4.1.9-stable review
Subsystems (9)
Greg Kroah-Hartman
maintainercover 4.1.9-stable review
This is the start of the stable review cycle for the 4.1.9 release. There are 159 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 Mon Sep 28 20:51:39 UTC 2015. Anything received after that time might be too late.
The whole patch series can be found in one patch at:
kernel.org/pub/linux/kernel/v4.x/stable-review/patch-4.1.9-rc1.gz
and the diffstat can be found below.
thanks,
greg k-h
-------------
Pseudo-Shortlog of commits:
Greg Kroah-Hartman <[email protected]>
Linux 4.1.9-rc1
Andy Whitcroft <[email protected]>
ipv4: off-by-one in continuation handling in /proc/net/route
Florian Fainelli <[email protected]>
net: dsa: Do not override PHY interface if already configured
Eric Dumazet <[email protected]>
inet: fix races with reqsk timers
Eric Dumazet <[email protected]>
inet: fix possible request socket leak
Daniel Borkmann <[email protected]>
netlink: make sure -EBUSY won't escape from netlink_insert
Ivan Vecera <[email protected]>
bna: fix interrupts storm caused by erroneous packets
Nikolay Aleksandrov <[email protected]>
bridge: netlink: account for the IFLA_BRPORT_PROXYARP_WIFI attribute size and policy
Nikolay Aleksandrov <[email protected]>
bridge: netlink: account for the IFLA_BRPORT_PROXYARP attribute size and policy
Eric Dumazet <[email protected]>
udp: fix dst races with multicast early demux
Dan Carpenter <[email protected]>
rds: fix an integer overflow test in rds_info_getsockopt()
Ido Schimmel <[email protected]>
rocker: free netdevice during netdevice removal
Daniel Borkmann <[email protected]>
net: sched: fix refcount imbalance in actions
Daniel Borkmann <[email protected]>
act_bpf: fix memory leaks when replacing bpf programs
Alexander Drozdov <[email protected]>
packet: tpacket_snd(): fix signed/unsigned comparison
Lars Westerhoff <[email protected]>
packet: missing dev_put() in packet_do_bind()
Alexander Duyck <[email protected]>
fib_trie: Drop unnecessary calls to leaf_pull_suffix
Jack Morgenstein <[email protected]>
net/mlx4_core: Fix wrong index in propagating port change event to VFs
Nikolay Aleksandrov <[email protected]>
bridge: netlink: fix slave_changelink/br_setport race conditions
Michael S. Tsirkin <[email protected]>
virtio_net: don't require ANY_LAYOUT with VERSION_1
Florian Westphal <[email protected]>
netlink: don't hold mutex in rcu callback when releasing mmapd ring
Edward Hyunkoo Jee <[email protected]>
inet: frags: fix defragmented packet's IP header for af_packet
Daniel Borkmann <[email protected]>
sched: cls_flow: fix panic on filter replace
Daniel Borkmann <[email protected]>
sched: cls_bpf: fix panic on filter replace
dingtianhong <[email protected]>
bonding: correct the MAC address for "follow" fail_over_mac policy
Herbert Xu <[email protected]>
Revert "sit: Add gro callbacks to sit_offload"
Nikolay Aleksandrov <[email protected]>
bonding: fix destruction of bond with devices different from arphrd_ether
Eric Dumazet <[email protected]>
ipv6: lock socket in ip6_datagram_connect()
Tilman Schmidt <[email protected]>
isdn/gigaset: reset tty->receive_room when attaching ser_gigaset
WANG Cong <[email protected]>
fq_codel: fix a use-after-free
Nikolay Aleksandrov <[email protected]>
bridge: mdb: fix double add notification
Herbert Xu <[email protected]>
net: Fix skb_set_peeked use-after-free bug
Herbert Xu <[email protected]>
net: Fix skb csum races when peeking
Herbert Xu <[email protected]>
net: Clone skb before setting peeked flag
Dan Carpenter <[email protected]>
net/xen-netback: off by one in BUG_ON() condition
Julian Anastasov <[email protected]>
net: call rcu_read_lock early in process_backlog
Julian Anastasov <[email protected]>
net: do not process device backlog during unregistration
Eric Dumazet <[email protected]>
bridge: fix potential crash in __netdev_pick_tx()
Oleg Nesterov <[email protected]>
net: pktgen: fix race between pktgen_thread_worker() and kthread_stop()
Nikolay Aleksandrov <[email protected]>
bridge: mdb: zero out the local br_ip variable before use
Stephen Smalley <[email protected]>
net/tipc: initialize security state for new connection socket
Timo Teräs <[email protected]>
ip_tunnel: fix ipv4 pmtu check to honor inner ip header df
Daniel Borkmann <[email protected]>
rtnetlink: verify IFLA_VF_INFO attributes before passing them to driver
Nicolas Dichtel <[email protected]>
Revert "dev: set iflink to 0 for virtual interfaces"
Eric Dumazet <[email protected]>
net: graceful exit from netif_alloc_netdev_queues()
Phil Sutter <[email protected]>
rhashtable: fix for resize events during table walk
Angga <[email protected]>
ipv6: Make MLD packets to only be processed locally
Jan Kara <[email protected]>
jbd2: avoid infinite loop when destroying aborted journal
Yinghai Lu <[email protected]>
lib/decompressors: use real out buf size for gunzip with kernel
Hin-Tak Leung <[email protected]>
hfs,hfsplus: cache pages correctly between bnode_create and bnode_free
Heiko Stübner <[email protected]>
net: stmmac: dwmac-rk: Fix clk rate when provided by soc
Alexey Brodkin <[email protected]>
stmmac: troubleshoot unexpected bits in des0 & des1
Alexey Brodkin <[email protected]>
stmmac: fix check for phydev being open
Ariel Nahum <[email protected]>
IB/mlx4: Fix incorrect cq flushing in error state
Noa Osherovich <[email protected]>
IB/mlx4: Use correct SL on AH query under RoCE
Jack Morgenstein <[email protected]>
IB/mlx4: Forbid using sysfs to change RoCE pkeys
Jack Morgenstein <[email protected]>
IB/mlx4: Fix potential deadlock when sending mad to wire
Haggai Eran <[email protected]>
IB/mlx5: avoid destroying a NULL mr in reg_user_mr error flow
Sagi Grimberg <[email protected]>
IB/iser: Fix possible bogus DMA unmapping
Sagi Grimberg <[email protected]>
IB/iser: Fix missing return status check in iser_send_data_out
Yishai Hadas <[email protected]>
IB/uverbs: Fix race between ib_uverbs_open and remove_one
Christoph Hellwig <[email protected]>
IB/uverbs: reject invalid or unknown opcodes
Mike Marciniszyn <[email protected]>
IB/qib: Change lkey table allocation to support more MRs
Bart Van Assche <[email protected]>
IB/srp: Stop the scsi_eh_<n> and scsi_tmf_<n> threads if login fails
Bart Van Assche <[email protected]>
IB/srp: Handle partial connection success correctly
Hans de Goede <[email protected]>
ideapad-laptop: Add Lenovo Yoga 3 14 to no_hw_rfkill dmi list
Hin-Tak Leung <[email protected]>
hfs: fix B-tree corruption after insertion at position 0
Tyler Hicks <[email protected]>
eCryptfs: Invalidate dcache entries when lower i_nlink is zero
Joerg Roedel <[email protected]>
iommu/vt-d: Really use upper context table when necessary
Thierry Reding <[email protected]>
iommu/tegra-smmu: Parameterize number of TLB lines
Will Deacon <[email protected]>
iommu/io-pgtable-arm: Unmap and free table when overwriting with block
Emil Medve <[email protected]>
iommu/fsl: Really fix init section(s) content
NeilBrown <[email protected]>
md: flush ->event_work before stopping array.
NeilBrown <[email protected]>
md/raid10: always set reshape_safe when initializing reshape_position.
NeilBrown <[email protected]>
md/raid5: don't let shrink_slab shrink too far.
NeilBrown <[email protected]>
md/raid5: avoid races when changing cache size.
Jialing Fu <[email protected]>
mmc: core: fix race condition in mmc_wait_data_done
Jisheng Zhang <[email protected]>
mmc: sdhci: also get preset value and driver type for MMC_DDR52
Adam Lee <[email protected]>
mmc: sdhci-pci: set the clear transfer mode register quirk for O2Micro
Jann Horn <[email protected]>
fs: Don't dump core if the corefile would become world-readable.
Jann Horn <[email protected]>
fs: if a coredump already exists, unlink and recreate with O_EXCL
Jaewon Kim <[email protected]>
vmscan: fix increasing nr_isolated incurred by putback unevictable pages
Helge Deller <[email protected]>
parisc: Filter out spurious interrupts in PA-RISC irq handler
John David Anglin <[email protected]>
parisc: Use double word condition in 64bit CAS operation
Helge Deller <[email protected]>
PCI,parisc: Enable 64-bit bus addresses on PA-RISC
Mitja Spes <[email protected]>
rtc: abx80x: fix RTC write bit
Joonyoung Shim <[email protected]>
rtc: s5m: fix to update ctrl register
Joonyoung Shim <[email protected]>
rtc: s3c: fix disabled clocks for alarm
Trond Myklebust <[email protected]>
SUNRPC: Lock the transport layer on shutdown
Trond Myklebust <[email protected]>
SUNRPC: Ensure that we wait for connections to complete before retrying
Trond Myklebust <[email protected]>
SUNRPC: xs_reset_transport must mark the connection as disconnected
Trond Myklebust <[email protected]>
SUNRPC: Fix a thinko in xs_connect()
Pratyush Anand <[email protected]>
net: sunrpc: fix tracepoint Warning: unknown op '->'
Trond Myklebust <[email protected]>
Revert "NFSv4: Remove incorrect check in can_open_delegated()"
Trond Myklebust <[email protected]>
NFSv4.1: Fix a protocol issue with CLOSE stateids
Trond Myklebust <[email protected]>
NFSv4.1/flexfiles: Fix a protocol error in layoutreturn
Peng Tao <[email protected]>
NFS41/flexfiles: zero out DS write wcc
Trond Myklebust <[email protected]>
NFSv4: Force a post-op attribute update when holding a delegation
Peng Tao <[email protected]>
NFS41/flexfiles: update inode after write finishes
Trond Myklebust <[email protected]>
NFS: nfs_set_pgio_error sometimes misses errors
Kinglong Mee <[email protected]>
NFS: Fix a NULL pointer dereference of migration recovery ops for v4.2 client
Trond Myklebust <[email protected]>
NFSv4.1/pNFS: Fix borken function _same_data_server_addrs_locked()
Trond Myklebust <[email protected]>
NFS: Don't let the ctime override attribute barriers.
NeilBrown <[email protected]>
NFSv4: don't set SETATTR for O_RDONLY|O_EXCL
Jeff Layton <[email protected]>
nfsd: ensure that delegation stateid hash references are only put once
Jeff Layton <[email protected]>
nfsd: ensure that the ol stateid hash reference is only put once
Kinglong Mee <[email protected]>
nfsd: Fix an FS_LAYOUT_TYPES/LAYOUT_TYPES encode bug
Trond Myklebust <[email protected]>
NFSv4/pnfs: Ensure we don't miss a file extension
Filipe Manana <[email protected]>
Btrfs: check if previous transaction aborted to avoid fs corruption
Benoit Parrot <[email protected]>
media: am437x-vpfe: Fix a race condition during release
Benoit Parrot <[email protected]>
media: am437x-vpfe: Requested frame size and fmt overwritten by current sensor setting
Sakari Ailus <[email protected]>
v4l: omap3isp: Fix sub-device power management code
David Härdeman <[email protected]>
rc-core: fix remove uevent generation
Michal Hocko <[email protected]>
mm: make page pfmemalloc check more robust
Minfei Huang <[email protected]>
x86/mm: Initialize pmd_idx in page_table_range_init_count()
Yinghai Lu <[email protected]>
mm: check if section present during memory block registering
Jeffery Miller <[email protected]>
Add radeon suspend/resume quirk for HP Compaq dc5750.
Jann Horn <[email protected]>
CIFS: fix type confusion in copy offload ioctl
Aneesh Kumar K.V <[email protected]>
powerpc/mm: Recompute hash value after a failed update
Benjamin Herrenschmidt <[email protected]>
powerpc/boot: Specify ABI v2 when building an LE boot wrapper
Leonidas Da Silva Barbosa <[email protected]>
crypto: vmx - Adding enable_kernel_vsx() to access VSX instructions
Leonidas Da Silva Barbosa <[email protected]>
powerpc: Uncomment and make enable_kernel_vsx() routine available
Thomas Huth <[email protected]>
powerpc/rtas: Introduce rtas_get_sensor_fast() for IRQ handlers
Michael Ellerman <[email protected]>
powerpc/mm: Fix pte_pagesize_index() crash on 4K w/64K hash
Gavin Shan <[email protected]>
powerpc/eeh: Fix fenced PHB caused by eeh_slot_error_detail()
Daniel Axtens <[email protected]>
powerpc/eeh: Probe after unbalanced kref check
Gavin Shan <[email protected]>
powerpc/pseries: Fix corrupted pdn list
David Dueck <[email protected]>
pinctrl: at91: fix null pointer dereference
Niranjan Sivakumar <[email protected]>
ALSA: hda - Fix white noise on Dell M3800
Woodrow Shen <[email protected]>
ALSA: hda - Add some FIXUP quirks for white noise on Dell laptop.
Takashi Iwai <[email protected]>
ALSA: hda - Use ALC880_FIXUP_FUJITSU for FSC Amilo M1437
Takashi Iwai <[email protected]>
ALSA: hda - Enable headphone jack detect on old Fujitsu laptops
Yao-Wen Mao <[email protected]>
ALSA: usb-audio: correct the value cache check.
Takashi Iwai <[email protected]>
Input: evdev - do not report errors form flush()
Dmitry Torokhov <[email protected]>
Input: synaptics - fix handling of disabling gesture mode
Marc Zyngier <[email protected]>
arm64: KVM: Disable virtual timer even if the guest is not using it
Will Deacon <[email protected]>
KVM: arm64: add workaround for Cortex-A57 erratum #852523
Pavel Fedin <[email protected]>
arm/arm64: KVM: vgic: Check for !irqchip_in_kernel() when mapping resources
Will Deacon <[email protected]>
arm64: errata: add module build workaround for erratum #843419
Will Deacon <[email protected]>
arm64: head.S: initialise mdcr_el2 in el2_setup
Will Deacon <[email protected]>
arm64: compat: fix vfp save/restore across signal handlers in big-endian
Ard Biesheuvel <[email protected]>
arm64: set MAX_MEMBLOCK_ADDR according to linear region size
Ard Biesheuvel <[email protected]>
of/fdt: make memblock maximum physical address arch configurable
Ard Biesheuvel <[email protected]>
arm64: flush FP/SIMD state correctly after execve()
Jeff Vander Stoep <[email protected]>
arm64: kconfig: Move LIST_POISON to a safe value
Theodore Ts'o <[email protected]>
Revert "ext4: remove block_device_ejected"
Eric Sandeen <[email protected]>
ext4: don't manipulate recovery flag when freezing no-journal fs
Daniel Axtens <[email protected]>
cxl: Fix unbalanced pci_dev_get in cxl_probe
Daniel Axtens <[email protected]>
cxl: Remove racy attempt to force EEH invocation in reset
Bob Copeland <[email protected]>
mac80211: enable assoc check for mesh interfaces
Markos Chandras <[email protected]>
MIPS: math-emu: Emulate missing BC1{EQ,NE}Z instructions
Markos Chandras <[email protected]>
MIPS: math-emu: Allow m{f,t}hc emulation on MIPS R6
Jean Delvare <[email protected]>
tg3: Fix temperature reporting
Shota Suzuki <[email protected]>
igb: Fix oops caused by missing queue pairing
Larry Finger <[email protected]>
rtlwifi: rtl8821ae: Fix an expression that is always false
Adrien Schildknecht <[email protected]>
rtlwifi: rtl8192cu: Add new device ID
Eric W. Biederman <[email protected]>
unshare: Unsharing a thread does not require unsharing a vm
Ming Lei <[email protected]>
blk-mq: fix buffer overflow when reading sysfs file of 'pending'
Christophe Ricard <[email protected]>
nfc: nci: hci: Add check on skb nci_hci_send_cmd parameter
Christophe Ricard <[email protected]>
NFC: st21nfca: fix use of uninitialized variables in error path
-------------
Diffstat:
Makefile | 4 +- arch/arm/boot/compressed/decompress.c | 2 +- arch/arm/kvm/arm.c | 2 +- arch/arm64/Kconfig | 20 +++ arch/arm64/Makefile | 4 + arch/arm64/include/asm/memory.h | 8 + arch/arm64/kernel/fpsimd.c | 1 + arch/arm64/kernel/head.S | 5 + arch/arm64/kernel/module.c | 2 + arch/arm64/kernel/signal32.c | 47 ++++-- arch/arm64/kvm/hyp.S | 9 +- arch/m32r/boot/compressed/misc.c | 3 +- arch/mips/boot/compressed/decompress.c | 4 +- arch/mips/math-emu/cp1emu.c | 24 ++- arch/parisc/kernel/irq.c | 8 +- arch/parisc/kernel/syscall.S | 2 +- arch/powerpc/boot/Makefile | 3 + arch/powerpc/include/asm/pgtable-ppc64.h | 14 +- arch/powerpc/include/asm/rtas.h | 1 + arch/powerpc/include/asm/switch_to.h | 1 + arch/powerpc/kernel/eeh.c | 27 +++- arch/powerpc/kernel/process.c | 3 - arch/powerpc/kernel/rtas.c | 17 ++ arch/powerpc/mm/hugepage-hash64.c | 3 +- arch/powerpc/platforms/pseries/ras.c | 3 +- arch/powerpc/platforms/pseries/setup.c | 5 + arch/s390/boot/compressed/misc.c | 2 +- arch/sh/boot/compressed/misc.c | 2 +- arch/unicore32/boot/compressed/misc.c | 4 +- arch/x86/boot/compressed/misc.c | 3 +- arch/x86/mm/init_32.c | 1 + block/blk-mq-sysfs.c | 25 ++- drivers/base/node.c | 10 ++ drivers/crypto/vmx/aes.c | 3 + drivers/crypto/vmx/aes_cbc.c | 3 + drivers/crypto/vmx/aes_ctr.c | 3 + drivers/crypto/vmx/ghash.c | 4 + drivers/gpu/drm/radeon/radeon_combios.c | 8 + drivers/infiniband/core/uverbs.h | 3 +- drivers/infiniband/core/uverbs_cmd.c | 10 +- drivers/infiniband/core/uverbs_main.c | 43 +++-- drivers/infiniband/hw/mlx4/ah.c | 6 +- drivers/infiniband/hw/mlx4/cq.c | 2 +- drivers/infiniband/hw/mlx4/mcg.c | 7 +- drivers/infiniband/hw/mlx4/sysfs.c | 5 +- drivers/infiniband/hw/mlx5/mr.c | 12 -- drivers/infiniband/hw/qib/qib_keys.c | 4 + drivers/infiniband/hw/qib/qib_verbs.c | 14 +- drivers/infiniband/hw/qib/qib_verbs.h | 2 + drivers/infiniband/ulp/iser/iscsi_iser.c | 12 +- drivers/infiniband/ulp/iser/iscsi_iser.h | 2 + drivers/infiniband/ulp/iser/iser_initiator.c | 8 +- drivers/infiniband/ulp/srp/ib_srp.c | 12 +- drivers/input/evdev.c | 13 +- drivers/input/mouse/synaptics.c | 12 +- drivers/iommu/fsl_pamu.c | 26 +-- drivers/iommu/intel-iommu.c | 2 +- drivers/iommu/io-pgtable-arm.c | 17 +- drivers/iommu/tegra-smmu.c | 9 +- drivers/isdn/gigaset/ser-gigaset.c | 11 +- drivers/md/md.c | 2 + drivers/md/raid10.c | 5 +- drivers/md/raid5.c | 34 +++- drivers/md/raid5.h | 3 +- drivers/media/platform/am437x/am437x-vpfe.c | 16 +- drivers/media/platform/omap3isp/isp.c | 4 +- drivers/media/rc/rc-main.c | 3 - drivers/memory/tegra/tegra114.c | 1 + drivers/memory/tegra/tegra124.c | 1 + drivers/memory/tegra/tegra30.c | 1 + drivers/misc/cxl/pci.c | 18 --- drivers/mmc/core/core.c | 6 +- drivers/mmc/host/sdhci-pci.c | 1 + drivers/mmc/host/sdhci.c | 4 +- drivers/net/bonding/bond_main.c | 21 +++ drivers/net/ethernet/broadcom/tg3.c | 2 +- drivers/net/ethernet/brocade/bna/bnad.c | 2 +- drivers/net/ethernet/intel/fm10k/fm10k_main.c | 2 +- drivers/net/ethernet/intel/igb/igb.h | 1 + drivers/net/ethernet/intel/igb/igb_ethtool.c | 5 +- drivers/net/ethernet/intel/igb/igb_main.c | 18 ++- drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 2 +- drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 2 +- drivers/net/ethernet/mellanox/mlx4/eq.c | 4 +- drivers/net/ethernet/rocker/rocker.c | 1 + drivers/net/ethernet/stmicro/stmmac/descs.h | 2 + drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c | 3 +- drivers/net/ethernet/stmicro/stmmac/enh_desc.c | 3 +- drivers/net/ethernet/stmicro/stmmac/norm_desc.c | 3 +- drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 49 +++--- drivers/net/virtio_net.c | 3 +- drivers/net/wireless/rtlwifi/rtl8192cu/sw.c | 1 + drivers/net/wireless/rtlwifi/rtl8821ae/hw.c | 2 +- drivers/net/wireless/rtlwifi/rtl8821ae/reg.h | 1 + drivers/net/xen-netback/netback.c | 6 +- drivers/nfc/st21nfca/st21nfca.c | 11 +- drivers/of/fdt.c | 12 +- drivers/parisc/lba_pci.c | 7 +- drivers/pci/Kconfig | 2 +- drivers/pinctrl/pinctrl-at91.c | 15 ++ drivers/platform/x86/ideapad-laptop.c | 7 + drivers/rtc/rtc-abx80x.c | 2 +- drivers/rtc/rtc-s3c.c | 24 ++- drivers/rtc/rtc-s5m.c | 10 ++ fs/btrfs/transaction.c | 3 + fs/cifs/ioctl.c | 6 + fs/coredump.c | 46 +++++- fs/ecryptfs/dentry.c | 16 +- fs/ext4/super.c | 30 +++- fs/hfs/bnode.c | 9 +- fs/hfs/brec.c | 20 +-- fs/hfsplus/bnode.c | 3 - fs/jbd2/checkpoint.c | 39 ++++- fs/jbd2/commit.c | 2 +- fs/jbd2/journal.c | 11 +- fs/nfs/flexfilelayout/flexfilelayout.c | 5 + fs/nfs/flexfilelayout/flexfilelayoutdev.c | 7 +- fs/nfs/inode.c | 15 +- fs/nfs/nfs4proc.c | 20 ++- fs/nfs/pagelist.c | 4 +- fs/nfs/pnfs_nfs.c | 33 ++-- fs/nfs/write.c | 15 +- fs/nfsd/nfs4state.c | 84 ++++++---- fs/nfsd/nfs4xdr.c | 45 ++++-- include/linux/jbd2.h | 3 +- include/linux/mm.h | 28 ++++ include/linux/mm_types.h | 9 -- include/linux/skbuff.h | 14 +- include/linux/sunrpc/xprtsock.h | 3 + include/net/act_api.h | 8 +- include/net/ip.h | 1 + include/soc/tegra/mc.h | 1 + include/trace/events/sunrpc.h | 21 ++- kernel/fork.c | 28 ++-- lib/decompress_bunzip2.c | 6 +- lib/decompress_inflate.c | 31 +++- lib/decompress_unlz4.c | 6 +- lib/decompress_unlzma.c | 7 +- lib/decompress_unlzo.c | 13 +- lib/decompress_unxz.c | 12 +- lib/rhashtable.c | 4 +- mm/page_alloc.c | 7 +- mm/slab.c | 4 +- mm/slub.c | 2 +- mm/vmscan.c | 2 +- net/bridge/br_forward.c | 1 + net/bridge/br_mdb.c | 3 +- net/bridge/br_netlink.c | 14 +- net/core/datagram.c | 57 +++++-- net/core/dev.c | 43 +++-- net/core/pktgen.c | 4 +- net/core/request_sock.c | 8 +- net/core/rtnetlink.c | 187 +++++++++++----------- net/core/skbuff.c | 2 +- net/dsa/slave.c | 3 +- net/ipv4/datagram.c | 16 +- net/ipv4/fib_trie.c | 6 +- net/ipv4/inet_connection_sock.c | 2 +- net/ipv4/ip_fragment.c | 7 +- net/ipv4/ip_tunnel.c | 8 +- net/ipv4/tcp_ipv4.c | 2 +- net/ipv4/udp.c | 13 +- net/ipv6/datagram.c | 20 ++- net/ipv6/ip6_input.c | 6 +- net/ipv6/ip6_offload.c | 2 - net/ipv6/tcp_ipv6.c | 2 +- net/mac80211/tx.c | 3 - net/netlink/af_netlink.c | 84 ++++++---- net/nfc/nci/hci.c | 2 +- net/packet/af_packet.c | 11 +- net/rds/info.c | 2 +- net/sched/act_api.c | 11 +- net/sched/act_bpf.c | 50 ++++-- net/sched/cls_bpf.c | 2 +- net/sched/cls_flow.c | 5 +- net/sched/sch_fq_codel.c | 2 +- net/sunrpc/xprt.c | 6 + net/sunrpc/xprtsock.c | 21 ++- net/tipc/socket.c | 1 + sound/pci/hda/patch_realtek.c | 12 +- sound/usb/mixer.c | 2 +- 181 files changed, 1431 insertions(+), 656 deletions(-)
Greg Kroah-Hartman
maintainerPATCH 1/159 NFC: st21nfca: fix use of uninitialized variables in error path
4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Christophe Ricard <[email protected]>
commit 5a3570061a131309143a49e4bbdbce7e23f261e7 upstream.
