps.has_gcm =1 ; dd->caps.has_xts =1 ; dd->caps.has_authenc =1 ;
@@ -2556,7 +2544,6 @@ static void atmel_aes_get_cap(struct atmcase 0x200 : dd->caps.has_dualbuff =1 ; dd->caps.has_cfb64 =1 ; - dd->caps.has_ctr32 =1 ; dd->caps.has_gcm =1 ; dd->caps.max_burst_size =4 ;break ;
Greg Kroah-Hartman
maintainerPATCH 66/173 crypto: api - Fix race condition in crypto_spawn_alg
In reply to Greg Kroah-HartmanFrom: Herbert Xu <[email protected]>
commit 73669cc556462f4e50376538d77ee312142e8a8a upstream.
The function crypto_spawn_alg is racy because it drops the lock before shooting the dying algorithm. The algorithm could disappear altogether before we shoot it.
This patch fixes it by moving the shooting into the locked section.
--- crypto/algapi.c | 16 +++++----------- crypto/api.c | 3 +-- crypto/internal.h | 1 - 3 files changed, 6 insertions(+), 14 deletions(-)
crypto/algapi.c +5 -11
@@ -662,22 +662,16 @@ EXPORT_SYMBOL_GPL(crypto_drop_spawn);static struct crypto_alg *crypto_spawn_alg (struct crypto_spawn *spawn) {struct crypto_alg *alg; -struct crypto_alg *alg2;down_read (&crypto_alg_sem); alg = spawn->alg; - alg2 = alg; -if (alg2) - alg2 =crypto_mod_get (alg2); -up_read (&crypto_alg_sem); - -if (!alg2) { -if (alg) -crypto_shoot_alg (alg); -return ERR_PTR (-EAGAIN ); +if (alg && !crypto_mod_get (alg)) { + alg->cra_flags |=CRYPTO_ALG_DYING ; + alg =NULL ; } +up_read (&crypto_alg_sem); -return alg; +return alg ?:ERR_PTR (-EAGAIN ); } struct crypto_tfm *crypto_spawn_tfm(struct crypto_spawn *spawn, u32 type,
crypto/api.c +1 -2
@@ -339,13 +339,12 @@ static unsigned int crypto_ctxsize(strucreturn len; } -void crypto_shoot_alg (struct crypto_alg *alg) +static void crypto_shoot_alg (struct crypto_alg *alg) {down_write (&crypto_alg_sem); alg->cra_flags |=CRYPTO_ALG_DYING ;up_write (&crypto_alg_sem); } -EXPORT_SYMBOL_GPL (crypto_shoot_alg);struct crypto_tfm *__crypto_alloc_tfm (struct crypto_alg *alg,u32 type,u32 mask)
crypto/internal.h +0 -1
@@ -84,7 +84,6 @@ void crypto_alg_tested(const char *name,void crypto_remove_spawns (struct crypto_alg *alg,struct list_head *list,struct crypto_alg *nalg);void crypto_remove_final (struct list_head *list); -void crypto_shoot_alg (struct crypto_alg *alg);struct crypto_tfm *__crypto_alloc_tfm (struct crypto_alg *alg,u32 type,u32 mask); void *crypto_create_tfm(struct crypto_alg *alg,
Greg Kroah-Hartman
maintainerPATCH 67/173 crypto: picoxcell - adjust the position of tasklet_init and fix missed tasklet_kill
In reply to Greg Kroah-HartmanFrom: Chuhong Yuan <[email protected]>
commit 7f8c36fe9be46862c4f3c5302f769378028a34fa upstream.
Since tasklet is needed to be initialized before registering IRQ handler, adjust the position of tasklet_init to fix the wrong order.
Besides, to fix the missed tasklet_kill, this patch adds a helper function and uses devm_add_action to kill the tasklet automatically.
--- drivers/crypto/picoxcell_crypto.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-)
drivers/crypto/picoxcell_crypto.c +13 -2
@@ -1616,6 +1616,11 @@ static const struct of_device_id spacc_oMODULE_DEVICE_TABLE (of, spacc_of_id_table);#endif /* CONFIG_OF */ +static void spacc_tasklet_kill (void *data) +{ +tasklet_kill (data); +} +static int spacc_probe (struct platform_device *pdev) {int i, err, ret = -EINVAL ;
@@ -1659,6 +1664,14 @@ static int spacc_probe(struct platform_dreturn -ENXIO ; } +tasklet_init (&engine->complete, spacc_spacc_complete, + (unsigned long )engine); + + ret =devm_add_action (&pdev->dev, spacc_tasklet_kill, + &engine->complete); +if (ret) +return ret; +if (devm_request_irq (&pdev->dev, irq->start, spacc_spacc_irq,0 , engine->name, engine)) {dev_err (engine->dev,"failed to request IRQ\n" );
@@ -1721,8 +1734,6 @@ static int spacc_probe(struct platform_dINIT_LIST_HEAD (&engine->completed);INIT_LIST_HEAD (&engine->in_progress); engine->in_flight =0 ; -tasklet_init (&engine->complete, spacc_spacc_complete, - (unsigned long )engine);platform_set_drvdata (pdev, engine);
Greg Kroah-Hartman
maintainerPATCH 68/173 scsi: qla2xxx: Fix unbound NVME response length
In reply to Greg Kroah-HartmanFrom: Arun Easi <[email protected]>
commit 00fe717ee1ea3c2979db4f94b1533c57aed8dea9 upstream.
On certain cases when response length is less than 32, NVME response data is supplied inline in IOCB. This is indicated by some combination of state flags. There was an instance when a high, and incorrect, response length was indicated causing driver to overrun buffers. Fix this by checking and limiting the response payload length.
--- drivers/scsi/qla2xxx/qla_dbg.c | 6 ------ drivers/scsi/qla2xxx/qla_dbg.h | 6 ++++++ drivers/scsi/qla2xxx/qla_isr.c | 12 ++++++++++++ 3 files changed, 18 insertions(+), 6 deletions(-)
drivers/scsi/qla2xxx/qla_dbg.c +0 -6
@@ -2517,12 +2517,6 @@ qla83xx_fw_dump_failed:/* Driver Debug Functions. */ /****************************************************************************/ -static inline int -ql_mask_match (uint32_t level) -{ -return (level & ql2xextended_error_logging) == level; -} -/* * This function is for formatting and logging debug information. * It is to be used when vha is available. It formats the message
drivers/scsi/qla2xxx/qla_dbg.h +6 -0
@@ -374,3 +374,9 @@ extern int qla24xx_dump_ram(struct qla_hextern void qla24xx_pause_risc (struct device_reg_24xx __iomem *,struct qla_hw_data *);extern int qla24xx_soft_reset (struct qla_hw_data *); + +static inline int +ql_mask_match (uint level) +{ +return (level & ql2xextended_error_logging) == level; +}
drivers/scsi/qla2xxx/qla_isr.c +12 -0
@@ -1853,6 +1853,18 @@ qla24xx_nvme_iocb_entry(scsi_qla_host_t inbuf = (uint32_t *)&sts->nvme_ersp_data; outbuf = (uint32_t *)fd->rspaddr; iocb->u.nvme.rsp_pyld_len =le16_to_cpu (sts->nvme_rsp_pyld_len); +if (unlikely (iocb->u.nvme.rsp_pyld_len > +sizeof (struct nvme_fc_ersp_iu ))) { +if (ql_mask_match (ql_dbg_io)) { +WARN_ONCE (1 ,"Unexpected response payload length %u.\n" , + iocb->u.nvme.rsp_pyld_len); +ql_log (ql_log_warn, fcport->vha,0x5100 , +"Unexpected response payload length %u.\n" , + iocb->u.nvme.rsp_pyld_len); + } + iocb->u.nvme.rsp_pyld_len = +sizeof (struct nvme_fc_ersp_iu ); + } iter = iocb->u.nvme.rsp_pyld_len >>2 ;for (; iter; iter--) *outbuf++ =swab32 (*inbuf++);
Greg Kroah-Hartman
maintainerPATCH 69/173 NFS: Fix memory leaks and corruption in readdir
In reply to Greg Kroah-HartmanFrom: Trond Myklebust <[email protected]>
commit 4b310319c6a8ce708f1033d57145e2aa027a883c upstream.
nfs_readdir_xdr_to_array() must not exit without having initialised the array, so that the page cache deletion routines can safely call nfs_readdir_clear_array().
Furthermore, we should ensure that if we exit nfs_readdir_filler() with an error, we free up any page contents to prevent a leak if we try to fill the page again.
--- fs/nfs/dir.c | 17 +++++++++++++++-- 1 file changed, 15 insertions(+), 2 deletions(-)
fs/nfs/dir.c +15 -2
@@ -169,6 +169,17 @@ typedef struct {bool eof; } nfs_readdir_descriptor_t; +static +void nfs_readdir_init_array (struct page *page) +{ +struct nfs_cache_array *array; + + array =kmap_atomic (page); +memset (array,0 ,sizeof (struct nfs_cache_array )); + array->eof_index =-1 ; +kunmap_atomic (array); +} +/* * we are freeing strings created by nfs_add_to_readdir_array() */
@@ -181,6 +192,7 @@ void nfs_readdir_clear_array(struct page array =kmap_atomic (page);for (i =0 ; i < array->size; i++)kfree (array->array[i].string.name); + array->size =0 ;kunmap_atomic (array); }
@@ -617,6 +629,8 @@ int nfs_readdir_xdr_to_array(nfs_readdirint status = -ENOMEM ;unsigned int array_size =ARRAY_SIZE (pages); +nfs_readdir_init_array (page); + entry.prev_cookie =0 ; entry.cookie = desc->last_cookie; entry.eof =0 ;
@@ -633,8 +647,6 @@ int nfs_readdir_xdr_to_array(nfs_readdir } array =kmap (page); -memset (array,0 ,sizeof (struct nfs_cache_array )); - array->eof_index =-1 ; status =nfs_readdir_alloc_pages (pages, array_size);if (status <0 )
@@ -688,6 +700,7 @@ int nfs_readdir_filler(nfs_readdir_descrunlock_page (page);return 0 ; error: +nfs_readdir_clear_array (page);unlock_page (page);return ret; }
Greg Kroah-Hartman
maintainerPATCH 70/173 NFS: Directory page cache pages need to be locked when read
In reply to Greg Kroah-HartmanFrom: Trond Myklebust <[email protected]>
commit 114de38225d9b300f027e2aec9afbb6e0def154b upstream.
When a NFS directory page cache page is removed from the page cache, its contents are freed through a call to nfs_readdir_clear_array(). To prevent the removal of the page cache entry until after we've finished reading it, we must take the page lock.
--- fs/nfs/dir.c | 30 +++++++++++++++++++----------- 1 file changed, 19 insertions(+), 11 deletions(-)
fs/nfs/dir.c +19 -11
@@ -708,8 +708,6 @@ int nfs_readdir_filler(nfs_readdir_descrstatic void cache_page_release (nfs_readdir_descriptor_t *desc) { -if (!desc->page->mapping) -nfs_readdir_clear_array (desc->page);put_page (desc->page); desc->page =NULL ; }
@@ -723,19 +721,28 @@ struct page *get_cache_page(nfs_readdir_/* * Returns 0 if desc->dir_cookie was found on page desc->page_index +* and locks the page to prevent removal from the page cache. */ static -int find_cache_page (nfs_readdir_descriptor_t *desc) +int find_and_lock_cache_page (nfs_readdir_descriptor_t *desc) {int res; desc->page =get_cache_page (desc);if (IS_ERR (desc->page))return PTR_ERR (desc->page); - - res =nfs_readdir_search_array (desc); + res =lock_page_killable (desc->page);if (res !=0 ) -cache_page_release (desc); + goto error; + res = -EAGAIN ; +if (desc->page->mapping !=NULL ) { + res =nfs_readdir_search_array (desc); +if (res ==0 ) +return 0 ; + } +unlock_page (desc->page); +error: +cache_page_release (desc);return res; }
@@ -750,7 +757,7 @@ int readdir_search_pagecache(nfs_readdir desc->last_cookie =0 ; }do { - res =find_cache_page (desc); + res =find_and_lock_cache_page (desc); }while (res == -EAGAIN );return res; }
@@ -789,7 +796,6 @@ int nfs_do_filldir(nfs_readdir_descripto desc->eof =1 ;kunmap (desc->page); -cache_page_release (desc);dfprintk (DIRCACHE ,"NFS: nfs_do_filldir() filling ended @ cookie %Lu; returning = %d\n" , (unsigned long long )*desc->dir_cookie, res);return res;
@@ -835,13 +841,13 @@ int uncached_readdir(nfs_readdir_descrip status =nfs_do_filldir (desc); + out_release: +nfs_readdir_clear_array (desc->page); +cache_page_release (desc); out:dfprintk (DIRCACHE ,"NFS: %s: returns %d\n" , __func__, status);return status; - out_release: -cache_page_release (desc); - goto out; }/* The file offset position represents the dirent entry number. A
@@ -906,6 +912,8 @@ static int nfs_readdir(struct file *filebreak ; res =nfs_do_filldir (desc); +unlock_page (desc->page); +cache_page_release (desc);if (res <0 )break ; }while (!desc->eof);
Greg Kroah-Hartman
maintainerPATCH 71/173 btrfs: set trans->drity in btrfs_commit_transaction
In reply to Greg Kroah-HartmanFrom: Josef Bacik <[email protected]>
commit d62b23c94952e78211a383b7d90ef0afbd9a3717 upstream.
If we abort a transaction we have the following sequence
if (!trans->dirty && list_empty(&trans->new_bgs))
return;
WRITE_ONCE(trans->transaction->aborted, err);
The idea being if we didn't modify anything with our trans handle then we don't really need to abort the whole transaction, maybe the other trans handles are fine and we can carry on.
However in the case of create_snapshot we add a pending_snapshot object to our transaction and then commit the transaction. We don't actually modify anything. sync() behaves the same way, attach to an existing transaction and commit it. This means that if we have an IO error in the right places we could abort the committing transaction with our trans->dirty being not set and thus not set transaction->aborted.
This is a problem because in the create_snapshot() case we depend on pending->error being set to something, or btrfs_commit_transaction returning an error.
If we are not the trans handle that gets to commit the transaction, and we're waiting on the commit to happen we get our return value from cur_trans->aborted. If this was not set to anything because sync() hit an error in the transaction commit before it could modify anything then cur_trans->aborted would be 0. Thus we'd return 0 from btrfs_commit_transaction() in create_snapshot.
This is a problem because we then try to do things with pending_snapshot->snap, which will be NULL because we didn't create the snapshot, and then we'll get a NULL pointer dereference like the following
"BUG: kernel NULL pointer dereference, address: 00000000000001f0"
RIP: 0010:btrfs_orphan_cleanup+0x2d/0x330 Call Trace: ? btrfs_mksubvol.isra.31+0x3f2/0x510 btrfs_mksubvol.isra.31+0x4bc/0x510 ? __sb_start_write+0xfa/0x200 ? mnt_want_write_file+0x24/0x50 btrfs_ioctl_snap_create_transid+0x16c/0x1a0 btrfs_ioctl_snap_create_v2+0x11e/0x1a0 btrfs_ioctl+0x1534/0x2c10 ? free_debug_processing+0x262/0x2a3 do_vfs_ioctl+0xa6/0x6b0 ? do_sys_open+0x188/0x220 ? syscall_trace_enter+0x1f8/0x330 ksys_ioctl+0x60/0x90 __x64_sys_ioctl+0x16/0x20 do_syscall_64+0x4a/0x1b0
In order to fix this we need to make sure anybody who calls commit_transaction has trans->dirty set so that they properly set the trans->transaction->aborted value properly so any waiters know bad things happened.
This was found while I was running generic/475 with my modified fsstress, it reproduced within a few runs. I ran with this patch all night and didn't see the problem again.
--- fs/btrfs/transaction.c | 8 ++++++++ 1 file changed, 8 insertions(+)
fs/btrfs/transaction.c +8 -0
@@ -1948,6 +1948,14 @@ int btrfs_commit_transaction(struct btrfstruct btrfs_transaction *prev_trans =NULL ;int ret; +/* +* Some places just start a transaction to commit it. We need to make +* sure that if this commit fails that the abort code actually marks the +* transaction as failed, so set trans->dirty to make the abort code do +* the right thing. +*/ + trans->dirty = true; +/* Stop the commit early if ->aborted is set */ if (unlikely (READ_ONCE (cur_trans->aborted))) { ret = cur_trans->aborted;
Greg Kroah-Hartman
maintainerPATCH 72/173 ARM: tegra: Enable PLLP bypass during Tegra124 LP1
In reply to Greg Kroah-HartmanFrom: Stephen Warren <[email protected]>
commit 1a3388d506bf5b45bb283e6a4c4706cfb4897333 upstream.
For a little over a year, U-Boot has configured the flow controller to perform automatic RAM re-repair on off->on power transitions of the CPU rail[1]. This is mandatory for correct operation of Tegra124. However, RAM re-repair relies on certain clocks, which the kernel must enable and leave running. PLLP is one of those clocks. This clock is shut down during LP1 in order to save power. Enable bypass (which I believe routes osc_div_clk, essentially the crystal clock, to the PLL output) so that this clock signal toggles even though the PLL is not active. This is required so that LP1 power mode (system suspend) operates correctly.
The bypass configuration must then be undone when resuming from LP1, so that all peripheral clocks run at the expected rate. Without this, many peripherals won't work correctly; for example, the UART baud rate would be incorrect.
NVIDIA's downstream kernel code only does this if not compiled for Tegra30, so the added code is made conditional upon the chip ID. NVIDIA's downstream code makes this change conditional upon the active CPU cluster. The upstream kernel currently doesn't support cluster switching, so this patch doesn't test the active CPU cluster ID.
[1] 3cc7942a4ae5 ARM: tegra: implement RAM repair
--- arch/arm/mach-tegra/sleep-tegra30.S | 11 +++++++++++ 1 file changed, 11 insertions(+)
arch/arm/mach-tegra/sleep-tegra30.S +11 -0
@@ -382,6 +382,14 @@ _pll_m_c_x_done: pll_locked r1, r0, CLK_RESET_PLLC_BASE pll_locked r1, r0, CLK_RESET_PLLX_BASE + tegra_get_soc_id TEGRA_APB_MISC_BASE, r1 + cmp r1, #TEGRA30 + beq 1f + ldr r1, [r0, #CLK_RESET_PLLP_BASE] + bic r1, r1, #(1<<31) @ disable PllP bypass + str r1, [r0, #CLK_RESET_PLLP_BASE] +1: + mov32 r7, TEGRA_TMRUS_BASE ldr r1, [r7] add r1, r1, #LOCK_DELAY
@@ -641,7 +649,10 @@ tegra30_switch_cpu_to_clk32k: str r0, [r4, #PMC_PLLP_WB0_OVERRIDE] /* disable PLLP, PLLA, PLLC and PLLX */ + tegra_get_soc_id TEGRA_APB_MISC_BASE, r1 + cmp r1, #TEGRA30 ldr r0, [r5, #CLK_RESET_PLLP_BASE] + orrne r0, r0, #(1 << 31) @ enable PllP bypass on fast cluster bic r0, r0, #(1 << 30) str r0, [r5, #CLK_RESET_PLLP_BASE] ldr r0, [r5, #CLK_RESET_PLLA_BASE]
Greg Kroah-Hartman
maintainerPATCH 73/173 iwlwifi: dont throw error when trying to remove IGTK
In reply to Greg Kroah-HartmanFrom: Luca Coelho <[email protected]>
commit 197288d5ba8a5289f22d3aeb4fca3824bfd9b4af upstream.
The IGTK keys are only removed by mac80211 after it has already removed the AP station. This causes the driver to throw an error because mac80211 is trying to remove the IGTK when the station doesn't exist anymore.
The firmware is aware that the station has been removed and can deal with it the next time we try to add an IGTK for a station, so we shouldn't try to remove the key if the station ID is IWL_MVM_INVALID_STA. Do this by removing the check for mvm_sta before calling iwl_mvm_send_sta_igtk() and check return from that function gracefully if the station ID is invalid.
--- drivers/net/wireless/intel/iwlwifi/mvm/sta.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-)
drivers/net/wireless/intel/iwlwifi/mvm/sta.c +7 -3
@@ -2981,6 +2981,10 @@ static int iwl_mvm_send_sta_igtk(struct igtk_cmd.sta_id =cpu_to_le32 (sta_id);if (remove_key) { +/* This is a valid situation for IGTK */ +if (sta_id ==IWL_MVM_INVALID_STA ) +return 0 ; + igtk_cmd.ctrl_flags |=cpu_to_le32 (STA_KEY_NOT_VALID ); }else {struct ieee80211_key_seq seq;
@@ -3285,9 +3289,9 @@ int iwl_mvm_remove_sta_key(struct iwl_mvIWL_DEBUG_WEP (mvm,"mvm remove dynamic key: idx=%d sta=%d\n" , keyconf->keyidx, sta_id); -if (mvm_sta && (keyconf->cipher ==WLAN_CIPHER_SUITE_AES_CMAC || - keyconf->cipher ==WLAN_CIPHER_SUITE_BIP_GMAC_128 || - keyconf->cipher ==WLAN_CIPHER_SUITE_BIP_GMAC_256 )) +if (keyconf->cipher ==WLAN_CIPHER_SUITE_AES_CMAC || + keyconf->cipher ==WLAN_CIPHER_SUITE_BIP_GMAC_128 || + keyconf->cipher ==WLAN_CIPHER_SUITE_BIP_GMAC_256 )return iwl_mvm_send_sta_igtk (mvm, keyconf, sta_id, true);if (!__test_and_clear_bit (keyconf->hw_key_idx, mvm->fw_key_table)) {
Greg Kroah-Hartman
maintainerPATCH 74/173 mwifiex: fix unbalanced locking in mwifiex_process_country_ie()
In reply to Greg Kroah-HartmanFrom: Brian Norris <[email protected]>
commit 65b1aae0d9d5962faccc06bdb8e91a2a0b09451c upstream.