st21nfca_hci_load_session() calls kfree_skb() on unitialized variables skb_pipe_info and skb_pipe_list if the call to nfc_hci_connect_gate() failed. Reword the error path to not use these variables when they are not initialized. While at it, there seemed to be a memory leak because skb_pipe_info was only freed once, after the for-loop, even though several ones were created by nfc_hci_send_cmd.
("NFC: st21nfca: Remove skb_pipe_list and skb_pipe_info useless allocation")
--- drivers/nfc/st21nfca/st21nfca.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-)
drivers/nfc/st21nfca/st21nfca.c +6 -5
@@ -148,14 +148,14 @@ static int st21nfca_hci_load_session(strST21NFCA_DEVICE_MGNT_GATE , ST21NFCA_DEVICE_MGNT_PIPE);if (r <0 ) - goto free_info; +return r;/* Get pipe list */ r =nfc_hci_send_cmd (hdev,ST21NFCA_DEVICE_MGNT_GATE ,ST21NFCA_DM_GETINFO , pipe_list,sizeof (pipe_list), &skb_pipe_list);if (r <0 ) - goto free_info; +return r;/* Complete the existing gate_pipe table */ for (i =0 ; i < skb_pipe_list->len; i++) {
@@ -181,6 +181,7 @@ static int st21nfca_hci_load_session(str info->src_host_id != ST21NFCA_ESE_HOST_ID) {pr_err ("Unexpected apdu_reader pipe on host %x\n" , info->src_host_id); +kfree_skb (skb_pipe_info);continue ; }
@@ -200,6 +201,7 @@ static int st21nfca_hci_load_session(str hdev->pipes[st21nfca_gates[j].pipe].dest_host = info->src_host_id; } +kfree_skb (skb_pipe_info); }/*
@@ -214,13 +216,12 @@ static int st21nfca_hci_load_session(str st21nfca_gates[i].gate, st21nfca_gates[i].pipe);if (r <0 ) - goto free_info; + goto free_list; } }memcpy (hdev->init_data.gates, st21nfca_gates,sizeof (st21nfca_gates)); -free_info: -kfree_skb (skb_pipe_info); +free_list:kfree_skb (skb_pipe_list);return r; }
Greg Kroah-Hartman
maintainerPATCH 2/159 nfc: nci: hci: Add check on skb nci_hci_send_cmd parameter
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Christophe Ricard <[email protected]>
commit 5a9e0ffc0f128ecdf7c770f76c268e4f9f3c9118 upstream.
skb can be NULL and may lead to a NULL pointer error.
Add a check condition before setting HCI rx buffer.
--- net/nfc/nci/hci.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
net/nfc/nci/hci.c +1 -1
@@ -233,7 +233,7 @@ int nci_hci_send_cmd(struct nci_dev *nde r =nci_request (ndev, nci_hci_send_data_req, (unsigned long )&data,msecs_to_jiffies (NCI_DATA_TIMEOUT )); -if (r ==NCI_STATUS_OK ) +if (r ==NCI_STATUS_OK && skb) *skb = conn_info->rx_skb;return r;
Greg Kroah-Hartman
maintainerPATCH 3/159 blk-mq: fix buffer overflow when reading sysfs file of pending
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Ming Lei <[email protected]>
commit 596f5aad2a704b72934e5abec1b1b4114c16f45b upstream.
There may be lots of pending requests so that the buffer of PAGE_SIZE can't hold them at all.
One typical example is scsi-mq, the queue depth(.can_queue) of scsi_host and blk-mq is quite big but scsi_device's queue_depth is a bit small(.cmd_per_lun), then it is quite easy to have lots of pending requests in hw queue.
This patch fixes the following warning and the related memory destruction.
[ 359.025101] fill_read_buffer: blk_mq_hw_sysfs_show+0x0/0x7d returned bad count^M [ 359.055595] irq event stamp: 15537^M [ 359.055606] general protection fault: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC ^M [ 359.055614] Dumping ftrace buffer:^M [ 359.055660] (ftrace buffer empty)^M [ 359.055672] Modules linked in: nbd ipv6 kvm_intel kvm serio_raw^M [ 359.055678] CPU: 4 PID: 21631 Comm: stress-ng-sysfs Not tainted 4.2.0-rc5-next-20150805 #434^M [ 359.055679] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011^M [ 359.055682] task: ffff8802161cc000 ti: ffff88021b4a8000 task.ti: ffff88021b4a8000^M [ 359.055693] RIP: 0010:[<ffffffff811541c5>] [<ffffffff811541c5>] __kmalloc+0xe8/0x152^M
--- block/blk-mq-sysfs.c | 21 ++++++++++++++++----- 1 file changed, 16 insertions(+), 5 deletions(-)
block/blk-mq-sysfs.c +16 -5
@@ -141,15 +141,26 @@ static ssize_t blk_mq_sysfs_completed_shstatic ssize_t sysfs_list_show (char *page,struct list_head *list,char *msg) { -char *start_page = page;struct request *rq; +int len =snprintf (page,PAGE_SIZE -1 ,"%s:\n" , msg); - page +=sprintf (page,"%s:\n" , msg); +list_for_each_entry (rq, list, queuelist) { +const int rq_len =2 *sizeof (rq) +2 ; -list_for_each_entry (rq, list, queuelist) - page +=sprintf (page,"\t%p\n" , rq); +/* if the output will be truncated */ +if (PAGE_SIZE -1 < len + rq_len) { +/* backspacing if it can't hold '\t...\n' */ +if (PAGE_SIZE -1 < len +5 ) + len -= rq_len; + len +=snprintf (page + len,PAGE_SIZE -1 - len, +"\t...\n" ); +break ; + } + len +=snprintf (page + len,PAGE_SIZE -1 - len, +"\t%p\n" , rq); + } -return page - start_page; +return len; }static ssize_t blk_mq_sysfs_rq_list_show (struct blk_mq_ctx *ctx,char *page)
Greg Kroah-Hartman
maintainerPATCH 4/159 unshare: Unsharing a thread does not require unsharing a vm
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: "Eric W. Biederman" <[email protected]>
commit 12c641ab8270f787dfcce08b5f20ce8b65008096 upstream.
In the logic in the initial commit of unshare made creating a new thread group for a process, contingent upon creating a new memory address space for that process. That is wrong. Two separate processes in different thread groups can share a memory address space and clone allows creation of such proceses.
This is significant because it was observed that mm_users > 1 does not mean that a process is multi-threaded, as reading /proc/PID/maps temporarily increments mm_users, which allows other processes to (accidentally) interfere with unshare() calls.
Correct the check in check_unshare_flags() to test for !thread_group_empty() for CLONE_THREAD, CLONE_SIGHAND, and CLONE_VM. For sighand->count > 1 for CLONE_SIGHAND and CLONE_VM.
For !current_is_single_threaded instead of mm_users > 1 for CLONE_VM.
By using the correct checks in unshare this removes the possibility of an accidental denial of service attack.
Additionally using the correct checks in unshare ensures that only an explicit unshare(CLONE_VM) can possibly trigger the slow path of current_is_single_threaded(). As an explict unshare(CLONE_VM) is pointless it is not expected there are many applications that make that call.
--- kernel/fork.c | 28 ++++++++++++++++++---------- 1 file changed, 18 insertions(+), 10 deletions(-)
kernel/fork.c +18 -10
@@ -1854,13 +1854,21 @@ static int check_unshare_flags(unsigned CLONE_NEWUSER|CLONE_NEWPID))return -EINVAL ;/* -* Not implemented, but pretend it works if there is nothing to -* unshare. Note that unsharing CLONE_THREAD or CLONE_SIGHAND -* needs to unshare vm. +* Not implemented, but pretend it works if there is nothing +* to unshare. Note that unsharing the address space or the +* signal handlers also need to unshare the signal queues (aka +* CLONE_THREAD). */ if (unshare_flags & (CLONE_THREAD |CLONE_SIGHAND |CLONE_VM )) { -/* FIXME: get_task_mm() increments ->mm_users */ -if (atomic_read (¤t->mm->mm_users) >1 ) +if (!thread_group_empty (current)) +return -EINVAL ; + } +if (unshare_flags & (CLONE_SIGHAND |CLONE_VM )) { +if (atomic_read (¤t->sighand->count) >1 ) +return -EINVAL ; + } +if (unshare_flags &CLONE_VM ) { +if (!current_is_single_threaded ())return -EINVAL ; }
@@ -1929,16 +1937,16 @@ SYSCALL_DEFINE1(unshare, unsigned long,if (unshare_flags &CLONE_NEWUSER ) unshare_flags |=CLONE_THREAD |CLONE_FS ;/* -* If unsharing a thread from a thread group, must also unshare vm. -*/ -if (unshare_flags &CLONE_THREAD ) - unshare_flags |=CLONE_VM ; -/* * If unsharing vm, must also unshare signal handlers. */ if (unshare_flags &CLONE_VM ) unshare_flags |=CLONE_SIGHAND ;/* +* If unsharing a signal handlers, must also unshare the signal queues. +*/ +if (unshare_flags &CLONE_SIGHAND ) + unshare_flags |=CLONE_THREAD ; +/* * If unsharing namespace, must also unshare filesystem information. */ if (unshare_flags &CLONE_NEWNS )
Greg Kroah-Hartman
maintainerPATCH 5/159 rtlwifi: rtl8192cu: Add new device ID
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Adrien Schildknecht <[email protected]>
commit 1642d09fb9b128e8e538b2a4179962a34f38dff9 upstream.
The v2 of NetGear WNA1000M uses a different idProduct: USB ID 0846:9043
--- drivers/net/wireless/rtlwifi/rtl8192cu/sw.c | 1 + 1 file changed, 1 insertion(+)
drivers/net/wireless/rtlwifi/rtl8192cu/sw.c +1 -0
@@ -321,6 +321,7 @@ static struct usb_device_id rtl8192c_usb {RTL_USB_DEVICE(0x07b8, 0x8188, rtl92cu_hal_cfg)}, /*Abocom - Abocom*/ {RTL_USB_DEVICE(0x07b8, 0x8189, rtl92cu_hal_cfg)}, /*Funai - Abocom*/ {RTL_USB_DEVICE(0x0846, 0x9041, rtl92cu_hal_cfg)}, /*NetGear WNA1000M*/ + {RTL_USB_DEVICE(0x0846, 0x9043, rtl92cu_hal_cfg)}, /*NG WNA1000Mv2*/ {RTL_USB_DEVICE(0x0b05, 0x17ba, rtl92cu_hal_cfg)}, /*ASUS-Edimax*/ {RTL_USB_DEVICE(0x0bda, 0x5088, rtl92cu_hal_cfg)}, /*Thinkware-CC&C*/ {RTL_USB_DEVICE(0x0df6, 0x0052, rtl92cu_hal_cfg)}, /*Sitecom - Edimax*/
Greg Kroah-Hartman
maintainerPATCH 6/159 rtlwifi: rtl8821ae: Fix an expression that is always false
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Larry Finger <[email protected]>
commit 251086f588720277a6f5782020a648ce32c4e00b upstream.
In routine _rtl8821ae_set_media_status(), an incorrect mask results in a test for AP status to always be false. Similar bugs were fixed in rtl8192cu and rtl8192de, but this instance was missed at that time.
--- drivers/net/wireless/rtlwifi/rtl8821ae/hw.c | 2 +- drivers/net/wireless/rtlwifi/rtl8821ae/reg.h | 1 + 2 files changed, 2 insertions(+), 1 deletion(-)
drivers/net/wireless/rtlwifi/rtl8821ae/hw.c +1 -1
@@ -2180,7 +2180,7 @@ static int _rtl8821ae_set_media_status(srtl_write_byte (rtlpriv,MSR , bt_msr); rtlpriv->cfg->ops->led_control (hw, ledaction); -if ((bt_msr &0xfc ) ==MSR_AP ) +if ((bt_msr &MSR_MASK ) ==MSR_AP )rtl_write_byte (rtlpriv,REG_BCNTCFG +1 ,0x00 );else rtl_write_byte (rtlpriv,REG_BCNTCFG +1 ,0x66 );
drivers/net/wireless/rtlwifi/rtl8821ae/reg.h +1 -0
@@ -429,6 +429,7 @@#define MSR_ADHOC 0x01#define MSR_INFRA 0x02#define MSR_AP 0x03 +#define MSR_MASK 0x03#define RRSR_RSC_OFFSET 21#define RRSR_SHORT_OFFSET 23
Greg Kroah-Hartman
maintainerPATCH 7/159 igb: Fix oops caused by missing queue pairing
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Shota Suzuki <[email protected]>
commit 72ddef0506da852dc82f078f37ced8ef4d74a2bf upstream.
When initializing igb driver (e.g. 82576, I350), IGB_FLAG_QUEUE_PAIRS is set if adapter->rss_queues exceeds half of max_rss_queues in igb_init_queue_configuration().
On the other hand, IGB_FLAG_QUEUE_PAIRS is not set even if the number of queues exceeds half of max_combined in igb_set_channels() when changing the number of queues by "ethtool -L".
In this case, if numvecs is larger than MAX_MSIX_ENTRIES (10), the size of adapter->msix_entries[], an overflow can occur in igb_set_interrupt_capability(), which in turn leads to an oops.
Fix this problem as follows:
- When changing the number of queues by "ethtool -L", set IGB_FLAG_QUEUE_PAIRS in the same way as initializing igb driver.
- When increasing the size of q_vector, reallocate it appropriately. (With IGB_FLAG_QUEUE_PAIRS set, the size of q_vector gets larger.)
Another possible way to fix this problem is to cap the queues at its initial number, which is the number of the initial online cpus. But this is not the optimal way because we cannot increase queues when another cpu becomes online.
Note that before commit cd14ef54d25b ("igb: Change to use statically allocated array for MSIx entries"), this problem did not cause oops but just made the number of queues become 1 because of entering msi_only mode in igb_set_interrupt_capability().
--- drivers/net/ethernet/intel/igb/igb.h | 1 + drivers/net/ethernet/intel/igb/igb_ethtool.c | 5 ++++- drivers/net/ethernet/intel/igb/igb_main.c | 16 ++++++++++++++-- 3 files changed, 19 insertions(+), 3 deletions(-)
drivers/net/ethernet/intel/igb/igb.h +1 -0
@@ -540,6 +540,7 @@ void igb_ptp_rx_pktstamp(struct igb_q_ve struct sk_buff *skb);int igb_ptp_set_ts_config (struct net_device *netdev,struct ifreq *ifr);int igb_ptp_get_ts_config (struct net_device *netdev,struct ifreq *ifr); +void igb_set_flag_queue_pairs (struct igb_adapter *,const u32 );#ifdef CONFIG_IGB_HWMON void igb_sysfs_exit (struct igb_adapter *adapter);int igb_sysfs_init (struct igb_adapter *adapter);
drivers/net/ethernet/intel/igb/igb_ethtool.c +4 -1
@@ -2991,6 +2991,7 @@ static int igb_set_channels(struct net_d {struct igb_adapter *adapter =netdev_priv (netdev);unsigned int count = ch->combined_count; +unsigned int max_combined =0 ;/* Verify they are not requesting separate vectors */ if (!count || ch->rx_count || ch->tx_count)
@@ -3001,11 +3002,13 @@ static int igb_set_channels(struct net_dreturn -EINVAL ;/* Verify the number of channels doesn't exceed hw limits */ -if (count >igb_max_channels (adapter)) + max_combined =igb_max_channels (adapter); +if (count > max_combined)return -EINVAL ;if (count != adapter->rss_queues) { adapter->rss_queues = count; +igb_set_flag_queue_pairs (adapter, max_combined);/* Hardware has to reinitialize queues and interrupts to * match the new configuration.
drivers/net/ethernet/intel/igb/igb_main.c +14 -2
@@ -1205,10 +1205,14 @@ static int igb_alloc_q_vector(struct igb/* allocate q_vector and rings */ q_vector = adapter->q_vector[v_idx]; -if (!q_vector) +if (!q_vector) { q_vector =kzalloc (size,GFP_KERNEL ); -else + }else if (size >ksize (q_vector)) { +kfree_rcu (q_vector, rcu); + q_vector =kzalloc (size,GFP_KERNEL ); + }else {memset (q_vector,0 , size); + }if (!q_vector)return -ENOMEM ;
@@ -2901,6 +2905,14 @@ static void igb_init_queue_configuration adapter->rss_queues =min_t (u32, max_rss_queues,num_online_cpus ()); +igb_set_flag_queue_pairs (adapter, max_rss_queues); +} + +void igb_set_flag_queue_pairs (struct igb_adapter *adapter, +const u32 max_rss_queues) +{ +struct e1000_hw *hw = &adapter->hw; +/* Determine if we need to pair queues. */ switch (hw->mac.type) {case e1000_82575:
Greg Kroah-Hartman
maintainerPATCH 8/159 tg3: Fix temperature reporting
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jean Delvare <[email protected]>
commit d3d11fe08ccc9bff174fc958722b5661f0932486 upstream.
The temperature registers appear to report values in degrees Celsius while the hwmon API mandates values to be exposed in millidegrees Celsius. Do the conversion so that the values reported by "sensors" are correct.
--- drivers/net/ethernet/broadcom/tg3.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/broadcom/tg3.c +1 -1
@@ -10757,7 +10757,7 @@ static ssize_t tg3_show_temp(struct devitg3_ape_scratchpad_read (tp, &temperature, attr->index,sizeof (temperature));spin_unlock_bh (&tp->lock); -return sprintf (buf,"%u\n" , temperature); +return sprintf (buf,"%u\n" , temperature *1000 ); }
Greg Kroah-Hartman
maintainerPATCH 9/159 MIPS: math-emu: Allow m{f,t}hc emulation on MIPS R6
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Markos Chandras <[email protected]>
commit e8f80cc1a6d80587136b015e989a12827e1fcfe5 upstream.
The mfhc/mthc instructions are supported on MIPS R6 so emulate them if needed.
Patchwork: https://patchwork.linux-mips.org/patch/10737/
--- arch/mips/math-emu/cp1emu.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
arch/mips/math-emu/cp1emu.c +2 -2
@@ -1137,7 +1137,7 @@ emul:break ;case mfhc_op: -if (!cpu_has_mips_r2) +if (!cpu_has_mips_r2_r6) goto sigill;/* copregister rd -> gpr[rt] */
@@ -1148,7 +1148,7 @@ emul:break ;case mthc_op: -if (!cpu_has_mips_r2) +if (!cpu_has_mips_r2_r6) goto sigill;/* copregister rd <- gpr[rt] */
Greg Kroah-Hartman
maintainerPATCH 10/159 MIPS: math-emu: Emulate missing BC1{EQ,NE}Z instructions
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Markos Chandras <[email protected]>
commit c909ca718e8f50cf484ef06a8dd935e738e8e53d upstream.
Commit c8a34581ec09 ("MIPS: Emulate the BC1{EQ,NE}Z FPU instructions") added support for emulating the new R6 BC1{EQ,NE}Z branches but it missed the case where the instruction that caused the exception was not on a DS.
Patchwork: https://patchwork.linux-mips.org/patch/10738/
--- arch/mips/math-emu/cp1emu.c | 20 +++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-)
arch/mips/math-emu/cp1emu.c +19 -1
@@ -1181,6 +1181,24 @@ emul: }break ; +case bc1eqz_op: +case bc1nez_op: +if (!cpu_has_mips_r6 ||delay_slot (xcp)) +return SIGILL ; + + cond = likely =0 ; +switch (MIPSInst_RS (ir)) { +case bc1eqz_op: +if (get_fpr32 (¤t->thread.fpu.fpr[MIPSInst_RT (ir)],0 ) &0x1 ) + cond =1 ; +break ; +case bc1nez_op: +if (!(get_fpr32 (¤t->thread.fpu.fpr[MIPSInst_RT (ir)],0 ) &0x1 )) + cond =1 ; +break ; + } + goto branch_common; +case bc_op:if (delay_slot (xcp))return SIGILL ;
@@ -1207,7 +1225,7 @@ emul:case bct_op:break ; } - +branch_common:set_delay_slot (xcp);if (cond) {/*
Greg Kroah-Hartman
maintainerPATCH 11/159 mac80211: enable assoc check for mesh interfaces
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Bob Copeland <[email protected]>
commit 3633ebebab2bbe88124388b7620442315c968e8f upstream.
We already set a station to be associated when peering completes, both in user space and in the kernel. Thus we should always have an associated sta before sending data frames to that station.
Failure to check assoc state can cause crashes in the lower-level driver due to transmitting unicast data frames before driver sta structures (e.g. ampdu state in ath9k) are initialized. This occurred when forwarding in the presence of fixed mesh paths: frames were transmitted to stations with whom we hadn't yet completed peering.
--- net/mac80211/tx.c | 3 --- 1 file changed, 3 deletions(-)
net/mac80211/tx.c +0 -3
@@ -301,9 +301,6 @@ ieee80211_tx_h_check_assoc(struct ieee80if (tx->sdata->vif.type ==NL80211_IFTYPE_WDS )return TX_CONTINUE ; -if (tx->sdata->vif.type ==NL80211_IFTYPE_MESH_POINT ) -return TX_CONTINUE ; -if (tx->flags &IEEE80211_TX_PS_BUFFERED )return TX_CONTINUE ;
Greg Kroah-Hartman
maintainerPATCH 12/159 cxl: Remove racy attempt to force EEH invocation in reset
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Daniel Axtens <[email protected]>
commit 9d8e27673c45927fee9e7d8992ffb325a6b0b0e4 upstream.
cxl_reset currently PERSTs the slot, and then repeatedly tries to read MMIO space in order to kick off EEH.
There are 2 problems with this: it's unnecessary, and it's racy.
It's unnecessary because the PERST will bring down the PHB link. That will be picked up by the CAPP, which will send out an HMI. Skiboot, noticing an HMI from the CAPP, will send an OPAL notification to the kernel, which will trigger EEH recovery.
It's also racy: the EEH recovery triggered by the CAPP will eventually cause the MMIO space to have its mapping invalidated and the pointer NULLed out. This races with our attempt to read the MMIO space. This is causing OOPSes in testing.
Simply drop all the attempts to force EEH detection, and trust that Skiboot will send the notification and that we'll act on it. The Skiboot code to send the EEH notification has been in Skiboot for as long as CAPP recovery has been supported, so we don't need to worry about breaking obscure setups with ancient firmware.
--- drivers/misc/cxl/pci.c | 16 ---------------- 1 file changed, 16 deletions(-)
drivers/misc/cxl/pci.c +0 -16
@@ -778,8 +778,6 @@ int cxl_reset(struct cxl *adapter) {struct pci_dev *dev =to_pci_dev (adapter->dev.parent);int rc; -int i; -u32 val;dev_info (&dev->dev,"CXL reset\n" );
@@ -794,20 +792,6 @@ int cxl_reset(struct cxl *adapter)return rc; } -/* the PERST done above fences the PHB. So, reset depends on EEH -* to unbind the driver, tell Sapphire to reinit the PHB, and rebind -* the driver. Do an mmio read explictly to ensure EEH notices the -* fenced PHB. Retry for a few seconds before giving up. */ - i =0 ; -while (((val =mmio_read32be (adapter->p1_mmio)) !=0xffffffff ) && - (i <5 )) { -msleep (500 ); - i++; - } - -if (val !=0xffffffff ) -dev_err (&dev->dev,"cxl: PERST failed to trigger EEH\n" ); -return rc; }
Greg Kroah-Hartman
maintainerPATCH 13/159 cxl: Fix unbalanced pci_dev_get in cxl_probe
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Daniel Axtens <[email protected]>
commit 2925c2fdf1e0eb642482f5b30577e9435aaa8edb upstream.
Currently the first thing we do in cxl_probe is to grab a reference on the pci device. Later on, we call device_register on our adapter. In our remove path, we call device_unregister, but we never call pci_dev_put. We therefore leak the device every time we do a reflash.
device_register/unregister is sufficient to hold the reference. Therefore, drop the call to pci_dev_get.
Here's why this is safe.
The proposed cxl_probe(pdev) calls cxl_adapter_init:
a) init calls cxl_adapter_alloc, which creates a struct cxl, conventionally called adapter. This struct contains a device entry, adapter->dev.
b) init calls cxl_configure_adapter, where we set adapter->dev.parent = &dev->dev (here dev is the pci dev)
So at this point, the cxl adapter's device's parent is the PCI device that I want to be refcounted properly.
c) init calls cxl_register_adapter
*) cxl_register_adapter calls device_register(&adapter->dev)
So now we're in device_register, where dev is the adapter device, and we want to know if the PCI device is safe after we return.
device_register(&adapter->dev) calls device_initialize() and then device_add().
device_add() does a get_device(). device_add() also explicitly grabs the device's parent, and calls get_device() on it:
parent = get_device(dev->parent);
So therefore, device_register() takes a lock on the parent PCI dev, which is what pci_dev_get() was guarding. pci_dev_get() can therefore be safely removed.
--- drivers/misc/cxl/pci.c | 2 -- 1 file changed, 2 deletions(-)
drivers/misc/cxl/pci.c +0 -2
@@ -1046,8 +1046,6 @@ static int cxl_probe(struct pci_dev *devint slice;int rc; -pci_dev_get (dev); -if (cxl_verbose)dump_cxl_config_space (dev);
Greg Kroah-Hartman
maintainerPATCH 14/159 ext4: dont manipulate recovery flag when freezing no-journal fs
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Eric Sandeen <[email protected]>
commit c642dc9e1aaed953597e7092d7df329e6234096e upstream.
At some point along this sequence of changes:
f6e63f9 ext4: fold ext4_nojournal_sops into ext4_sops
bb04457 ext4: support freezing ext2 (nojournal) file systems 9ca9238 ext4: Use separate super_operations structure for no_journal filesystems
ext4 started setting needs_recovery on filesystems without journals when they are unfrozen. This makes no sense, and in fact confuses blkid to the point where it doesn't recognize the filesystem at all.
(freeze ext2; unfreeze ext2; run blkid; see no output; run dumpe2fs, see needs_recovery set on fs w/ no journal).
To fix this, don't manipulate the INCOMPAT_RECOVER feature on filesystems without journals.
--- fs/ext4/super.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-)
fs/ext4/super.c +8 -4
@@ -4807,10 +4807,11 @@ static int ext4_freeze(struct super_bloc error =jbd2_journal_flush (journal);if (error <0 ) goto out; + +/* Journal blocked and flushed, clear needs_recovery flag. */ +EXT4_CLEAR_INCOMPAT_FEATURE (sb,EXT4_FEATURE_INCOMPAT_RECOVER ); } -/* Journal blocked and flushed, clear needs_recovery flag. */ -EXT4_CLEAR_INCOMPAT_FEATURE (sb,EXT4_FEATURE_INCOMPAT_RECOVER ); error =ext4_commit_super (sb,1 ); out:if (journal)
@@ -4828,8 +4829,11 @@ static int ext4_unfreeze(struct super_blif (sb->s_flags &MS_RDONLY )return 0 ; -/* Reset the needs_recovery flag before the fs is unlocked. */ -EXT4_SET_INCOMPAT_FEATURE (sb,EXT4_FEATURE_INCOMPAT_RECOVER ); +if (EXT4_SB (sb)->s_journal) { +/* Reset the needs_recovery flag before the fs is unlocked. */ +EXT4_SET_INCOMPAT_FEATURE (sb,EXT4_FEATURE_INCOMPAT_RECOVER ); + } +ext4_commit_super (sb,1 );return 0 ; }
Greg Kroah-Hartman
maintainerPATCH 15/159 Revert "ext4: remove block_device_ejected"
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Theodore Ts'o <[email protected]>
commit bdfe0cbd746aa9b2509c2f6d6be17193cf7facd7 upstream.
This reverts commit 08439fec266c3cc5702953b4f54bdf5649357de0.
Unfortunately we still need to test for bdi->dev to avoid a crash when a USB stick is yanked out while a file system is mounted:
usb 2-2: USB disconnect, device number 2
Buffer I/O error on dev sdb1, logical block 15237120, lost sync page write
JBD2: Error -5 detected when updating journal superblock for sdb1-8.
BUG: unable to handle kernel paging request at 34beb000
IP: [<c136ce88>] __percpu_counter_add+0x18/0xc0
*pdpt = 0000000023db9001 *pde = 0000000000000000
Oops: 0000 [#1] SMP
CPU: 0 PID: 4083 Comm: umount Tainted: G U OE 4.1.1-040101-generic #201507011435
Hardware name: LENOVO 7675CTO/7675CTO, BIOS 7NETC2WW (2.22 ) 03/22/2011
task: ebf06b50 ti: ebebc000 task.ti: ebebc000
EIP: 0060:[<c136ce88>] EFLAGS: 00010082 CPU: 0
EIP is at __percpu_counter_add+0x18/0xc0
EAX: f21c8e88 EBX: f21c8e88 ECX: 00000000 EDX: 00000001
ESI: 00000001 EDI: 00000000 EBP: ebebde60 ESP: ebebde40
DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
CR0: 8005003b CR2: 34beb000 CR3: 33354200 CR4: 000007f0
Stack:
c1abe100 edcb0098 edcb00ec ffffffff f21c8e68 ffffffff f21c8e68 f286d160
ebebde84 c1160454 00000010 00000282 f72a77f8 00000984 f72a77f8 f286d160
f286d170 ebebdea0 c11e613f 00000000 00000282 f72a77f8 edd7f4d0 00000000
Call Trace:
[<c1160454>] account_page_dirtied+0x74/0x110
[<c11e613f>] __set_page_dirty+0x3f/0xb0
[<c11e6203>] mark_buffer_dirty+0x53/0xc0
[<c124a0cb>] ext4_commit_super+0x17b/0x250
[<c124ac71>] ext4_put_super+0xc1/0x320
[<c11f04ba>] ? fsnotify_unmount_inodes+0x1aa/0x1c0
[<c11cfeda>] ? evict_inodes+0xca/0xe0
[<c11b925a>] generic_shutdown_super+0x6a/0xe0
[<c10a1df0>] ? prepare_to_wait_event+0xd0/0xd0
[<c1165a50>] ? unregister_shrinker+0x40/0x50
[<c11b92f6>] kill_block_super+0x26/0x70
[<c11b94f5>] deactivate_locked_super+0x45/0x80
[<c11ba007>] deactivate_super+0x47/0x60
[<c11d2b39>] cleanup_mnt+0x39/0x80
[<c11d2bc0>] __cleanup_mnt+0x10/0x20
[<c1080b51>] task_work_run+0x91/0xd0
[<c1011e3c>] do_notify_resume+0x7c/0x90
[<c1720da5>] work_notify
Code: 8b 55 e8 e9 f4 fe ff ff 90 90 90 90 90 90 90 90 90 90 90 55 89 e5 83 ec 20 89 5d f4 89 c3 89 75 f8 89 d6 89 7d fc 89 cf 8b 48 14 <64> 8b 01 89 45 ec 89 c2 8b 45 08 c1 fa 1f 01 75 ec 89 55 f0 89
EIP: [<c136ce88>] __percpu_counter_add+0x18/0xc0 SS:ESP 0068:ebebde40
CR2: 0000000034beb000
---[ end trace dd564a7bea834ecd ]---Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=101011
--- fs/ext4/super.c | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-)
fs/ext4/super.c +17 -1
@@ -324,6 +324,22 @@ static void save_error_info(struct superext4_commit_super (sb,1 ); } +/* +* The del_gendisk() function uninitializes the disk-specific data +* structures, including the bdi structure, without telling anyone +* else. Once this happens, any attempt to call mark_buffer_dirty() +* (for example, by ext4_commit_super), will cause a kernel OOPS. +* This is a kludge to prevent these oops until we can put in a proper +* hook in del_gendisk() to inform the VFS and file system layers. +*/ +static int block_device_ejected (struct super_block *sb) +{ +struct inode *bd_inode = sb->s_bdev->bd_inode; +struct backing_dev_info *bdi =inode_to_bdi (bd_inode); + +return bdi->dev ==NULL ; +} +static void ext4_journal_commit_callback (journal_t *journal,transaction_t *txn) {struct super_block *sb = journal->j_private;
@@ -4591,7 +4607,7 @@ static int ext4_commit_super(struct supestruct buffer_head *sbh =EXT4_SB (sb)->s_sbh;int error =0 ; -if (!sbh) +if (!sbh ||block_device_ejected (sb))return error;if (buffer_write_io_error (sbh)) {/*
Greg Kroah-Hartman
maintainerPATCH 16/159 arm64: kconfig: Move LIST_POISON to a safe value
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jeff Vander Stoep <[email protected]>
commit bf0c4e04732479f650ff59d1ee82de761c0071f0 upstream.
Move the poison pointer offset to 0xdead000000000000, a recognized value that is not mappable by user-space exploits.
--- arch/arm64/Kconfig | 4 ++++ 1 file changed, 4 insertions(+)
arch/arm64/Kconfig +4 -0
@@ -101,6 +101,10 @@ config NO_IOPORT_MAP config STACKTRACE_SUPPORT def_bool y +config ILLEGAL_POINTER_VALUE + hex + default 0xdead000000000000 + config LOCKDEP_SUPPORT def_bool y
Greg Kroah-Hartman
maintainerPATCH 17/159 arm64: flush FP/SIMD state correctly after execve()
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Ard Biesheuvel <[email protected]>
commit 674c242c9323d3c293fc4f9a3a3a619fe3063290 upstream.
When a task calls execve(), its FP/SIMD state is flushed so that none of the original program state is observeable by the incoming program.
However, since this flushing consists of setting the in-memory copy of the FP/SIMD state to all zeroes, the CPU field is set to CPU 0 as well, which indicates to the lazy FP/SIMD preserve/restore code that the FP/SIMD state does not need to be reread from memory if the task is scheduled again on CPU 0 without any other tasks having entered userland (or used the FP/SIMD in kernel mode) on the same CPU in the mean time. If this happens, the FP/SIMD state of the old program will still be present in the registers when the new program starts.
So set the CPU field to the invalid value of NR_CPUS when performing the flush, by calling fpsimd_flush_task_state().
--- arch/arm64/kernel/fpsimd.c | 1 + 1 file changed, 1 insertion(+)
arch/arm64/kernel/fpsimd.c +1 -0
@@ -157,6 +157,7 @@ void fpsimd_thread_switch(struct task_stvoid fpsimd_flush_thread (void ) {memset (¤t->thread.fpsimd_state,0 ,sizeof (struct fpsimd_state )); +fpsimd_flush_task_state (current);set_thread_flag (TIF_FOREIGN_FPSTATE ); }
Greg Kroah-Hartman
maintainerPATCH 18/159 of/fdt: make memblock maximum physical address arch configurable
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Ard Biesheuvel <[email protected]>
commit 8eafeb48022816513abc4f440bdad4c350fe81a3 upstream.
When parsing the memory nodes to populate the memblock memory table, we check against high and low limits and clip any memory that exceeds either one of them.
However, for arm64, the high limit of (phys_addr_t)~0 is not very meaningful, since phys_addr_t is 64 bits (i.e., no limit) but there may be other constraints that limit the memory ranges that we can support.
So rename MAX_PHYS_ADDR to MAX_MEMBLOCK_ADDR (for clarity) and only define it if the arch does not supply a definition of its own.
--- drivers/of/fdt.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-)
drivers/of/fdt.c +7 -5
@@ -955,7 +955,9 @@ int __init early_init_dt_scan_chosen(uns }#ifdef CONFIG_HAVE_MEMBLOCK -#define MAX_PHYS_ADDR ((phys_addr_t)~0) +#ifndef MAX_MEMBLOCK_ADDR +#define MAX_MEMBLOCK_ADDR ((phys_addr_t)~0) +#endif void __init __weak early_init_dt_add_memory_arch (u64 base,u64 size) {
@@ -972,16 +974,16 @@ void __init __weak early_init_dt_add_mem } size &=PAGE_MASK ; -if (base >MAX_PHYS_ADDR ) { +if (base >MAX_MEMBLOCK_ADDR ) {pr_warning ("Ignoring memory block 0x%llx - 0x%llx\n" , base, base + size);return ; } -if (base + size -1 >MAX_PHYS_ADDR ) { +if (base + size -1 >MAX_MEMBLOCK_ADDR ) {pr_warning ("Ignoring memory range 0x%llx - 0x%llx\n" , - ((u64 )MAX_PHYS_ADDR ) +1 , base + size); - size =MAX_PHYS_ADDR - base +1 ; + ((u64 )MAX_MEMBLOCK_ADDR ) +1 , base + size); + size =MAX_MEMBLOCK_ADDR - base +1 ; }if (base + size < phys_offset) {
Greg Kroah-Hartman
maintainerPATCH 19/159 arm64: set MAX_MEMBLOCK_ADDR according to linear region size
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Ard Biesheuvel <[email protected]>
commit 34ba2c4247e5c4b1542b1106e156af324660c4f0 upstream.
The linear region size of a 39-bit VA kernel is only 256 GB, which may be insufficient to cover all of system RAM, even on platforms that have much less than 256 GB of memory but which is laid out very sparsely.
So make sure we clip the memory we will not be able to map before installing it into the memblock memory table, by setting MAX_MEMBLOCK_ADDR accordingly.
--- arch/arm64/include/asm/memory.h | 8 ++++++++ 1 file changed, 8 insertions(+)
arch/arm64/include/asm/memory.h +8 -0
@@ -114,6 +114,14 @@ extern phys_addr_t memstart_addr;#define PHYS_OFFSET ({ memstart_addr; })/* +* The maximum physical address that the linear direct mapping +* of system RAM can cover. (PAGE_OFFSET can be interpreted as +* a 2's complement signed quantity and negated to derive the +* maximum size of the linear mapping.) +*/ +#define MAX_MEMBLOCK_ADDR ({ memstart_addr - PAGE_OFFSET - 1; }) + +/* * PFNs are used to describe any physical page; this means * PFN 0 == physical address 0. *
Greg Kroah-Hartman
maintainerPATCH 20/159 arm64: compat: fix vfp save/restore across signal handlers in big-endian
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Will Deacon <[email protected]>
commit bdec97a855ef1e239f130f7a11584721c9a1bf04 upstream.
When saving/restoring the VFP registers from a compat (AArch32) signal frame, we rely on the compat registers forming a prefix of the native register file and therefore make use of copy_{to,from}_user to transfer between the native fpsimd_state and the compat_vfp_sigframe.
Unfortunately, this doesn't work so well in a big-endian environment. Our fpsimd save/restore code operates directly on 128-bit quantities (Q registers) whereas the compat_vfp_sigframe represents the registers as an array of 64-bit (D) registers. The architecture packs the compat D registers into the Q registers, with the least significant bytes holding the lower register. Consequently, we need to swap the 64-bit halves when converting between these two representations on a big-endian machine.
This patch replaces the __copy_{to,from}_user invocations in our compat VFP signal handling code with explicit __put_user loops that operate on 64-bit values and swap them accordingly.
--- arch/arm64/kernel/signal32.c | 47 ++++++++++++++++++++++++++++++++----------- 1 file changed, 36 insertions(+), 11 deletions(-)
arch/arm64/kernel/signal32.c +36 -11
@@ -212,14 +212,32 @@ int copy_siginfo_from_user32(siginfo_t */* * VFP save/restore code. +* +* We have to be careful with endianness, since the fpsimd context-switch +* code operates on 128-bit (Q) register values whereas the compat ABI +* uses an array of 64-bit (D) registers. Consequently, we need to swap +* the two halves of each Q register when running on a big-endian CPU. */ +union __fpsimd_vreg { +__uint128_t raw; +struct { +#ifdef __AARCH64EB__ +u64 hi; +u64 lo; +#else +u64 lo; +u64 hi; +#endif + }; +}; +static int compat_preserve_vfp_context (struct compat_vfp_sigframe __user *frame) {struct fpsimd_state *fpsimd = ¤t->thread.fpsimd_state;compat_ulong_t magic =VFP_MAGIC ;compat_ulong_t size =VFP_STORAGE_SIZE ;compat_ulong_t fpscr, fpexc; -int err =0 ; +int i, err =0 ;/* * Save the hardware registers to the fpsimd_state structure.
@@ -235,10 +253,15 @@ static int compat_preserve_vfp_context(s/* * Now copy the FP registers. Since the registers are packed, * we can copy the prefix we want (V0-V15) as it is. -* FIXME: Won't work if big endian. */ - err |=__copy_to_user (&frame->ufp.fpregs, fpsimd->vregs, -sizeof (frame->ufp.fpregs)); +for (i =0 ; i <ARRAY_SIZE (frame->ufp.fpregs); i +=2 ) { +union __fpsimd_vreg vreg = { + .raw = fpsimd->vregs[i >>1 ], + }; + +__put_user_error (vreg.lo, &frame->ufp.fpregs[i], err); +__put_user_error (vreg.hi, &frame->ufp.fpregs[i +1 ], err); + }/* Create an AArch32 fpscr from the fpsr and the fpcr. */ fpscr = (fpsimd->fpsr & VFP_FPSCR_STAT_MASK) |
@@ -263,7 +286,7 @@ static int compat_restore_vfp_context(stcompat_ulong_t magic =VFP_MAGIC ;compat_ulong_t size =VFP_STORAGE_SIZE ;compat_ulong_t fpscr; -int err =0 ; +int i, err =0 ;__get_user_error (magic, &frame->magic, err);__get_user_error (size, &frame->size, err);
@@ -273,12 +296,14 @@ static int compat_restore_vfp_context(stif (magic !=VFP_MAGIC || size !=VFP_STORAGE_SIZE )return -EINVAL ; -/* -* Copy the FP registers into the start of the fpsimd_state. -* FIXME: Won't work if big endian. -*/ - err |=__copy_from_user (fpsimd.vregs, frame->ufp.fpregs, -sizeof (frame->ufp.fpregs)); +/* Copy the FP registers into the start of the fpsimd_state. */ +for (i =0 ; i <ARRAY_SIZE (frame->ufp.fpregs); i +=2 ) { +union __fpsimd_vreg vreg; + +__get_user_error (vreg.lo, &frame->ufp.fpregs[i], err); +__get_user_error (vreg.hi, &frame->ufp.fpregs[i +1 ], err); + fpsimd.vregs[i >>1 ] = vreg.raw; + }/* Extract the fpsr and the fpcr from the fpscr */ __get_user_error (fpscr, &frame->ufp.fpscr, err);
Greg Kroah-Hartman
maintainerPATCH 21/159 arm64: head.S: initialise mdcr_el2 in el2_setup
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Will Deacon <[email protected]>
commit d10bcd473301888f957ec4b6b12aa3621be78d59 upstream.