We called rcu_read_lock(), so we need to call rcu_read_unlock() before we return.
--- drivers/net/wireless/marvell/mwifiex/sta_ioctl.c | 1 + 1 file changed, 1 insertion(+)
drivers/net/wireless/marvell/mwifiex/sta_ioctl.c +1 -0
@@ -274,6 +274,7 @@ static int mwifiex_process_country_ie(stif (country_ie_len > (IEEE80211_COUNTRY_STRING_LEN +MWIFIEX_MAX_TRIPLET_802_11D )) { +rcu_read_unlock ();mwifiex_dbg (priv->adapter,ERROR ,"11D: country_ie_len overflow!, deauth AP\n" );return -EINVAL ;
Greg Kroah-Hartman
maintainerPATCH 75/173 sunrpc: expiry_time should be seconds not timeval
In reply to Greg Kroah-HartmanFrom: Roberto Bergantinos Corpas <[email protected]>
commit 3d96208c30f84d6edf9ab4fac813306ac0d20c10 upstream.
When upcalling gssproxy, cache_head.expiry_time is set as a timeval, not seconds since boot. As such, RPC cache expiry logic will not clean expired objects created under auth.rpcsec.context cache.
This has proven to cause kernel memory leaks on field. Using 64 bit variants of getboottime/timespec
Expiration times have worked this way since 2010's c5b29f885afe "sunrpc:
use seconds since boot in expiry cache". The gssproxy code introduced in 2012 added gss_proxy_save_rsc and introduced the bug. That's a while for this to lurk, but it required a bit of an extreme case to make it obvious.
--- net/sunrpc/auth_gss/svcauth_gss.c | 4 ++++ 1 file changed, 4 insertions(+)
net/sunrpc/auth_gss/svcauth_gss.c +4 -0
@@ -1188,6 +1188,7 @@ static int gss_proxy_save_rsc(struct cacdprintk ("RPC: No creds found!\n" ); goto out; }else { +struct timespec64 boot;/* steal creds */ rsci.cred = ud->creds;
@@ -1208,6 +1209,9 @@ static int gss_proxy_save_rsc(struct cac &expiry, GFP_KERNEL);if (status) goto out; + +getboottime64 (&boot); + expiry -= boot.tv_sec; } rsci.h.expiry_time = expiry;
Greg Kroah-Hartman
maintainerPATCH 76/173 tools/kvm_stat: Fix kvm_exit filter name
In reply to Greg Kroah-HartmanFrom: Gavin Shan <[email protected]>
commit 5fcf3a55a62afb0760ccb6f391d62f20bce4a42f upstream.
The filter name is fixed to "exit_reason" for some kvm_exit events, no matter what architect we have. Actually, the filter name ("exit_reason") is only applicable to x86, meaning it's broken on other architects including aarch64.
This fixes the issue by providing various kvm_exit filter names, depending on architect we're on. Afterwards, the variable filter name is picked and applied through ioctl(fd, SET_FILTER).
--- tools/kvm/kvm_stat/kvm_stat | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
tools/kvm/kvm_stat/kvm_stat +6 -2
@@ -261,6 +261,7 @@ class ArchX86(Arch): def __init__(self, exit_reasons): self.sc_perf_evt_open = 298 self.ioctl_numbers = IOCTL_NUMBERS + self.exit_reason_field = 'exit_reason' self.exit_reasons = exit_reasons
@@ -276,6 +277,7 @@ class ArchPPC(Arch): # numbers depend on the wordsize. char_ptr_size = ctypes.sizeof(ctypes.c_char_p) self.ioctl_numbers['SET_FILTER'] = 0x80002406 | char_ptr_size << 16 + self.exit_reason_field = 'exit_nr' self.exit_reasons = {}
@@ -283,6 +285,7 @@ class ArchA64(Arch): def __init__(self): self.sc_perf_evt_open = 241 self.ioctl_numbers = IOCTL_NUMBERS + self.exit_reason_field = 'esr_ec' self.exit_reasons = AARCH64_EXIT_REASONS
@@ -290,6 +293,7 @@ class ArchS390(Arch): def __init__(self): self.sc_perf_evt_open = 331 self.ioctl_numbers = IOCTL_NUMBERS + self.exit_reason_field = None self.exit_reasons = None ARCH = Arch.get_arch()
@@ -513,8 +517,8 @@ class TracepointProvider(Provider): """ filters = {} filters['kvm_userspace_exit'] = ('reason', USERSPACE_EXIT_REASONS) - if ARCH.exit_reasons: - filters['kvm_exit'] = ('exit_reason', ARCH.exit_reasons) + if ARCH.exit_reason_field and ARCH.exit_reasons: + filters['kvm_exit'] = (ARCH.exit_reason_field, ARCH.exit_reasons) return filters def get_available_fields(self):
Greg Kroah-Hartman
maintainerPATCH 77/173 xen/balloon: Support xend-based toolstack take two
In reply to Greg Kroah-HartmanFrom: Juergen Gross <[email protected]>
commit eda4eabf86fd6806eaabc23fb90dd056fdac037b upstream.
Commit 3aa6c19d2f38be ("xen/balloon: Support xend-based toolstack") tried to fix a regression with running on rather ancient Xen versions. Unfortunately the fix was based on the assumption that xend would just use another Xenstore node, but in reality only some downstream versions of xend are doing that. The upstream xend does not write that Xenstore node at all, so the problem must be fixed in another way.
The easiest way to achieve that is to fall back to the behavior before commit 96edd61dcf4436 ("xen/balloon: don't online new memory initially") in case the static memory maximum can't be read.
This is achieved by setting static_max to the current number of memory pages known by the system resulting in target_diff becoming zero.
--- drivers/xen/xen-balloon.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/xen/xen-balloon.c +1 -1
@@ -82,7 +82,7 @@ static void watch_target(struct xenbus_w "%llu", &static_max) == 1)) static_max >>=PAGE_SHIFT -10 ; else - static_max = new_target; + static_max = balloon_stats.current_pages; target_diff = (xen_pv_domain () ||xen_initial_domain ()) ?0 : static_max - balloon_stats.target_pages;
Greg Kroah-Hartman
maintainerPATCH 78/173 KVM: x86: Refactor picdev_write() to prevent Spectre-v1/L1TF attacks
In reply to Greg Kroah-HartmanFrom: Marios Pomonis <[email protected]>
commit 14e32321f3606e4b0970200b6e5e47ee6f1e6410 upstream.
This fixes a Spectre-v1/L1TF vulnerability in picdev_write(). It replaces index computations based on the (attacked-controlled) port number with constants through a minor refactoring.
--- arch/x86/kvm/i8259.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
arch/x86/kvm/i8259.c +5 -1
@@ -460,10 +460,14 @@ static int picdev_write(struct kvm_pic *switch (addr) {case 0x20 :case 0x21 : +pic_lock (s); +pic_ioport_write (&s->pics[0 ], addr, data); +pic_unlock (s); +break ;case 0xa0 :case 0xa1 :pic_lock (s); -pic_ioport_write (&s->pics[addr >>7 ], addr, data); +pic_ioport_write (&s->pics[1 ], addr, data);pic_unlock (s);break ;case 0x4d0 :
Greg Kroah-Hartman
maintainerPATCH 79/173 KVM: x86: Refactor prefix decoding to prevent Spectre-v1/L1TF attacks
In reply to Greg Kroah-HartmanFrom: Marios Pomonis <[email protected]>
commit 125ffc5e0a56a3eded608dc51e09d5ebf72cf652 upstream.
This fixes Spectre-v1/L1TF vulnerabilities in vmx_read_guest_seg_selector(), vmx_read_guest_seg_base(), vmx_read_guest_seg_limit() and vmx_read_guest_seg_ar(). When invoked from emulation, these functions contain index computations based on the (attacker-influenced) segment value. Using constants prevents the attack.
--- arch/x86/kvm/emulate.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-)
arch/x86/kvm/emulate.c +14 -2
@@ -5094,16 +5094,28 @@ int x86_decode_insn(struct x86_emulate_c ctxt->ad_bytes = def_ad_bytes ^6 ;break ;case 0x26 :/* ES override */ + has_seg_override = true; + ctxt->seg_override =VCPU_SREG_ES ; +break ;case 0x2e :/* CS override */ + has_seg_override = true; + ctxt->seg_override =VCPU_SREG_CS ; +break ;case 0x36 :/* SS override */ + has_seg_override = true; + ctxt->seg_override =VCPU_SREG_SS ; +break ;case 0x3e :/* DS override */ has_seg_override = true; - ctxt->seg_override = (ctxt->b >>3 ) &3 ; + ctxt->seg_override =VCPU_SREG_DS ;break ;case 0x64 :/* FS override */ + has_seg_override = true; + ctxt->seg_override =VCPU_SREG_FS ; +break ;case 0x65 :/* GS override */ has_seg_override = true; - ctxt->seg_override = ctxt->b &7 ; + ctxt->seg_override =VCPU_SREG_GS ;break ; case 0x40 ... 0x4f: /* REX */if (mode !=X86EMUL_MODE_PROT64 )
Greg Kroah-Hartman
maintainerPATCH 80/173 KVM: x86: Protect DR-based index computations from Spectre-v1/L1TF attacks
In reply to Greg Kroah-HartmanFrom: Marios Pomonis <[email protected]>
commit ea740059ecb37807ba47b84b33d1447435a8d868 upstream.
This fixes a Spectre-v1/L1TF vulnerability in __kvm_set_dr() and kvm_get_dr().
Both kvm_get_dr() and kvm_set_dr() (a wrapper of __kvm_set_dr()) are exported symbols so KVM should tream them conservatively from a security perspective.
--- arch/x86/kvm/x86.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
arch/x86/kvm/x86.c +6 -2
@@ -924,9 +924,11 @@ static u64 kvm_dr6_fixed(struct kvm_vcpustatic int __kvm_set_dr (struct kvm_vcpu *vcpu,int dr,unsigned long val) { +size_t size =ARRAY_SIZE (vcpu->arch.db); +switch (dr) { case 0 ... 3: - vcpu->arch.db[dr] = val; + vcpu->arch.db[array_index_nospec (dr, size)] = val;if (!(vcpu->guest_debug &KVM_GUESTDBG_USE_HW_BP )) vcpu->arch.eff_db[dr] = val;break ;
@@ -963,9 +965,11 @@ EXPORT_SYMBOL_GPL(kvm_set_dr);int kvm_get_dr (struct kvm_vcpu *vcpu,int dr,unsigned long *val) { +size_t size =ARRAY_SIZE (vcpu->arch.db); +switch (dr) { case 0 ... 3: - *val = vcpu->arch.db[dr]; + *val = vcpu->arch.db[array_index_nospec (dr, size)];break ;case 4 :/* fall through */
Greg Kroah-Hartman
maintainerPATCH 81/173 KVM: x86: Protect kvm_lapic_reg_write() from Spectre-v1/L1TF attacks
In reply to Greg Kroah-HartmanFrom: Marios Pomonis <[email protected]>
commit 4bf79cb089f6b1c6c632492c0271054ce52ad766 upstream.
This fixes a Spectre-v1/L1TF vulnerability in kvm_lapic_reg_write(). This function contains index computations based on the (attacker-controlled) MSR number.
--- arch/x86/kvm/lapic.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-)
arch/x86/kvm/lapic.c +9 -4
@@ -1754,15 +1754,20 @@ int kvm_lapic_reg_write(struct kvm_lapiccase APIC_LVTTHMR :case APIC_LVTPC :case APIC_LVT1 : -case APIC_LVTERR : +case APIC_LVTERR : {/* TODO: Check vector */ +size_t size; +u32 index; +if (!kvm_apic_sw_enabled (apic)) val |=APIC_LVT_MASKED ; - - val &= apic_lvt_mask[(reg -APIC_LVTT ) >>4 ]; + size =ARRAY_SIZE (apic_lvt_mask); + index =array_index_nospec ( + (reg -APIC_LVTT ) >>4 , size); + val &= apic_lvt_mask[index];kvm_lapic_set_reg (apic, reg, val); -break ; + }case APIC_LVTT :if (!kvm_apic_sw_enabled (apic))
Greg Kroah-Hartman
maintainerPATCH 82/173 KVM: x86: Protect kvm_hv_msr_[get|set]_crash_data() from Spectre-v1/L1TF attacks
In reply to Greg Kroah-HartmanFrom: Marios Pomonis <[email protected]>
commit 8618793750071d66028584a83ed0b4fa7eb4f607 upstream.
This fixes Spectre-v1/L1TF vulnerabilities in kvm_hv_msr_get_crash_data() and kvm_hv_msr_set_crash_data().
These functions contain index computations that use the (attacker-controlled) MSR number.
--- arch/x86/kvm/hyperv.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-)
arch/x86/kvm/hyperv.c +6 -4
@@ -747,11 +747,12 @@ static int kvm_hv_msr_get_crash_data(str u32 index, u64 *pdata) { struct kvm_hv *hv = &vcpu->kvm->arch.hyperv; +size_t size =ARRAY_SIZE (hv->hv_crash_param); -if (WARN_ON_ONCE (index >=ARRAY_SIZE (hv->hv_crash_param))) +if (WARN_ON_ONCE (index >= size))return -EINVAL ; - *pdata = hv->hv_crash_param[index]; + *pdata = hv->hv_crash_param[array_index_nospec (index, size)];return 0 ; }
@@ -790,11 +791,12 @@ static int kvm_hv_msr_set_crash_data(str u32 index, u64 data) {struct kvm_hv *hv = &vcpu->kvm->arch.hyperv; +size_t size =ARRAY_SIZE (hv->hv_crash_param); -if (WARN_ON_ONCE (index >=ARRAY_SIZE (hv->hv_crash_param))) +if (WARN_ON_ONCE (index >= size))return -EINVAL ; - hv->hv_crash_param[index] = data; + hv->hv_crash_param[array_index_nospec (index, size)] = data;return 0 ; }
Greg Kroah-Hartman
maintainerPATCH 83/173 KVM: x86: Protect ioapic_write_indirect() from Spectre-v1/L1TF attacks
In reply to Greg Kroah-HartmanFrom: Marios Pomonis <[email protected]>
commit 670564559ca35b439c8d8861fc399451ddf95137 upstream.
This fixes a Spectre-v1/L1TF vulnerability in ioapic_write_indirect(). This function contains index computations based on the (attacker-controlled) IOREGSEL register.
This patch depends on patch
"KVM: x86: Protect ioapic_read_indirect() from Spectre-v1/L1TF attacks".
--- arch/x86/kvm/ioapic.c | 1 + 1 file changed, 1 insertion(+)
arch/x86/kvm/ioapic.c +1 -0
@@ -297,6 +297,7 @@ static void ioapic_write_indirect(structioapic_debug ("change redir index %x val %x\n" , index, val);if (index >=IOAPIC_NUM_PINS )return ; + index =array_index_nospec (index,IOAPIC_NUM_PINS ); e = &ioapic->redirtbl[index]; mask_before = e->fields.mask;/* Preserve read-only fields */
Greg Kroah-Hartman
maintainerPATCH 84/173 KVM: x86: Protect MSR-based index computations in pmu.h from Spectre-v1/L1TF attacks
In reply to Greg Kroah-HartmanFrom: Marios Pomonis <[email protected]>
commit 13c5183a4e643cc2b03a22d0e582c8e17bb7457d upstream.
This fixes a Spectre-v1/L1TF vulnerability in the get_gp_pmc() and get_fixed_pmc() functions.
They both contain index computations based on the (attacker-controlled) MSR number.
--- arch/x86/kvm/pmu.h | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-)
arch/x86/kvm/pmu.h +14 -4
@@ -2,6 +2,8 @@#ifndef __KVM_X86_PMU_H#define __KVM_X86_PMU_H +#include <linux/nospec.h> +#define vcpu_to_pmu (vcpu) (&(vcpu)->arch.pmu)#define pmu_to_vcpu (pmu) (container_of((pmu), struct kvm_vcpu, arch.pmu))#define pmc_to_pmu (pmc) (&(pmc)->vcpu->arch.pmu)
@@ -81,8 +83,12 @@ static inline bool pmc_is_enabled(structstatic inline struct kvm_pmc *get_gp_pmc (struct kvm_pmu *pmu,u32 msr,u32 base) { -if (msr >= base && msr < base + pmu->nr_arch_gp_counters) -return &pmu->gp_counters[msr - base]; +if (msr >= base && msr < base + pmu->nr_arch_gp_counters) { +u32 index =array_index_nospec (msr - base, + pmu->nr_arch_gp_counters); + +return &pmu->gp_counters[index]; + }return NULL ; }
@@ -92,8 +98,12 @@ static inline struct kvm_pmc *get_fixed_ {int base =MSR_CORE_PERF_FIXED_CTR0 ; -if (msr >= base && msr < base + pmu->nr_arch_fixed_counters) -return &pmu->fixed_counters[msr - base]; +if (msr >= base && msr < base + pmu->nr_arch_fixed_counters) { +u32 index =array_index_nospec (msr - base, + pmu->nr_arch_fixed_counters); + +return &pmu->fixed_counters[index]; + }return NULL ; }
Greg Kroah-Hartman
maintainerPATCH 85/173 KVM: x86: Protect ioapic_read_indirect() from Spectre-v1/L1TF attacks
In reply to Greg Kroah-HartmanFrom: Marios Pomonis <[email protected]>
commit 8c86405f606ca8508b8d9280680166ca26723695 upstream.
This fixes a Spectre-v1/L1TF vulnerability in ioapic_read_indirect(). This function contains index computations based on the (attacker-controlled) IOREGSEL register.
--- arch/x86/kvm/ioapic.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-)
arch/x86/kvm/ioapic.c +8 -6
@@ -36,6 +36,7 @@#include <linux/io.h> #include <linux/slab.h> #include <linux/export.h> +#include <linux/nospec.h> #include <asm/processor.h> #include <asm/page.h> #include <asm/current.h>
@@ -73,13 +74,14 @@ static unsigned long ioapic_read_indirecdefault : {u32 redir_index = (ioapic->ioregsel -0x10 ) >>1 ; -u64 redir_content; +u64 redir_content = ~0ULL ; -if (redir_index <IOAPIC_NUM_PINS ) - redir_content = - ioapic->redirtbl[redir_index].bits; -else - redir_content = ~0ULL ; +if (redir_index <IOAPIC_NUM_PINS ) { +u32 index =array_index_nospec ( + redir_index,IOAPIC_NUM_PINS ); + + redir_content = ioapic->redirtbl[index].bits; + } result = (ioapic->ioregsel &0x1 ) ? (redir_content >> 32) & 0xffffffff :
Greg Kroah-Hartman
maintainerPATCH 86/173 KVM: x86: Protect MSR-based index computations from Spectre-v1/L1TF attacks in x86.c
In reply to Greg Kroah-HartmanFrom: Marios Pomonis <[email protected]>
commit 6ec4c5eee1750d5d17951c4e1960d953376a0dda upstream.
This fixes a Spectre-v1/L1TF vulnerability in set_msr_mce() and get_msr_mce().
Both functions contain index computations based on the (attacker-controlled) MSR number.
--- arch/x86/kvm/x86.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)
arch/x86/kvm/x86.c +8 -2
@@ -2165,7 +2165,10 @@ static int set_msr_mce(struct kvm_vcpu *default :if (msr >=MSR_IA32_MC0_CTL && msr <MSR_IA32_MCx_CTL (bank_num)) { -u32 offset = msr -MSR_IA32_MC0_CTL ; +u32 offset =array_index_nospec ( + msr -MSR_IA32_MC0_CTL , +MSR_IA32_MCx_CTL (bank_num) -MSR_IA32_MC0_CTL ); +/* only 0 or all 1s can be written to IA32_MCi_CTL * some Linux kernels though clear bit 10 in bank 4 to * workaround a BIOS/GART TBL issue on AMD K8s, ignore
@@ -2549,7 +2552,10 @@ static int get_msr_mce(struct kvm_vcpu *default :if (msr >=MSR_IA32_MC0_CTL && msr <MSR_IA32_MCx_CTL (bank_num)) { -u32 offset = msr -MSR_IA32_MC0_CTL ; +u32 offset =array_index_nospec ( + msr -MSR_IA32_MC0_CTL , +MSR_IA32_MCx_CTL (bank_num) -MSR_IA32_MC0_CTL ); + data = vcpu->arch.mce_banks[offset];break ; }
Greg Kroah-Hartman
maintainerPATCH 87/173 KVM: x86: Protect x86_decode_insn from Spectre-v1/L1TF attacks
In reply to Greg Kroah-HartmanFrom: Marios Pomonis <[email protected]>
commit 3c9053a2cae7ba2ba73766a34cea41baa70f57f7 upstream.
This fixes a Spectre-v1/L1TF vulnerability in x86_decode_insn(). kvm_emulate_instruction() (an ancestor of x86_decode_insn()) is an exported symbol, so KVM should treat it conservatively from a security perspective.