When entering the kernel at EL2, we fail to initialise the MDCR_EL2 register which controls debug access and PMU capabilities at EL1.
This patch ensures that the register is initialised so that all traps are disabled and all the PMU counters are available to the host. When a guest is scheduled, KVM takes care to configure trapping appropriately.
--- arch/arm64/kernel/head.S | 5 +++++ 1 file changed, 5 insertions(+)
arch/arm64/kernel/head.S +5 -0
@@ -565,6 +565,11 @@ CPU_LE( movk x0, #0x30d0, lsl #16 ) // C msr hstr_el2, xzr // Disable CP15 traps to EL2 #endif + /* EL2 debug */ + mrs x0, pmcr_el0 // Disable debug access traps + ubfx x0, x0, #11, #5 // to EL2 and allow access to + msr mdcr_el2, x0 // all PMU counters from EL1 + /* Stage-2 translation */ msr vttbr_el2, xzr
Greg Kroah-Hartman
maintainerPATCH 22/159 arm64: errata: add module build workaround for erratum #843419
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Will Deacon <[email protected]>
commit df057cc7b4fa59e9b55f07ffdb6c62bf02e99a00 upstream.
Cortex-A53 processors <= r0p4 are affected by erratum #843419 which can lead to a memory access using an incorrect address in certain sequences headed by an ADRP instruction.
There is a linker fix to generate veneers for ADRP instructions, but this doesn't work for kernel modules which are built as unlinked ELF objects.
This patch adds a new config option for the erratum which, when enabled, builds kernel modules with the mcmodel=large flag. This uses absolute addressing for all kernel symbols, thereby removing the use of ADRP as a PC-relative form of addressing. The ADRP relocs are removed from the module loader so that we fail to load any potentially affected modules.
--- arch/arm64/Kconfig | 16 ++++++++++++++++ arch/arm64/Makefile | 4 ++++ arch/arm64/kernel/module.c | 2 ++ 3 files changed, 22 insertions(+)
arch/arm64/Kconfig +16 -0
@@ -413,6 +413,22 @@ config ARM64_ERRATUM_845719 If unsure, say Y. +config ARM64_ERRATUM_843419 + bool "Cortex-A53: 843419: A load or store might access an incorrect address" + depends on MODULES + default y + help + This option builds kernel modules using the large memory model in + order to avoid the use of the ADRP instruction, which can cause + a subsequent memory access to use an incorrect address on Cortex-A53 + parts up to r0p4. + + Note that the kernel itself must be linked with a version of ld + which fixes potentially affected ADRP instructions through the + use of veneers. + + If unsure, say Y. + endmenu
arch/arm64/Makefile +4 -0
@@ -30,6 +30,10 @@ endif CHECKFLAGS += -D__aarch64__ +ifeq ($(CONFIG_ARM64_ERRATUM_843419), y) +CFLAGS_MODULE += -mcmodel=large +endif + # Default value head-y := arch/arm64/kernel/head.o
arch/arm64/kernel/module.c +2 -0
@@ -332,12 +332,14 @@ int apply_relocate_add(Elf64_Shdr *sechd ovf =reloc_insn_imm (RELOC_OP_PREL , loc, val,0 ,21 ,AARCH64_INSN_IMM_ADR );break ; +#ifndef CONFIG_ARM64_ERRATUM_843419 case R_AARCH64_ADR_PREL_PG_HI21_NC : overflow_check = false;case R_AARCH64_ADR_PREL_PG_HI21 : ovf =reloc_insn_imm (RELOC_OP_PAGE , loc, val,12 ,21 ,AARCH64_INSN_IMM_ADR );break ; +#endif case R_AARCH64_ADD_ABS_LO12_NC :case R_AARCH64_LDST8_ABS_LO12_NC : overflow_check = false;
Greg Kroah-Hartman
maintainerPATCH 23/159 arm/arm64: KVM: vgic: Check for !irqchip_in_kernel() when mapping resources
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Pavel Fedin <[email protected]>
commit c2f58514cfb374d5368c9da945f1765cd48eb0da upstream.
Until b26e5fdac43c ("arm/arm64: KVM: introduce per-VM ops"), kvm_vgic_map_resources() used to include a check on irqchip_in_kernel(), and vgic_v2_map_resources() still has it.
But now vm_ops are not initialized until we call kvm_vgic_create(). Therefore kvm_vgic_map_resources() can being called without a VGIC, and we die because vm_ops.map_resources is NULL.
Fixing this restores QEMU's kernel-irqchip=off option to a working state, allowing to use GIC emulation in userspace.
--- arch/arm/kvm/arm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/kvm/arm.c +1 -1
@@ -450,7 +450,7 @@ static int kvm_vcpu_first_run_init(struc* Map the VGIC hardware resources before running a vcpu the first * time on this VM. */ -if (unlikely (!vgic_ready (kvm))) { +if (unlikely (irqchip_in_kernel (kvm) && !vgic_ready (kvm))) { ret =kvm_vgic_map_resources (kvm);if (ret)return ret;
Greg Kroah-Hartman
maintainerPATCH 24/159 KVM: arm64: add workaround for Cortex-A57 erratum #852523
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Will Deacon <[email protected]>
commit 43297dda0a51e4ffed0888ce727c218cfb7474b6 upstream.
When restoring the system register state for an AArch32 guest at EL2, writes to DACR32_EL2 may not be correctly synchronised by Cortex-A57, which can lead to the guest effectively running with junk in the DACR and running into unexpected domain faults.
This patch works around the issue by re-ordering our restoration of the AArch32 register aliases so that they happen before the AArch64 system registers. Ensuring that the registers are restored in this order guarantees that they will be correctly synchronised by the core.
--- arch/arm64/kvm/hyp.S | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
arch/arm64/kvm/hyp.S +3 -1
@@ -947,13 +947,15 @@ ENTRY(__kvm_vcpu_run) // Guest context add x2, x0, #VCPU_CONTEXT + // We must restore the 32-bit state before the sysregs, thanks + // to Cortex-A57 erratum #852523. + restore_guest_32bit_state bl __restore_sysregs bl __restore_fpsimd skip_debug_state x3, 1f bl __restore_debug 1: - restore_guest_32bit_state restore_guest_regs // That's it, no more messing around.
Greg Kroah-Hartman
maintainerPATCH 25/159 arm64: KVM: Disable virtual timer even if the guest is not using it
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Marc Zyngier <[email protected]>
commit c4cbba9fa078f55d9f6d081dbb4aec7cf969e7c7 upstream.
When running a guest with the architected timer disabled (with QEMU and the kernel_irqchip=off option, for example), it is important to make sure the timer gets turned off. Otherwise, the guest may try to enable it anyway, leading to a screaming HW interrupt.
The fix is to unconditionally turn off the virtual timer on guest exit.
--- arch/arm64/kvm/hyp.S | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
arch/arm64/kvm/hyp.S +3 -2
@@ -844,8 +844,6 @@ mrs x3, cntv_ctl_el0 and x3, x3, #3 str w3, [x0, #VCPU_TIMER_CNTV_CTL] - bic x3, x3, #1 // Clear Enable - msr cntv_ctl_el0, x3 isb
@@ -853,6 +851,9 @@ str x3, [x0, #VCPU_TIMER_CNTV_CVAL] 1: + // Disable the virtual timer + msr cntv_ctl_el0, xzr + // Allow physical timer/counter access for the host mrs x2, cnthctl_el2 orr x2, x2, #3
Greg Kroah-Hartman
maintainerPATCH 26/159 Input: synaptics - fix handling of disabling gesture mode
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Dmitry Torokhov <[email protected]>
commit e51e38494a8ecc18650efb0c840600637891de2c upstream.
Bit 2 of the mode byte has dual meaning: it can disable reporting of gestures when touchpad works in Relative mode or normal Absolute mode, or it can enable so called Extended W-Mode when touchpad uses enhanced Absolute mode (W-mode). The extended W-Mode confuses our driver and causes missing button presses on some Thinkpads (x250, T450s), so let's make sure we do not enable it.
Also, according to the spec W mode "... bit is defined only in Absolute mode on pads whose capExtended capability bit is set. In Relative mode and in TouchPads without this capability, the bit is reserved and should be left at 0.", so let's make sure we respect this requirement as well.
--- drivers/input/mouse/synaptics.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-)
drivers/input/mouse/synaptics.c +8 -4
@@ -519,14 +519,18 @@ static int synaptics_set_mode(struct psmstruct synaptics_data *priv = psmouse->private; priv->mode =0 ; -if (priv->absolute_mode) + +if (priv->absolute_mode) { priv->mode |=SYN_BIT_ABSOLUTE_MODE ; -if (priv->disable_gesture) +if (SYN_CAP_EXTENDED (priv->capabilities)) + priv->mode |=SYN_BIT_W_MODE ; + } + +if (!SYN_MODE_WMODE (priv->mode) && priv->disable_gesture) priv->mode |=SYN_BIT_DISABLE_GESTURE ; +if (psmouse->rate >=80 ) priv->mode |=SYN_BIT_HIGH_RATE ; -if (SYN_CAP_EXTENDED (priv->capabilities)) - priv->mode |=SYN_BIT_W_MODE ;if (synaptics_mode_cmd (psmouse, priv->mode))return -1 ;
Re: [PATCH 4.1 026/159] Input: synaptics - fix handling of disabling gesture mode
Quoted 20 lines · Greg Kroah-Hartman
4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Dmitry Torokhov <[email protected]>
commit
e51e38494a8ecc18650efb0c840600637891de2cupstream.Bit 2 of the mode byte has dual meaning: it can disable reporting of gestures when touchpad works in Relative mode or normal Absolute mode, or it can enable so called Extended W-Mode when touchpad uses enhanced Absolute mode (W-mode). The extended W-Mode confuses our driver and causes missing button presses on some Thinkpads (x250, T450s), so let's make sure we do not enable it.
Also, according to the spec W mode "... bit is defined only in Absolute mode on pads whose capExtended capability bit is set. In Relative mode and in TouchPads without this capability, the bit is reserved and should be left at 0.", so let's make sure we respect this requirement as well.
Quoted 6 lines
Reported-by Nick Bowler <[email protected]>Suggested-by Gabor Balla <[email protected]>Tested-by Gabor Balla <[email protected]>Tested-by Nick Bowler <[email protected]>Signed-off-by Dmitry Torokhov <[email protected]>Signed-off-by Greg Kroah-Hartman <[email protected]>
I believe Dmitry is going to revert this commit very shortly. See
http://www.spinics.net/lists/linux-input/msg41176.html
You might want to leave this one out of both 4.2.y and 4.1.1y.
josh
Quoted 36 lines
--- drivers/input/mouse/synaptics.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-)drivers/input/mouse/synaptics.c +8
-5@@ -519,14 +519,18 @@ static int synaptics_set_mode(struct psmstruct synaptics_data *priv = psmouse->private; priv->mode =0 ; -if (priv->absolute_mode) + +if (priv->absolute_mode) { priv->mode |=SYN_BIT_ABSOLUTE_MODE ; -if (priv->disable_gesture) +if (SYN_CAP_EXTENDED (priv->capabilities)) + priv->mode |=SYN_BIT_W_MODE ; + } + +if (!SYN_MODE_WMODE (priv->mode) && priv->disable_gesture) priv->mode |=SYN_BIT_DISABLE_GESTURE ; +if (psmouse->rate >=80 ) priv->mode |=SYN_BIT_HIGH_RATE ; -if (SYN_CAP_EXTENDED (priv->capabilities)) - priv->mode |=SYN_BIT_W_MODE ;if (synaptics_mode_cmd (psmouse, priv->mode))return -1 ; --To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Greg Kroah-Hartman
maintainerRe: [PATCH 4.1 026/159] Input: synaptics - fix handling of disabling gesture mode
Quoted 29 lines · Josh Boyer
Quoted 20 lines · Greg Kroah-Hartman
4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Dmitry Torokhov <[email protected]>
commit
e51e38494a8ecc18650efb0c840600637891de2cupstream.Bit 2 of the mode byte has dual meaning: it can disable reporting of gestures when touchpad works in Relative mode or normal Absolute mode, or it can enable so called Extended W-Mode when touchpad uses enhanced Absolute mode (W-mode). The extended W-Mode confuses our driver and causes missing button presses on some Thinkpads (x250, T450s), so let's make sure we do not enable it.
Also, according to the spec W mode "... bit is defined only in Absolute mode on pads whose capExtended capability bit is set. In Relative mode and in TouchPads without this capability, the bit is reserved and should be left at 0.", so let's make sure we respect this requirement as well.
Quoted 6 lines
Reported-by Nick Bowler <[email protected]>Suggested-by Gabor Balla <[email protected]>Tested-by Gabor Balla <[email protected]>Tested-by Nick Bowler <[email protected]>Signed-off-by Dmitry Torokhov <[email protected]>Signed-off-by Greg Kroah-Hartman <[email protected]>
Quoted 5 lines
I believe Dmitry is going to revert this commit very shortly. See
http://www.spinics.net/lists/linux-input/msg41176.html
You might want to leave this one out of both 4.2.y and 4.1.1y.
I prefer to wait for stuff like this to hit Linus's tree to keep in sync, bugs at all at times.
thanks,
greg k-h
Re: [PATCH 4.1 026/159] Input: synaptics - fix handling of disabling gesture mode
Quoted 30 lines · Greg Kroah-Hartman
Quoted 29 lines · Josh Boyer
Quoted 26 lines · Greg Kroah-Hartman
4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Dmitry Torokhov <[email protected]>
commit
e51e38494a8ecc18650efb0c840600637891de2cupstream.Bit 2 of the mode byte has dual meaning: it can disable reporting of gestures when touchpad works in Relative mode or normal Absolute mode, or it can enable so called Extended W-Mode when touchpad uses enhanced Absolute mode (W-mode). The extended W-Mode confuses our driver and causes missing button presses on some Thinkpads (x250, T450s), so let's make sure we do not enable it.
Also, according to the spec W mode "... bit is defined only in Absolute mode on pads whose capExtended capability bit is set. In Relative mode and in TouchPads without this capability, the bit is reserved and should be left at 0.", so let's make sure we respect this requirement as well.
Reported-by Nick Bowler <[email protected]>Suggested-by Gabor Balla <[email protected]>Tested-by Gabor Balla <[email protected]>Tested-by Nick Bowler <[email protected]>Signed-off-by Dmitry Torokhov <[email protected]>Signed-off-by Greg Kroah-Hartman <[email protected]>
Quoted 8 lines
Quoted 5 lines
I believe Dmitry is going to revert this commit very shortly. See
http://www.spinics.net/lists/linux-input/msg41176.html
You might want to leave this one out of both 4.2.y and 4.1.1y.
I prefer to wait for stuff like this to hit Linus's tree to keep in sync, bugs at all at times.
Wait, what? You're going to release a stable kernel with a patch that is known to be buggy just to keep it in sync with a buggy upstream Linus tree? That doesn't make sense to me. I would maybe understand if the upstream solution wasn't "revert this" and instead had a follow on patch, but knowing upstream is going to revert and still including it is confusing.
At any rate, distros may wish to queue up the revert on top of 4.1.9 and 4.2.2.
josh
Greg Kroah-Hartman
maintainerRe: [PATCH 4.1 026/159] Input: synaptics - fix handling of disabling gesture mode
Quoted 41 lines · Josh Boyer
Quoted 30 lines · Greg Kroah-Hartman
Quoted 29 lines · Josh Boyer
Quoted 26 lines · Greg Kroah-Hartman
4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Dmitry Torokhov <[email protected]>
commit
e51e38494a8ecc18650efb0c840600637891de2cupstream.Bit 2 of the mode byte has dual meaning: it can disable reporting of gestures when touchpad works in Relative mode or normal Absolute mode, or it can enable so called Extended W-Mode when touchpad uses enhanced Absolute mode (W-mode). The extended W-Mode confuses our driver and causes missing button presses on some Thinkpads (x250, T450s), so let's make sure we do not enable it.
Also, according to the spec W mode "... bit is defined only in Absolute mode on pads whose capExtended capability bit is set. In Relative mode and in TouchPads without this capability, the bit is reserved and should be left at 0.", so let's make sure we respect this requirement as well.
Reported-by Nick Bowler <[email protected]>Suggested-by Gabor Balla <[email protected]>Tested-by Gabor Balla <[email protected]>Tested-by Nick Bowler <[email protected]>Signed-off-by Dmitry Torokhov <[email protected]>Signed-off-by Greg Kroah-Hartman <[email protected]>Quoted 8 lines
Quoted 5 lines
I believe Dmitry is going to revert this commit very shortly. See
http://www.spinics.net/lists/linux-input/msg41176.html
You might want to leave this one out of both 4.2.y and 4.1.1y.
I prefer to wait for stuff like this to hit Linus's tree to keep in sync, bugs at all at times.
Quoted 6 lines
Wait, what? You're going to release a stable kernel with a patch that is known to be buggy just to keep it in sync with a buggy upstream Linus tree? That doesn't make sense to me. I would maybe understand if the upstream solution wasn't "revert this" and instead had a follow on patch, but knowing upstream is going to revert and still including it is confusing.
We do this all the time, as the patch usually takes a while to get reverted in Linus's tree, and there doesn't seem to be any "rush" at the moment to get it reverted, I usually just leave things as-is if for no other reason than to wake the maintainer up :)
Unless the maintainer asks me not to include it, then I'll reconsider, otherwise I have to trust that a random person says that the patch will be reverted some unknown time in the future by some other person, and that's nothing I can really count on.
Again, keeping the trees in sync, even for bugs, makes things easier overall.
thanks,
greg k-h
Re: [PATCH 4.1 026/159] Input: synaptics - fix handling of disabling gesture mode
Quoted 67 lines · Greg Kroah-Hartman
Quoted 6 lines · Josh Boyer
Quoted 4 lines · Greg Kroah-Hartman
Quoted 3 lines · Josh Boyer
Quoted 1 line · Greg Kroah-Hartman
4.1-stable review patch. If anyone has any objections, please
let me know.
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
------------------
From: Dmitry Torokhov <[email protected]>
commit
e51e38494a8ecc18650efb0c840600637891de2cupstream.Bit 2 of the mode byte has dual meaning: it can disable
reporting of
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
gestures when touchpad works in Relative mode or normal Absolute
mode,
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
or it can enable so called Extended W-Mode when touchpad uses
enhanced
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Absolute mode (W-mode). The extended W-Mode confuses our driver
and
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
causes missing button presses on some Thinkpads (x250, T450s),
so let's
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
make sure we do not enable it.
Also, according to the spec W mode "... bit is defined only in
Absolute
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
mode on pads whose capExtended capability bit is set. In
Relative mode and
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
in TouchPads without this capability, the bit is reserved and
should be
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
left at 0.", so let's make sure we respect this requirement as
well.
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 7 lines
Reported-by Nick Bowler <[email protected]>Suggested-by Gabor Balla <[email protected]>Tested-by Gabor Balla <[email protected]>Tested-by Nick Bowler <[email protected]>Signed-off-by Dmitry Torokhov <[email protected]>Signed-off-by Greg Kroah-Hartman <[email protected]>I believe Dmitry is going to revert this commit very shortly. See
Quoted 8 lines
Quoted 6 lines
Quoted 3 lines
http://www.spinics.net/lists/linux-input/msg41176.html
You might want to leave this one out of both 4.2.y and 4.1.1y.
I prefer to wait for stuff like this to hit Linus's tree to keep in sync, bugs at all at times.
Wait, what? You're going to release a stable kernel with a patch
that
Quoted 3 lines
is known to be buggy just to keep it in sync with a buggy upstream Linus tree? That doesn't make sense to me. I would maybe understand if the upstream solution wasn't "revert this" and instead had a
follow
Quoted 2 lines
on patch, but knowing upstream is going to revert and still including it is confusing.
We do this all the time, as the patch usually takes a while to get reverted in Linus's tree, and there doesn't seem to be any "rush" at the
moment to get it reverted, I usually just leave things as-is if for no other reason than to wake the maintainer up :)
Quoted 7 lines
Unless the maintainer asks me not to include it, then I'll reconsider, otherwise I have to trust that a random person says that the patch will be reverted some unknown time in the future by some other person, and that's nothing I can really count on.
Again, keeping the trees in sync, even for bugs, makes things easier overall.
Greg,
I'd rather we did not merge this path into stable just yet. I'm not sure if I'll revert it outright or if there will be a fixup, but as is it messes up touchpad behavior for many people :(
Thanks.
Signature · 1 line
Dmitry
Greg Kroah-Hartman
maintainerPATCH 27/159 Input: evdev - do not report errors form flush()
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Takashi Iwai <[email protected]>
commit eb38f3a4f6e86f8bb10a3217ebd85ecc5d763aae upstream.
We've got bug reports showing the old systemd-logind (at least system-210) aborting unexpectedly, and this turned out to be because of an invalid error code from close() call to evdev devices. close() is supposed to return only either EINTR or EBADFD, while the device returned ENODEV. logind was overreacting to it and decided to kill itself when an unexpected error code was received. What a tragedy.
The bad error code comes from flush fops, and actually evdev_flush() returns ENODEV when device is disconnected or client's access to it is revoked. But in these cases the fact that flush did not actually happen is not an error, but rather normal behavior. For non-disconnected devices result of flush is also not that interesting as there is no potential of data loss and even if it fails application has no way of handling the error. Because of that we are better off always returning success from evdev_flush().
Also returning EINTR from flush()/close() is discouraged (as it is not clear how application should handle this error), so let's stop taking evdev->mutex interruptibly.
Bugzilla: http://bugzilla.suse.com/show_bug.cgi?id=939834
--- drivers/input/evdev.c | 13 ++++--------- 1 file changed, 4 insertions(+), 9 deletions(-)
drivers/input/evdev.c +4 -9
@@ -290,19 +290,14 @@ static int evdev_flush(struct file *file {struct evdev_client *client = file->private_data;struct evdev *evdev = client->evdev; -int retval; - retval =mutex_lock_interruptible (&evdev->mutex); -if (retval) -return retval; +mutex_lock (&evdev->mutex); -if (!evdev->exist || client->revoked) - retval = -ENODEV ; -else - retval =input_flush_device (&evdev->handle, file); +if (evdev->exist && !client->revoked) +input_flush_device (&evdev->handle, file);mutex_unlock (&evdev->mutex); -return retval; +return 0 ; }static void evdev_free (struct device *dev)
Greg Kroah-Hartman
maintainerPATCH 28/159 ALSA: usb-audio: correct the value cache check.