--- arch/x86/kvm/emulate.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-)
arch/x86/kvm/emulate.c +8 -3
@@ -5199,10 +5199,15 @@ done_prefixes: }break ;case Escape: -if (ctxt->modrm >0xbf ) - opcode = opcode.u.esc->high[ctxt->modrm -0xc0 ]; -else +if (ctxt->modrm >0xbf ) { +size_t size =ARRAY_SIZE (opcode.u.esc->high); +u32 index =array_index_nospec ( + ctxt->modrm -0xc0 , size); + + opcode = opcode.u.esc->high[index]; + }else { opcode = opcode.u.esc->op[(ctxt->modrm >>3 ) &7 ]; + }break ;case InstrDual:if ((ctxt->modrm >>6 ) ==3 )
Greg Kroah-Hartman
maintainerPATCH 88/173 KVM: x86: Protect MSR-based index computations in fixed_msr_to_seg_unit() from Spectre-v1/L1TF attacks
In reply to Greg Kroah-HartmanFrom: Marios Pomonis <[email protected]>
commit 25a5edea71b7c154b6a0b8cec14c711cafa31d26 upstream.
This fixes a Spectre-v1/L1TF vulnerability in fixed_msr_to_seg_unit(). This function contains index computations based on the (attacker-controlled) MSR number.
--- arch/x86/kvm/mtrr.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
arch/x86/kvm/mtrr.c +6 -2
@@ -202,11 +202,15 @@ static bool fixed_msr_to_seg_unit(u32 msbreak ; case MSR_MTRRfix16K_80000 ... MSR_MTRRfix16K_A0000: *seg =1 ; - *unit = msr - MSR_MTRRfix16K_80000; + *unit =array_index_nospec ( + msr - MSR_MTRRfix16K_80000, + MSR_MTRRfix16K_A0000 - MSR_MTRRfix16K_80000 +1 );break ; case MSR_MTRRfix4K_C0000 ... MSR_MTRRfix4K_F8000: *seg =2 ; - *unit = msr - MSR_MTRRfix4K_C0000; + *unit =array_index_nospec ( + msr - MSR_MTRRfix4K_C0000, + MSR_MTRRfix4K_F8000 - MSR_MTRRfix4K_C0000 +1 );break ;default :return false;
Greg Kroah-Hartman
maintainerPATCH 89/173 KVM: PPC: Book3S HV: Uninit vCPU if vcore creation fails
In reply to Greg Kroah-HartmanFrom: Sean Christopherson <[email protected]>
commit 1a978d9d3e72ddfa40ac60d26301b154247ee0bc upstream.
Call kvm_vcpu_uninit() if vcore creation fails to avoid leaking any resources allocated by kvm_vcpu_init(), i.e. the vcpu->run page.
--- arch/powerpc/kvm/book3s_hv.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
arch/powerpc/kvm/book3s_hv.c +3 -1
@@ -1997,7 +1997,7 @@ static struct kvm_vcpu *kvmppc_core_vcpumutex_unlock (&kvm->lock);if (!vcore) - goto free_vcpu; + goto uninit_vcpu;spin_lock (&vcore->lock); ++vcore->num_threads;
@@ -2014,6 +2014,8 @@ static struct kvm_vcpu *kvmppc_core_vcpureturn vcpu; +uninit_vcpu: +kvm_vcpu_uninit (vcpu); free_vcpu:kmem_cache_free (kvm_vcpu_cache, vcpu); out:
Greg Kroah-Hartman
maintainerPATCH 90/173 KVM: PPC: Book3S PR: Free shared page if mmu initialization fails
In reply to Greg Kroah-HartmanFrom: Sean Christopherson <[email protected]>
commit cb10bf9194f4d2c5d830eddca861f7ca0fecdbb4 upstream.
Explicitly free the shared page if kvmppc_mmu_init() fails during kvmppc_core_vcpu_create(), as the page is freed only in kvmppc_core_vcpu_free(), which is not reached via kvm_vcpu_uninit().
--- arch/powerpc/kvm/book3s_pr.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
arch/powerpc/kvm/book3s_pr.c +3 -1
@@ -1482,10 +1482,12 @@ static struct kvm_vcpu *kvmppc_core_vcpu err =kvmppc_mmu_init (vcpu);if (err <0 ) - goto uninit_vcpu; + goto free_shared_page;return vcpu; +free_shared_page: +free_page ((unsigned long )vcpu->arch.shared); uninit_vcpu:kvm_vcpu_uninit (vcpu); free_shadow_vcpu:
Greg Kroah-Hartman
maintainerPATCH 91/173 KVM: x86: Free wbinvd_dirty_mask if vCPU creation fails
In reply to Greg Kroah-HartmanFrom: Sean Christopherson <[email protected]>
commit 16be9ddea268ad841457a59109963fff8c9de38d upstream.
Free the vCPU's wbinvd_dirty_mask if vCPU creation fails after kvm_arch_vcpu_init(), e.g. when installing the vCPU's file descriptor. Do the freeing by calling kvm_arch_vcpu_free() instead of open coding the freeing. This adds a likely superfluous, but ultimately harmless, call to kvmclock_reset(), which only clears vcpu->arch.pv_time_enabled. Using kvm_arch_vcpu_free() allows for additional cleanup in the future.
--- arch/x86/kvm/x86.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kvm/x86.c +1 -1
@@ -8063,7 +8063,7 @@ void kvm_arch_vcpu_destroy(struct kvm_vckvm_mmu_unload (vcpu);vcpu_put (vcpu); - kvm_x86_ops->vcpu_free (vcpu); +kvm_arch_vcpu_free (vcpu); }void kvm_vcpu_reset (struct kvm_vcpu *vcpu,bool init_event)
Greg Kroah-Hartman
maintainerPATCH 92/173 clk: tegra: Mark fuse clock as critical
In reply to Greg Kroah-HartmanFrom: Stephen Warren <[email protected]>
commit bf83b96f87ae2abb1e535306ea53608e8de5dfbb upstream.
For a little over a year, U-Boot on Tegra124 has configured the flow controller to perform automatic RAM re-repair on off->on power transitions of the CPU rail[1]. This is mandatory for correct operation of Tegra124. However, RAM re-repair relies on certain clocks, which the kernel must enable and leave running. The fuse clock is one of those clocks. Mark this clock as critical so that LP1 power mode (system suspend) operates correctly.
[1] 3cc7942a4ae5 ARM: tegra: implement RAM repair
--- drivers/clk/tegra/clk-tegra-periph.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/clk/tegra/clk-tegra-periph.c +5 -1
@@ -825,7 +825,11 @@ static struct tegra_periph_init_data gatGATE ("vcp" ,"clk_m" ,29 ,0 , tegra_clk_vcp,0 ),GATE ("apbdma" ,"clk_m" ,34 ,0 , tegra_clk_apbdma,0 ),GATE ("kbc" ,"clk_32k" ,36 ,TEGRA_PERIPH_ON_APB |TEGRA_PERIPH_NO_RESET , tegra_clk_kbc,0 ), -GATE ("fuse" ,"clk_m" ,39 ,TEGRA_PERIPH_ON_APB , tegra_clk_fuse,0 ), +/* +* Critical for RAM re-repair operation, which must occur on resume +* from LP1 system suspend and as part of CCPLEX cluster switching. +*/ +GATE ("fuse" ,"clk_m" ,39 ,TEGRA_PERIPH_ON_APB , tegra_clk_fuse,CLK_IS_CRITICAL ),GATE ("fuse_burn" ,"clk_m" ,39 ,TEGRA_PERIPH_ON_APB , tegra_clk_fuse_burn,0 ),GATE ("kfuse" ,"clk_m" ,40 ,TEGRA_PERIPH_ON_APB , tegra_clk_kfuse,0 ), GATE("apbif", "clk_m", 107, TEGRA_PERIPH_ON_APB, tegra_clk_apbif, 0),
Greg Kroah-Hartman
maintainerPATCH 93/173 scsi: qla2xxx: Fix the endianness of the qla82xx_get_fw_size() return type
In reply to Greg Kroah-HartmanFrom: Bart Van Assche <[email protected]>
commit 3f5f7335e5e234e340b48ecb24c2aba98a61f934 upstream.
Since qla82xx_get_fw_size() returns a number in CPU-endian format, change its return type from __le32 into u32. This patch does not change any functionality.
--- drivers/scsi/qla2xxx/qla_nx.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-)
drivers/scsi/qla2xxx/qla_nx.c +3 -4
@@ -1605,8 +1605,7 @@ qla82xx_get_bootld_offset(struct qla_hw_return (u8 *)&ha->hablob->fw->data[offset]; } -static __le32 -qla82xx_get_fw_size (struct qla_hw_data *ha) +static u32 qla82xx_get_fw_size (struct qla_hw_data *ha) {struct qla82xx_uri_data_desc *uri_desc =NULL ;
@@ -1617,7 +1616,7 @@ qla82xx_get_fw_size(struct qla_hw_data *return cpu_to_le32 (uri_desc->size); } -return cpu_to_le32 (*(u32 *)&ha->hablob->fw->data[FW_SIZE_OFFSET ]); +return get_unaligned_le32 (&ha->hablob->fw->data[FW_SIZE_OFFSET ]); } static u8 *
@@ -1808,7 +1807,7 @@ qla82xx_fw_load_from_blob(struct qla_hw_ } flashaddr =FLASH_ADDR_START ; - size = (__force u32 )qla82xx_get_fw_size (ha) /8 ; + size =qla82xx_get_fw_size (ha) /8 ; ptr64 = (u64 *)qla82xx_get_fw_offs (ha);for (i =0 ; i < size; i++) {
Greg Kroah-Hartman
maintainerPATCH 94/173 scsi: csiostor: Adjust indentation in csio_device_reset
In reply to Greg Kroah-HartmanFrom: Nathan Chancellor <[email protected]>
commit a808a04c861782e31fc30e342a619c144aaee14a upstream.
Clang warns:
../drivers/scsi/csiostor/csio_scsi.c:1386:3: warning: misleading indentation; statement is not part of the previous 'if' [-Wmisleading-indentation]
csio_lnodes_exit(hw, 1);
^
../drivers/scsi/csiostor/csio_scsi.c:1382:2: note: previous statement is
here
if (*buf != '1')
^1 warning generated.
This warning occurs because there is a space after the tab on this line. Remove it so that the indentation is consistent with the Linux kernel coding style and clang no longer warns.
--- drivers/scsi/csiostor/csio_scsi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/scsi/csiostor/csio_scsi.c +1 -1
@@ -1383,7 +1383,7 @@ csio_device_reset(struct device *dev,return -EINVAL ;/* Delete NPIV lnodes */ -csio_lnodes_exit (hw,1 ); +csio_lnodes_exit (hw,1 );/* Block upper IOs */ csio_lnodes_block_request (hw);
Greg Kroah-Hartman
maintainerPATCH 95/173 scsi: qla4xxx: Adjust indentation in qla4xxx_mem_free
In reply to Greg Kroah-HartmanFrom: Nathan Chancellor <[email protected]>
commit aa8679736a82386551eb9f3ea0e6ebe2c0e99104 upstream.
Clang warns:
../drivers/scsi/qla4xxx/ql4_os.c:4148:3: warning: misleading indentation; statement is not part of the previous 'if' [-Wmisleading-indentation]
if (ha->fw_dump)
^
../drivers/scsi/qla4xxx/ql4_os.c:4144:2: note: previous statement is
here
if (ha->queues)
^1 warning generated.
This warning occurs because there is a space after the tab on this line. Remove it so that the indentation is consistent with the Linux kernel coding style and clang no longer warns.
--- drivers/scsi/qla4xxx/ql4_os.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/scsi/qla4xxx/ql4_os.c +1 -1
@@ -4150,7 +4150,7 @@ static void qla4xxx_mem_free(struct scsidma_free_coherent (&ha->pdev->dev, ha->queues_len, ha->queues, ha->queues_dma); -if (ha->fw_dump) +if (ha->fw_dump)vfree (ha->fw_dump); ha->queues_len =0 ;
Greg Kroah-Hartman
maintainerPATCH 96/173 scsi: ufs: Recheck bkops level if bkops is disabled
In reply to Greg Kroah-HartmanFrom: Asutosh Das <[email protected]>
commit 24366c2afbb0539fb14eff330d4e3a5db5c0a3ef upstream.
bkops level should be rechecked upon receiving an exception. Currently the level is being cached and never updated.
Update bkops each time the level is checked. Also do not use the cached bkops level value if it is disabled and then enabled.
--- drivers/scsi/ufs/ufshcd.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/scsi/ufs/ufshcd.c +3 -0
@@ -4812,6 +4812,7 @@ static int ufshcd_disable_auto_bkops(str hba->auto_bkops_enabled = false;trace_ufshcd_auto_bkops_state (dev_name (hba->dev),"Disabled" ); + hba->is_urgent_bkops_lvl_checked = false; out:return err; }
@@ -4836,6 +4837,7 @@ static void ufshcd_force_reset_auto_bkop hba->ee_ctrl_mask &= ~MASK_EE_URGENT_BKOPS ;ufshcd_disable_auto_bkops (hba); } + hba->is_urgent_bkops_lvl_checked = false; }static inline int ufshcd_get_bkops_status (struct ufs_hba *hba,u32 *status)
@@ -4882,6 +4884,7 @@ static int ufshcd_bkops_ctrl(struct ufs_ err =ufshcd_enable_auto_bkops (hba); else err =ufshcd_disable_auto_bkops (hba); + hba->urgent_bkops_lvl = curr_status; out:return err; }
Greg Kroah-Hartman
maintainerPATCH 97/173 phy: qualcomm: Adjust indentation in read_poll_timeout
In reply to Greg Kroah-HartmanFrom: Nathan Chancellor <[email protected]>
commit a89806c998ee123bb9c0f18526e55afd12c0c0ab upstream.
Clang warns:
../drivers/phy/qualcomm/phy-qcom-apq8064-sata.c:83:4: warning:
misleading indentation; statement is not part of the previous 'if' [-Wmisleading-indentation]
usleep_range(DELAY_INTERVAL_US, DELAY_INTERVAL_US + 50);
^../drivers/phy/qualcomm/phy-qcom-apq8064-sata.c:80:3: note: previous statement is here
if (readl_relaxed(addr) & mask)
^
1 warning generated.
This warning occurs because there is a space after the tab on this line. Remove it so that the indentation is consistent with the Linux kernel coding style and clang no longer warns.
--- drivers/phy/qualcomm/phy-qcom-apq8064-sata.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/phy/qualcomm/phy-qcom-apq8064-sata.c +1 -1
@@ -88,7 +88,7 @@ static int read_poll_timeout(void __iomeif (readl_relaxed (addr) & mask)return 0 ; -usleep_range (DELAY_INTERVAL_US ,DELAY_INTERVAL_US +50 ); +usleep_range (DELAY_INTERVAL_US ,DELAY_INTERVAL_US +50 ); }while (!time_after (jiffies, timeout));return (readl_relaxed (addr) & mask) ?0 : -ETIMEDOUT ;
Greg Kroah-Hartman
maintainerPATCH 98/173 ext2: Adjust indentation in ext2_fill_super
In reply to Greg Kroah-HartmanFrom: Nathan Chancellor <[email protected]>
commit d9e9866803f7b6c3fdd35d345e97fb0b2908bbbc upstream.
Clang warns:
../fs/ext2/super.c:1076:3: warning: misleading indentation; statement is not part of the previous 'if' [-Wmisleading-indentation]
sbi->s_groups_count = ((le32_to_cpu(es->s_blocks_count) -
^
../fs/ext2/super.c:1074:2: note: previous statement is here
if (EXT2_BLOCKS_PER_GROUP(sb) == 0)
^1 warning generated.
This warning occurs because there is a space before the tab on this line. Remove it so that the indentation is consistent with the Linux kernel coding style and clang no longer warns.
--- fs/ext2/super.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
fs/ext2/super.c +3 -3
@@ -1077,9 +1077,9 @@ static int ext2_fill_super(struct super_if (EXT2_BLOCKS_PER_GROUP (sb) ==0 ) goto cantfind_ext2; - sbi->s_groups_count = ((le32_to_cpu (es->s_blocks_count) - -le32_to_cpu (es->s_first_data_block) -1 ) - /EXT2_BLOCKS_PER_GROUP (sb)) +1 ; + sbi->s_groups_count = ((le32_to_cpu (es->s_blocks_count) - +le32_to_cpu (es->s_first_data_block) -1 ) + /EXT2_BLOCKS_PER_GROUP (sb)) +1 ; db_count = (sbi->s_groups_count +EXT2_DESC_PER_BLOCK (sb) -1 ) /EXT2_DESC_PER_BLOCK (sb); sbi->s_group_desc =kmalloc (db_count *sizeof (struct buffer_head *),GFP_KERNEL );
Greg Kroah-Hartman
maintainerPATCH 99/173 powerpc/44x: Adjust indentation in ibm4xx_denali_fixup_memsize
In reply to Greg Kroah-HartmanFrom: Nathan Chancellor <[email protected]>
commit c3aae14e5d468d18dbb5d7c0c8c7e2968cc14aad upstream.
Clang warns:
../arch/powerpc/boot/4xx.c:231:3: warning: misleading indentation; statement is not part of the previous 'else' [-Wmisleading-indentation]
val = SDRAM0_READ(DDR0_42);
^
../arch/powerpc/boot/4xx.c:227:2: note: previous statement is here
else
^This is because there is a space at the beginning of this line; remove it so that the indentation is consistent according to the Linux kernel coding style and clang no longer warns.
--- arch/powerpc/boot/4xx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/powerpc/boot/4xx.c +1 -1
@@ -232,7 +232,7 @@ void ibm4xx_denali_fixup_memsize(void) dpath =8 ;/* 64 bits */ /* get address pins (rows) */ - val =SDRAM0_READ (DDR0_42 ); + val =SDRAM0_READ (DDR0_42 ); row =DDR_GET_VAL (val,DDR_APIN ,DDR_APIN_SHIFT );if (row > max_row)
Greg Kroah-Hartman
maintainerPATCH 100/173 NFC: pn544: Adjust indentation in pn544_hci_check_presence
In reply to Greg Kroah-HartmanFrom: Nathan Chancellor <[email protected]>
commit 5080832627b65e3772a35d1dced68c64e2b24442 upstream.
Clang warns
../drivers/nfc/pn544/pn544.c:696:4: warning: misleading indentation; statement is not part of the previous 'if' [-Wmisleading-indentation]
return nfc_hci_send_cmd(hdev, NFC_HCI_RF_READER_A_GATE, ^../drivers/nfc/pn544/pn544.c:692:3: note: previous statement is here
if (target->nfcid1_len != 4 && target->nfcid1_len != 7 &&
^1 warning generated.
This warning occurs because there is a space after the tab on this line. Remove it so that the indentation is consistent with the Linux kernel coding style and clang no longer warns.
--- drivers/nfc/pn544/pn544.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/nfc/pn544/pn544.c +1 -1
@@ -704,7 +704,7 @@ static int pn544_hci_check_presence(stru target->nfcid1_len != 10)return -EOPNOTSUPP ; -return nfc_hci_send_cmd (hdev,NFC_HCI_RF_READER_A_GATE , +return nfc_hci_send_cmd (hdev,NFC_HCI_RF_READER_A_GATE ,PN544_RF_READER_CMD_ACTIVATE_NEXT , target->nfcid1, target->nfcid1_len,NULL ); } else if (target->supported_protocols & (NFC_PROTO_JEWEL_MASK |
Greg Kroah-Hartman
maintainerPATCH 101/173 ppp: Adjust indentation into ppp_async_input
In reply to Greg Kroah-HartmanFrom: Nathan Chancellor <[email protected]>
commit 08cbc75f96029d3092664213a844a5e25523aa35 upstream.
Clang warns:
../drivers/net/ppp/ppp_async.c:877:6: warning: misleading indentation; statement is not part of the previous 'if' [-Wmisleading-indentation]
ap->rpkt = skb;
^
../drivers/net/ppp/ppp_async.c:875:5: note: previous statement is here
if (!skb)
^1 warning generated.
This warning occurs because there is a space before the tab on this line. Clean up this entire block's indentation so that it is consistent with the Linux kernel coding style and clang no longer warns.
--- drivers/net/ppp/ppp_async.c | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-)
drivers/net/ppp/ppp_async.c +9 -9
@@ -878,15 +878,15 @@ ppp_async_input(struct asyncppp *ap, con skb =dev_alloc_skb (ap->mru +PPP_HDRLEN +2 );if (!skb) goto nomem; - ap->rpkt = skb; - } -if (skb->len ==0 ) { -/* Try to get the payload 4-byte aligned. -* This should match the -* PPP_ALLSTATIONS/PPP_UI/compressed tests in -* process_input_packet, but we do not have -* enough chars here to test buf[1] and buf[2]. -*/ + ap->rpkt = skb; + } +if (skb->len ==0 ) { +/* Try to get the payload 4-byte aligned. +* This should match the +* PPP_ALLSTATIONS/PPP_UI/compressed tests in +* process_input_packet, but we do not have +* enough chars here to test buf[1] and buf[2]. +*/ if (buf[0 ] !=PPP_ALLSTATIONS )skb_reserve (skb,2 + (buf[0 ] &1 )); }
Greg Kroah-Hartman
maintainerPATCH 102/173 net: smc911x: Adjust indentation in smc911x_phy_configure
In reply to Greg Kroah-HartmanFrom: Nathan Chancellor <[email protected]>
commit 5c61e223004b3b5c3f1dd25718e979bc17a3b12d upstream.
Clang warns:
../drivers/net/ethernet/smsc/smc911x.c:939:3: warning: misleading indentation; statement is not part of the previous 'if' [-Wmisleading-indentation]
if (!lp->ctl_rfduplx)
^
../drivers/net/ethernet/smsc/smc911x.c:936:2: note: previous statement
is here
if (lp->ctl_rspeed != 100)
^1 warning generated.
This warning occurs because there is a space after the tab on this line. Remove it so that the indentation is consistent with the Linux kernel coding style and clang no longer warns.