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Yao-Wen Mao <[email protected]>
commit 6aa6925cad06159dc6e25857991bbc4960821242 upstream.
The check of cval->cached should be zero-based (including master channel).
--- sound/usb/mixer.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
sound/usb/mixer.c +1 -1
@@ -2522,7 +2522,7 @@ static int restore_mixer_value(struct usfor (c =0 ; c <MAX_CHANNELS ; c++) {if (!(cval->cmask & (1 << c)))continue ; -if (cval->cached & (1 << c)) { +if (cval->cached & (1 << (c +1 ))) { err =snd_usb_set_cur_mix_value (cval, c +1 , idx, cval->cache_val[idx]);if (err <0 )
Greg Kroah-Hartman
maintainerPATCH 29/159 ALSA: hda - Enable headphone jack detect on old Fujitsu laptops
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Takashi Iwai <[email protected]>
commit bb148bdeb0ab16fc0ae8009799471e4d7180073b upstream.
According to the bug report, FSC Amilo laptops with ALC880 can detect the headphone jack but currently the driver disables it. It's partly intentionally, as non-working jack detect was reported in the past. Let's enable now.
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=102501
--- sound/pci/hda/patch_realtek.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
sound/pci/hda/patch_realtek.c +2 -2
@@ -1134,7 +1134,7 @@ static const struct hda_fixup alc880_fix/* override all pins as BIOS on old Amilo is broken */ .type = HDA_FIXUP_PINS, .v.pins = (const struct hda_pintbl []) { - {0x14 ,0x0121411f },/* HP */ + {0x14 ,0x0121401f },/* HP */ {0x15 ,0x99030120 },/* speaker */ {0x16 ,0x99030130 },/* bass speaker */ {0x17 ,0x411111f0 },/* N/A */
@@ -1154,7 +1154,7 @@ static const struct hda_fixup alc880_fix/* almost compatible with FUJITSU, but no bass and SPDIF */ .type = HDA_FIXUP_PINS, .v.pins = (const struct hda_pintbl []) { - {0x14 ,0x0121411f },/* HP */ + {0x14 ,0x0121401f },/* HP */ {0x15 ,0x99030120 },/* speaker */ {0x16 ,0x411111f0 },/* N/A */ {0x17 ,0x411111f0 },/* N/A */
Greg Kroah-Hartman
maintainerPATCH 30/159 ALSA: hda - Use ALC880_FIXUP_FUJITSU for FSC Amilo M1437
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Takashi Iwai <[email protected]>
commit a161574e200ae63a5042120e0d8c36830e81bde3 upstream.
It turned out that the machine has a bass speaker, so take a correct fixup entry.
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=102501
--- sound/pci/hda/patch_realtek.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
sound/pci/hda/patch_realtek.c +1 -1
@@ -1363,7 +1363,7 @@ static const struct snd_pci_quirk alc880SND_PCI_QUIRK (0x161f ,0x203d ,"W810" ,ALC880_FIXUP_W810 ),SND_PCI_QUIRK (0x161f ,0x205d ,"Medion Rim 2150" ,ALC880_FIXUP_MEDION_RIM ),SND_PCI_QUIRK (0x1631 ,0xe011 ,"PB 13201056" ,ALC880_FIXUP_6ST_AUTOMUTE ), -SND_PCI_QUIRK (0x1734 ,0x107c ,"FSC F1734" ,ALC880_FIXUP_F1734 ), +SND_PCI_QUIRK (0x1734 ,0x107c ,"FSC Amilo M1437" ,ALC880_FIXUP_FUJITSU ),SND_PCI_QUIRK (0x1734 ,0x1094 ,"FSC Amilo M1451G" ,ALC880_FIXUP_FUJITSU ),SND_PCI_QUIRK (0x1734 ,0x10ac ,"FSC AMILO Xi 1526" ,ALC880_FIXUP_F1734 ), SND_PCI_QUIRK(0x1734, 0x10b0, "FSC Amilo Pi1556", ALC880_FIXUP_FUJITSU),
Greg Kroah-Hartman
maintainerPATCH 31/159 ALSA: hda - Add some FIXUP quirks for white noise on Dell laptop.
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Woodrow Shen <[email protected]>
commit 1adecc6755e1e4193b5618ddb2e107f6d6e88f4b upstream.
Dell laptop has a series model to use the same codec but different subsystem ID. At the same time they happens the white noise by login screen and headphone;
for fixing them together, I only can add these IDs to FIXUP function ALC292_FIXUP_DISABLE_AAMIX, then try to solve such the similar issues.
Codec: Realtek ALC3235
Vendor Id: 0x10ec0293
Subsystem Id: 0x102806dd
Subsystem Id: 0x102806df
Subsystem Id: 0x102806e0
--- sound/pci/hda/patch_realtek.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
sound/pci/hda/patch_realtek.c +4 -1
@@ -5118,8 +5118,11 @@ static const struct snd_pci_quirk alc269SND_PCI_QUIRK (0x1028 ,0x06c7 ,"Dell" ,ALC255_FIXUP_DELL1_MIC_NO_PRESENCE ),SND_PCI_QUIRK (0x1028 ,0x06d9 ,"Dell" ,ALC293_FIXUP_DELL1_MIC_NO_PRESENCE ),SND_PCI_QUIRK (0x1028 ,0x06da ,"Dell" ,ALC293_FIXUP_DELL1_MIC_NO_PRESENCE ), -SND_PCI_QUIRK (0x1028 ,0x06de ,"Dell" ,ALC292_FIXUP_DISABLE_AAMIX ),SND_PCI_QUIRK (0x1028 ,0x06db ,"Dell" ,ALC292_FIXUP_DISABLE_AAMIX ), +SND_PCI_QUIRK (0x1028 ,0x06dd ,"Dell" ,ALC292_FIXUP_DISABLE_AAMIX ), +SND_PCI_QUIRK (0x1028 ,0x06de ,"Dell" ,ALC292_FIXUP_DISABLE_AAMIX ), +SND_PCI_QUIRK (0x1028 ,0x06df ,"Dell" ,ALC292_FIXUP_DISABLE_AAMIX ), +SND_PCI_QUIRK (0x1028 ,0x06e0 ,"Dell" ,ALC292_FIXUP_DISABLE_AAMIX ),SND_PCI_QUIRK (0x1028 ,0x164a ,"Dell" ,ALC293_FIXUP_DELL1_MIC_NO_PRESENCE ),SND_PCI_QUIRK (0x1028 ,0x164b ,"Dell" ,ALC293_FIXUP_DELL1_MIC_NO_PRESENCE ), SND_PCI_QUIRK(0x103c, 0x1586, "HP", ALC269_FIXUP_HP_MUTE_LED_MIC2),
Greg Kroah-Hartman
maintainerPATCH 32/159 ALSA: hda - Fix white noise on Dell M3800
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Niranjan Sivakumar <[email protected]>
commit 467e1436ba85f78b8c4610c4549eb255a8211c42 upstream.
The M3800 is very minor workstation variant of the XPS 15 which has already been patched for this issue. I figured it's probably more important for this version of the laptop to be patched than the regular XPS as Dell sells is pre-configured with Ubuntu to be used as a Linux workstation. I have tested the patch on my the hardware on Linux 4.2.0.
--- sound/pci/hda/patch_realtek.c | 1 + 1 file changed, 1 insertion(+)
sound/pci/hda/patch_realtek.c +1 -0
@@ -6457,6 +6457,7 @@ static const struct snd_pci_quirk alc662SND_PCI_QUIRK (0x1028 ,0x05db ,"Dell" ,ALC668_FIXUP_DELL_MIC_NO_PRESENCE ),SND_PCI_QUIRK (0x1028 ,0x05fe ,"Dell XPS 15" ,ALC668_FIXUP_DELL_XPS13 ),SND_PCI_QUIRK (0x1028 ,0x060a ,"Dell XPS 13" ,ALC668_FIXUP_DELL_XPS13 ), +SND_PCI_QUIRK (0x1028 ,0x060d ,"Dell M3800" ,ALC668_FIXUP_DELL_XPS13 ),SND_PCI_QUIRK (0x1028 ,0x0625 ,"Dell" ,ALC668_FIXUP_DELL_MIC_NO_PRESENCE ),SND_PCI_QUIRK (0x1028 ,0x0626 ,"Dell" ,ALC668_FIXUP_DELL_MIC_NO_PRESENCE ), SND_PCI_QUIRK(0x1028, 0x0696, "Dell", ALC668_FIXUP_DELL_MIC_NO_PRESENCE),
Greg Kroah-Hartman
maintainerPATCH 33/159 pinctrl: at91: fix null pointer dereference
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: David Dueck <[email protected]>
commit 1ab36387ea4face01aac3560b396b1e2ce07c4ff upstream.
Not all gpio banks are necessarily enabled, in the current code this can lead to null pointer dereferences.
[ 51.130000] Unable to handle kernel NULL pointer dereference at virtual address 00000058 [ 51.130000] pgd = dee04000 [ 51.130000] [00000058] *pgd=3f66d831, *pte=00000000, *ppte=00000000 [ 51.140000] Internal error: Oops: 17 [#1] ARM [ 51.140000] Modules linked in: [ 51.140000] CPU: 0 PID: 1664 Comm: cat Not tainted 4.1.1+ #6 [ 51.140000] Hardware name: Atmel SAMA5 [ 51.140000] task: df6dd880 ti: dec60000 task.ti: dec60000 [ 51.140000] PC is at at91_pinconf_get+0xb4/0x200 [ 51.140000] LR is at at91_pinconf_get+0xb4/0x200 [ 51.140000] pc : [<c01e71a0>] lr : [<c01e71a0>] psr: 600f0013 sp : dec61e48 ip : 600f0013 fp : df522538 [ 51.140000] r10: df52250c r9 : 00000058 r8 : 00000068 [ 51.140000] r7 : 00000000 r6 : df53c910 r5 : 00000000 r4 : dec61e7c [ 51.140000] r3 : 00000000 r2 : c06746d4 r1 : 00000000 r0 : 00000003 [ 51.140000] Flags: nZCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user [ 51.140000] Control: 10c53c7d Table: 3ee04059 DAC: 00000015 [ 51.140000] Process cat (pid: 1664, stack limit = 0xdec60208) [ 51.140000] Stack: (0xdec61e48 to 0xdec62000) [ 51.140000] 1e40: 00000358 00000000 df522500 ded15f80 c05a9d08 ded15f80 [ 51.140000] 1e60: 0000048c 00000061 df522500 ded15f80 c05a9d08 c01e7304 ded15f80 00000000 [ 51.140000] 1e80: c01e6008 00000060 0000048c c01e6034 c01e5f6c ded15f80 dec61ec0 00000000 [ 51.140000] 1ea0: 00020000 ded6f280 dec61f80 00000001 00000001 c00ae0b8 b6e80000 ded15fb0 [ 51.140000] 1ec0: 00000000 00000000 df4bc974 00000055 00000800 ded6f280 b6e80000 ded6f280 [ 51.140000] 1ee0: ded6f280 00020000 b6e80000 00000000 00020000 c0090dec c0671e1c dec61fb0 [ 51.140000] 1f00: b6f8b510 00000001 00004201 c000924c 00000000 00000003 00000003 00000000 [ 51.140000] 1f20: df4bc940 00022000 00000022 c066e188 b6e7f000 c00836f4 000b6e7f ded6f280 [ 51.140000] 1f40: ded6f280 b6e80000 dec61f80 ded6f280 00020000 c0091508 00000000 00000003 [ 51.140000] 1f60: 00022000 00000000 00000000 ded6f280 ded6f280 00020000 b6e80000 c0091d9c [ 51.140000] 1f80: 00000000 00000000 ffffffff 00020000 00020000 b6e80000 00000003 c000f124 [ 51.140000] 1fa0: dec60000 c000efa0 00020000 00020000 00000003 b6e80000 00020000 000271c4 [ 51.140000] 1fc0: 00020000 00020000 b6e80000 00000003 7fffe000 00000000 00000000 00020000 [ 51.140000] 1fe0: 00000000 bef50b64 00013835 b6f29c76 400f0030 00000003 00000000 00000000 [ 51.140000] [<c01e71a0>] (at91_pinconf_get) from [<c01e7304>] (at91_pinconf_dbg_show+0x18/0x2c0) [ 51.140000] [<c01e7304>] (at91_pinconf_dbg_show) from [<c01e6034>] (pinconf_pins_show+0xc8/0xf8) [ 51.140000] [<c01e6034>] (pinconf_pins_show) from [<c00ae0b8>] (seq_read+0x1a0/0x464) [ 51.140000] [<c00ae0b8>] (seq_read) from [<c0090dec>] (__vfs_read+0x20/0xd0) [ 51.140000] [<c0090dec>] (__vfs_read) from [<c0091508>] (vfs_read+0x7c/0x108) [ 51.140000] [<c0091508>] (vfs_read) from [<c0091d9c>] (SyS_read+0x40/0x94) [ 51.140000] [<c0091d9c>] (SyS_read) from [<c000efa0>] (ret_fast_syscall+0x0/0x3c) [ 51.140000] Code: eb010ec2 e30a0d08 e34c005a eb0ae5a7 (e5993000) [ 51.150000] ---[ end trace fb3c370da3ea4794 ]---
--- drivers/pinctrl/pinctrl-at91.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+)
drivers/pinctrl/pinctrl-at91.c +15 -0
@@ -320,6 +320,9 @@ static const struct pinctrl_ops at91_pctstatic void __iomem *pin_to_controller (struct at91_pinctrl *info,unsigned int bank) { +if (!gpio_chips[bank]) +return NULL ; +return gpio_chips[bank]->regbase; }
@@ -729,6 +732,10 @@ static int at91_pmx_set(struct pinctrl_d pin = &pins_conf[i];at91_pin_dbg (info->dev, pin); pio =pin_to_controller (info, pin->bank); + +if (!pio) +continue ; + mask =pin_to_mask (pin->pin);at91_mux_disable_interrupt (pio, mask);switch (pin->mux) {
@@ -848,6 +855,10 @@ static int at91_pinconf_get(struct pinct *config =0 ;dev_dbg (info->dev,"%s:%d, pin_id=%d" , __func__, __LINE__, pin_id); pio =pin_to_controller (info,pin_to_bank (pin_id)); + +if (!pio) +return -EINVAL ; + pin = pin_id %MAX_NB_GPIO_PER_BANK ;if (at91_mux_get_multidrive (pio, pin))
@@ -889,6 +900,10 @@ static int at91_pinconf_set(struct pinct"%s:%d, pin_id=%d, config=0x%lx" , __func__, __LINE__, pin_id, config); pio =pin_to_controller (info,pin_to_bank (pin_id)); + +if (!pio) +return -EINVAL ; + pin = pin_id %MAX_NB_GPIO_PER_BANK ; mask =pin_to_mask (pin);
Greg Kroah-Hartman
maintainerPATCH 34/159 powerpc/pseries: Fix corrupted pdn list
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Gavin Shan <[email protected]>
commit 590c7567a2895f939525ead57b0334c6d47986f0 upstream.
Commit cca87d30 ("powerpc/pci: Refactor pci_dn") introduced pdn list for SRIOV VFs. It means the pdn is be put into the child list of its parent pdn when the pdn is created. When doing PCI hot unplugging on pSeries, the PCI device node as well as its pdn are released through procfs entry "powerpc/ofdt". Some one else grabs the memory chunk of the pdn and update it accordingly. At the same time, the pdn is still tracked in the child list of parent pdn. It leads to corrupted child list in the parent pdn.
This fixes above issue by removing the pdn from the child list of its parent pdn when the device node is detached from the system. Note the pdn is free'd when the device node is released if the device node is dynamic one. Otherwise, the device node as well as the pdn won't be released.
--- arch/powerpc/platforms/pseries/setup.c | 5 +++++ 1 file changed, 5 insertions(+)
arch/powerpc/platforms/pseries/setup.c +5 -0
@@ -268,6 +268,11 @@ static int pci_dn_reconfig_notifier(strueeh_dev_init (PCI_DN (np), pci->phb); }break ; +case OF_RECONFIG_DETACH_NODE : + pci =PCI_DN (np); +if (pci) +list_del (&pci->list); +break ;default : err =NOTIFY_DONE ;break ;
Greg Kroah-Hartman
maintainerPATCH 35/159 powerpc/eeh: Probe after unbalanced kref check
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Daniel Axtens <[email protected]>
commit e642d11bdbfe8eb10116ab3959a2b5d75efda832 upstream.
In the complete hotplug case, EEH PEs are supposed to be released and set to NULL. Normally, this is done by eeh_remove_device(), which is called from pcibios_release_device().
However, if something is holding a kref to the device, it will not be released, and the PE will remain. eeh_add_device_late() has a check for this which will explictly destroy the PE in this case.
This check in eeh_add_device_late() occurs after a call to eeh_ops->probe(). On PowerNV, probe is a pointer to pnv_eeh_probe(), which will exit without probing if there is an existing PE.
This means that on PowerNV, devices with outstanding krefs will not be rediscovered by EEH correctly after a complete hotplug. This is affecting CXL (CAPI) devices in the field.
Put the probe after the kref check so that the PE is destroyed and affected devices are correctly rediscovered by EEH.
--- arch/powerpc/kernel/eeh.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
arch/powerpc/kernel/eeh.c +3 -3
@@ -1118,9 +1118,6 @@ void eeh_add_device_late(struct pci_devreturn ; } -if (eeh_has_flag (EEH_PROBE_MODE_DEV )) - eeh_ops->probe (pdn,NULL ); -/* * The EEH cache might not be removed correctly because of * unbalanced kref to the device during unplug time, which
@@ -1144,6 +1141,9 @@ void eeh_add_device_late(struct pci_dev dev->dev.archdata.edev =NULL ; } +if (eeh_has_flag (EEH_PROBE_MODE_DEV )) + eeh_ops->probe (pdn,NULL ); + edev->pdev = dev; dev->dev.archdata.edev = edev;
Greg Kroah-Hartman
maintainerPATCH 36/159 powerpc/eeh: Fix fenced PHB caused by eeh_slot_error_detail()
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Gavin Shan <[email protected]>
commit 259800135c654a098d9f0adfdd3d1f20eef1f231 upstream.
The config space of some PCI devices can't be accessed when their PEs are in frozen state. Otherwise, fenced PHB might be seen. Those PEs are identified with flag EEH_PE_CFG_RESTRICTED, meaing EEH_PE_CFG_BLOCKED is set automatically when the PE is put to frozen state (EEH_PE_ISOLATED). eeh_slot_error_detail() restores PCI device BARs with eeh_pe_restore_bars(), which then calls eeh_ops->restore_config() to reinitialize the PCI device in (OPAL) firmware. eeh_ops->restore_config() produces PCI config access that causes fenced PHB. The problem was reported on below adapter:
0001:01:00.0 0200: 14e4:168e (rev 10)
0001:01:00.0 Ethernet controller: Broadcom Corporation \
NetXtreme II BCM57810 10 Gigabit Ethernet (rev 10)
This fixes the issue by skipping eeh_pe_restore_bars() in eeh_slot_error_detail() when EEH_PE_CFG_BLOCKED is set for the PE.
--- arch/powerpc/kernel/eeh.c | 21 ++++++++++++++++++--- 1 file changed, 18 insertions(+), 3 deletions(-)
arch/powerpc/kernel/eeh.c +18 -3
@@ -310,11 +310,26 @@ void eeh_slot_error_detail(struct eeh_peif (!(pe->type &EEH_PE_PHB )) {if (eeh_has_flag (EEH_ENABLE_IO_FOR_LOG ))eeh_pci_enable (pe,EEH_OPT_THAW_MMIO ); + +/* +* The config space of some PCI devices can't be accessed +* when their PEs are in frozen state. Otherwise, fenced +* PHB might be seen. Those PEs are identified with flag +* EEH_PE_CFG_RESTRICTED, indicating EEH_PE_CFG_BLOCKED +* is set automatically when the PE is put to EEH_PE_ISOLATED. +* +* Restoring BARs possibly triggers PCI config access in +* (OPAL) firmware and then causes fenced PHB. If the +* PCI config is blocked with flag EEH_PE_CFG_BLOCKED, it's +* pointless to restore BARs and dump config space. +*/ eeh_ops->configure_bridge (pe); -eeh_pe_restore_bars (pe); +if (!(pe->state &EEH_PE_CFG_BLOCKED )) { +eeh_pe_restore_bars (pe); - pci_regs_buf[0 ] =0 ; -eeh_pe_traverse (pe, eeh_dump_pe_log, &loglen); + pci_regs_buf[0 ] =0 ; +eeh_pe_traverse (pe, eeh_dump_pe_log, &loglen); + } } eeh_ops->get_log (pe, severity, pci_regs_buf, loglen);
Greg Kroah-Hartman
maintainerPATCH 37/159 powerpc/mm: Fix pte_pagesize_index() crash on 4K w/64K hash
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Michael Ellerman <[email protected]>
commit 74b5037baa2011a2799e2c43adde7d171b072f9e upstream.
The powerpc kernel can be built to have either a 4K PAGE_SIZE or a 64K PAGE_SIZE.
However when built with a 4K PAGE_SIZE there is an additional config option which can be enabled, PPC_HAS_HASH_64K, which means the kernel also knows how to hash a 64K page even though the base PAGE_SIZE is 4K.
This is used in one obscure configuration, to support 64K pages for SPU local store on the Cell processor when the rest of the kernel is using 4K pages.
In this configuration, pte_pagesize_index() is defined to just pass through its arguments to get_slice_psize(). However pte_pagesize_index() is called for both user and kernel addresses, whereas get_slice_psize() only knows how to handle user addresses.
This has been broken forever, however until recently it happened to work. That was because in get_slice_psize() the large kernel address would cause the right shift of the slice mask to return zero.
However in commit 7aa0727f3302 ("powerpc/mm: Increase the slice range to 64TB"), the get_slice_psize() code was changed so that instead of a right shift we do an array lookup based on the address. When passed a kernel address this means we index way off the end of the slice array and return random junk.
That is only fatal if we happen to hit something non-zero, but when we do return a non-zero value we confuse the MMU code and eventually cause a check stop.
This fix is ugly, but simple. When we're called for a kernel address we return 4K, which is always correct in this configuration, otherwise we use the slice mask.
--- arch/powerpc/include/asm/pgtable-ppc64.h | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-)
arch/powerpc/include/asm/pgtable-ppc64.h +13 -1
@@ -135,7 +135,19 @@#define pte_iterate_hashed_end () } while(0)#ifdef CONFIG_PPC_HAS_HASH_64K -#define pte_pagesize_index (mm, addr, pte) get_slice_psize(mm, addr) +/* +* We expect this to be called only for user addresses or kernel virtual +* addresses other than the linear mapping. +*/ +#define pte_pagesize_index (mm, addr, pte) \ + ({ \ + unsigned int psize; \ + if (is_kernel_addr(addr)) \ + psize = MMU_PAGE_4K; \ + else \ + psize = get_slice_psize(mm, addr); \ + psize; \ + })#else #define pte_pagesize_index (mm, addr, pte) MMU_PAGE_4K#endif
Greg Kroah-Hartman
maintainerPATCH 38/159 powerpc/rtas: Introduce rtas_get_sensor_fast() for IRQ handlers
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Thomas Huth <[email protected]>
commit 1c2cb594441d02815d304cccec9742ff5c707495 upstream.