--- drivers/net/ethernet/smsc/smc911x.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/smsc/smc911x.c +1 -1
@@ -948,7 +948,7 @@ static void smc911x_phy_configure(structif (lp->ctl_rspeed !=100 ) my_ad_caps &= ~(ADVERTISE_100BASE4 |ADVERTISE_100FULL |ADVERTISE_100HALF ); -if (!lp->ctl_rfduplx) +if (!lp->ctl_rfduplx) my_ad_caps &= ~(ADVERTISE_100FULL |ADVERTISE_10FULL );/* Update our Auto-Neg Advertisement Register */
Greg Kroah-Hartman
maintainerPATCH 103/173 net: tulip: Adjust indentation in {dmfe, uli526x}_init_module
In reply to Greg Kroah-HartmanFrom: Nathan Chancellor <[email protected]>
commit fe06bf3d83ef0d92f35a24e03297172e92ce9ce3 upstream.
Clang warns:
../drivers/net/ethernet/dec/tulip/uli526x.c:1812:3: warning: misleading indentation; statement is not part of the previous 'if' [-Wmisleading-indentation]
switch (mode) {
^
../drivers/net/ethernet/dec/tulip/uli526x.c:1809:2: note: previous
statement is here
if (cr6set)
^1 warning generated.
../drivers/net/ethernet/dec/tulip/dmfe.c:2217:3: warning: misleading indentation; statement is not part of the previous 'if' [-Wmisleading-indentation]
switch(mode) {
^
../drivers/net/ethernet/dec/tulip/dmfe.c:2214:2: note: previous
statement is here
if (cr6set)
^1 warning generated.
This warning occurs because there is a space before the tab on these lines. Remove them so that the indentation is consistent with the Linux kernel coding style and clang no longer warns.
While we are here, adjust the default block in dmfe_init_module to have a proper break between the label and assignment and add a space between the switch and opening parentheses to avoid a checkpatch warning.
--- drivers/net/ethernet/dec/tulip/dmfe.c | 7 ++++--- drivers/net/ethernet/dec/tulip/uli526x.c | 4 ++-- 2 files changed, 6 insertions(+), 5 deletions(-)
drivers/net/ethernet/dec/tulip/dmfe.c +4 -3
@@ -2224,15 +2224,16 @@ static int __init dmfe_init_module(void)if (cr6set) dmfe_cr6_user_set = cr6set; -switch (mode) { -case DMFE_10MHF : +switch (mode) { +case DMFE_10MHF :case DMFE_100MHF :case DMFE_10MFD :case DMFE_100MFD :case DMFE_1M_HPNA : dmfe_media_mode = mode;break ; -default :dmfe_media_mode =DMFE_AUTO ; +default : + dmfe_media_mode =DMFE_AUTO ;break ; }
drivers/net/ethernet/dec/tulip/uli526x.c +2 -2
@@ -1819,8 +1819,8 @@ static int __init uli526x_init_module(voif (cr6set) uli526x_cr6_user_set = cr6set; -switch (mode) { -case ULI526X_10MHF : +switch (mode) { +case ULI526X_10MHF :case ULI526X_100MHF :case ULI526X_10MFD :case ULI526X_100MFD :
Greg Kroah-Hartman
maintainerPATCH 104/173 IB/mlx5: Fix outstanding_pi index for GSI qps
In reply to Greg Kroah-HartmanFrom: Prabhath Sajeepa <[email protected]>
commit b5671afe5e39ed71e94eae788bacdcceec69db09 upstream.
Commit b0ffeb537f3a ("IB/mlx5: Fix iteration overrun in GSI qps") changed the way outstanding WRs are tracked for the GSI QP. But the fix did not cover the case when a call to ib_post_send() fails and updates index to track outstanding.
Since the prior commmit outstanding_pi should not be bounded otherwise the loop generate_completions() will fail.
--- drivers/infiniband/hw/mlx5/gsi.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/infiniband/hw/mlx5/gsi.c +1 -2
@@ -507,8 +507,7 @@ int mlx5_ib_gsi_post_send(struct ib_qp * ret =ib_post_send (tx_qp, &cur_wr.wr, bad_wr);if (ret) {/* Undo the effect of adding the outstanding wr */ - gsi->outstanding_pi = (gsi->outstanding_pi -1 ) % - gsi->cap.max_send_wr; + gsi->outstanding_pi--; goto err; }spin_unlock_irqrestore (&gsi->lock, flags);
Greg Kroah-Hartman
maintainerPATCH 105/173 IB/core: Fix ODP get user pages flow
In reply to Greg Kroah-HartmanFrom: Yishai Hadas <[email protected]>
commit d07de8bd1709a80a282963ad7b2535148678a9e4 upstream.
The nr_pages argument of get_user_pages_remote() should always be in terms of the system page size, not the MR page size. Use PAGE_SIZE instead of umem_odp->page_shift.
--- drivers/infiniband/core/umem_odp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/infiniband/core/umem_odp.c +1 -1
@@ -637,7 +637,7 @@ int ib_umem_odp_map_dma_pages(struct ib_while (bcnt >0 ) {const size_t gup_num_pages =min_t (size_t, - (bcnt +BIT (page_shift) -1 ) >> page_shift, +ALIGN (bcnt,PAGE_SIZE ) /PAGE_SIZE ,PAGE_SIZE /sizeof (struct page *));down_read (&owning_mm->mmap_sem);
Greg Kroah-Hartman
maintainerPATCH 106/173 nfsd: fix delay timer on 32-bit architectures
In reply to Greg Kroah-HartmanFrom: Arnd Bergmann <[email protected]>
commit 2561c92b12f4f4e386d453556685f75775c0938b upstream.
The nfsd4_cb_layout_done() function takes a 'time_t' value, multiplied by NSEC_PER_SEC*2 to get a nanosecond value.
This works fine on 64-bit architectures, but on 32-bit, any value over 1 second results in a signed integer overflow with unexpected results.
Cast one input to a 64-bit type in order to produce the same result that we have on 64-bit architectures, regarless of the type of nfsd4_lease.
--- fs/nfsd/nfs4layouts.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
fs/nfsd/nfs4layouts.c +1 -1
@@ -683,7 +683,7 @@ nfsd4_cb_layout_done(struct nfsd4_callba/* Client gets 2 lease periods to return it */ cutoff =ktime_add_ns (task->tk_start, - nn->nfsd4_lease *NSEC_PER_SEC *2 ); + (u64 )nn->nfsd4_lease *NSEC_PER_SEC *2 );if (ktime_before (now, cutoff)) {rpc_delay (task,HZ /100 );/* 10 mili-seconds */
Greg Kroah-Hartman
maintainerPATCH 107/173 nfsd: fix jiffies/time_t mixup in LRU list
In reply to Greg Kroah-HartmanFrom: Arnd Bergmann <[email protected]>
commit 9594497f2c78993cb66b696122f7c65528ace985 upstream.
The nfsd4_blocked_lock->nbl_time timestamp is recorded in jiffies, but then compared to a CLOCK_REALTIME timestamp later on, which makes no sense.
For consistency with the other timestamps, change this to use a time_t.
This is a change in behavior, which may cause regressions, but the current code is not sensible. On a system with CONFIG_HZ=1000, the 'time_after((unsigned long)nbl->nbl_time, (unsigned long)cutoff))' check is false for roughly the first 18 days of uptime and then true for the next 49 days.
--- fs/nfsd/nfs4state.c | 2 +- fs/nfsd/state.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)
fs/nfsd/nfs4state.c +1 -1
@@ -6040,7 +6040,7 @@ nfsd4_lock(struct svc_rqst *rqstp, struc }if (fl_flags &FL_SLEEP ) { - nbl->nbl_time = jiffies; + nbl->nbl_time =get_seconds ();spin_lock (&nn->blocked_locks_lock);list_add_tail (&nbl->nbl_list, &lock_sop->lo_blocked);list_add_tail (&nbl->nbl_lru, &nn->blocked_locks_lru);
fs/nfsd/state.h +1 -1
@@ -591,7 +591,7 @@ static inline bool nfsd4_stateid_generatstruct nfsd4_blocked_lock {struct list_head nbl_list;struct list_head nbl_lru; -unsigned long nbl_time; +time_t nbl_time;struct file_lock nbl_lock;struct knfsd_fh nbl_fh;struct nfsd4_callback nbl_cb;
Greg Kroah-Hartman
maintainerPATCH 108/173 ubi: fastmap: Fix inverted logic in seen selfcheck
In reply to Greg Kroah-HartmanFrom: Sascha Hauer <[email protected]>
commit ef5aafb6e4e9942a28cd300bdcda21ce6cbaf045 upstream.
set_seen() sets the bit corresponding to the PEB number in the bitmap, so when self_check_seen() wants to find PEBs that haven't been seen we have to print the PEBs that have their bit cleared, not the ones which have it set.
--- drivers/mtd/ubi/fastmap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/mtd/ubi/fastmap.c +1 -1
@@ -73,7 +73,7 @@ static int self_check_seen(struct ubi_dereturn 0 ;for (pnum =0 ; pnum < ubi->peb_count; pnum++) { -if (test_bit (pnum, seen) && ubi->lookuptbl[pnum]) { +if (!test_bit (pnum, seen) && ubi->lookuptbl[pnum]) {ubi_err (ubi,"self-check failed for PEB %d, fastmap didn't see it" , pnum); ret = -EINVAL ; }
Greg Kroah-Hartman
maintainerPATCH 109/173 ubi: Fix an error pointer dereference in error handling code
In reply to Greg Kroah-HartmanFrom: Dan Carpenter <[email protected]>
commit 5d3805af279c93ef49a64701f35254676d709622 upstream.
If "seen_pebs = init_seen(ubi);" fails then "seen_pebs" is an error pointer and we try to kfree() it which results in an Oops.
This patch re-arranges the error handling so now it only frees things which have been allocated successfully.
--- drivers/mtd/ubi/fastmap.c | 21 ++++++++++++--------- 1 file changed, 12 insertions(+), 9 deletions(-)
drivers/mtd/ubi/fastmap.c +12 -9
@@ -1147,7 +1147,7 @@ static int ubi_write_fastmap(struct ubi_struct rb_node *tmp_rb;int ret, i, j, free_peb_count, used_peb_count, vol_count;int scrub_peb_count, erase_peb_count; -unsigned long *seen_pebs =NULL ; +unsigned long *seen_pebs; fm_raw = ubi->fm_buf;memset (ubi->fm_buf,0 , ubi->fm_size);
@@ -1161,7 +1161,7 @@ static int ubi_write_fastmap(struct ubi_ dvbuf =new_fm_vbuf (ubi,UBI_FM_DATA_VOLUME_ID );if (!dvbuf) { ret = -ENOMEM ; - goto out_kfree; + goto out_free_avbuf; } avhdr =ubi_get_vid_hdr (avbuf);
@@ -1170,7 +1170,7 @@ static int ubi_write_fastmap(struct ubi_ seen_pebs =init_seen (ubi);if (IS_ERR (seen_pebs)) { ret =PTR_ERR (seen_pebs); - goto out_kfree; + goto out_free_dvbuf; }spin_lock (&ubi->volumes_lock);
@@ -1338,7 +1338,7 @@ static int ubi_write_fastmap(struct ubi_ ret =ubi_io_write_vid_hdr (ubi, new_fm->e[0 ]->pnum, avbuf);if (ret) {ubi_err (ubi,"unable to write vid_hdr to fastmap SB!" ); - goto out_kfree; + goto out_free_seen; }for (i =0 ; i < new_fm->used_blocks; i++) {
@@ -1360,7 +1360,7 @@ static int ubi_write_fastmap(struct ubi_if (ret) {ubi_err (ubi,"unable to write vid_hdr to PEB %i!" , new_fm->e[i]->pnum); - goto out_kfree; + goto out_free_seen; } }
@@ -1370,7 +1370,7 @@ static int ubi_write_fastmap(struct ubi_if (ret) {ubi_err (ubi,"unable to write fastmap to PEB %i!" , new_fm->e[i]->pnum); - goto out_kfree; + goto out_free_seen; } }
@@ -1380,10 +1380,13 @@ static int ubi_write_fastmap(struct ubi_ ret =self_check_seen (ubi, seen_pebs);dbg_bld ("fastmap written!" ); -out_kfree: -ubi_free_vid_buf (avbuf); -ubi_free_vid_buf (dvbuf); +out_free_seen:free_seen (seen_pebs); +out_free_dvbuf: +ubi_free_vid_buf (dvbuf); +out_free_avbuf: +ubi_free_vid_buf (avbuf); + out:return ret; }
Greg Kroah-Hartman
maintainerPATCH 110/173 mfd: da9062: Fix watchdog compatible string
In reply to Greg Kroah-HartmanFrom: Marco Felsch <[email protected]>
commit 1112ba02ff1190ca9c15a912f9269e54b46d2d82 upstream.
The watchdog driver compatible is "dlg,da9062-watchdog" and not "dlg,da9062-wdt". Therefore the mfd-core can't populate the of_node and fwnode. As result the watchdog driver can't parse the devicetree.
--- drivers/mfd/da9062-core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/mfd/da9062-core.c +1 -1
@@ -257,7 +257,7 @@ static const struct mfd_cell da9062_devs .name = "da9062-watchdog", .num_resources = ARRAY_SIZE(da9062_wdt_resources), .resources = da9062_wdt_resources, - .of_compatible = "dlg,da9062-wdt", + .of_compatible = "dlg,da9062-watchdog", }, { .name = "da9062-thermal",
Greg Kroah-Hartman
maintainerPATCH 111/173 mfd: rn5t618: Mark ADC control register volatile
In reply to Greg Kroah-HartmanFrom: Andreas Kemnade <[email protected]>
commit 2f3dc25c0118de03a00ddc88b61f7216854f534d upstream.
There is a bit which gets cleared after conversion.
--- drivers/mfd/rn5t618.c | 1 + 1 file changed, 1 insertion(+)
drivers/mfd/rn5t618.c +1 -0
@@ -32,6 +32,7 @@ static bool rn5t618_volatile_reg(structcase RN5T618_WATCHDOGCNT :case RN5T618_DCIRQ : case RN5T618_ILIMDATAH ... RN5T618_AIN0DATAL: +case RN5T618_ADCCNT3 : case RN5T618_IR_ADC1 ... RN5T618_IR_ADC3:case RN5T618_IR_GPR :case RN5T618_IR_GPF :
Greg Kroah-Hartman
maintainerPATCH 112/173 net: dsa: bcm_sf2: Only 7278 supports 2Gb/sec IMP port
In reply to Greg Kroah-HartmanFrom: Florian Fainelli <[email protected]>
[ Upstream commit de34d7084edd069dac5aa010cfe32bd8c4619fa6 ]
The 7445 switch clocking profiles do not allow us to run the IMP port at 2Gb/sec in a way that it is reliable and consistent. Make sure that the setting is only applied to the 7278 family.
--- drivers/net/dsa/bcm_sf2.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/net/dsa/bcm_sf2.c +3 -1
@@ -137,7 +137,9 @@ static void bcm_sf2_imp_setup(struct dsa/* Force link status for IMP port */ reg =core_readl (priv, offset); - reg |= (MII_SW_OR |LINK_STS |GMII_SPEED_UP_2G ); + reg |= (MII_SW_OR |LINK_STS ); +if (priv->type ==BCM7278_DEVICE_ID ) + reg |=GMII_SPEED_UP_2G ;core_writel (priv, reg, offset);/* Enable Broadcast, Multicast, Unicast forwarding to IMP port */
Greg Kroah-Hartman
maintainerPATCH 113/173 net_sched: fix a resource leak in tcindex_set_parms()
In reply to Greg Kroah-HartmanFrom: Cong Wang <[email protected]>
[ Upstream commit 52b5ae501c045010aeeb1d5ac0373ff161a88291 ]
Jakub noticed there is a potential resource leak in tcindex_set_parms(): when tcindex_filter_result_init() fails and it jumps to 'errout1' which doesn't release the memory and resources allocated by tcindex_alloc_perfect_hash().
We should just jump to 'errout_alloc' which callstcindex_free_perfect_hash().
--- net/sched/cls_tcindex.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
net/sched/cls_tcindex.c +1 -2
@@ -383,7 +383,7 @@ tcindex_set_parms(struct net *net, struc err =tcindex_filter_result_init (&new_filter_result);if (err <0 ) - goto errout1; + goto errout_alloc;if (old_r) cr = r->res;
@@ -502,7 +502,6 @@ errout_alloc:tcindex_free_perfect_hash (cp); else if (balloc == 2) kfree(cp->h); -errout1:tcf_exts_destroy (&new_filter_result.exts); errout:kfree (cp);
Greg Kroah-Hartman
maintainerPATCH 114/173 net: systemport: Avoid RBUF stuck in Wake-on-LAN mode
In reply to Greg Kroah-HartmanFrom: Florian Fainelli <[email protected]>
[ Upstream commit 263a425a482fc495d6d3f9a29b9103a664c38b69 ]
After a number of suspend and resume cycles, it is possible for the RBUF to be stuck in Wake-on-LAN mode, despite the MPD enable bit being cleared which instructed the RBUF to exit that mode.
Avoid creating that problematic condition by clearing the RX_EN and TX_EN bits in the UniMAC prior to disable the Magic Packet Detector logic which is guaranteed to make the RBUF exit Wake-on-LAN mode.
--- drivers/net/ethernet/broadcom/bcmsysport.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/net/ethernet/broadcom/bcmsysport.c +3 -0
@@ -2329,6 +2329,9 @@ static int bcm_sysport_resume(struct devumac_reset (priv); +/* Disable the UniMAC RX/TX */ +umac_enable_set (priv,CMD_RX_EN |CMD_TX_EN ,0 ); +/* We may have been suspended and never received a WOL event that * would turn off MPD detection, take care of that now */
Greg Kroah-Hartman
maintainerPATCH 115/173 net: macb: Remove unnecessary alignment check for TSO
In reply to Greg Kroah-HartmanFrom: Harini Katakam <[email protected]>
[ Upstream commit 41c1ef978c8d0259c6636e6d2d854777e92650eb ]
The IP TSO implementation does NOT require the length to be a multiple of 8. That is only a requirement for UFO as per IP documentation. Hence, exit macb_features_check function in the beginning if the protocol is not UDP. Only when it is UDP, proceed further to the alignment checks. Update comments to reflect the same. Also remove dead code checking for protocol TCP when calculating header length.
--- drivers/net/ethernet/cadence/macb_main.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-)
drivers/net/ethernet/cadence/macb_main.c +3 -5
@@ -1577,16 +1577,14 @@ static netdev_features_t macb_features_c/* Validate LSO compatibility */ -/* there is only one buffer */ -if (!skb_is_nonlinear (skb)) +/* there is only one buffer or protocol is not UDP */ +if (!skb_is_nonlinear (skb) || (ip_hdr (skb)->protocol !=IPPROTO_UDP ))return features;/* length of header */ hdrlen =skb_transport_offset (skb); -if (ip_hdr (skb)->protocol ==IPPROTO_TCP ) - hdrlen +=tcp_hdrlen (skb); -/* For LSO: +/* For UFO only: * When software supplies two or more payload buffers all payload buffers * apart from the last must be a multiple of 8 bytes in size. */
Greg Kroah-Hartman
maintainerPATCH 116/173 net: macb: Limit maximum GEM TX length in TSO
In reply to Greg Kroah-HartmanFrom: Harini Katakam <[email protected]>
[ Upstream commit f822e9c4ffa511a5c681cf866287d9383a3b6f1b ]
GEM_MAX_TX_LEN currently resolves to 0x3FF8 for any IP version supporting TSO with full 14bits of length field in payload descriptor. But an IP errata causes false amba_error (bit 6 of ISR) when length in payload descriptors is specified above 16387. The error occurs because the DMA falsely concludes that there is not enough space in SRAM for incoming payload. These errors were observed continuously under stress of large packets using iperf on a version where SRAM was 16K for each queue. This errata will be documented shortly and affects all versions since TSO functionality was added. Hence limit the max length to 0x3FC0 (rounded).
--- drivers/net/ethernet/cadence/macb_main.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/net/ethernet/cadence/macb_main.c +5 -1
@@ -66,7 +66,11 @@/* Max length of transmit frame must be a multiple of 8 bytes */ #define MACB_TX_LEN_ALIGN 8#define MACB_MAX_TX_LEN ((unsigned int)((1 << MACB_TX_FRMLEN_SIZE) - 1) & ~((unsigned int)(MACB_TX_LEN_ALIGN - 1))) -#define GEM_MAX_TX_LEN ((unsigned int)((1 << GEM_TX_FRMLEN_SIZE) - 1) & ~((unsigned int)(MACB_TX_LEN_ALIGN - 1))) +/* Limit maximum TX length as per Cadence TSO errata. This is to avoid a +* false amba_error in TX path from the DMA assuming there is not enough +* space in the SRAM (16KB) even when there is. +*/ +#define GEM_MAX_TX_LEN (unsigned int)(0x3FC0)#define GEM_MTU_MIN_SIZE ETH_MIN_MTU#define MACB_NETIF_LSO NETIF_F_TSO
Greg Kroah-Hartman
maintainerPATCH 117/173 bonding/alb: properly access headers in bond_alb_xmit()
In reply to Greg Kroah-HartmanFrom: Eric Dumazet <[email protected]>
[ Upstream commit 38f88c45404293bbc027b956def6c10cbd45c616 ]
syzbot managed to send an IPX packet through bond_alb_xmit() and af_packet and triggered a use-after-free.
First, bond_alb_xmit() was using ipx_hdr() helper to reach the IPX header, but ipx_hdr() was using the transport offset instead of the network offset. In the particular syzbot report transport offset was 0xFFFF
This patch removes ipx_hdr() since it was only (mis)used from bonding.
Then we need to make sure IPv4/IPv6/IPX headers are pulled in skb->head before dereferencing anything.