The EPOW interrupt handler uses rtas_get_sensor(), which in turn uses rtas_busy_delay() to wait for RTAS becoming ready in case it is necessary. But rtas_busy_delay() is annotated with might_sleep() and thus may not be used by interrupts handlers like the EPOW handler! This leads to the following BUG when CONFIG_DEBUG_ATOMIC_SLEEP is enabled:
BUG: sleeping function called from invalid context at arch/powerpc/kernel/rtas.c:496 in_atomic(): 1, irqs_disabled(): 1, pid: 0, name: swapper/1 CPU: 1 PID: 0 Comm: swapper/1 Not tainted 4.2.0-rc2-thuth #6 Call Trace: [c00000007ffe7b90] [c000000000807670] dump_stack+0xa0/0xdc (unreliable) [c00000007ffe7bc0] [c0000000000e1f14] ___might_sleep+0x134/0x180 [c00000007ffe7c20] [c00000000002aec0] rtas_busy_delay+0x30/0xd0 [c00000007ffe7c50] [c00000000002bde4] rtas_get_sensor+0x74/0xe0 [c00000007ffe7ce0] [c000000000083264] ras_epow_interrupt+0x44/0x450 [c00000007ffe7d90] [c000000000120260] handle_irq_event_percpu+0xa0/0x300 [c00000007ffe7e70] [c000000000120524] handle_irq_event+0x64/0xc0 [c00000007ffe7eb0] [c000000000124dbc] handle_fasteoi_irq+0xec/0x260 [c00000007ffe7ef0] [c00000000011f4f0] generic_handle_irq+0x50/0x80 [c00000007ffe7f20] [c000000000010f3c] __do_irq+0x8c/0x200 [c00000007ffe7f90] [c0000000000236cc] call_do_irq+0x14/0x24 [c00000007e6f39e0] [c000000000011144] do_IRQ+0x94/0x110 [c00000007e6f3a30] [c000000000002594] hardware_interrupt_common+0x114/0x180
Fix this issue by introducing a new rtas_get_sensor_fast() function that does not use rtas_busy_delay() - and thus can only be used for sensors that do not cause a BUSY condition - known as "fast" sensors.
The EPOW sensor is defined to be "fast" in sPAPR - mpe.
--- arch/powerpc/include/asm/rtas.h | 1 + arch/powerpc/kernel/rtas.c | 17 +++++++++++++++++ arch/powerpc/platforms/pseries/ras.c | 3 ++- 3 files changed, 20 insertions(+), 1 deletion(-)
arch/powerpc/include/asm/rtas.h +1 -0
@@ -343,6 +343,7 @@ extern void rtas_power_off(void);extern void rtas_halt (void );extern void rtas_os_term (char *str);extern int rtas_get_sensor (int sensor,int index,int *state); +extern int rtas_get_sensor_fast (int sensor,int index,int *state);extern int rtas_get_power_level (int powerdomain,int *level);extern int rtas_set_power_level (int powerdomain,int level,int *setlevel);extern bool rtas_indicator_present (int token,int *maxindex);
arch/powerpc/kernel/rtas.c +17 -0
@@ -584,6 +584,23 @@ int rtas_get_sensor(int sensor, int inde }EXPORT_SYMBOL (rtas_get_sensor); +int rtas_get_sensor_fast (int sensor,int index,int *state) +{ +int token =rtas_token ("get-sensor-state" ); +int rc; + +if (token ==RTAS_UNKNOWN_SERVICE ) +return -ENOENT ; + + rc =rtas_call (token,2 ,2 , state, sensor, index); +WARN_ON (rc ==RTAS_BUSY || (rc >=RTAS_EXTENDED_DELAY_MIN && + rc <=RTAS_EXTENDED_DELAY_MAX )); + +if (rc <0 ) +return rtas_error_rc (rc); +return rc; +} +bool rtas_indicator_present (int token,int *maxindex) {int proplen, count, i;
arch/powerpc/platforms/pseries/ras.c +2 -1
@@ -189,7 +189,8 @@ static irqreturn_t ras_epow_interrupt(inint state;int critical; - status =rtas_get_sensor (EPOW_SENSOR_TOKEN ,EPOW_SENSOR_INDEX , &state); + status =rtas_get_sensor_fast (EPOW_SENSOR_TOKEN ,EPOW_SENSOR_INDEX , + &state);if (state >3 ) critical =1 ;/* Time Critical */
Greg Kroah-Hartman
maintainerPATCH 39/159 powerpc: Uncomment and make enable_kernel_vsx() routine available
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Leonidas Da Silva Barbosa <[email protected]>
commit 72cd7b44bc99376b3f3c93cedcd052663fcdf705 upstream.
enable_kernel_vsx() function was commented since anything was using it. However, vmx-crypto driver uses VSX instructions which are only available if VSX is enable. Otherwise it rises an exception oops.
This patch uncomment enable_kernel_vsx() routine and makes it available.
--- arch/powerpc/include/asm/switch_to.h | 1 + arch/powerpc/kernel/process.c | 3 --- 2 files changed, 1 insertion(+), 3 deletions(-)
arch/powerpc/include/asm/switch_to.h +1 -0
@@ -29,6 +29,7 @@ static inline void save_early_sprs(strucextern void enable_kernel_fp (void );extern void enable_kernel_altivec (void ); +extern void enable_kernel_vsx (void );extern int emulate_altivec (struct pt_regs *);extern void __giveup_vsx (struct task_struct *);extern void giveup_vsx (struct task_struct *);
arch/powerpc/kernel/process.c +0 -3
@@ -204,8 +204,6 @@ EXPORT_SYMBOL_GPL(flush_altivec_to_threa#endif /* CONFIG_ALTIVEC */ #ifdef CONFIG_VSX -#if 0 -/* not currently used, but some crazy RAID module might want to later */void enable_kernel_vsx (void ) {WARN_ON (preemptible ());
@@ -220,7 +218,6 @@ void enable_kernel_vsx(void)#endif /* CONFIG_SMP */ }EXPORT_SYMBOL (enable_kernel_vsx); -#endif void giveup_vsx (struct task_struct *tsk) {
Greg Kroah-Hartman
maintainerPATCH 40/159 crypto: vmx - Adding enable_kernel_vsx() to access VSX instructions
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Leonidas Da Silva Barbosa <[email protected]>
commit 2d6f0600b2cd755959527230ef5a6fba97bb762a upstream.
vmx-crypto driver make use of some VSX instructions which are only available if VSX is enabled. Running in cases where VSX are not enabled vmx-crypto fails in a VSX exception.
In order to fix this enable_kernel_vsx() was added to turn on VSX instructions for vmx-crypto.
--- drivers/crypto/vmx/aes.c | 3 +++ drivers/crypto/vmx/aes_cbc.c | 3 +++ drivers/crypto/vmx/aes_ctr.c | 3 +++ drivers/crypto/vmx/ghash.c | 4 ++++ 4 files changed, 13 insertions(+)
drivers/crypto/vmx/aes.c +3 -0
@@ -80,6 +80,7 @@ static int p8_aes_setkey(struct crypto_tpagefault_disable ();enable_kernel_altivec (); +enable_kernel_vsx (); ret =aes_p8_set_encrypt_key (key, keylen *8 , &ctx->enc_key); ret +=aes_p8_set_decrypt_key (key, keylen *8 , &ctx->dec_key);pagefault_enable ();
@@ -97,6 +98,7 @@ static void p8_aes_encrypt(struct crypto }else {pagefault_disable ();enable_kernel_altivec (); +enable_kernel_vsx ();aes_p8_encrypt (src, dst, &ctx->enc_key);pagefault_enable (); }
@@ -111,6 +113,7 @@ static void p8_aes_decrypt(struct crypto }else {pagefault_disable ();enable_kernel_altivec (); +enable_kernel_vsx ();aes_p8_decrypt (src, dst, &ctx->dec_key);pagefault_enable (); }
drivers/crypto/vmx/aes_cbc.c +3 -0
@@ -81,6 +81,7 @@ static int p8_aes_cbc_setkey(struct cryppagefault_disable ();enable_kernel_altivec (); +enable_kernel_vsx (); ret =aes_p8_set_encrypt_key (key, keylen *8 , &ctx->enc_key); ret +=aes_p8_set_decrypt_key (key, keylen *8 , &ctx->dec_key);pagefault_enable ();
@@ -108,6 +109,7 @@ static int p8_aes_cbc_encrypt(struct blk }else {pagefault_disable ();enable_kernel_altivec (); +enable_kernel_vsx ();blkcipher_walk_init (&walk, dst, src, nbytes); ret =blkcipher_walk_virt (desc, &walk);
@@ -143,6 +145,7 @@ static int p8_aes_cbc_decrypt(struct blk }else {pagefault_disable ();enable_kernel_altivec (); +enable_kernel_vsx ();blkcipher_walk_init (&walk, dst, src, nbytes); ret =blkcipher_walk_virt (desc, &walk);
drivers/crypto/vmx/aes_ctr.c +3 -0
@@ -79,6 +79,7 @@ static int p8_aes_ctr_setkey(struct cryppagefault_disable ();enable_kernel_altivec (); +enable_kernel_vsx (); ret =aes_p8_set_encrypt_key (key, keylen *8 , &ctx->enc_key);pagefault_enable ();
@@ -97,6 +98,7 @@ static void p8_aes_ctr_final(struct p8_apagefault_disable ();enable_kernel_altivec (); +enable_kernel_vsx ();aes_p8_encrypt (ctrblk, keystream, &ctx->enc_key);pagefault_enable ();
@@ -127,6 +129,7 @@ static int p8_aes_ctr_crypt(struct blkciwhile ((nbytes = walk.nbytes) >=AES_BLOCK_SIZE ) {pagefault_disable ();enable_kernel_altivec (); +enable_kernel_vsx ();aes_p8_ctr32_encrypt_blocks (walk.src.virt.addr, walk.dst.virt.addr, (nbytes &AES_BLOCK_MASK )/AES_BLOCK_SIZE , &ctx->enc_key, walk.iv);pagefault_enable ();
drivers/crypto/vmx/ghash.c +4 -0
@@ -116,6 +116,7 @@ static int p8_ghash_setkey(struct cryptopagefault_disable ();enable_kernel_altivec (); +enable_kernel_vsx ();enable_kernel_fp ();gcm_init_p8 (ctx->htable, (const u64 *) key);pagefault_enable ();
@@ -142,6 +143,7 @@ static int p8_ghash_update(struct shash_ GHASH_DIGEST_SIZE - dctx->bytes);pagefault_disable ();enable_kernel_altivec (); +enable_kernel_vsx ();enable_kernel_fp ();gcm_ghash_p8 (dctx->shash, ctx->htable, dctx->buffer,GHASH_DIGEST_SIZE );
@@ -154,6 +156,7 @@ static int p8_ghash_update(struct shash_if (len) {pagefault_disable ();enable_kernel_altivec (); +enable_kernel_vsx ();enable_kernel_fp ();gcm_ghash_p8 (dctx->shash, ctx->htable, src, len);pagefault_enable ();
@@ -182,6 +185,7 @@ static int p8_ghash_final(struct shash_d dctx->buffer[i] =0 ;pagefault_disable ();enable_kernel_altivec (); +enable_kernel_vsx ();enable_kernel_fp ();gcm_ghash_p8 (dctx->shash, ctx->htable, dctx->buffer,GHASH_DIGEST_SIZE );
Greg Kroah-Hartman
maintainerPATCH 41/159 powerpc/boot: Specify ABI v2 when building an LE boot wrapper
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Benjamin Herrenschmidt <[email protected]>
commit 655471f54c2e395ba29ae4156ba0f49928177cc1 upstream.
The kernel does it, not the boot wrapper, which breaks with some cross compilers that still default to ABI v1.
--- arch/powerpc/boot/Makefile | 3 +++ 1 file changed, 3 insertions(+)
arch/powerpc/boot/Makefile +3 -0
@@ -28,6 +28,9 @@ BOOTCFLAGS += -m64 endif ifdef CONFIG_CPU_BIG_ENDIAN BOOTCFLAGS += -mbig-endian +else +BOOTCFLAGS += -mlittle-endian +BOOTCFLAGS += $(call cc-option,-mabi=elfv2) endif BOOTAFLAGS := -D__ASSEMBLY__ $(BOOTCFLAGS) -traditional -nostdinc
Greg Kroah-Hartman
maintainerPATCH 42/159 powerpc/mm: Recompute hash value after a failed update
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: "Aneesh Kumar K.V" <[email protected]>
commit 36b35d5d807b7e57aff7d08e63de8b17731ee211 upstream.
If we had secondary hash flag set, we ended up modifying hash value in the updatepp code path. Hence with a failed updatepp we will be using a wrong hash value for the following hash insert. Fix this by recomputing hash before insert.
Without this patch we can end up with using wrong slot number in linux pte. That can result in us missing an hash pte update or invalidate which can cause memory corruption or even machine check.
--- arch/powerpc/mm/hugepage-hash64.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
arch/powerpc/mm/hugepage-hash64.c +2 -1
@@ -85,7 +85,6 @@ int __hash_page_thp(unsigned long ea, unBUG_ON (index >=4096 ); vpn =hpt_vpn (ea, vsid, ssize); - hash =hpt_hash (vpn, shift, ssize); hpte_slot_array =get_hpte_slot_array (pmdp);if (psize ==MMU_PAGE_4K ) {/*
@@ -101,6 +100,7 @@ int __hash_page_thp(unsigned long ea, un valid =hpte_valid (hpte_slot_array, index);if (valid) {/* update the hpte bits */ + hash =hpt_hash (vpn, shift, ssize); hidx =hpte_hash_index (hpte_slot_array, index);if (hidx & _PTEIDX_SECONDARY) hash = ~hash;
@@ -126,6 +126,7 @@ int __hash_page_thp(unsigned long ea, unif (!valid) {unsigned long hpte_group; + hash =hpt_hash (vpn, shift, ssize);/* insert new entry */ pa =pmd_pfn (__pmd (old_pmd)) <<PAGE_SHIFT ; new_pmd |= _PAGE_HASHPTE;
Greg Kroah-Hartman
maintainerPATCH 43/159 CIFS: fix type confusion in copy offload ioctl
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jann Horn <[email protected]>
commit 4c17a6d56bb0cad3066a714e94f7185a24b40f49 upstream.
This might lead to local privilege escalation (code execution as kernel) for systems where the following conditions are met:
- CONFIG_CIFS_SMB2 and CONFIG_CIFS_POSIX are enabled
- a cifs filesystem is mounted where:
- the mount option "vers" was used and set to a value >=2.0
- the attacker has write access to at least one file on the filesystem
To attack this, an attacker would have to guess the target_tcon pointer (but guessing wrong doesn't cause a crash, it just returns an error code) and win a narrow race.
--- fs/cifs/ioctl.c | 6 ++++++ 1 file changed, 6 insertions(+)
fs/cifs/ioctl.c +6 -0
@@ -67,6 +67,12 @@ static long cifs_ioctl_clone(unsigned in goto out_drop_write; } +if (src_file.file->f_op->unlocked_ioctl != cifs_ioctl) { + rc = -EBADF ; +cifs_dbg (VFS ,"src file seems to be from a different filesystem type\n" ); + goto out_fput; + } +if ((!src_file.file->private_data) || (!dst_file->private_data)) { rc = -EBADF ;cifs_dbg (VFS ,"missing cifsFileInfo on copy range src file\n" );
Greg Kroah-Hartman
maintainerPATCH 44/159 Add radeon suspend/resume quirk for HP Compaq dc5750.
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jeffery Miller <[email protected]>
commit 09bfda10e6efd7b65bcc29237bee1765ed779657 upstream.
With the radeon driver loaded the HP Compaq dc5750 Small Form Factor machine fails to resume from suspend. Adding a quirk similar to other devices avoids the problem and the system resumes properly.
--- drivers/gpu/drm/radeon/radeon_combios.c | 8 ++++++++ 1 file changed, 8 insertions(+)
drivers/gpu/drm/radeon/radeon_combios.c +8 -0
@@ -3387,6 +3387,14 @@ void radeon_combios_asic_init(struct drm rdev->pdev->subsystem_device == 0x30ae)return ; +/* quirk for rs4xx HP Compaq dc5750 Small Form Factor to make it resume +* - it hangs on resume inside the dynclk 1 table. +*/ +if (rdev->family ==CHIP_RS480 && + rdev->pdev->subsystem_vendor ==0x103c && + rdev->pdev->subsystem_device ==0x280a ) +return ; +/* DYN CLK 1 */ table =combios_get_table_offset (dev,COMBIOS_DYN_CLK_1_TABLE );if (table)
Greg Kroah-Hartman
maintainerPATCH 45/159 mm: check if section present during memory block registering
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Yinghai Lu <[email protected]>
commit 04697858d89e4bf2650364f8d6956e2554e8ef88 upstream.
Tony Luck found on his setup, if memory block size 512M will cause crash during booting.
BUG: unable to handle kernel paging request at ffffea0074000020
IP: get_nid_for_pfn+0x17/0x40
PGD 128ffcb067 PUD 128ffc9067 PMD 0
Oops: 0000 [#1] SMP
Modules linked in:
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.2.0-rc8 #1
...
Call Trace:
? register_mem_sect_under_node+0x66/0xe0
register_one_node+0x17b/0x240
? pci_iommu_alloc+0x6e/0x6e
topology_init+0x3c/0x95
do_one_initcall+0xcd/0x1f0The system has non continuous RAM address:
BIOS-e820: [mem 0x0000001300000000-0x0000001cffffffff] usable
BIOS-e820: [mem 0x0000001d70000000-0x0000001ec7ffefff] usable
BIOS-e820: [mem 0x0000001f00000000-0x0000002bffffffff] usable
BIOS-e820: [mem 0x0000002c18000000-0x0000002d6fffefff] usable
BIOS-e820: [mem 0x0000002e00000000-0x00000039ffffffff] usable
So there are start sections in memory block not present. For example:
memory block : [0x2c18000000, 0x2c20000000) 512M
first three sections are not present.
The current register_mem_sect_under_node() assume first section is present, but memory block section number range [start_section_nr, end_section_nr] would include not present section.
For arch that support vmemmap, we don't setup memmap for struct page area within not present sections area.
So skip the pfn range that belong to absent section.
--- drivers/base/node.c | 10 ++++++++++ 1 file changed, 10 insertions(+)
drivers/base/node.c +10 -0
@@ -388,6 +388,16 @@ int register_mem_sect_under_node(structfor (pfn = sect_start_pfn; pfn <= sect_end_pfn; pfn++) {int page_nid; +/* +* memory block could have several absent sections from start. +* skip pfn range from absent section +*/ +if (!pfn_present (pfn)) { + pfn =round_down (pfn +PAGES_PER_SECTION , +PAGES_PER_SECTION ) -1 ; +continue ; + } + page_nid =get_nid_for_pfn (pfn);if (page_nid <0 )continue ;
Greg Kroah-Hartman
maintainerPATCH 46/159 x86/mm: Initialize pmd_idx in page_table_range_init_count()
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Minfei Huang <[email protected]>
commit 9962eea9e55f797f05f20ba6448929cab2a9f018 upstream.
The variable pmd_idx is not initialized for the first iteration of the for loop.
Assign the proper value which indexes the start address.
--- arch/x86/mm/init_32.c | 1 + 1 file changed, 1 insertion(+)
arch/x86/mm/init_32.c +1 -0
@@ -137,6 +137,7 @@ page_table_range_init_count(unsigned lon vaddr = start; pgd_idx =pgd_index (vaddr); + pmd_idx =pmd_index (vaddr);for ( ; (pgd_idx <PTRS_PER_PGD ) && (vaddr != end); pgd_idx++) { for (; (pmd_idx < PTRS_PER_PMD) && (vaddr != end);
Greg Kroah-Hartman
maintainerPATCH 47/159 mm: make page pfmemalloc check more robust
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Michal Hocko <[email protected]>
commit 2f064f3485cd29633ad1b3cfb00cc519509a3d72 upstream.
Commit c48a11c7ad26 ("netvm: propagate page->pfmemalloc to skb") added checks for page->pfmemalloc to __skb_fill_page_desc():
if (page->pfmemalloc && !page->mapping)
skb->pfmemalloc = true;It assumes page->mapping == NULL implies that page->pfmemalloc can be trusted. However, __delete_from_page_cache() can set set page->mapping to NULL and leave page->index value alone. Due to being in union, a non-zero page->index will be interpreted as true page->pfmemalloc.
So the assumption is invalid if the networking code can see such a page. And it seems it can. We have encountered this with a NFS over loopback setup when such a page is attached to a new skbuf. There is no copying going on in this case so the page confuses __skb_fill_page_desc which interprets the index as pfmemalloc flag and the network stack drops packets that have been allocated using the reserves unless they are to be queued on sockets handling the swapping which is the case here and that leads to hangs when the nfs client waits for a response from the server which has been dropped and thus never arrive.
The struct page is already heavily packed so rather than finding another hole to put it in, let's do a trick instead. We can reuse the index again but define it to an impossible value (-1UL). This is the page index so it should never see the value that large. Replace all direct users of page->pfmemalloc by page_is_pfmemalloc which will hide this nastiness from unspoiled eyes.
The information will get lost if somebody wants to use page->index obviously but that was the case before and the original code expected that the information should be persisted somewhere else if that is really needed (e.g. what SLAB and SLUB do).