BUG: KASAN: use-after-free in bond_alb_xmit+0x153a/0x1590 drivers/net/bonding/bond_alb.c:1452 Read of size 2 at addr ffff8801ce56dfff by task syz-executor.2/18108 (if (ipx_hdr(skb)->ipx_checksum != IPX_NO_CHECKSUM) ...) Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: [<ffffffff8441fc42>] __dump_stack lib/dump_stack.c:17 [inline] [<ffffffff8441fc42>] dump_stack+0x14d/0x20b lib/dump_stack.c:53 [<ffffffff81a7dec4>] print_address_description+0x6f/0x20b mm/kasan/report.c:282 [<ffffffff81a7e0ec>] kasan_report_error mm/kasan/report.c:380 [inline] [<ffffffff81a7e0ec>] kasan_report mm/kasan/report.c:438 [inline] [<ffffffff81a7e0ec>] kasan_report.cold+0x8c/0x2a0 mm/kasan/report.c:422 [<ffffffff81a7dc4f>] __asan_report_load_n_noabort+0xf/0x20 mm/kasan/report.c:469 [<ffffffff82c8c00a>] bond_alb_xmit+0x153a/0x1590 drivers/net/bonding/bond_alb.c:1452 [<ffffffff82c60c74>] __bond_start_xmit drivers/net/bonding/bond_main.c:4199 [inline] [<ffffffff82c60c74>] bond_start_xmit+0x4f4/0x1570 drivers/net/bonding/bond_main.c:4224 [<ffffffff83baa558>] __netdev_start_xmit include/linux/netdevice.h:4525 [inline] [<ffffffff83baa558>] netdev_start_xmit include/linux/netdevice.h:4539 [inline] [<ffffffff83baa558>] xmit_one net/core/dev.c:3611 [inline] [<ffffffff83baa558>] dev_hard_start_xmit+0x168/0x910 net/core/dev.c:3627 [<ffffffff83bacf35>] __dev_queue_xmit+0x1f55/0x33b0 net/core/dev.c:4238 [<ffffffff83bae3a8>] dev_queue_xmit+0x18/0x20 net/core/dev.c:4278 [<ffffffff84339189>] packet_snd net/packet/af_packet.c:3226 [inline] [<ffffffff84339189>] packet_sendmsg+0x4919/0x70b0 net/packet/af_packet.c:3252 [<ffffffff83b1ac0c>] sock_sendmsg_nosec net/socket.c:673 [inline] [<ffffffff83b1ac0c>] sock_sendmsg+0x12c/0x160 net/socket.c:684 [<ffffffff83b1f5a2>] __sys_sendto+0x262/0x380 net/socket.c:1996 [<ffffffff83b1f700>] SYSC_sendto net/socket.c:2008 [inline] [<ffffffff83b1f700>] SyS_sendto+0x40/0x60 net/socket.c:2004
--- drivers/net/bonding/bond_alb.c | 44 +++++++++++++++++++++++++++++------------ 1 file changed, 32 insertions(+), 12 deletions(-)
drivers/net/bonding/bond_alb.c +32 -12
@@ -1403,26 +1403,31 @@ int bond_alb_xmit(struct sk_buff *skb, sbool do_tx_balance = true;u32 hash_index =0 ;const u8 *hash_start =NULL ; -struct ipv6hdr *ip6hdr;skb_reset_mac_header (skb); eth_data =eth_hdr (skb);switch (ntohs (skb->protocol)) {case ETH_P_IP : { -const struct iphdr *iph =ip_hdr (skb); +const struct iphdr *iph;if (ether_addr_equal_64bits (eth_data->h_dest, mac_bcast) || - (iph->daddr == ip_bcast) || - (iph->protocol ==IPPROTO_IGMP )) { + (!pskb_network_may_pull (skb,sizeof (*iph)))) { + do_tx_balance = false; +break ; + } + iph =ip_hdr (skb); +if (iph->daddr == ip_bcast || iph->protocol ==IPPROTO_IGMP ) { do_tx_balance = false;break ; } hash_start = (char *)&(iph->daddr); hash_size =sizeof (iph->daddr); - }break ; -case ETH_P_IPV6 : + } +case ETH_P_IPV6 : { +const struct ipv6hdr *ip6hdr; +/* IPv6 doesn't really use broadcast mac address, but leave * that here just in case. */
@@ -1439,7 +1444,11 @@ int bond_alb_xmit(struct sk_buff *skb, sbreak ; } -/* Additianally, DAD probes should not be tx-balanced as that +if (!pskb_network_may_pull (skb,sizeof (*ip6hdr))) { + do_tx_balance = false; +break ; + } +/* Additionally, DAD probes should not be tx-balanced as that * will lead to false positives for duplicate addresses and * prevent address configuration from working. */
@@ -1449,17 +1458,26 @@ int bond_alb_xmit(struct sk_buff *skb, sbreak ; } - hash_start = (char *)&(ipv6_hdr (skb)->daddr); - hash_size =sizeof (ipv6_hdr (skb)->daddr); + hash_start = (char *)&ip6hdr->daddr; + hash_size =sizeof (ip6hdr->daddr);break ; -case ETH_P_IPX : -if (ipx_hdr (skb)->ipx_checksum !=IPX_NO_CHECKSUM ) { + } +case ETH_P_IPX : { +const struct ipxhdr *ipxhdr; + +if (pskb_network_may_pull (skb,sizeof (*ipxhdr))) { + do_tx_balance = false; +break ; + } + ipxhdr = (struct ipxhdr *)skb_network_header (skb); + +if (ipxhdr->ipx_checksum !=IPX_NO_CHECKSUM ) {/* something is wrong with this packet */ do_tx_balance = false;break ; } -if (ipx_hdr (skb)->ipx_type !=IPX_TYPE_NCP ) { +if (ipxhdr->ipx_type !=IPX_TYPE_NCP ) {/* The only protocol worth balancing in * this family since it has an "ARP" like * mechanism
@@ -1468,9 +1486,11 @@ int bond_alb_xmit(struct sk_buff *skb, sbreak ; } + eth_data =eth_hdr (skb); hash_start = (char *)eth_data->h_dest; hash_size =ETH_ALEN ;break ; + }case ETH_P_ARP : do_tx_balance = false;if (bond_info->rlb_enabled)
Greg Kroah-Hartman
maintainerPATCH 118/173 ext4: fix deadlock allocating crypto bounce page from mempool
In reply to Greg Kroah-HartmanFrom: Eric Biggers <[email protected]>
[ Upstream commit 547c556f4db7c09447ecf5f833ab6aaae0c5ab58 ]
ext4_writepages() on an encrypted file has to encrypt the data, but it can't modify the pagecache pages in-place, so it encrypts the data into bounce pages and writes those instead. All bounce pages are allocated from a mempool using GFP_NOFS.
This is not correct use of a mempool, and it can deadlock. This is because GFP_NOFS includes __GFP_DIRECT_RECLAIM, which enables the "never fail" mode for mempool_alloc() where a failed allocation will fall back to waiting for one of the preallocated elements in the pool.
But since this mode is used for all a bio's pages and not just the first, it can deadlock waiting for pages already in the bio to be freed.
This deadlock can be reproduced by patching mempool_alloc() to pretend that pool->alloc() always fails (so that it always falls back to the preallocations), and then creating an encrypted file of size > 128 KiB.
Fix it by only using GFP_NOFS for the first page in the bio. For subsequent pages just use GFP_NOWAIT, and if any of those fail, just submit the bio and start a new one.
This will need to be fixed in f2fs too, but that's less straightforward.
--- fs/ext4/page-io.c | 19 ++++++++++++++----- 1 file changed, 14 insertions(+), 5 deletions(-)
fs/ext4/page-io.c +14 -5
@@ -481,17 +481,26 @@ int ext4_bio_write_page(struct ext4_io_submit *io, nr_to_submit) {gfp_t gfp_flags =GFP_NOFS ; +/* +* Since bounce page allocation uses a mempool, we can only use +* a waiting mask (i.e. request guaranteed allocation) on the +* first page of the bio. Otherwise it can deadlock. +*/ +if (io->io_bio) + gfp_flags =GFP_NOWAIT | __GFP_NOWARN; retry_encrypt: data_page =fscrypt_encrypt_page (inode, page,PAGE_SIZE ,0 , page->index, gfp_flags);if (IS_ERR (data_page)) { ret =PTR_ERR (data_page); -if (ret == -ENOMEM && wbc->sync_mode ==WB_SYNC_ALL ) { -if (io->io_bio) { +if (ret == -ENOMEM && + (io->io_bio || wbc->sync_mode ==WB_SYNC_ALL )) { + gfp_flags =GFP_NOFS ; +if (io->io_bio)ext4_io_submit (io); -congestion_wait (BLK_RW_ASYNC ,HZ /50 ); - } - gfp_flags |= __GFP_NOFAIL; +else + gfp_flags |= __GFP_NOFAIL; +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto retry_encrypt; } data_page =NULL ;
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 119/173 btrfs: Get rid of the confusing btrfs_file_extent_inline_len
In reply to Greg Kroah-HartmanFrom: Qu Wenruo <[email protected]>
[ Upstream commit e41ca5897489b1c18af75ff0cc8f5c80260b3281 ]
We used to call btrfs_file_extent_inline_len() to get the uncompressed data size of an inlined extent.
However this function is hiding evil, for compressed extent, it has no choice but to directly read out ram_bytes from btrfs_file_extent_item. While for uncompressed extent, it uses item size to calculate the real data size, and ignoring ram_bytes completely.
In fact, for corrupted ram_bytes, due to above behavior kernel btrfs_print_leaf() can't even print correct ram_bytes to expose the bug.
Since we have the tree-checker to verify all EXTENT_DATA, such mismatch can be detected pretty easily, thus we can trust ram_bytes without the evil btrfs_file_extent_inline_len().
--- fs/btrfs/ctree.h | 26 -------------------------- fs/btrfs/file-item.c | 2 +- fs/btrfs/file.c | 3 +-- fs/btrfs/inode.c | 12 ++++++------ fs/btrfs/print-tree.c | 4 ++-- fs/btrfs/send.c | 17 +++++++---------- fs/btrfs/tree-log.c | 12 ++++-------- include/trace/events/btrfs.h | 2 +- 8 files changed, 22 insertions(+), 56 deletions(-)
fs/btrfs/ctree.h +0 -26
@@ -2408,32 +2408,6 @@ static inline u32 btrfs_file_extent_inline_item_len(return btrfs_item_size (eb, e) -BTRFS_FILE_EXTENT_INLINE_DATA_START ; } -/* this returns the number of file bytes represented by the inline item. -* If an item is compressed, this is the uncompressed size -*/ -static inline u32 btrfs_file_extent_inline_len (const struct extent_buffer *eb, -int slot, -const struct btrfs_file_extent_item *fi) -{ -struct btrfs_map_token token; - -btrfs_init_map_token (&token); -/* -* return the space used on disk if this item isn't -* compressed or encoded -*/ -if (btrfs_token_file_extent_compression (eb, fi, &token) ==0 && -btrfs_token_file_extent_encryption (eb, fi, &token) ==0 && -btrfs_token_file_extent_other_encoding (eb, fi, &token) ==0 ) { -return btrfs_file_extent_inline_item_len (eb, -btrfs_item_nr (slot)); - } - -/* otherwise use the ram bytes field */ -return btrfs_token_file_extent_ram_bytes (eb, fi, &token); -} - -/* btrfs_dev_stats_item */ static inline u64 btrfs_dev_stats_value(const struct extent_buffer *eb, const struct btrfs_dev_stats_item *ptr,
fs/btrfs/file-item.c +1 -1
@@ -955,7 +955,7 @@ void btrfs_extent_item_to_extent_map(struct btrfs_inode *inode,btrfs_file_extent_num_bytes (leaf, fi); }else if (type ==BTRFS_FILE_EXTENT_INLINE ) {size_t size; - size =btrfs_file_extent_inline_len (leaf, slot, fi); + size =btrfs_file_extent_ram_bytes (leaf, fi); extent_end =ALIGN (extent_start + size, fs_info->sectorsize); }
fs/btrfs/file.c +1 -2
@@ -784,8 +784,7 @@ int __btrfs_drop_extents(struct btrfs_trans_handle *trans,btrfs_file_extent_num_bytes (leaf, fi); }else if (extent_type ==BTRFS_FILE_EXTENT_INLINE ) { extent_end = key.offset + -btrfs_file_extent_inline_len (leaf, - path->slots[0 ], fi); +btrfs_file_extent_ram_bytes (leaf, fi); }else {/* can't happen */ BUG ();
fs/btrfs/inode.c +6 -6
@@ -1476,8 +1476,7 @@ static noinline int run_delalloc_nocow(struct inode *inode, nocow =1 ; }else if (extent_type ==BTRFS_FILE_EXTENT_INLINE ) { extent_end = found_key.offset + -btrfs_file_extent_inline_len (leaf, - path->slots[0 ], fi); +btrfs_file_extent_ram_bytes (leaf, fi); extent_end =ALIGN (extent_end, fs_info->sectorsize); }else {
@@ -4651,8 +4650,8 @@ int btrfs_truncate_inode_items(struct btrfs_trans_handle *trans,BTRFS_I (inode), leaf, fi, found_key.offset); }else if (extent_type ==BTRFS_FILE_EXTENT_INLINE ) { - item_end +=btrfs_file_extent_inline_len (leaf, - path->slots[0 ], fi); + item_end +=btrfs_file_extent_ram_bytes (leaf, + fi); trace_btrfs_truncate_show_fi_inline( BTRFS_I(inode), leaf, fi, path->slots[0],
@@ -7167,7 +7166,8 @@ struct extent_map *btrfs_get_extent(struct btrfs_inode *inode, extent_start); }else if (found_type ==BTRFS_FILE_EXTENT_INLINE ) {size_t size; - size =btrfs_file_extent_inline_len (leaf, path->slots[0 ], item); + + size =btrfs_file_extent_ram_bytes (leaf, item); extent_end =ALIGN (extent_start + size, fs_info->sectorsize);
@@ -7218,7 +7218,7 @@ struct extent_map *btrfs_get_extent(struct btrfs_inode *inode,if (new_inline) goto out; - size =btrfs_file_extent_inline_len (leaf, path->slots[0 ], item); + size =btrfs_file_extent_ram_bytes (leaf, item); extent_offset =page_offset (page) + pg_offset - extent_start; copy_size =min_t (u64,PAGE_SIZE - pg_offset, size - extent_offset);
fs/btrfs/print-tree.c +2 -2
@@ -259,8 +259,8 @@ void btrfs_print_leaf(struct extent_buffer *l) struct btrfs_file_extent_item);if (btrfs_file_extent_type (l, fi) ==BTRFS_FILE_EXTENT_INLINE ) { -pr_info ("\t\tinline extent data size %u\n" , -btrfs_file_extent_inline_len (l, i, fi)); +pr_info ("\t\tinline extent data size %llu\n" , +btrfs_file_extent_ram_bytes (l, fi));break ; } pr_info("\t\textent data disk bytenr %llu nr %llu\n",
fs/btrfs/send.c +7 -10
@@ -1545,7 +1545,7 @@ static int read_symlink(struct btrfs_root *root,BUG_ON (compression); off =btrfs_file_extent_inline_start (ei); - len =btrfs_file_extent_inline_len (path->nodes[0 ], path->slots[0 ], ei); + len =btrfs_file_extent_ram_bytes (path->nodes[0 ], ei); ret =fs_path_add_from_extent_buffer (dest, path->nodes[0 ], off, len);
@@ -5195,7 +5195,7 @@ static int clone_range(struct send_ctx *sctx, ei = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item); type =btrfs_file_extent_type (leaf, ei);if (type ==BTRFS_FILE_EXTENT_INLINE ) { - ext_len =btrfs_file_extent_inline_len (leaf, slot, ei); + ext_len =btrfs_file_extent_ram_bytes (leaf, ei); ext_len =PAGE_ALIGN (ext_len); }else { ext_len =btrfs_file_extent_num_bytes (leaf, ei);
@@ -5271,8 +5271,7 @@ static int send_write_or_clone(struct send_ctx *sctx, struct btrfs_file_extent_item); type =btrfs_file_extent_type (path->nodes[0 ], ei);if (type ==BTRFS_FILE_EXTENT_INLINE ) { - len =btrfs_file_extent_inline_len (path->nodes[0 ], - path->slots[0 ], ei); + len =btrfs_file_extent_ram_bytes (path->nodes[0 ], ei);/* * it is possible the inline item won't cover the whole page, * but there may be items after this page. Make
@@ -5405,7 +5404,7 @@ static int is_extent_unchanged(struct send_ctx *sctx, }if (right_type ==BTRFS_FILE_EXTENT_INLINE ) { - right_len =btrfs_file_extent_inline_len (eb, slot, ei); + right_len =btrfs_file_extent_ram_bytes (eb, ei); right_len =PAGE_ALIGN (right_len); }else { right_len =btrfs_file_extent_num_bytes (eb, ei);
@@ -5526,8 +5525,7 @@ static int get_last_extent(struct send_ctx *sctx, u64 offset) struct btrfs_file_extent_item); type =btrfs_file_extent_type (path->nodes[0 ], fi);if (type ==BTRFS_FILE_EXTENT_INLINE ) { -u64 size =btrfs_file_extent_inline_len (path->nodes[0 ], - path->slots[0 ], fi); +u64 size =btrfs_file_extent_ram_bytes (path->nodes[0 ], fi); extent_end =ALIGN (key.offset + size, sctx->send_root->fs_info->sectorsize); }else {
@@ -5590,7 +5588,7 @@ static int range_is_hole_in_parent(struct send_ctx *sctx, fi = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item);if (btrfs_file_extent_type (leaf, fi) ==BTRFS_FILE_EXTENT_INLINE ) { -u64 size =btrfs_file_extent_inline_len (leaf, slot, fi); +u64 size =btrfs_file_extent_ram_bytes (leaf, fi); extent_end =ALIGN (key.offset + size, root->fs_info->sectorsize);
@@ -5636,8 +5634,7 @@ static int maybe_send_hole(struct send_ctx *sctx, struct btrfs_path *path, struct btrfs_file_extent_item); type =btrfs_file_extent_type (path->nodes[0 ], fi);if (type ==BTRFS_FILE_EXTENT_INLINE ) { -u64 size =btrfs_file_extent_inline_len (path->nodes[0 ], - path->slots[0 ], fi); +u64 size =btrfs_file_extent_ram_bytes (path->nodes[0 ], fi); extent_end =ALIGN (key->offset + size, sctx->send_root->fs_info->sectorsize); }else {
fs/btrfs/tree-log.c +4 -8
@@ -619,7 +619,7 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans,if (btrfs_file_extent_disk_bytenr (eb, item) ==0 ) nbytes =0 ; }else if (found_type ==BTRFS_FILE_EXTENT_INLINE ) { - size =btrfs_file_extent_inline_len (eb, slot, item); + size =btrfs_file_extent_ram_bytes (eb, item); nbytes =btrfs_file_extent_ram_bytes (eb, item); extent_end =ALIGN (start + size, fs_info->sectorsize);
@@ -3943,9 +3943,7 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, struct btrfs_file_extent_item);if (btrfs_file_extent_type (src, extent) ==BTRFS_FILE_EXTENT_INLINE ) { - len =btrfs_file_extent_inline_len (src, - src_path->slots[0 ], - extent); + len =btrfs_file_extent_ram_bytes (src, extent); *last_extent =ALIGN (key.offset + len, fs_info->sectorsize); }else {
@@ -4010,7 +4008,7 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, extent = btrfs_item_ptr(src, i, struct btrfs_file_extent_item);if (btrfs_file_extent_type (src, extent) ==BTRFS_FILE_EXTENT_INLINE ) { - len =btrfs_file_extent_inline_len (src, i, extent); + len =btrfs_file_extent_ram_bytes (src, extent); extent_end =ALIGN (key.offset + len, fs_info->sectorsize); }else {
@@ -4730,9 +4728,7 @@ static int btrfs_log_trailing_hole(struct btrfs_trans_handle *trans, if (btrfs_file_extent_type(leaf, extent) == BTRFS_FILE_EXTENT_INLINE) { - len = btrfs_file_extent_inline_len(leaf, - path->slots[0], - extent); + len = btrfs_file_extent_ram_bytes(leaf, extent); ASSERT(len == i_size || (len == fs_info->sectorsize && btrfs_file_extent_compression(leaf, extent) !=
include/trace/events/btrfs.h +1 -1
@@ -325,7 +325,7 @@ DECLARE_EVENT_CLASS( __entry->extent_type =btrfs_file_extent_type (l, fi); __entry->compression =btrfs_file_extent_compression (l, fi); __entry->extent_start = start; - __entry->extent_end = (start +btrfs_file_extent_inline_len (l, slot, fi)); + __entry->extent_end = (start +btrfs_file_extent_ram_bytes (l, fi)); ), TP_printk_btrfs(
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 120/173 Btrfs: fix assertion failure on fsync with NO_HOLES enabled
In reply to Greg Kroah-HartmanFrom: Filipe Manana <[email protected]>
[ Upstream commit 0ccc3876e4b2a1559a4dbe3126dda4459d38a83b ]
Back in commit a89ca6f24ffe4 ("Btrfs: fix fsync after truncate when no_holes feature is enabled") I added an assertion that is triggered when an inline extent is found to assert that the length of the (uncompressed) data the extent represents is the same as the i_size of the inode, since that is true most of the time I couldn't find or didn't remembered about any exception at that time. Later on the assertion was expanded twice to deal with a case of a compressed inline extent representing a range that matches the sector size followed by an expanding truncate, and another case where fallocate can update the i_size of the inode without adding or updating existing extents (if the fallocate range falls entirely within the first block of the file). These two expansion/fixes of the assertion were done by commit 7ed586d0a8241 ("Btrfs: fix assertion on fsync of regular file when using no-holes feature") and commit 6399fb5a0b69a ("Btrfs: fix assertion failure during fsync in no-holes mode"). These however missed the case where an falloc expands the i_size of an inode to exactly the sector size and inline extent exists, for example:
$ mkfs.btrfs -f -O no-holes /dev/sdc $ mount /dev/sdc /mnt
$ xfs_io -f -c "pwrite -S 0xab 0 1096" /mnt/foobar wrote 1096/1096 bytes at offset 0 1 KiB, 1 ops; 0.0002 sec (4.448 MiB/sec and 4255.3191 ops/sec)
$ xfs_io -c "falloc 1096 3000" /mnt/foobar $ xfs_io -c "fsync" /mnt/foobar Segmentation fault
$ dmesg [701253.602385] assertion failed: len == i_size || (len == fs_info->sectorsize && btrfs_file_extent_compression(leaf, extent) != BTRFS_COMPRESS_NONE) || (len < i_size && i_size < fs_info->sectorsize), file: fs/btrfs/tree-log.c, line: 4727 [701253.602962] ------------[ cut here ]------------ [701253.603224] kernel BUG at fs/btrfs/ctree.h:3533! [701253.603503] invalid opcode: 0000 [#1] SMP DEBUG_PAGEALLOC PTI [701253.603774] CPU: 2 PID: 7192 Comm: xfs_io Tainted: G W 5.0.0-rc8-btrfs-next-45 #1 [701253.604054] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.11.2-0-gf9626ccb91-prebuilt.qemu-project.org 04/01/2014 [701253.604650] RIP: 0010:assfail.constprop.23+0x18/0x1a [btrfs] (...) [701253.605591] RSP: 0018:ffffbb48c186bc48 EFLAGS: 00010286 [701253.605914] RAX: 00000000000000de RBX: ffff921d0a7afc08 RCX: 0000000000000000 [701253.606244] RDX: 0000000000000000 RSI: ffff921d36b16868 RDI: ffff921d36b16868 [701253.606580] RBP: ffffbb48c186bcf0 R08: 0000000000000000 R09: 0000000000000000 [701253.606913] R10: 0000000000000003 R11: 0000000000000000 R12: ffff921d05d2de18 [701253.607247] R13: ffff921d03b54000 R14: 0000000000000448 R15: ffff921d059ecf80 [701253.607769] FS: 00007f14da906700(0000) GS:ffff921d36b00000(0000) knlGS:0000000000000000 [701253.608163] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [701253.608516] CR2: 000056087ea9f278 CR3: 00000002268e8001 CR4: 00000000003606e0 [701253.608880] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [701253.609250] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [701253.609608] Call Trace: [701253.609994] btrfs_log_inode+0xdfb/0xe40 [btrfs] [701253.610383] btrfs_log_inode_parent+0x2be/0xa60 [btrfs] [701253.610770] ? do_raw_spin_unlock+0x49/0xc0 [701253.611150] btrfs_log_dentry_safe+0x4a/0x70 [btrfs] [701253.611537] btrfs_sync_file+0x3b2/0x440 [btrfs] [701253.612010] ? do_sysinfo+0xb0/0xf0 [701253.612552] do_fsync+0x38/0x60 [701253.612988] __x64_sys_fsync+0x10/0x20 [701253.613360] do_syscall_64+0x60/0x1b0 [701253.613733] entry_SYSCALL_64_after_hwframe+0x49/0xbe [701253.614103] RIP: 0033:0x7f14da4e66d0 (...) [701253.615250] RSP: 002b:00007fffa670fdb8 EFLAGS: 00000246 ORIG_RAX: 000000000000004a [701253.615647] RAX: ffffffffffffffda RBX: 0000000000000001 RCX: 00007f14da4e66d0 [701253.616047] RDX: 000056087ea9c260 RSI: 000056087ea9c260 RDI: 0000000000000003 [701253.616450] RBP: 0000000000000001 R08: 0000000000000020 R09: 0000000000000010 [701253.616854] R10: 000000000000009b R11: 0000000000000246 R12: 000056087ea9c260 [701253.617257] R13: 000056087ea9c240 R14: 0000000000000000 R15: 000056087ea9dd10 (...) [701253.619941] ---[ end trace e088d74f132b6da5 ]---
Updating the assertion again to allow for this particular case would result in a meaningless assertion, plus there is currently no risk of logging content that would result in any corruption after a log replay if the size of the data encoded in an inline extent is greater than the inode's i_size (which is not currently possibe either with or without compression), therefore just remove the assertion.