--- drivers/net/ethernet/intel/fm10k/fm10k_main.c | 2 - drivers/net/ethernet/intel/igb/igb_main.c | 2 - drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 2 - drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 2 - include/linux/mm.h | 28 ++++++++++++++++++++++ include/linux/mm_types.h | 9 ------- include/linux/skbuff.h | 14 +++-------- mm/page_alloc.c | 7 +++-- mm/slab.c | 4 +-- mm/slub.c | 2 - net/core/skbuff.c | 2 - 11 files changed, 46 insertions(+), 28 deletions(-)
drivers/net/ethernet/intel/fm10k/fm10k_main.c +1 -1
@@ -216,7 +216,7 @@ static void fm10k_reuse_rx_page(struct fstatic inline bool fm10k_page_is_reserved (struct page *page) { -return (page_to_nid (page) !=numa_mem_id ()) || page->pfmemalloc; +return (page_to_nid (page) !=numa_mem_id ()) ||page_is_pfmemalloc (page); } static bool fm10k_can_reuse_rx_page(struct fm10k_rx_buffer *rx_buffer,
drivers/net/ethernet/intel/igb/igb_main.c +1 -1
@@ -6596,7 +6596,7 @@ static void igb_reuse_rx_page(struct igbstatic inline bool igb_page_is_reserved (struct page *page) { -return (page_to_nid (page) !=numa_mem_id ()) || page->pfmemalloc; +return (page_to_nid (page) !=numa_mem_id ()) ||page_is_pfmemalloc (page); } static bool igb_can_reuse_rx_page(struct igb_rx_buffer *rx_buffer,
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c +1 -1
@@ -1829,7 +1829,7 @@ static void ixgbe_reuse_rx_page(struct istatic inline bool ixgbe_page_is_reserved (struct page *page) { -return (page_to_nid (page) !=numa_mem_id ()) || page->pfmemalloc; +return (page_to_nid (page) !=numa_mem_id ()) ||page_is_pfmemalloc (page); }/**
drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c +1 -1
@@ -765,7 +765,7 @@ static void ixgbevf_reuse_rx_page(structstatic inline bool ixgbevf_page_is_reserved (struct page *page) { -return (page_to_nid (page) !=numa_mem_id ()) || page->pfmemalloc; +return (page_to_nid (page) !=numa_mem_id ()) ||page_is_pfmemalloc (page); }/**
include/linux/mm.h +28 -0
@@ -1002,6 +1002,34 @@ static inline int page_mapped(struct pag }/* +* Return true only if the page has been allocated with +* ALLOC_NO_WATERMARKS and the low watermark was not +* met implying that the system is under some pressure. +*/ +static inline bool page_is_pfmemalloc (struct page *page) +{ +/* +* Page index cannot be this large so this must be +* a pfmemalloc page. +*/ +return page->index ==-1UL ; +} + +/* +* Only to be called by the page allocator on a freshly allocated +* page. +*/ +static inline void set_page_pfmemalloc (struct page *page) +{ + page->index =-1UL ; +} + +static inline void clear_page_pfmemalloc (struct page *page) +{ + page->index =0 ; +} + +/* * Different kinds of faults, as returned by handle_mm_fault(). * Used to decide whether a process gets delivered SIGBUS or * just gets major/minor fault counters bumped up.
include/linux/mm_types.h +0 -9
@@ -63,15 +63,6 @@ struct page {union {pgoff_t index;/* Our offset within mapping. */ void *freelist;/* sl[aou]b first free object */ -bool pfmemalloc;/* If set by the page allocator, -* ALLOC_NO_WATERMARKS was set -* and the low watermark was not -* met implying that the system -* is under some pressure. The -* caller should try ensure -* this page is only used to -* free other pages. -*/ };union {
include/linux/skbuff.h +5 -9
@@ -1590,20 +1590,16 @@ static inline void __skb_fill_page_desc(skb_frag_t *frag = &skb_shinfo (skb)->frags[i];/* -* Propagate page->pfmemalloc to the skb if we can. The problem is -* that not all callers have unique ownership of the page. If -* pfmemalloc is set, we check the mapping as a mapping implies -* page->index is set (index and pfmemalloc share space). -* If it's a valid mapping, we cannot use page->pfmemalloc but we -* do not lose pfmemalloc information as the pages would not be -* allocated using __GFP_MEMALLOC. +* Propagate page pfmemalloc to the skb if we can. The problem is +* that not all callers have unique ownership of the page but rely +* on page_is_pfmemalloc doing the right thing(tm). */ frag->page.p = page; frag->page_offset = off;skb_frag_size_set (frag, size); page =compound_head (page); -if (page->pfmemalloc && !page->mapping) +if (page_is_pfmemalloc (page)) skb->pfmemalloc = true; }
@@ -2250,7 +2246,7 @@ static inline struct page *dev_alloc_pagstatic inline void skb_propagate_pfmemalloc (struct page *page,struct sk_buff *skb) { -if (page && page->pfmemalloc) +if (page_is_pfmemalloc (page)) skb->pfmemalloc = true; }
mm/page_alloc.c +5 -2
@@ -983,12 +983,15 @@ static int prep_new_page(struct page *paset_page_owner (page, order, gfp_flags);/* -* page->pfmemalloc is set when ALLOC_NO_WATERMARKS was necessary to +* page is set pfmemalloc when ALLOC_NO_WATERMARKS was necessary to * allocate the page. The expectation is that the caller is taking * steps that will free more memory. The caller should avoid the page * being used for !PFMEMALLOC purposes. */ - page->pfmemalloc = !!(alloc_flags &ALLOC_NO_WATERMARKS ); +if (alloc_flags &ALLOC_NO_WATERMARKS ) +set_page_pfmemalloc (page); +else +clear_page_pfmemalloc (page);return 0 ; }
mm/slab.c +2 -2
@@ -1602,7 +1602,7 @@ static struct page *kmem_getpages(struct }/* Record if ALLOC_NO_WATERMARKS was set when allocating the slab */ -if (unlikely (page->pfmemalloc)) +if (page_is_pfmemalloc (page)) pfmemalloc_active = true; nr_pages = (1 << cachep->gfporder);
@@ -1613,7 +1613,7 @@ static struct page *kmem_getpages(structadd_zone_page_state (page_zone (page),NR_SLAB_UNRECLAIMABLE , nr_pages);__SetPageSlab (page); -if (page->pfmemalloc) +if (page_is_pfmemalloc (page))SetPageSlabPfmemalloc (page);if (kmemcheck_enabled && !(cachep->flags &SLAB_NOTRACK )) {
mm/slub.c +1 -1
@@ -1427,7 +1427,7 @@ static struct page *new_slab(struct kmeminc_slabs_node (s,page_to_nid (page), page->objects); page->slab_cache = s;__SetPageSlab (page); -if (page->pfmemalloc) +if (page_is_pfmemalloc (page))SetPageSlabPfmemalloc (page); start =page_address (page);
net/core/skbuff.c +1 -1
@@ -340,7 +340,7 @@ struct sk_buff *build_skb(void *data, unif (skb && frag_size) { skb->head_frag =1 ; -if (virt_to_head_page (data)->pfmemalloc) +if (page_is_pfmemalloc (virt_to_head_page (data))) skb->pfmemalloc =1 ; }return skb;
Greg Kroah-Hartman
maintainerPATCH 49/159 v4l: omap3isp: Fix sub-device power management code
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sakari Ailus <[email protected]>
commit 9d39f05490115bf145e5ea03c0b7ec9d3d015b01 upstream.
Commit 813f5c0ac5cc ("media: Change media device link_notify behaviour") modified the media controller link setup notification API and updated the OMAP3 ISP driver accordingly. As a side effect it introduced a bug by turning power on after setting the link instead of before. This results in sub-devices not being powered down in some cases when they should be. Fix it.
--- drivers/media/platform/omap3isp/isp.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/media/platform/omap3isp/isp.c +2 -2
@@ -829,14 +829,14 @@ static int isp_pipeline_link_notify(struint ret;if (notification ==MEDIA_DEV_NOTIFY_POST_LINK_CH && - !(link->flags &MEDIA_LNK_FL_ENABLED )) { + !(flags &MEDIA_LNK_FL_ENABLED )) {/* Powering off entities is assumed to never fail. */ isp_pipeline_pm_power (source, -sink_use);isp_pipeline_pm_power (sink, -source_use);return 0 ; } -if (notification ==MEDIA_DEV_NOTIFY_POST_LINK_CH && +if (notification ==MEDIA_DEV_NOTIFY_PRE_LINK_CH && (flags &MEDIA_LNK_FL_ENABLED )) { ret =isp_pipeline_pm_power (source, sink_use);
Greg Kroah-Hartman
maintainerPATCH 50/159 media: am437x-vpfe: Requested frame size and fmt overwritten by current sensor setting
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Benoit Parrot <[email protected]>
commit f47c9045643f91e76d8a9030828b9fe1cf4a6bcf upstream.
Upon a S_FMT the input/requested frame size and pixel format is overwritten by the current sub-device settings.
Fix this so application can actually set the frame size and format.
--- drivers/media/platform/am437x/am437x-vpfe.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/platform/am437x/am437x-vpfe.c +1 -1
@@ -1577,7 +1577,7 @@ static int vpfe_s_fmt(struct file *file,return -EBUSY ; } - ret =vpfe_try_fmt (file, priv, fmt); + ret =vpfe_try_fmt (file, priv, &format);if (ret)return ret;
Greg Kroah-Hartman
maintainerPATCH 51/159 media: am437x-vpfe: Fix a race condition during release
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Benoit Parrot <[email protected]>
commit c99235fa3ef833c3c23926085f2bb68851c8460a upstream.
There was a race condition where during cleanup/release operation on-going streaming would cause a kernel panic because the hardware module was disabled prematurely with IRQ still pending.
--- drivers/media/platform/am437x/am437x-vpfe.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-)
drivers/media/platform/am437x/am437x-vpfe.c +12 -2
@@ -1185,14 +1185,24 @@ static int vpfe_initialize_device(structstatic int vpfe_release (struct file *file) {struct vpfe_device *vpfe =video_drvdata (file); +bool fh_singular;int ret;mutex_lock (&vpfe->lock); -if (v4l2_fh_is_singular_file (file)) -vpfe_ccdc_close (&vpfe->ccdc, vpfe->pdev); +/* Save the singular status before we call the clean-up helper */ + fh_singular =v4l2_fh_is_singular_file (file); + +/* the release helper will cleanup any on-going streaming */ ret =_vb2_fop_release (file,NULL ); +/* +* If this was the last open file. +* Then de-initialize hw module. +*/ +if (fh_singular) +vpfe_ccdc_close (&vpfe->ccdc, vpfe->pdev); +mutex_unlock (&vpfe->lock);return ret;
Greg Kroah-Hartman
maintainerPATCH 52/159 Btrfs: check if previous transaction aborted to avoid fs corruption
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Filipe Manana <[email protected]>
commit 1f9b8c8fbc9a4d029760b16f477b9d15500e3a34 upstream.
While we are committing a transaction, it's possible the previous one is still finishing its commit and therefore we wait for it to finish first. However we were not checking if that previous transaction ended up getting aborted after we waited for it to commit, so we ended up committing the current transaction which can lead to fs corruption because the new superblock can point to trees that have had one or more nodes/leafs that were never durably persisted.
The following sequence diagram exemplifies how this is possible:
CPU 0 CPU 1
transaction N starts
(...)
btrfs_commit_transaction(N)
cur_trans->state = TRANS_STATE_COMMIT_START;
(...)
cur_trans->state = TRANS_STATE_COMMIT_DOING;
(...)
cur_trans->state = TRANS_STATE_UNBLOCKED;
root->fs_info->running_transaction = NULL;
btrfs_start_transaction()
--> starts transaction N + 1
btrfs_write_and_wait_transaction(trans, root);
--> starts writing all new or COWed ebs created
at transaction N
creates some new ebs, COWs some
existing ebs but doesn't COW or
deletes eb X
btrfs_commit_transaction(N + 1)
(...)
cur_trans->state = TRANS_STATE_COMMIT_START;
(...)
wait_for_commit(root, prev_trans);
--> prev_trans == transaction N
btrfs_write_and_wait_transaction() continues
writing ebs
--> fails writing eb X, we abort transaction N
and set bit BTRFS_FS_STATE_ERROR on
fs_info->fs_state, so no new transactions
can start after setting that bit
cleanup_transaction()
btrfs_cleanup_one_transaction()
wakes up task at CPU 1
continues, doesn't abort because
cur_trans->aborted (transaction N + 1)
is zero, and no checks for bit
BTRFS_FS_STATE_ERROR in fs_info->fs_state
are made
btrfs_write_and_wait_transaction(trans, root);
--> succeeds, no errors during writeback
write_ctree_super(trans, root, 0);
--> succeeds
--> we have now a superblock that points us
to some root that uses eb X, which was
never written to diskIn this scenario future attempts to read eb X from disk results in an error message like "parent transid verify failed on X wanted Y found Z".
So fix this by aborting the current transaction if after waiting for the previous transaction we verify that it was aborted.
--- fs/btrfs/transaction.c | 3 +++ 1 file changed, 3 insertions(+)
fs/btrfs/transaction.c +3 -0
@@ -1875,8 +1875,11 @@ int btrfs_commit_transaction(struct btrfspin_unlock (&root->fs_info->trans_lock);wait_for_commit (root, prev_trans); + ret = prev_trans->aborted;btrfs_put_transaction (prev_trans); +if (ret) + goto cleanup_transaction; }else {spin_unlock (&root->fs_info->trans_lock); }
Greg Kroah-Hartman
maintainerPATCH 53/159 NFSv4/pnfs: Ensure we dont miss a file extension
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Trond Myklebust <[email protected]>
commit 2b83d3de4c18af49800e0b26ae013db4fcf43a4a upstream.
pNFS writes don't return attributes, however that doesn't mean that we should ignore the fact that they may be extending the file. This patch ensures that if a write is seen to extend the file, then we always set an attribute barrier, and update the cached file size.
--- fs/nfs/write.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-)
fs/nfs/write.c +9 -6
@@ -1383,24 +1383,27 @@ static void nfs_writeback_check_extend(s {struct nfs_pgio_args *argp = &hdr->args;struct nfs_pgio_res *resp = &hdr->res; +u64 size = argp->offset + resp->count;if (!(fattr->valid &NFS_ATTR_FATTR_SIZE )) + fattr->size = size; +if (nfs_size_to_loff_t (fattr->size) <i_size_read (hdr->inode)) { + fattr->valid &= ~NFS_ATTR_FATTR_SIZE ;return ; -if (argp->offset + resp->count != fattr->size) -return ; -if (nfs_size_to_loff_t (fattr->size) <i_size_read (hdr->inode)) + } +if (size != fattr->size)return ;/* Set attribute barrier */ nfs_fattr_set_barrier (fattr); +/* ...and update size */ + fattr->valid |=NFS_ATTR_FATTR_SIZE ; }void nfs_writeback_update_inode (struct nfs_pgio_header *hdr) { -struct nfs_fattr *fattr = hdr->res.fattr; +struct nfs_fattr *fattr = &hdr->fattr;struct inode *inode = hdr->inode; -if (fattr ==NULL ) -return ;spin_lock (&inode->i_lock);nfs_writeback_check_extend (hdr, fattr);nfs_post_op_update_inode_force_wcc_locked (inode, fattr);
Greg Kroah-Hartman
maintainerPATCH 54/159 nfsd: Fix an FS_LAYOUT_TYPES/LAYOUT_TYPES encode bug
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Kinglong Mee <[email protected]>
commit 6896f15aabde505b35888039af93d1d182a0108a upstream.
Currently we'll respond correctly to a request for either FS_LAYOUT_TYPES or LAYOUT_TYPES, but not to a request for both attributes simultaneously.
--- fs/nfsd/nfs4xdr.c | 45 +++++++++++++++++++++++++++++++-------------- 1 file changed, 31 insertions(+), 14 deletions(-)
fs/nfsd/nfs4xdr.c +31 -14
@@ -2139,6 +2139,27 @@ nfsd4_encode_aclname(struct xdr_stream *return nfsd4_encode_user (xdr, rqstp, ace->who_uid); } +static inline __be32 +nfsd4_encode_layout_type (struct xdr_stream *xdr,enum pnfs_layouttype layout_type) +{ +__be32 *p; + +if (layout_type) { + p =xdr_reserve_space (xdr,8 ); +if (!p) +return nfserr_resource; + *p++ =cpu_to_be32 (1 ); + *p++ =cpu_to_be32 (layout_type); + }else { + p =xdr_reserve_space (xdr,4 ); +if (!p) +return nfserr_resource; + *p++ =cpu_to_be32 (0 ); + } + +return 0 ; +} +#define WORD0_ABSENT_FS_ATTRS (FATTR4_WORD0_FS_LOCATIONS | FATTR4_WORD0_FSID | \ FATTR4_WORD0_RDATTR_ERROR)#define WORD1_ABSENT_FS_ATTRS FATTR4_WORD1_MOUNTED_ON_FILEID
@@ -2692,20 +2713,16 @@ out_acl: p =xdr_encode_hyper (p, stat.ino); }#ifdef CONFIG_NFSD_PNFS -if ((bmval1 &FATTR4_WORD1_FS_LAYOUT_TYPES ) || - (bmval2 &FATTR4_WORD2_LAYOUT_TYPES )) { -if (exp->ex_layout_type) { - p =xdr_reserve_space (xdr,8 ); -if (!p) - goto out_resource; - *p++ =cpu_to_be32 (1 ); - *p++ =cpu_to_be32 (exp->ex_layout_type); - }else { - p =xdr_reserve_space (xdr,4 ); -if (!p) - goto out_resource; - *p++ =cpu_to_be32 (0 ); - } +if (bmval1 &FATTR4_WORD1_FS_LAYOUT_TYPES ) { + status =nfsd4_encode_layout_type (xdr, exp->ex_layout_type); +if (status) + goto out; + } + +if (bmval2 &FATTR4_WORD2_LAYOUT_TYPES ) { + status =nfsd4_encode_layout_type (xdr, exp->ex_layout_type); +if (status) + goto out; }if (bmval2 &FATTR4_WORD2_LAYOUT_BLKSIZE ) {
Greg Kroah-Hartman
maintainerPATCH 55/159 nfsd: ensure that the ol stateid hash reference is only put once
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jeff Layton <[email protected]>
commit e85687393f3ee0a77ccca016f903d1558bb69258 upstream.
When an open or lock stateid is hashed, we take an extra reference to it. When we unhash it, we drop that reference. The code however does not properly account for the case where we have two callers concurrently trying to unhash the stateid. This can lead to list corruption and the hash reference being put more than once.
Fix this by having unhash_ol_stateid use list_del_init on the st_perfile list_head, and then testing to see if that list_head is empty before releasing the hash reference. This means that some of the unhashing wrappers now become bool return functions so we can test to see whether the stateid was unhashed before we put the reference.
--- fs/nfsd/nfs4state.c | 58 ++++++++++++++++++++++++++++++++-------------------- 1 file changed, 36 insertions(+), 22 deletions(-)
fs/nfsd/nfs4state.c +36 -22
@@ -1004,16 +1004,20 @@ static void nfs4_put_stateowner(struct n sop->so_ops->so_free (sop); } -static void unhash_ol_stateid (struct nfs4_ol_stateid *stp) +static bool unhash_ol_stateid (struct nfs4_ol_stateid *stp) {struct nfs4_file *fp = stp->st_stid.sc_file;lockdep_assert_held (&stp->st_stateowner->so_client->cl_lock); +if (list_empty (&stp->st_perfile)) +return false; +spin_lock (&fp->fi_lock); -list_del (&stp->st_perfile); +list_del_init (&stp->st_perfile);spin_unlock (&fp->fi_lock);list_del (&stp->st_perstateowner); +return true; }static void nfs4_free_ol_stateid (struct nfs4_stid *stid)
@@ -1063,25 +1067,27 @@ static void put_ol_stateid_locked(structlist_add (&stp->st_locks, reaplist); } -static void unhash_lock_stateid (struct nfs4_ol_stateid *stp) +static bool unhash_lock_stateid (struct nfs4_ol_stateid *stp) {struct nfs4_openowner *oo =openowner (stp->st_openstp->st_stateowner);lockdep_assert_held (&oo->oo_owner.so_client->cl_lock);list_del_init (&stp->st_locks); -unhash_ol_stateid (stp);nfs4_unhash_stid (&stp->st_stid); +return unhash_ol_stateid (stp); }static void release_lock_stateid (struct nfs4_ol_stateid *stp) {struct nfs4_openowner *oo =openowner (stp->st_openstp->st_stateowner); +bool unhashed;spin_lock (&oo->oo_owner.so_client->cl_lock); -unhash_lock_stateid (stp); + unhashed =unhash_lock_stateid (stp);spin_unlock (&oo->oo_owner.so_client->cl_lock); -nfs4_put_stid (&stp->st_stid); +if (unhashed) +nfs4_put_stid (&stp->st_stid); }static void unhash_lockowner_locked (struct nfs4_lockowner *lo)
@@ -1129,7 +1135,7 @@ static void release_lockowner(struct nfswhile (!list_empty (&lo->lo_owner.so_stateids)) { stp =list_first_entry (&lo->lo_owner.so_stateids, struct nfs4_ol_stateid, st_perstateowner); -unhash_lock_stateid (stp); +WARN_ON (!unhash_lock_stateid (stp));put_ol_stateid_locked (stp, &reaplist); }spin_unlock (&clp->cl_lock);
@@ -1142,21 +1148,26 @@ static void release_open_stateid_locks(s {struct nfs4_ol_stateid *stp; +lockdep_assert_held (&open_stp->st_stid.sc_client->cl_lock); +while (!list_empty (&open_stp->st_locks)) { stp =list_entry (open_stp->st_locks.next, struct nfs4_ol_stateid, st_locks); -unhash_lock_stateid (stp); +WARN_ON (!unhash_lock_stateid (stp));put_ol_stateid_locked (stp, reaplist); } } -static void unhash_open_stateid (struct nfs4_ol_stateid *stp, +static bool unhash_open_stateid (struct nfs4_ol_stateid *stp,struct list_head *reaplist) { +bool unhashed; +lockdep_assert_held (&stp->st_stid.sc_client->cl_lock); -unhash_ol_stateid (stp); + unhashed =unhash_ol_stateid (stp);release_open_stateid_locks (stp, reaplist); +return unhashed; }static void release_open_stateid (struct nfs4_ol_stateid *stp)
@@ -1164,8 +1175,8 @@ static void release_open_stateid(structLIST_HEAD (reaplist);spin_lock (&stp->st_stid.sc_client->cl_lock); -unhash_open_stateid (stp, &reaplist); -put_ol_stateid_locked (stp, &reaplist); +if (unhash_open_stateid (stp, &reaplist)) +put_ol_stateid_locked (stp, &reaplist);spin_unlock (&stp->st_stid.sc_client->cl_lock);free_ol_stateid_reaplist (&reaplist); }
@@ -1210,8 +1221,8 @@ static void release_openowner(struct nfswhile (!list_empty (&oo->oo_owner.so_stateids)) { stp =list_first_entry (&oo->oo_owner.so_stateids, struct nfs4_ol_stateid, st_perstateowner); -unhash_open_stateid (stp, &reaplist); -put_ol_stateid_locked (stp, &reaplist); +if (unhash_open_stateid (stp, &reaplist)) +put_ol_stateid_locked (stp, &reaplist); }spin_unlock (&clp->cl_lock);free_ol_stateid_reaplist (&reaplist);
@@ -4714,7 +4725,7 @@ nfsd4_free_stateid(struct svc_rqst *rqstif (check_for_locks (stp->st_stid.sc_file,lockowner (stp->st_stateowner)))break ; -unhash_lock_stateid (stp); +WARN_ON (!unhash_lock_stateid (stp));spin_unlock (&cl->cl_lock);nfs4_put_stid (s); ret = nfs_ok;
@@ -4930,20 +4941,23 @@ out:static void nfsd4_close_open_stateid (struct nfs4_ol_stateid *s) {struct nfs4_client *clp = s->st_stid.sc_client; +bool unhashed;LIST_HEAD (reaplist); s->st_stid.sc_type =NFS4_CLOSED_STID ;spin_lock (&clp->cl_lock); -unhash_open_stateid (s, &reaplist); + unhashed =unhash_open_stateid (s, &reaplist);if (clp->cl_minorversion) { -put_ol_stateid_locked (s, &reaplist); +if (unhashed) +put_ol_stateid_locked (s, &reaplist);spin_unlock (&clp->cl_lock);free_ol_stateid_reaplist (&reaplist); }else {spin_unlock (&clp->cl_lock);free_ol_stateid_reaplist (&reaplist); -move_to_close_lru (s, clp->net); +if (unhashed) +move_to_close_lru (s, clp->net); } }
@@ -5982,7 +5996,7 @@ nfsd_inject_add_lock_to_list(struct nfs4static u64 nfsd_foreach_client_lock (struct nfs4_client *clp,u64 max,struct list_head *collect, -void (*func)(struct nfs4_ol_stateid *)) +bool (*func)(struct nfs4_ol_stateid *)) {struct nfs4_openowner *oop;struct nfs4_ol_stateid *stp, *st_next;
@@ -5996,9 +6010,9 @@ static u64 nfsd_foreach_client_lock(strulist_for_each_entry_safe (lst, lst_next, &stp->st_locks, st_locks) {if (func) { -func (lst); -nfsd_inject_add_lock_to_list (lst, - collect); +if (func (lst)) +nfsd_inject_add_lock_to_list (lst, + collect); } ++count;/*
Greg Kroah-Hartman
maintainerPATCH 56/159 nfsd: ensure that delegation stateid hash references are only put once
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jeff Layton <[email protected]>
commit 3fcbbd244ed1d20dc0eb7d48d729503992fa9b7d upstream.