--- fs/btrfs/tree-log.c | 9 +-------- 1 file changed, 1 insertion(+), 8 deletions(-)
fs/btrfs/tree-log.c +1 -8
@@ -4727,15 +4727,8 @@ static int btrfs_log_trailing_hole(struct btrfs_trans_handle *trans, struct btrfs_file_extent_item);if (btrfs_file_extent_type (leaf, extent) == -BTRFS_FILE_EXTENT_INLINE ) { - len =btrfs_file_extent_ram_bytes (leaf, extent); -ASSERT (len == i_size || - (len == fs_info->sectorsize && -btrfs_file_extent_compression (leaf, extent) != -BTRFS_COMPRESS_NONE ) || - (len < i_size && i_size < fs_info->sectorsize)); +BTRFS_FILE_EXTENT_INLINE )return 0 ; - } len =btrfs_file_extent_num_bytes (leaf, extent);/* Last extent goes beyond i_size, no need to log a hole. */
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 121/173 Btrfs: fix missing hole after hole punching and fsync when using NO_HOLES
In reply to Greg Kroah-HartmanFrom: Filipe Manana <[email protected]>
[ Upstream commit 0e56315ca147b3e60c7bf240233a301d3c7fb508 ]
When using the NO_HOLES feature, if we punch a hole into a file and then fsync it, there are cases where a subsequent fsync will miss the fact that a hole was punched, resulting in the holes not existing after replaying the log tree.
Essentially these cases all imply that, tree-log.c:copy_items(), is not invoked for the leafs that delimit holes, because nothing changed those leafs in the current transaction. And it's precisely copy_items() where we currenly detect and log holes, which works as long as the holes are between file extent items in the input leaf or between the beginning of input leaf and the previous leaf or between the last item in the leaf and the next leaf.
First example where we miss a hole:
*) The extent items of the inode span multiple leafs;
*) The punched hole covers a range that affects only the extent items of
the first leaf;
*) The fsync operation is done in full mode (BTRFS_INODE_NEEDS_FULL_SYNC
is set in the inode's runtime flags).
That results in the hole not existing after replaying the log tree.
For example, if the fs/subvolume tree has the following layout for a particular inode:
Leaf N, generation 10:
[ ... INODE_ITEM INODE_REF EXTENT_ITEM (0 64K) EXTENT_ITEM (64K 128K) ]
Leaf N + 1, generation 10:
[ EXTENT_ITEM (128K 64K) ... ]
If at transaction 11 we punch a hole coverting the range [0, 128K[, we end up dropping the two extent items from leaf N, but we don't touch the other leaf, so we end up in the following state:
Leaf N, generation 11:
[ ... INODE_ITEM INODE_REF ]
Leaf N + 1, generation 10:
[ EXTENT_ITEM (128K 64K) ... ]
A full fsync after punching the hole will only process leaf N because it was modified in the current transaction, but not leaf N + 1, since it was not modified in the current transaction (generation 10 and not 11). As a result the fsync will not log any holes, because it didn't process any leaf with extent items.
Second example where we will miss a hole:
*) An inode as its items spanning 5 (or more) leafs;
*) A hole is punched and it covers only the extents items of the 3rd
leaf. This resulsts in deleting the entire leaf and not touching any of the other leafs.
So the only leaf that is modified in the current transaction, when punching the hole, is the first leaf, which contains the inode item. During the full fsync, the only leaf that is passed to copy_items() is that first leaf, and that's not enough for the hole detection code in copy_items() to determine there's a hole between the last file extent item in the 2nd leaf and the first file extent item in the 3rd leaf (which was the 4th leaf before punching the hole).
Fix this by scanning all leafs and punch holes as necessary when doing a full fsync (less common than a non-full fsync) when the NO_HOLES feature is enabled. The lack of explicit file extent items to mark holes makes it necessary to scan existing extents to determine if holes exist.
A test case for fstests follows soon.
--- fs/btrfs/tree-log.c | 388 ++++++++++++-------------------------------- 1 file changed, 100 insertions(+), 288 deletions(-)
fs/btrfs/tree-log.c +100 -288
@@ -3758,7 +3758,7 @@ static int log_inode_item(struct btrfs_trans_handle *trans, static noinline int copy_items(struct btrfs_trans_handle *trans,struct btrfs_inode *inode,struct btrfs_path *dst_path, -struct btrfs_path *src_path,u64 *last_extent, +struct btrfs_path *src_path,int start_slot,int nr,int inode_only,u64 logged_isize) {
@@ -3769,7 +3769,6 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,struct btrfs_file_extent_item *extent;struct btrfs_inode_item *inode_item;struct extent_buffer *src = src_path->nodes[0 ]; -struct btrfs_key first_key, last_key, key;int ret;struct btrfs_key *ins_keys;u32 *ins_sizes;
@@ -3777,9 +3776,6 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,int i;struct list_head ordered_sums;int skip_csum = inode->flags &BTRFS_INODE_NODATASUM ; -bool has_extents = false; -bool need_find_last_extent = true; -bool done = false;INIT_LIST_HEAD (&ordered_sums);
@@ -3788,8 +3784,6 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,if (!ins_data)return -ENOMEM ; - first_key.objectid = (u64)-1 ; - ins_sizes = (u32 *)ins_data; ins_keys = (struct btrfs_key *)(ins_data + nr *sizeof (u32));
@@ -3810,9 +3804,6 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, src_offset =btrfs_item_ptr_offset (src, start_slot + i); -if (i == nr -1 ) - last_key = ins_keys[i]; -if (ins_keys[i].type ==BTRFS_INODE_ITEM_KEY ) { inode_item = btrfs_item_ptr(dst_path->nodes[0], dst_path->slots[0],
@@ -3826,20 +3817,6 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, src_offset, ins_sizes[i]); } -/* -* We set need_find_last_extent here in case we know we were -* processing other items and then walk into the first extent in -* the inode. If we don't hit an extent then nothing changes, -* we'll do the last search the next time around. -*/ -if (ins_keys[i].type ==BTRFS_EXTENT_DATA_KEY ) { - has_extents = true; -if (first_key.objectid == (u64)-1 ) - first_key = ins_keys[i]; - }else { - need_find_last_extent = false; - } -/* take a reference on file data extents so that truncates * or deletes of this inode don't have to relog the inode * again
@@ -3905,167 +3882,6 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,kfree (sums); } -if (!has_extents) -return ret; - -if (need_find_last_extent && *last_extent == first_key.offset) { -/* -* We don't have any leafs between our current one and the one -* we processed before that can have file extent items for our -* inode (and have a generation number smaller than our current -* transaction id). -*/ - need_find_last_extent = false; - } - -/* -* Because we use btrfs_search_forward we could skip leaves that were -* not modified and then assume *last_extent is valid when it really -* isn't. So back up to the previous leaf and read the end of the last -* extent before we go and fill in holes. -*/ -if (need_find_last_extent) { -u64 len; - - ret =btrfs_prev_leaf (inode->root, src_path); -if (ret <0 ) -return ret; -if (ret) - goto fill_holes; -if (src_path->slots[0 ]) - src_path->slots[0 ]--; - src = src_path->nodes[0 ]; -btrfs_item_key_to_cpu (src, &key, src_path->slots[0 ]); -if (key.objectid !=btrfs_ino (inode) || - key.type !=BTRFS_EXTENT_DATA_KEY ) - goto fill_holes; - extent =btrfs_item_ptr (src, src_path->slots[0 ], - struct btrfs_file_extent_item); -if (btrfs_file_extent_type (src, extent) == -BTRFS_FILE_EXTENT_INLINE ) { - len =btrfs_file_extent_ram_bytes (src, extent); - *last_extent =ALIGN (key.offset + len, - fs_info->sectorsize); - }else { - len =btrfs_file_extent_num_bytes (src, extent); - *last_extent = key.offset + len; - } - } -fill_holes: -/* So we did prev_leaf, now we need to move to the next leaf, but a few -* things could have happened -* -* 1) A merge could have happened, so we could currently be on a leaf -* that holds what we were copying in the first place. -* 2) A split could have happened, and now not all of the items we want -* are on the same leaf. -* -* So we need to adjust how we search for holes, we need to drop the -* path and re-search for the first extent key we found, and then walk -* forward until we hit the last one we copied. -*/ -if (need_find_last_extent) { -/* btrfs_prev_leaf could return 1 without releasing the path */ -btrfs_release_path (src_path); - ret =btrfs_search_slot (NULL , inode->root, &first_key, - src_path,0 ,0 ); -if (ret <0 ) -return ret; -ASSERT (ret ==0 ); - src = src_path->nodes[0 ]; - i = src_path->slots[0 ]; - }else { - i = start_slot; - } - -/* -* Ok so here we need to go through and fill in any holes we may have -* to make sure that holes are punched for those areas in case they had -* extents previously. -*/ -while (!done) { -u64 offset, len; -u64 extent_end; - -if (i >=btrfs_header_nritems (src_path->nodes[0 ])) { - ret =btrfs_next_leaf (inode->root, src_path); -if (ret <0 ) -return ret; -ASSERT (ret ==0 ); - src = src_path->nodes[0 ]; - i =0 ; - need_find_last_extent = true; - } - -btrfs_item_key_to_cpu (src, &key, i); -if (!btrfs_comp_cpu_keys (&key, &last_key)) - done = true; -if (key.objectid !=btrfs_ino (inode) || - key.type !=BTRFS_EXTENT_DATA_KEY ) { - i++; -continue ; - } - extent = btrfs_item_ptr(src, i, struct btrfs_file_extent_item); -if (btrfs_file_extent_type (src, extent) == -BTRFS_FILE_EXTENT_INLINE ) { - len =btrfs_file_extent_ram_bytes (src, extent); - extent_end =ALIGN (key.offset + len, - fs_info->sectorsize); - }else { - len =btrfs_file_extent_num_bytes (src, extent); - extent_end = key.offset + len; - } - i++; - -if (*last_extent == key.offset) { - *last_extent = extent_end; -continue ; - } - offset = *last_extent; - len = key.offset - *last_extent; - ret =btrfs_insert_file_extent (trans, log,btrfs_ino (inode), - offset,0 ,0 , len,0 , len,0 ,0 ,0 ); -if (ret) -break ; - *last_extent = extent_end; - } - -/* -* Check if there is a hole between the last extent found in our leaf -* and the first extent in the next leaf. If there is one, we need to -* log an explicit hole so that at replay time we can punch the hole. -*/ -if (ret ==0 && - key.objectid ==btrfs_ino (inode) && - key.type ==BTRFS_EXTENT_DATA_KEY && - i ==btrfs_header_nritems (src_path->nodes[0 ])) { - ret =btrfs_next_leaf (inode->root, src_path); - need_find_last_extent = true; -if (ret >0 ) { - ret =0 ; - }else if (ret ==0 ) { -btrfs_item_key_to_cpu (src_path->nodes[0 ], &key, - src_path->slots[0 ]); -if (key.objectid ==btrfs_ino (inode) && - key.type ==BTRFS_EXTENT_DATA_KEY && - *last_extent < key.offset) { -const u64 len = key.offset - *last_extent; - - ret =btrfs_insert_file_extent (trans, log, -btrfs_ino (inode), - *last_extent,0 , -0 , len,0 , len, -0 ,0 ,0 ); - *last_extent += len; - } - } - } -/* -* Need to let the callers know we dropped the path so they should -* re-search. -*/ -if (!ret && need_find_last_extent) - ret =1 ;return ret; }
@@ -4338,7 +4154,7 @@ static int btrfs_log_prealloc_extents(struct btrfs_trans_handle *trans,const u64 i_size =i_size_read (&inode->vfs_inode);const u64 ino =btrfs_ino (inode);struct btrfs_path *dst_path =NULL ; -u64 last_extent = (u64)-1 ; +bool dropped_extents = false;int ins_nr =0 ;int start_slot;int ret;
@@ -4360,8 +4176,7 @@ static int btrfs_log_prealloc_extents(struct btrfs_trans_handle *trans,if (slot >=btrfs_header_nritems (leaf)) {if (ins_nr >0 ) { ret =copy_items (trans, inode, dst_path, path, - &last_extent, start_slot, - ins_nr,1 ,0 ); + start_slot, ins_nr,1 ,0 );if (ret <0 ) goto out; ins_nr =0 ;
@@ -4385,8 +4200,7 @@ static int btrfs_log_prealloc_extents(struct btrfs_trans_handle *trans, path->slots[0 ]++;continue ; } -if (last_extent == (u64)-1 ) { - last_extent = key.offset; +if (!dropped_extents) {/* * Avoid logging extent items logged in past fsync calls * and leading to duplicate keys in the log tree.
@@ -4400,6 +4214,7 @@ static int btrfs_log_prealloc_extents(struct btrfs_trans_handle *trans, }while (ret == -EAGAIN );if (ret) goto out; + dropped_extents = true; }if (ins_nr ==0 ) start_slot = slot;
@@ -4414,7 +4229,7 @@ static int btrfs_log_prealloc_extents(struct btrfs_trans_handle *trans, } }if (ins_nr >0 ) { - ret =copy_items (trans, inode, dst_path, path, &last_extent, + ret =copy_items (trans, inode, dst_path, path, start_slot, ins_nr,1 ,0 );if (ret >0 ) ret =0 ;
@@ -4608,13 +4423,8 @@ static int btrfs_log_all_xattrs(struct btrfs_trans_handle *trans,if (slot >= nritems) {if (ins_nr >0 ) { -u64 last_extent =0 ; - ret =copy_items (trans, inode, dst_path, path, - &last_extent, start_slot, - ins_nr,1 ,0 ); -/* can't be 1, extent items aren't processed */ -ASSERT (ret <=0 ); + start_slot, ins_nr,1 ,0 );if (ret <0 )return ret; ins_nr =0 ;
@@ -4638,13 +4448,8 @@ static int btrfs_log_all_xattrs(struct btrfs_trans_handle *trans,cond_resched (); }if (ins_nr >0 ) { -u64 last_extent =0 ; - ret =copy_items (trans, inode, dst_path, path, - &last_extent, start_slot, - ins_nr,1 ,0 ); -/* can't be 1, extent items aren't processed */ -ASSERT (ret <=0 ); + start_slot, ins_nr,1 ,0 );if (ret <0 )return ret; }
@@ -4653,100 +4458,119 @@ static int btrfs_log_all_xattrs(struct btrfs_trans_handle *trans, }/* -* If the no holes feature is enabled we need to make sure any hole between the -* last extent and the i_size of our inode is explicitly marked in the log. This -* is to make sure that doing something like: -* -* 1) create file with 128Kb of data -* 2) truncate file to 64Kb -* 3) truncate file to 256Kb -* 4) fsync file -* 5) <crash/power failure> -* 6) mount fs and trigger log replay -* -* Will give us a file with a size of 256Kb, the first 64Kb of data match what -* the file had in its first 64Kb of data at step 1 and the last 192Kb of the -* file correspond to a hole. The presence of explicit holes in a log tree is -* what guarantees that log replay will remove/adjust file extent items in the -* fs/subvol tree. -* -* Here we do not need to care about holes between extents, that is already done -* by copy_items(). We also only need to do this in the full sync path, where we -* lookup for extents from the fs/subvol tree only. In the fast path case, we -* lookup the list of modified extent maps and if any represents a hole, we -* insert a corresponding extent representing a hole in the log tree. +* When using the NO_HOLES feature if we punched a hole that causes the +* deletion of entire leafs or all the extent items of the first leaf (the one +* that contains the inode item and references) we may end up not processing +* any extents, because there are no leafs with a generation matching the +* current transaction that have extent items for our inode. So we need to find +* if any holes exist and then log them. We also need to log holes after any +* truncate operation that changes the inode's size. */ -static int btrfs_log_trailing_hole (struct btrfs_trans_handle *trans, -struct btrfs_root *root, -struct btrfs_inode *inode, -struct btrfs_path *path) +static int btrfs_log_holes (struct btrfs_trans_handle *trans, +struct btrfs_root *root, +struct btrfs_inode *inode, +struct btrfs_path *path) {struct btrfs_fs_info *fs_info = root->fs_info; -int ret;struct btrfs_key key; -u64 hole_start; -u64 hole_size; -struct extent_buffer *leaf; -struct btrfs_root *log = root->log_root;const u64 ino =btrfs_ino (inode);const u64 i_size =i_size_read (&inode->vfs_inode); +u64 prev_extent_end =0 ; +int ret; -if (!btrfs_fs_incompat (fs_info,NO_HOLES )) +if (!btrfs_fs_incompat (fs_info,NO_HOLES ) || i_size ==0 )return 0 ; key.objectid = ino; key.type =BTRFS_EXTENT_DATA_KEY ; - key.offset = (u64)-1 ; + key.offset =0 ; ret =btrfs_search_slot (NULL , root, &key, path,0 ,0 ); -ASSERT (ret !=0 );if (ret <0 )return ret; -ASSERT (path->slots[0 ] >0 ); - path->slots[0 ]--; - leaf = path->nodes[0 ]; -btrfs_item_key_to_cpu (leaf, &key, path->slots[0 ]); - -if (key.objectid != ino || key.type !=BTRFS_EXTENT_DATA_KEY ) { -/* inode does not have any extents */ - hole_start =0 ; - hole_size = i_size; - }else { +while (true) {struct btrfs_file_extent_item *extent; +struct extent_buffer *leaf = path->nodes[0 ];u64 len; -/* -* If there's an extent beyond i_size, an explicit hole was -* already inserted by copy_items(). -*/ -if (key.offset >= i_size) -return 0 ; +if (path->slots[0 ] >=btrfs_header_nritems (path->nodes[0 ])) { + ret =btrfs_next_leaf (root, path); +if (ret <0 ) +return ret; +if (ret >0 ) { + ret =0 ; +break ; + } + leaf = path->nodes[0 ]; + } + +btrfs_item_key_to_cpu (leaf, &key, path->slots[0 ]); +if (key.objectid != ino || key.type !=BTRFS_EXTENT_DATA_KEY ) +break ; + +/* We have a hole, log it. */ +if (prev_extent_end < key.offset) { +const u64 hole_len = key.offset - prev_extent_end; + +/* +* Release the path to avoid deadlocks with other code +* paths that search the root while holding locks on +* leafs from the log root. +*/ +btrfs_release_path (path); + ret =btrfs_insert_file_extent (trans, root->log_root, + ino, prev_extent_end,0 , +0 , hole_len,0 , hole_len, +0 ,0 ,0 ); +if (ret <0 ) +return ret; + +/* +* Search for the same key again in the root. Since it's +* an extent item and we are holding the inode lock, the +* key must still exist. If it doesn't just emit warning +* and return an error to fall back to a transaction +* commit. +*/ + ret =btrfs_search_slot (NULL , root, &key, path,0 ,0 ); +if (ret <0 ) +return ret; +if (WARN_ON (ret >0 )) +return -ENOENT ; + leaf = path->nodes[0 ]; + } extent =btrfs_item_ptr (leaf, path->slots[0 ], struct btrfs_file_extent_item); -if (btrfs_file_extent_type (leaf, extent) == -BTRFS_FILE_EXTENT_INLINE ) -return 0 ; +BTRFS_FILE_EXTENT_INLINE ) { + len =btrfs_file_extent_ram_bytes (leaf, extent); + prev_extent_end =ALIGN (key.offset + len, + fs_info->sectorsize); + }else { + len =btrfs_file_extent_num_bytes (leaf, extent); + prev_extent_end = key.offset + len; + } - len =btrfs_file_extent_num_bytes (leaf, extent); -/* Last extent goes beyond i_size, no need to log a hole. */ -if (key.offset + len > i_size) -return 0 ; - hole_start = key.offset + len; - hole_size = i_size - hole_start; + path->slots[0 ]++; +cond_resched (); } -btrfs_release_path (path); -/* Last extent ends at i_size. */ -if (hole_size ==0 ) -return 0 ; +if (prev_extent_end < i_size) { +u64 hole_len; - hole_size =ALIGN (hole_size, fs_info->sectorsize); - ret =btrfs_insert_file_extent (trans, log, ino, hole_start,0 ,0 , - hole_size,0 , hole_size,0 ,0 ,0 ); -return ret; +btrfs_release_path (path); + hole_len =ALIGN (i_size - prev_extent_end, fs_info->sectorsize); + ret =btrfs_insert_file_extent (trans, root->log_root, + ino, prev_extent_end,0 ,0 , + hole_len,0 , hole_len, +0 ,0 ,0 ); +if (ret <0 ) +return ret; + } + +return 0 ; }/*
@@ -4914,7 +4738,6 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,struct btrfs_root *log = root->log_root;struct extent_buffer *src =NULL ;LIST_HEAD (logged_list); -u64 last_extent =0 ;int err =0 ;int ret;int nritems;
@@ -5088,7 +4911,7 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, ins_start_slot = path->slots[0 ]; } ret =copy_items (trans, inode, dst_path, path, - &last_extent, ins_start_slot, + ins_start_slot, ins_nr, inode_only, logged_isize);if (ret <0 ) {
@@ -5142,17 +4965,13 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,if (ins_nr ==0 ) goto next_slot; ret =copy_items (trans, inode, dst_path, path, - &last_extent, ins_start_slot, + ins_start_slot, ins_nr, inode_only, logged_isize);if (ret <0 ) { err = ret; goto out_unlock; } ins_nr =0 ; -if (ret) { -btrfs_release_path (path); -continue ; - } goto next_slot; }
@@ -5166,18 +4985,13 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, goto next_slot; } - ret =copy_items (trans, inode, dst_path, path, &last_extent, + ret =copy_items (trans, inode, dst_path, path, ins_start_slot, ins_nr, inode_only, logged_isize);if (ret <0 ) { err = ret; goto out_unlock; } -if (ret) { - ins_nr =0 ; -btrfs_release_path (path); -continue ; - } ins_nr =1 ; ins_start_slot = path->slots[0 ]; next_slot:
@@ -5191,13 +5005,12 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, }if (ins_nr) { ret =copy_items (trans, inode, dst_path, path, - &last_extent, ins_start_slot, + ins_start_slot, ins_nr, inode_only, logged_isize);if (ret <0 ) { err = ret; goto out_unlock; } - ret =0 ; ins_nr =0 ; }btrfs_release_path (path);
@@ -5212,14 +5025,13 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, } }if (ins_nr) { - ret =copy_items (trans, inode, dst_path, path, &last_extent, + ret =copy_items (trans, inode, dst_path, path, ins_start_slot, ins_nr, inode_only, logged_isize);if (ret <0 ) { err = ret; goto out_unlock; } - ret =0 ; ins_nr =0 ; }
@@ -5232,7 +5044,7 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,if (max_key.type >=BTRFS_EXTENT_DATA_KEY && !fast_search) {btrfs_release_path (path);btrfs_release_path (dst_path); - err =btrfs_log_trailing_hole (trans, root, inode, path); + err =btrfs_log_holes (trans, root, inode, path);if (err) goto out_unlock; }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 122/173 btrfs: use bool argument in free_root_pointers()
In reply to Greg Kroah-HartmanFrom: Anand Jain <[email protected]>
[ Upstream commit 4273eaff9b8d5e141113a5bdf9628c02acf3afe5 ]
We don't need int argument bool shall do in free_root_pointers(). And rename the argument as it confused two people.
--- fs/btrfs/disk-io.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)
fs/btrfs/disk-io.c +5 -5
@@ -2051,7 +2051,7 @@ static void free_root_extent_buffers(struct btrfs_root *root) }/* helper to cleanup tree roots */ -static void free_root_pointers (struct btrfs_fs_info *info,int chunk_root) +static void free_root_pointers (struct btrfs_fs_info *info,bool free_chunk_root) {free_root_extent_buffers (info->tree_root);
@@ -2060,7 +2060,7 @@ static void free_root_pointers(struct btrfs_fs_info *info, int chunk_root)free_root_extent_buffers (info->csum_root);free_root_extent_buffers (info->quota_root);free_root_extent_buffers (info->uuid_root); -if (chunk_root) +if (free_chunk_root)free_root_extent_buffers (info->chunk_root);free_root_extent_buffers (info->free_space_root); }
@@ -3069,7 +3069,7 @@ int open_ctree(struct super_block *sb,btrfs_put_block_group_cache (fs_info); fail_tree_roots: -free_root_pointers (fs_info,1 ); +free_root_pointers (fs_info, true);invalidate_inode_pages2 (fs_info->btree_inode->i_mapping); fail_sb_buffer:
@@ -3097,7 +3097,7 @@ int open_ctree(struct super_block *sb,if (!btrfs_test_opt (fs_info,USEBACKUPROOT )) goto fail_tree_roots; -free_root_pointers (fs_info,0 ); +free_root_pointers (fs_info, false);/* don't use the log in recovery mode, it won't be valid */ btrfs_set_super_log_root (disk_super,0 );
@@ -3764,7 +3764,7 @@ void close_ctree(struct btrfs_fs_info *fs_info)btrfs_free_block_groups (fs_info);clear_bit (BTRFS_FS_OPEN , &fs_info->flags); -free_root_pointers (fs_info,1 ); +free_root_pointers (fs_info, true);iput (fs_info->btree_inode);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 123/173 btrfs: free block groups after freeing fs trees
In reply to Greg Kroah-HartmanFrom: Josef Bacik <[email protected]>
[ Upstream commit 4e19443da1941050b346f8fc4c368aa68413bc88 ]
Sometimes when running generic/475 we would trip the WARN_ON(cache->reserved) check when free'ing the block groups on umount. This is because sometimes we don't commit the transaction because of IO errors and thus do not cleanup the tree logs until at umount time.
These blocks are still reserved until they are cleaned up, but they aren't cleaned up until _after_ we do the free block groups work. Fix this by moving the free after free'ing the fs roots, that way all of the tree logs are cleaned up and we have a properly cleaned fs. A bunch of loops of generic/475 confirmed this fixes the problem.
--- fs/btrfs/disk-io.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-)
fs/btrfs/disk-io.c +9 -2
@@ -3761,11 +3761,18 @@ void close_ctree(struct btrfs_fs_info *fs_info)invalidate_inode_pages2 (fs_info->btree_inode->i_mapping);btrfs_stop_all_workers (fs_info); -btrfs_free_block_groups (fs_info); -clear_bit (BTRFS_FS_OPEN , &fs_info->flags);free_root_pointers (fs_info, true); +/* +* We must free the block groups after dropping the fs_roots as we could +* have had an IO error and have left over tree log blocks that aren't +* cleaned up until the fs roots are freed. This makes the block group +* accounting appear to be wrong because there's pending reserved bytes, +* so make sure we do the block group cleanup afterwards. +*/ +btrfs_free_block_groups (fs_info); +iput (fs_info->btree_inode);#ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 124/173 btrfs: remove trivial locking wrappers of tree mod log
In reply to Greg Kroah-HartmanFrom: David Sterba <[email protected]>
[ Upstream commit b1a09f1ec540408abf3a50d15dff5d9506932693 ]
The wrappers are trivial and do not bring any extra value on top of the plain locking primitives.
--- fs/btrfs/ctree.c | 58 ++++++++++++++++-------------------------------- 1 file changed, 19 insertions(+), 39 deletions(-)
fs/btrfs/ctree.c +19 -39
@@ -334,26 +334,6 @@ struct tree_mod_elem {struct tree_mod_root old_root; }; -static inline void tree_mod_log_read_lock (struct btrfs_fs_info *fs_info) -{ -read_lock (&fs_info->tree_mod_log_lock); -} - -static inline void tree_mod_log_read_unlock (struct btrfs_fs_info *fs_info) -{ -read_unlock (&fs_info->tree_mod_log_lock); -} - -static inline void tree_mod_log_write_lock (struct btrfs_fs_info *fs_info) -{ -write_lock (&fs_info->tree_mod_log_lock); -} - -static inline void tree_mod_log_write_unlock (struct btrfs_fs_info *fs_info) -{ -write_unlock (&fs_info->tree_mod_log_lock); -} -/* * Pull a new tree mod seq number for our operation. */
@@ -373,14 +353,14 @@ static inline u64 btrfs_inc_tree_mod_seq(struct btrfs_fs_info *fs_info)u64 btrfs_get_tree_mod_seq (struct btrfs_fs_info *fs_info,struct seq_list *elem) { -tree_mod_log_write_lock (fs_info); +write_lock (&fs_info->tree_mod_log_lock);spin_lock (&fs_info->tree_mod_seq_lock);if (!elem->seq) { elem->seq =btrfs_inc_tree_mod_seq (fs_info);list_add_tail (&elem->list, &fs_info->tree_mod_seq_list); }spin_unlock (&fs_info->tree_mod_seq_lock); -tree_mod_log_write_unlock (fs_info); +write_unlock (&fs_info->tree_mod_log_lock);return elem->seq; }
@@ -422,7 +402,7 @@ void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info,* anything that's lower than the lowest existing (read: blocked) * sequence number can be removed from the tree. */ -tree_mod_log_write_lock (fs_info); +write_lock (&fs_info->tree_mod_log_lock); tm_root = &fs_info->tree_mod_log;for (node =rb_first (tm_root); node; node = next) { next =rb_next (node);
@@ -432,7 +412,7 @@ void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info,rb_erase (node, tm_root);kfree (tm); } -tree_mod_log_write_unlock (fs_info); +write_unlock (&fs_info->tree_mod_log_lock); }/*
@@ -443,7 +423,7 @@ void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info,* for root replace operations, or the logical address of the affected * block for all other operations. * -* Note: must be called with write lock (tree_mod_log_write_lock). +* Note: must be called with write lock for fs_info::tree_mod_log_lock. */ static noinline int__tree_mod_log_insert (struct btrfs_fs_info *fs_info,struct tree_mod_elem *tm)
@@ -481,7 +461,7 @@ __tree_mod_log_insert(struct btrfs_fs_info *fs_info, struct tree_mod_elem *tm)* Determines if logging can be omitted. Returns 1 if it can. Otherwise, it * returns zero with the tree_mod_log_lock acquired. The caller must hold * this until all tree mod log insertions are recorded in the rb tree and then -* call tree_mod_log_write_unlock() to release. +* write unlock fs_info::tree_mod_log_lock. */ static inline int tree_mod_dont_log (struct btrfs_fs_info *fs_info,struct extent_buffer *eb) {
@@ -491,9 +471,9 @@ static inline int tree_mod_dont_log(struct btrfs_fs_info *fs_info,if (eb &&btrfs_header_level (eb) ==0 )return 1 ; -tree_mod_log_write_lock (fs_info); +write_lock (&fs_info->tree_mod_log_lock);if (list_empty (&(fs_info)->tree_mod_seq_list)) { -tree_mod_log_write_unlock (fs_info); +write_unlock (&fs_info->tree_mod_log_lock);return 1 ; }
@@ -557,7 +537,7 @@ tree_mod_log_insert_key(struct btrfs_fs_info *fs_info, } ret =__tree_mod_log_insert (fs_info, tm); -tree_mod_log_write_unlock (fs_info); +write_unlock (&eb->fs_info->tree_mod_log_lock);if (ret)kfree (tm);
@@ -621,7 +601,7 @@ tree_mod_log_insert_move(struct btrfs_fs_info *fs_info, ret =__tree_mod_log_insert (fs_info, tm);if (ret) goto free_tms; -tree_mod_log_write_unlock (fs_info); +write_unlock (&eb->fs_info->tree_mod_log_lock);kfree (tm_list);return 0 ;
@@ -632,7 +612,7 @@ tree_mod_log_insert_move(struct btrfs_fs_info *fs_info,kfree (tm_list[i]); }if (locked) -tree_mod_log_write_unlock (fs_info); +write_unlock (&eb->fs_info->tree_mod_log_lock);kfree (tm_list);kfree (tm);
@@ -713,7 +693,7 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info,if (!ret) ret =__tree_mod_log_insert (fs_info, tm); -tree_mod_log_write_unlock (fs_info); +write_unlock (&fs_info->tree_mod_log_lock);if (ret) goto free_tms;kfree (tm_list);
@@ -740,7 +720,7 @@ __tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq,struct tree_mod_elem *cur =NULL ;struct tree_mod_elem *found =NULL ; -tree_mod_log_read_lock (fs_info); +read_lock (&fs_info->tree_mod_log_lock); tm_root = &fs_info->tree_mod_log; node = tm_root->rb_node;while (node) {
@@ -768,7 +748,7 @@ __tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq,break ; } } -tree_mod_log_read_unlock (fs_info); +read_unlock (&fs_info->tree_mod_log_lock);return found; }
@@ -849,7 +829,7 @@ tree_mod_log_eb_copy(struct btrfs_fs_info *fs_info, struct extent_buffer *dst, goto free_tms; } -tree_mod_log_write_unlock (fs_info); +write_unlock (&fs_info->tree_mod_log_lock);kfree (tm_list);return 0 ;
@@ -861,7 +841,7 @@ tree_mod_log_eb_copy(struct btrfs_fs_info *fs_info, struct extent_buffer *dst,kfree (tm_list[i]); }if (locked) -tree_mod_log_write_unlock (fs_info); +write_unlock (&fs_info->tree_mod_log_lock);kfree (tm_list);return ret;
@@ -921,7 +901,7 @@ tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, struct extent_buffer *eb) goto free_tms; ret =__tree_mod_log_free_eb (fs_info, tm_list, nritems); -tree_mod_log_write_unlock (fs_info); +write_unlock (&eb->fs_info->tree_mod_log_lock);if (ret) goto free_tms;kfree (tm_list);
@@ -1279,7 +1259,7 @@ __tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,unsigned long p_size =sizeof (struct btrfs_key_ptr ); n =btrfs_header_nritems (eb); -tree_mod_log_read_lock (fs_info); +read_lock (&fs_info->tree_mod_log_lock);while (tm && tm->seq >= time_seq) {/* * all the operations are recorded with the operator used for
@@ -1334,7 +1314,7 @@ __tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,if (tm->logical != first_tm->logical)break ; } -tree_mod_log_read_unlock (fs_info); +read_unlock (&fs_info->tree_mod_log_lock);btrfs_set_header_nritems (eb, n); }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 125/173 Btrfs: fix race between adding and putting tree mod seq elements and nodes
In reply to Greg Kroah-HartmanFrom: Filipe Manana <[email protected]>
[ Upstream commit 7227ff4de55d931bbdc156c8ef0ce4f100c78a5b ]
There is a race between adding and removing elements to the tree mod log list and rbtree that can lead to use-after-free problems.
Consider the following example that explains how/why the problems happens:
1) Task A has mod log element with sequence number 200. It currently is the only element in the mod log list;
2) Task A calls btrfs_put_tree_mod_seq() because it no longer needs to access the tree mod log. When it enters the function, it initializes 'min_seq' to (u64)-1. Then it acquires the lock 'tree_mod_seq_lock' before checking if there are other elements in the mod seq list. Since the list it empty, 'min_seq' remains set to (u64)-1. Then it unlocks the lock 'tree_mod_seq_lock';
3) Before task A acquires the lock 'tree_mod_log_lock', task B adds itself to the mod seq list through btrfs_get_tree_mod_seq() and gets a sequence number of 201;
4) Some other task, name it task C, modifies a btree and because there elements in the mod seq list, it adds a tree mod elem to the tree mod log rbtree. That node added to the mod log rbtree is assigned a sequence number of 202;
5) Task B, which is doing fiemap and resolving indirect back references, calls btrfs get_old_root(), with 'time_seq' == 201, which in turn calls tree_mod_log_search() - the search returns the mod log node from the rbtree with sequence number 202, created by task C;
6) Task A now acquires the lock 'tree_mod_log_lock', starts iterating the mod log rbtree and finds the node with sequence number 202. Since 202 is less than the previously computed 'min_seq', (u64)-1, it removes the node and frees it;
7) Task B still has a pointer to the node with sequence number 202, and it dereferences the pointer itself and through the call to __tree_mod_log_rewind(), resulting in a use-after-free problem.
This issue can be triggered sporadically with the test case generic/561 from fstests, and it happens more frequently with a higher number of duperemove processes. When it happens to me, it either freezes the VM or it produces a trace like the following before crashing:
[ 1245.321140] general protection fault: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC PTI [ 1245.321200] CPU: 1 PID: 26997 Comm: pool Not tainted 5.5.0-rc6-btrfs-next-52 #1 [ 1245.321235] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.0-0-ga698c8995f-prebuilt.qemu.org 04/01/2014 [ 1245.321287] RIP: 0010:rb_next+0x16/0x50 [ 1245.321307] Code: .... [ 1245.321372] RSP: 0018:ffffa151c4d039b0 EFLAGS: 00010202 [ 1245.321388] RAX: 6b6b6b6b6b6b6b6b RBX: ffff8ae221363c80 RCX: 6b6b6b6b6b6b6b6b [ 1245.321409] RDX: 0000000000000001 RSI: 0000000000000000 RDI: ffff8ae221363c80 [ 1245.321439] RBP: ffff8ae20fcc4688 R08: 0000000000000002 R09: 0000000000000000 [ 1245.321475] R10: ffff8ae20b120910 R11: 00000000243f8bb1 R12: 0000000000000038 [ 1245.321506] R13: ffff8ae221363c80 R14: 000000000000075f R15: ffff8ae223f762b8 [ 1245.321539] FS: 00007fdee1ec7700(0000) GS:ffff8ae236c80000(0000) knlGS:0000000000000000 [ 1245.321591] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 1245.321614] CR2: 00007fded4030c48 CR3: 000000021da16003 CR4: 00000000003606e0 [ 1245.321642] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 1245.321668] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 1245.321706] Call Trace: [ 1245.321798] __tree_mod_log_rewind+0xbf/0x280 [btrfs] [ 1245.321841] btrfs_search_old_slot+0x105/0xd00 [btrfs] [ 1245.321877] resolve_indirect_refs+0x1eb/0xc60 [btrfs] [ 1245.321912] find_parent_nodes+0x3dc/0x11b0 [btrfs] [ 1245.321947] btrfs_check_shared+0x115/0x1c0 [btrfs] [ 1245.321980] ? extent_fiemap+0x59d/0x6d0 [btrfs] [ 1245.322029] extent_fiemap+0x59d/0x6d0 [btrfs] [ 1245.322066] do_vfs_ioctl+0x45a/0x750 [ 1245.322081] ksys_ioctl+0x70/0x80 [ 1245.322092] ? trace_hardirqs_off_thunk+0x1a/0x1c [ 1245.322113] __x64_sys_ioctl+0x16/0x20 [ 1245.322126] do_syscall_64+0x5c/0x280 [ 1245.322139] entry_SYSCALL_64_after_hwframe+0x49/0xbe [ 1245.322155] RIP: 0033:0x7fdee3942dd7 [ 1245.322177] Code: .... [ 1245.322258] RSP: 002b:00007fdee1ec6c88 EFLAGS: 00000246 ORIG_RAX: 0000000000000010 [ 1245.322294] RAX: ffffffffffffffda RBX: 00007fded40210d8 RCX: 00007fdee3942dd7 [ 1245.322314] RDX: 00007fded40210d8 RSI: 00000000c020660b RDI: 0000000000000004 [ 1245.322337] RBP: 0000562aa89e7510 R08: 0000000000000000 R09: 00007fdee1ec6d44 [ 1245.322369] R10: 0000000000000073 R11: 0000000000000246 R12: 00007fdee1ec6d48 [ 1245.322390] R13: 00007fdee1ec6d40 R14: 00007fded40210d0 R15: 00007fdee1ec6d50 [ 1245.322423] Modules linked in: .... [ 1245.323443] ---[ end trace 01de1e9ec5dff3cd ]---
Fix this by ensuring that btrfs_put_tree_mod_seq() computes the minimum sequence number and iterates the rbtree while holding the lock 'tree_mod_log_lock' in write mode. Also get rid of the 'tree_mod_seq_lock' lock, since it is now redundant.
--- fs/btrfs/ctree.c | 8 ++------ fs/btrfs/ctree.h | 6 ++---- fs/btrfs/delayed-ref.c | 8 ++++---- fs/btrfs/disk-io.c | 1 - fs/btrfs/tests/btrfs-tests.c | 1 - 5 files changed, 8 insertions(+), 16 deletions(-)
fs/btrfs/ctree.c +2 -6
@@ -354,12 +354,10 @@ u64 btrfs_get_tree_mod_seq(struct btrfs_fs_info *fs_info, struct seq_list *elem) {write_lock (&fs_info->tree_mod_log_lock); -spin_lock (&fs_info->tree_mod_seq_lock);if (!elem->seq) { elem->seq =btrfs_inc_tree_mod_seq (fs_info);list_add_tail (&elem->list, &fs_info->tree_mod_seq_list); } -spin_unlock (&fs_info->tree_mod_seq_lock);write_unlock (&fs_info->tree_mod_log_lock);return elem->seq;
@@ -379,7 +377,7 @@ void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info,if (!seq_putting)return ; -spin_lock (&fs_info->tree_mod_seq_lock); +write_lock (&fs_info->tree_mod_log_lock);list_del (&elem->list); elem->seq =0 ;
@@ -390,19 +388,17 @@ void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info,* blocker with lower sequence number exists, we * cannot remove anything from the log */ -spin_unlock (&fs_info->tree_mod_seq_lock); +write_unlock (&fs_info->tree_mod_log_lock);return ; } min_seq = cur_elem->seq; } } -spin_unlock (&fs_info->tree_mod_seq_lock);/* * anything that's lower than the lowest existing (read: blocked) * sequence number can be removed from the tree. */ -write_lock (&fs_info->tree_mod_log_lock); tm_root = &fs_info->tree_mod_log;for (node =rb_first (tm_root); node; node = next) { next =rb_next (node);
fs/btrfs/ctree.h +2 -4
@@ -869,14 +869,12 @@ struct btrfs_fs_info {struct list_head delayed_iputs;struct mutex cleaner_delayed_iput_mutex; -/* this protects tree_mod_seq_list */ -spinlock_t tree_mod_seq_lock;atomic64_t tree_mod_seq; -struct list_head tree_mod_seq_list; -/* this protects tree_mod_log */ +/* this protects tree_mod_log and tree_mod_seq_list */ rwlock_t tree_mod_log_lock;struct rb_root tree_mod_log; +struct list_head tree_mod_seq_list;atomic_t nr_async_submits;atomic_t async_submit_draining;
fs/btrfs/delayed-ref.c +4 -4
@@ -281,7 +281,7 @@ void btrfs_merge_delayed_refs(struct btrfs_trans_handle *trans,if (head->is_data)return ; -spin_lock (&fs_info->tree_mod_seq_lock); +read_lock (&fs_info->tree_mod_log_lock);if (!list_empty (&fs_info->tree_mod_seq_list)) {struct seq_list *elem;
@@ -289,7 +289,7 @@ void btrfs_merge_delayed_refs(struct btrfs_trans_handle *trans, struct seq_list, list); seq = elem->seq; } -spin_unlock (&fs_info->tree_mod_seq_lock); +read_unlock (&fs_info->tree_mod_log_lock); ref = list_first_entry(&head->ref_list, struct btrfs_delayed_ref_node, list);
@@ -317,7 +317,7 @@ int btrfs_check_delayed_seq(struct btrfs_fs_info *fs_info,struct seq_list *elem;int ret =0 ; -spin_lock (&fs_info->tree_mod_seq_lock); +read_lock (&fs_info->tree_mod_log_lock);if (!list_empty (&fs_info->tree_mod_seq_list)) { elem =list_first_entry (&fs_info->tree_mod_seq_list, struct seq_list, list);
@@ -331,7 +331,7 @@ int btrfs_check_delayed_seq(struct btrfs_fs_info *fs_info, } } -spin_unlock (&fs_info->tree_mod_seq_lock); +read_unlock (&fs_info->tree_mod_log_lock);return ret; }
fs/btrfs/disk-io.c +0 -1
@@ -2455,7 +2455,6 @@ int open_ctree(struct super_block *sb,spin_lock_init (&fs_info->fs_roots_radix_lock);spin_lock_init (&fs_info->delayed_iput_lock);spin_lock_init (&fs_info->defrag_inodes_lock); -spin_lock_init (&fs_info->tree_mod_seq_lock);spin_lock_init (&fs_info->super_lock);spin_lock_init (&fs_info->qgroup_op_lock);spin_lock_init (&fs_info->buffer_lock);
fs/btrfs/tests/btrfs-tests.c +0 -1
@@ -115,7 +115,6 @@ struct btrfs_fs_info *btrfs_alloc_dummy_fs_info(u32 nodesize, u32 sectorsize)spin_lock_init (&fs_info->qgroup_op_lock);spin_lock_init (&fs_info->super_lock);spin_lock_init (&fs_info->fs_roots_radix_lock); -spin_lock_init (&fs_info->tree_mod_seq_lock);mutex_init (&fs_info->qgroup_ioctl_lock);mutex_init (&fs_info->qgroup_rescan_lock);rwlock_init (&fs_info->tree_mod_log_lock);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 126/173 drm: atmel-hlcdc: enable clock before configuring timing engine
In reply to Greg Kroah-HartmanFrom: Claudiu Beznea <[email protected]>
[ Upstream commit 2c1fb9d86f6820abbfaa38a6836157c76ccb4e7b ]
Changing pixel clock source without having this clock source enabled will block the timing engine and the next operations after (in this case setting ATMEL_HLCDC_CFG(5) settings in atmel_hlcdc_crtc_mode_set_nofb() will fail). It is recomended (although in datasheet this is not present) to actually enabled pixel clock source before doing any changes on timing enginge (only SAM9X60 datasheet specifies that the peripheral clock and pixel clock must be enabled before using LCD controller).
--- drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_crtc.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)
drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_crtc.c +7 -1
@@ -79,7 +79,11 @@ static void atmel_hlcdc_crtc_mode_set_nofb(struct drm_crtc *c)struct videomode vm;unsigned long prate;unsigned int cfg; -int div; +int div, ret; + + ret =clk_prepare_enable (crtc->dc->hlcdc->sys_clk); +if (ret) +return ; vm.vfront_porch = adj->crtc_vsync_start - adj->crtc_vdisplay; vm.vback_porch = adj->crtc_vtotal - adj->crtc_vsync_end;
@@ -138,6 +142,8 @@ static void atmel_hlcdc_crtc_mode_set_nofb(struct drm_crtc *c)ATMEL_HLCDC_VSPSU |ATMEL_HLCDC_VSPHO |ATMEL_HLCDC_GUARDTIME_MASK |ATMEL_HLCDC_MODE_MASK , cfg); + +clk_disable_unprepare (crtc->dc->hlcdc->sys_clk); } static enum drm_mode_status
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 127/173 drm/dp_mst: Remove VCPI while disabling topology mgr
In reply to Greg Kroah-HartmanFrom: Wayne Lin <[email protected]>
[ Upstream commit 64e62bdf04ab8529f45ed0a85122c703035dec3a ]
[Why]
This patch is trying to address the issue observed when hotplug DP daisy chain monitors.
e.g.
src-mstb-mstb-sst -> src (unplug) mstb-mstb-sst -> src-mstb-mstb-sst (plug in again)
Once unplug a DP MST capable device, driver will call drm_dp_mst_topology_mgr_set_mst() to disable MST. In this function, it cleans data of topology manager while disabling mst_state. However, it doesn't clean up the proposed_vcpis of topology manager. If proposed_vcpi is not reset, once plug in MST daisy chain monitors later, code will fail at checking port validation while trying to allocate payloads.
When MST capable device is plugged in again and try to allocate payloads by calling drm_dp_update_payload_part1(), this function will iterate over all proposed virtual channels to see if any proposed VCPI's num_slots is greater than 0. If any proposed VCPI's num_slots is greater than 0 and the port which the specific virtual channel directed to is not in the topology, code then fails at the port validation. Since there are stale VCPI allocations from the previous topology enablement in proposed_vcpi[], code will fail at port validation and reurn EINVAL.
[How]
Clean up the data of stale proposed_vcpi[] and reset mgr->proposed_vcpis to NULL while disabling mst in drm_dp_mst_topology_mgr_set_mst().
Changes since v1:
*Add on more details in commit message to describe the issue which the patch is trying to fix
--- drivers/gpu/drm/drm_dp_mst_topology.c | 12 ++++++++++++ 1 file changed, 12 insertions(+)
drivers/gpu/drm/drm_dp_mst_topology.c +12 -0
@@ -2046,6 +2046,7 @@ static bool drm_dp_get_vc_payload_bw(int dp_link_bw,int drm_dp_mst_topology_mgr_set_mst (struct drm_dp_mst_topology_mgr *mgr,bool mst_state) {int ret =0 ; +int i =0 ;struct drm_dp_mst_branch *mstb =NULL ;mutex_lock (&mgr->lock);
@@ -2106,10 +2107,21 @@ int drm_dp_mst_topology_mgr_set_mst(struct drm_dp_mst_topology_mgr *mgr, bool ms/* this can fail if the device is gone */ drm_dp_dpcd_writeb (mgr->aux,DP_MSTM_CTRL ,0 ); ret =0 ; +mutex_lock (&mgr->payload_lock);memset (mgr->payloads,0 , mgr->max_payloads *sizeof (struct drm_dp_payload )); mgr->payload_mask =0 ;set_bit (0 , &mgr->payload_mask); +for (i =0 ; i < mgr->max_payloads; i++) { +struct drm_dp_vcpi *vcpi = mgr->proposed_vcpis[i]; + +if (vcpi) { + vcpi->vcpi =0 ; + vcpi->num_slots =0 ; + } + mgr->proposed_vcpis[i] =NULL ; + } mgr->vcpi_mask =0 ; +mutex_unlock (&mgr->payload_lock); } out_unlock:
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 128/173 KVM: x86: Protect pmu_intel.c from Spectre-v1/L1TF attacks
In reply to Greg Kroah-HartmanFrom: Marios Pomonis <[email protected]>
[ Upstream commit 66061740f1a487f4ed54fde75e724709f805da53 ]
This fixes Spectre-v1/L1TF vulnerabilities in intel_find_fixed_event() and intel_rdpmc_ecx_to_pmc().kvm_rdpmc() (ancestor of intel_find_fixed_event()) and reprogram_fixed_counter() (ancestor of intel_rdpmc_ecx_to_pmc()) are exported symbols so KVM should treat them conservatively from a security perspective.
--- arch/x86/kvm/pmu_intel.c | 24 ++++++++++++++++-------- 1 file changed, 16 insertions(+), 8 deletions(-)
arch/x86/kvm/pmu_intel.c +16 -8
@@ -87,10 +87,14 @@ static unsigned intel_find_arch_event(struct kvm_pmu *pmu,static unsigned intel_find_fixed_event (int idx) { -if (idx >=ARRAY_SIZE (fixed_pmc_events)) +u32 event; +size_t size =ARRAY_SIZE (fixed_pmc_events); + +if (idx >= size)return PERF_COUNT_HW_MAX ; -return intel_arch_events[fixed_pmc_events[idx]].event_type; + event = fixed_pmc_events[array_index_nospec (idx, size)]; +return intel_arch_events[event].event_type; }/* check if a PMC is enabled by comparing it with globl_ctrl bits. */
@@ -131,15 +135,19 @@ static struct kvm_pmc *intel_msr_idx_to_pmc(struct kvm_vcpu *vcpu,struct kvm_pmu *pmu =vcpu_to_pmu (vcpu);bool fixed = idx & (1u <<30 );struct kvm_pmc *counters; +unsigned int num_counters; idx &= ~(3u <<30 ); -if (!fixed && idx >= pmu->nr_arch_gp_counters) -return NULL ; -if (fixed && idx >= pmu->nr_arch_fixed_counters) +if (fixed) { + counters = pmu->fixed_counters; + num_counters = pmu->nr_arch_fixed_counters; + }else { + counters = pmu->gp_counters; + num_counters = pmu->nr_arch_gp_counters; + } +if (idx >= num_counters)return NULL ; - counters = fixed ? pmu->fixed_counters : pmu->gp_counters; - -return &counters[idx]; +return &counters[array_index_nospec (idx, num_counters)]; }static bool intel_is_valid_msr (struct kvm_vcpu *vcpu,u32 msr)
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 129/173 btrfs: flush write bio if we loop in extent_write_cache_pages
In reply to Greg Kroah-HartmanFrom: Josef Bacik <[email protected]>
[ Upstream commit 96bf313ecb33567af4cb53928b0c951254a02759 ]
There exists a deadlock with range_cyclic that has existed forever. If we loop around with a bio already built we could deadlock with a writer who has the page locked that we're attempting to write but is waiting on a page in our bio to be written out. The task traces are as follows
PID: 1329874 TASK: ffff889ebcdf3800 CPU: 33 COMMAND: "kworker/u113:5"
#0 [ffffc900297bb658] __schedule at ffffffff81a4c33f
#1 [ffffc900297bb6e0] schedule at ffffffff81a4c6e3
#2 [ffffc900297bb6f8] io_schedule at ffffffff81a4ca42
#3 [ffffc900297bb708] __lock_page at ffffffff811f145b
#4 [ffffc900297bb798] __process_pages_contig at ffffffff814bc502 #5 [ffffc900297bb8c8] lock_delalloc_pages at ffffffff814bc684 #6 [ffffc900297bb900] find_lock_delalloc_range at ffffffff814be9ff #7 [ffffc900297bb9a0] writepage_delalloc at ffffffff814bebd0 #8 [ffffc900297bba18] __extent_writepage at ffffffff814bfbf2 #9 [ffffc900297bba98] extent_write_cache_pages at ffffffff814bffbd
PID: 2167901 TASK: ffff889dc6a59c00 CPU: 14 COMMAND:
"aio-dio-invalid"
#0 [ffffc9003b50bb18] __schedule at ffffffff81a4c33f #1 [ffffc9003b50bba0] schedule at ffffffff81a4c6e3 #2 [ffffc9003b50bbb8] io_schedule at ffffffff81a4ca42 #3 [ffffc9003b50bbc8] wait_on_page_bit at ffffffff811f24d6 #4 [ffffc9003b50bc60] prepare_pages at ffffffff814b05a7 #5 [ffffc9003b50bcd8] btrfs_buffered_write at ffffffff814b1359 #6 [ffffc9003b50bdb0] btrfs_file_write_iter at ffffffff814b5933 #7 [ffffc9003b50be38] new_sync_write at ffffffff8128f6a8 #8 [ffffc9003b50bec8] vfs_write at ffffffff81292b9d #9 [ffffc9003b50bf00] ksys_pwrite64 at ffffffff81293032
I used drgn to find the respective pages we were stuck on
page_entry.page 0xffffea00fbfc7500 index 8148 bit 15 pid 2167901 page_entry.page 0xffffea00f9bb7400 index 7680 bit 0 pid 1329874
As you can see the kworker is waiting for bit 0 (PG_locked) on index 7680, and aio-dio-invalid is waiting for bit 15 (PG_writeback) on index 8148. aio-dio-invalid has 7680, and the kworker epd looks like the following
crash> struct extent_page_data ffffc900297bbbb0
struct extent_page_data {
bio = 0xffff889f747ed830,
tree = 0xffff889eed6ba448,
extent_locked = 0,
sync_io = 0
}Probably worth mentioning as well that it waits for writeback of the page to complete while holding a lock on it (at prepare_pages()).
Using drgn I walked the bio pages looking for page 0xffffea00fbfc7500 which is the one we're waiting for writeback on
bio = Object(prog, 'struct bio', address=0xffff889f747ed830) for i in range(0, bio.bi_vcnt.value_()):
bv = bio.bi_io_vec[i]
if bv.bv_page.value_() == 0xffffea00fbfc7500:
print("FOUND IT")which validated what I suspected.
The fix for this is simple, flush the epd before we loop back around to the beginning of the file during writeout.
--- fs/btrfs/extent_io.c | 8 ++++++++ 1 file changed, 8 insertions(+)
fs/btrfs/extent_io.c +8 -0
@@ -4048,6 +4048,14 @@ static int extent_write_cache_pages(struct address_space *mapping,*/ scanned =1 ; index =0 ; + +/* +* If we're looping we could run into a page that is locked by a +* writer and that writer could be waiting on writeback for a +* page in our current bio, and thus deadlock, so flush the +* write bio here. +*/ +flush_write_bio (data); goto retry; }
Signature · 1 line
2.20.1
Re: [PATCH 4.14 129/173] btrfs: flush write bio if we loop in extent_write_cache_pages
Quoted 3 lines · Greg Kroah-Hartman
From: Josef Bacik <[email protected]>
[ Upstream commit
96bf313ecb33567af4cb53928b0c951254a02759]
Same comment as for the 4.9 backport, correct commit id is 42ffb0bf584ae5b6b38f7.
Greg Kroah-Hartman
maintainerRe: [PATCH 4.14 129/173] btrfs: flush write bio if we loop in extent_write_cache_pages
Quoted 7 lines · David Sterba
Quoted 3 lines · Greg Kroah-Hartman
From: Josef Bacik <[email protected]>
[ Upstream commit
96bf313ecb33567af4cb53928b0c951254a02759]Same comment as for the 4.9 backport, correct commit id is
42ffb0bf584ae5b6b38f7.
Now fixed up for all 3 places (4.14, 4.9, and 4.4)
thanks,
greg k-h
Greg Kroah-Hartman
maintainerPATCH 130/173 KVM: x86: Fix potential put_fpu() w/o load_fpu() on MPX platform
In reply to Greg Kroah-HartmanFrom: Sean Christopherson <[email protected]>
[ Upstream commit f958bd2314d117f8c29f4821401bc1925bc2e5ef ]
Unlike most state managed by XSAVE, MPX is initialized to zero on INIT. Because INITs are usually recognized in the context of a VCPU_RUN call, kvm_vcpu_reset() puts the guest's FPU so that the FPU state is resident in memory, zeros the MPX state, and reloads FPU state to hardware. But, in the unlikely event that an INIT is recognized during kvm_arch_vcpu_ioctl_get_mpstate() via kvm_apic_accept_events(), kvm_vcpu_reset() will call kvm_put_guest_fpu() without a preceding kvm_load_guest_fpu() and corrupt the guest's FPU state (and possibly userspace's FPU state as well).
Given that MPX is being removed from the kernel[*], fix the bug with the simple-but-ugly approach of loading the guest's FPU during KVM_GET_MP_STATE.
[*] See commit f240652b6032b ("x86/mpx: Remove MPX APIs").
--- arch/x86/kvm/x86.c | 5 +++++ 1 file changed, 5 insertions(+)
arch/x86/kvm/x86.c +5 -0
@@ -7677,6 +7677,9 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,int kvm_arch_vcpu_ioctl_get_mpstate (struct kvm_vcpu *vcpu,struct kvm_mp_state *mp_state) { +if (kvm_mpx_supported ()) +kvm_load_guest_fpu (vcpu); +kvm_apic_accept_events (vcpu);if (vcpu->arch.mp_state ==KVM_MP_STATE_HALTED && vcpu->arch.pv.pv_unhalted)
@@ -7684,6 +7687,8 @@ int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,else mp_state->mp_state = vcpu->arch.mp_state; +if (kvm_mpx_supported ()) +kvm_put_guest_fpu (vcpu);return 0 ; }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 131/173 KVM: x86/mmu: Apply max PA check for MMIO sptes to 32-bit KVM
In reply to Greg Kroah-HartmanFrom: Sean Christopherson <[email protected]>
[ Upstream commit e30a7d623dccdb3f880fbcad980b0cb589a1da45 ]
Remove the bogus 64-bit only condition from the check that disables MMIO spte optimization when the system supports the max PA, i.e. doesn't have any reserved PA bits. 32-bit KVM always uses PAE paging for the shadow MMU, and per Intel's SDM:
PAE paging translates 32-bit linear addresses to 52-bit physical addresses.
The kernel's restrictions on max physical addresses are limits on how much memory the kernel can reasonably use, not what physical addresses are supported by hardware.
--- arch/x86/kvm/x86.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kvm/x86.c +1 -1
@@ -6314,7 +6314,7 @@ static void kvm_set_mmio_spte_mask(void)* If reserved bit is not supported, clear the present bit to disable * mmio page fault. */ -if (IS_ENABLED (CONFIG_X86_64 ) && maxphyaddr ==52 ) +if (maxphyaddr ==52 ) mask &= ~1ull ;kvm_mmu_set_mmio_spte_mask (mask, mask);
Signature · 1 line
2.20.1