It's possible that a DELEGRETURN could race with (e.g.) client expiry, in which case we could end up putting the delegation hash reference more than once.
Have unhash_delegation_locked return a bool that indicates whether it was already unhashed. In the case of destroy_delegation we only conditionally put the hash reference if that returns true.
The other callers of unhash_delegation_locked call it while walking list_heads that shouldn't yet be detached. If we find that it doesn't return true in those cases, then throw a WARN_ON as that indicates that we have a partially hashed delegation, and that something is likely very wrong.
--- fs/nfsd/nfs4state.c | 26 +++++++++++++++++--------- 1 file changed, 17 insertions(+), 9 deletions(-)
fs/nfsd/nfs4state.c +17 -9
@@ -777,13 +777,16 @@ hash_delegation_locked(struct nfs4_deleglist_add (&dp->dl_perclnt, &dp->dl_stid.sc_client->cl_delegations); } -static void +static bool unhash_delegation_locked (struct nfs4_delegation *dp) {struct nfs4_file *fp = dp->dl_stid.sc_file;lockdep_assert_held (&state_lock); +if (list_empty (&dp->dl_perfile)) +return false; + dp->dl_stid.sc_type =NFS4_CLOSED_DELEG_STID ;/* Ensure that deleg break won't try to requeue it */ ++dp->dl_time;
@@ -792,16 +795,21 @@ unhash_delegation_locked(struct nfs4_dellist_del_init (&dp->dl_recall_lru);list_del_init (&dp->dl_perfile);spin_unlock (&fp->fi_lock); +return true; }static void destroy_delegation (struct nfs4_delegation *dp) { +bool unhashed; +spin_lock (&state_lock); -unhash_delegation_locked (dp); + unhashed =unhash_delegation_locked (dp);spin_unlock (&state_lock); -put_clnt_odstate (dp->dl_clnt_odstate); -nfs4_put_deleg_lease (dp->dl_stid.sc_file); -nfs4_put_stid (&dp->dl_stid); +if (unhashed) { +put_clnt_odstate (dp->dl_clnt_odstate); +nfs4_put_deleg_lease (dp->dl_stid.sc_file); +nfs4_put_stid (&dp->dl_stid); + } }static void revoke_delegation (struct nfs4_delegation *dp)
@@ -1725,7 +1733,7 @@ __destroy_client(struct nfs4_client *clpspin_lock (&state_lock);while (!list_empty (&clp->cl_delegations)) { dp = list_entry(clp->cl_delegations.next, struct nfs4_delegation, dl_perclnt); -unhash_delegation_locked (dp); +WARN_ON (!unhash_delegation_locked (dp));list_add (&dp->dl_recall_lru, &reaplist); }spin_unlock (&state_lock);
@@ -4357,7 +4365,7 @@ nfs4_laundromat(struct nfsd_net *nn) new_timeo =min (new_timeo, t);break ; } -unhash_delegation_locked (dp); +WARN_ON (!unhash_delegation_locked (dp));list_add (&dp->dl_recall_lru, &reaplist); }spin_unlock (&state_lock);
@@ -6282,7 +6290,7 @@ static u64 nfsd_find_all_delegations(strcontinue ;atomic_inc (&clp->cl_refcount); -unhash_delegation_locked (dp); +WARN_ON (!unhash_delegation_locked (dp));list_add (&dp->dl_recall_lru, victims); } ++count;
@@ -6612,7 +6620,7 @@ nfs4_state_shutdown_net(struct net *net)spin_lock (&state_lock);list_for_each_safe (pos, next, &nn->del_recall_lru) { dp = list_entry (pos, struct nfs4_delegation, dl_recall_lru); -unhash_delegation_locked (dp); +WARN_ON (!unhash_delegation_locked (dp));list_add (&dp->dl_recall_lru, &reaplist); }spin_unlock (&state_lock);
Greg Kroah-Hartman
maintainerPATCH 57/159 NFSv4: dont set SETATTR for O_RDONLY|O_EXCL
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: NeilBrown <[email protected]>
commit efcbc04e16dfa95fef76309f89710dd1d99a5453 upstream.
It is unusual to combine the open flags O_RDONLY and O_EXCL, but it appears that libre-office does just that.
[pid 3250] stat("/home/USER/.config", {st_mode=S_IFDIR|0700, st_size=8192, ...}) = 0
[pid 3250] open("/home/USER/.config/libreoffice/4-suse/user/extensions/buildid", O_RDONLY|O_EXCL <unfinished ...>
NFSv4 takes O_EXCL as a sign that a setattr command should be sent, probably to reset the timestamps.
When it was an O_RDONLY open, the SETATTR command does not identify any actual attributes to change.
If no delegation was provided to the open, the SETATTR uses the all-zeros stateid and the request is accepted (at least by the Linux NFS server - no harm, no foul).
If a read-delegation was provided, this is used in the SETATTR request, and a Netapp filer will justifiably claim NFS4ERR_BAD_STATEID, which the Linux client takes as a sign to retry - indefinitely.
So only treat O_EXCL specially if O_CREAT was also given.
--- fs/nfs/nfs4proc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
fs/nfs/nfs4proc.c +1 -1
@@ -2413,7 +2413,7 @@ static int _nfs4_do_open(struct inode *d goto err_free_label; state = ctx->state; -if ((opendata->o_arg.open_flags &O_EXCL ) && +if ((opendata->o_arg.open_flags & (O_CREAT |O_EXCL )) == (O_CREAT |O_EXCL ) && (opendata->o_arg.createmode !=NFS4_CREATE_GUARDED )) {nfs4_exclusive_attrset (opendata, sattr);
Greg Kroah-Hartman
maintainerPATCH 58/159 NFS: Dont let the ctime override attribute barriers.
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Trond Myklebust <[email protected]>
commit 7c2dad99d60c86ec686b3bfdcb787c450a7ea89f upstream.
Chuck reports seeing cases where a GETATTR that happens to race with an asynchronous WRITE is overriding the file size, despite the attribute barrier being set by the writeback code.
The culprit turns out to be the check in nfs_ctime_need_update(), which sees that the ctime is newer than the cached ctime, and assumes that it is safe to override the attribute barrier. This patch removes that override, and ensures that attribute barriers are always respected.
--- fs/nfs/inode.c | 8 -------- 1 file changed, 8 deletions(-)
fs/nfs/inode.c +0 -8
@@ -1270,13 +1270,6 @@ static int nfs_check_inode_attributes(streturn 0 ; } -static int nfs_ctime_need_update (const struct inode *inode,const struct nfs_fattr *fattr) -{ -if (!(fattr->valid &NFS_ATTR_FATTR_CTIME )) -return 0 ; -return timespec_compare (&fattr->ctime, &inode->i_ctime) >0 ; -} -static atomic_long_t nfs_attr_generation_counter;static unsigned long nfs_read_attr_generation_counter (void )
@@ -1425,7 +1418,6 @@ static int nfs_inode_attrs_need_update(cconst struct nfs_inode *nfsi =NFS_I (inode);return ((long )fattr->gencount - (long )nfsi->attr_gencount) >0 || -nfs_ctime_need_update (inode, fattr) || ((long )nfsi->attr_gencount - (long )nfs_read_attr_generation_counter () >0 ); }
Greg Kroah-Hartman
maintainerPATCH 59/159 NFSv4.1/pNFS: Fix borken function _same_data_server_addrs_locked()
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Trond Myklebust <[email protected]>
commit 6f536936b79bd4b5cea8fb0e5b8b0bce8cd1ea4a upstream.
- Switch back to using list_for_each_entry(). Fixes an incorrect test for list NULL termination.
- Do not assume that lists are sorted.
- Finally, consider an existing entry to match if it consists of a subset of the addresses in the new entry.
--- fs/nfs/pnfs_nfs.c | 31 ++++++++++++++++++------------- 1 file changed, 18 insertions(+), 13 deletions(-)
fs/nfs/pnfs_nfs.c +18 -13
@@ -359,26 +359,31 @@ same_sockaddr(struct sockaddr *addr1, streturn false; } +/* +* Checks if 'dsaddrs1' contains a subset of 'dsaddrs2'. If it does, +* declare a match. +*/ static bool _same_data_server_addrs_locked (const struct list_head *dsaddrs1,const struct list_head *dsaddrs2) {struct nfs4_pnfs_ds_addr *da1, *da2; +struct sockaddr *sa1, *sa2; +bool match = false; -/* step through both lists, comparing as we go */ -for (da1 =list_first_entry (dsaddrs1,typeof (*da1), da_node), - da2 =list_first_entry (dsaddrs2,typeof (*da2), da_node); - da1 !=NULL && da2 !=NULL ; - da1 =list_entry (da1->da_node.next,typeof (*da1), da_node), - da2 =list_entry (da2->da_node.next,typeof (*da2), da_node)) { -if (!same_sockaddr ((struct sockaddr *)&da1->da_addr, - (struct sockaddr *)&da2->da_addr)) -return false; +list_for_each_entry (da1, dsaddrs1, da_node) { + sa1 = (struct sockaddr *)&da1->da_addr; + match = false; +list_for_each_entry (da2, dsaddrs2, da_node) { + sa2 = (struct sockaddr *)&da2->da_addr; + match =same_sockaddr (sa1, sa2); +if (match) +break ; + } +if (!match) +break ; } -if (da1 ==NULL && da2 ==NULL ) -return true; - -return false; +return match; }/*
Greg Kroah-Hartman
maintainerPATCH 60/159 NFS: Fix a NULL pointer dereference of migration recovery ops for v4.2 client
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Kinglong Mee <[email protected]>
commit 18e3b739fdc826481c6a1335ce0c5b19b3d415da upstream.
---Steps to Reproduce--
<nfs-server>
# cat /etc/exports /nfs/referal *(rw,insecure,no_subtree_check,no_root_squash,crossmnt) /nfs/old *(ro,insecure,subtree_check,root_squash,crossmnt)
<nfs-client>
# mount -t nfs nfs-server:/nfs/ /mnt/ # ll /mnt/*/
<nfs-server>
# cat /etc/exports /nfs/referal *(rw,insecure,no_subtree_check,no_root_squash,crossmnt,refer=/nfs/old/@nfs-server) /nfs/old *(ro,insecure,subtree_check,root_squash,crossmnt) # service nfs restart
<nfs-client>
# ll /mnt/*/ --->>>>> oops here
[ 5123.102925] BUG: unable to handle kernel NULL pointer dereference at (null) [ 5123.103363] IP: [<ffffffffa03ed38b>] nfs4_proc_get_locations+0x9b/0x120 [nfsv4] [ 5123.103752] PGD 587b9067 PUD 3cbf5067 PMD 0 [ 5123.104131] Oops: 0000 [#1] [ 5123.104529] Modules linked in: nfsv4(OE) nfs(OE) fscache(E) nfsd(OE) xfs libcrc32c iscsi_tcp libiscsi_tcp libiscsi scsi_transport_iscsi coretemp crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel ppdev vmw_balloon parport_pc parport i2c_piix4 shpchp auth_rpcgss nfs_acl vmw_vmci lockd grace sunrpc vmwgfx drm_kms_helper ttm drm mptspi serio_raw scsi_transport_spi e1000 mptscsih mptbase ata_generic pata_acpi [last unloaded: nfsd] [ 5123.105887] CPU: 0 PID: 15853 Comm: ::1-manager Tainted: G OE 4.2.0-rc6+ #214 [ 5123.106358] Hardware name: VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform, BIOS 6.00 05/20/2014 [ 5123.106860] task: ffff88007620f300 ti: ffff88005877c000 task.ti: ffff88005877c000 [ 5123.107363] RIP: 0010:[<ffffffffa03ed38b>] [<ffffffffa03ed38b>] nfs4_proc_get_locations+0x9b/0x120 [nfsv4] [ 5123.107909] RSP: 0018:ffff88005877fdb8 EFLAGS: 00010246 [ 5123.108435] RAX: ffff880053f3bc00 RBX: ffff88006ce6c908 RCX: ffff880053a0d240 [ 5123.108968] RDX: ffffea0000e6d940 RSI: ffff8800399a0000 RDI: ffff88006ce6c908 [ 5123.109503] RBP: ffff88005877fe28 R08: ffffffff81c708a0 R09: 0000000000000000 [ 5123.110045] R10: 00000000000001a2 R11: ffff88003ba7f5c8 R12: ffff880054c55800 [ 5123.110618] R13: 0000000000000000 R14: ffff880053a0d240 R15: ffff880053a0d240 [ 5123.111169] FS: 0000000000000000(0000) GS:ffffffff81c27000(0000) knlGS:0000000000000000 [ 5123.111726] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 5123.112286] CR2: 0000000000000000 CR3: 0000000054cac000 CR4: 00000000001406f0 [ 5123.112888] Stack: [ 5123.113458] ffffea0000e6d940 ffff8800399a0000 00000000000167d0 0000000000000000 [ 5123.114049] 0000000000000000 0000000000000000 0000000000000000 00000000a7ec82c6 [ 5123.114662] ffff88005877fe18 ffffea0000e6d940 ffff8800399a0000 ffff880054c55800 [ 5123.115264] Call Trace: [ 5123.115868] [<ffffffffa03fb44b>] nfs4_try_migration+0xbb/0x220 [nfsv4] [ 5123.116487] [<ffffffffa03fcb3b>] nfs4_run_state_manager+0x4ab/0x7b0 [nfsv4] [ 5123.117104] [<ffffffffa03fc690>] ? nfs4_do_reclaim+0x510/0x510 [nfsv4] [ 5123.117813] [<ffffffff810a4527>] kthread+0xd7/0xf0 [ 5123.118456] [<ffffffff810a4450>] ? kthread_worker_fn+0x160/0x160 [ 5123.119108] [<ffffffff816d9cdf>] ret_from_fork+0x3f/0x70 [ 5123.119723] [<ffffffff810a4450>] ? kthread_worker_fn+0x160/0x160 [ 5123.120329] Code: 4c 8b 6a 58 74 17 eb 52 48 8d 55 a8 89 c6 4c 89 e7 e8 4a b5 ff ff 8b 45 b0 85 c0 74 1c 4c 89 f9 48 8b 55 90 48 8b 75 98 48 89 df <41> ff 55 00 3d e8 d8 ff ff 41 89 c6 74 cf 48 8b 4d c8 65 48 33 [ 5123.121643] RIP [<ffffffffa03ed38b>] nfs4_proc_get_locations+0x9b/0x120 [nfsv4] [ 5123.122308] RSP <ffff88005877fdb8> [ 5123.122942] CR2: 0000000000000000
--- fs/nfs/nfs4proc.c | 1 + 1 file changed, 1 insertion(+)
fs/nfs/nfs4proc.c +1 -0
@@ -8571,6 +8571,7 @@ static const struct nfs4_minor_version_o .reboot_recovery_ops = &nfs41_reboot_recovery_ops, .nograce_recovery_ops = &nfs41_nograce_recovery_ops, .state_renewal_ops = &nfs41_state_renewal_ops, + .mig_recovery_ops = &nfs41_mig_recovery_ops, };#endif
Greg Kroah-Hartman
maintainerPATCH 61/159 NFS: nfs_set_pgio_error sometimes misses errors
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Trond Myklebust <[email protected]>
commit e9ae58aeee8842a50f7e199d602a5ccb2e41a95f upstream.
We should ensure that we always set the pgio_header's error field if a READ or WRITE RPC call returns an error. The current code depends on 'hdr->good_bytes' always being initialised to a large value, which is not always done correctly by callers.
When this happens, applications may end up missing important errors.
--- fs/nfs/pagelist.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
fs/nfs/pagelist.c +2 -2
@@ -77,8 +77,8 @@ EXPORT_SYMBOL_GPL(nfs_pgheader_init);void nfs_set_pgio_error (struct nfs_pgio_header *hdr,int error,loff_t pos) {spin_lock (&hdr->lock); -if (pos < hdr->io_start + hdr->good_bytes) { -set_bit (NFS_IOHDR_ERROR , &hdr->flags); +if (!test_and_set_bit (NFS_IOHDR_ERROR , &hdr->flags) + || pos < hdr->io_start + hdr->good_bytes) {clear_bit (NFS_IOHDR_EOF , &hdr->flags); hdr->good_bytes = pos - hdr->io_start; hdr->error = error;
Greg Kroah-Hartman
maintainerPATCH 62/159 NFS41/flexfiles: update inode after write finishes
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Peng Tao <[email protected]>
commit 69f230d907e8c1ca3f9bd528993eeb98f712b0dd upstream.
Otherwise we break fstest case tests/read_write/mctime.t
Does files layout need the same fix as well?
--- fs/nfs/flexfilelayout/flexfilelayout.c | 3 +++ 1 file changed, 3 insertions(+)
fs/nfs/flexfilelayout/flexfilelayout.c +3 -0
@@ -1039,6 +1039,9 @@ static int ff_layout_write_done_cb(struc hdr->res.verf->committed == NFS_DATA_SYNC)ff_layout_set_layoutcommit (hdr); +if (task->tk_status >=0 ) +nfs_writeback_update_inode (hdr); +return 0 ; }
Greg Kroah-Hartman
maintainerPATCH 63/159 NFSv4: Force a post-op attribute update when holding a delegation
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Trond Myklebust <[email protected]>
commit aaae3f00d3f67f681a1f3cb7af999e976e8a24ce upstream.
If the ctime or mtime or change attribute have changed because of an operation we initiated, we should make sure that we force an attribute update. However we do not want to mark the page cache for revalidation.
--- fs/nfs/inode.c | 7 +++++++ 1 file changed, 7 insertions(+)
fs/nfs/inode.c +7 -0
@@ -1480,6 +1480,13 @@ static int nfs_post_op_update_inode_lock {unsigned long invalid =NFS_INO_INVALID_ATTR |NFS_INO_REVAL_PAGECACHE ; +/* +* Don't revalidate the pagecache if we hold a delegation, but do +* force an attribute update +*/ +if (NFS_PROTO (inode)->have_delegation (inode,FMODE_READ )) + invalid =NFS_INO_INVALID_ATTR |NFS_INO_REVAL_FORCED ; +if (S_ISDIR (inode->i_mode)) invalid |=NFS_INO_INVALID_DATA ;nfs_set_cache_invalid (inode, invalid);
Greg Kroah-Hartman
maintainerPATCH 64/159 NFS41/flexfiles: zero out DS write wcc
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Peng Tao <[email protected]>
commit 5420401079e152ff68a8024f6a375804b1c21505 upstream.
We do not want to update inode attributes with DS values.
--- fs/nfs/flexfilelayout/flexfilelayout.c | 2 ++ 1 file changed, 2 insertions(+)
fs/nfs/flexfilelayout/flexfilelayout.c +2 -0
@@ -1039,6 +1039,8 @@ static int ff_layout_write_done_cb(struc hdr->res.verf->committed == NFS_DATA_SYNC)ff_layout_set_layoutcommit (hdr); +/* zero out fattr since we don't care DS attr at all */ + hdr->fattr.valid =0 ;if (task->tk_status >=0 )nfs_writeback_update_inode (hdr);
Greg Kroah-Hartman
maintainerPATCH 65/159 NFSv4.1/flexfiles: Fix a protocol error in layoutreturn
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Trond Myklebust <[email protected]>
commit d13549074cf066d6d5bb29903d044beffea342d3 upstream.
According to the flexfiles protocol, the layoutreturn should specify an array of errors in the following format:
struct ff_ioerr4 {
offset4 ffie_offset;
length4 ffie_length;
stateid4 ffie_stateid;
device_error4 ffie_errors<>;
};This patch fixes up the code to ensure that our ffie_errors is indeed encoded as an array (albeit with only a single entry).
--- fs/nfs/flexfilelayout/flexfilelayoutdev.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)
fs/nfs/flexfilelayout/flexfilelayoutdev.c +5 -2
@@ -500,16 +500,19 @@ int ff_layout_encode_ds_ioerr(struct nfs range->offset, range->length))continue ;/* offset(8) + length(8) + stateid(NFS4_STATEID_SIZE) -* + deviceid(NFS4_DEVICEID4_SIZE) + status(4) + opnum(4) +* + array length + deviceid(NFS4_DEVICEID4_SIZE) +* + status(4) + opnum(4) */ p =xdr_reserve_space (xdr, -24 +NFS4_STATEID_SIZE +NFS4_DEVICEID4_SIZE ); +28 +NFS4_STATEID_SIZE +NFS4_DEVICEID4_SIZE );if (unlikely (!p))return -ENOBUFS ; p =xdr_encode_hyper (p, err->offset); p =xdr_encode_hyper (p, err->length); p =xdr_encode_opaque_fixed (p, &err->stateid,NFS4_STATEID_SIZE ); +/* Encode 1 error */ + *p++ =cpu_to_be32 (1 ); p =xdr_encode_opaque_fixed (p, &err->deviceid,NFS4_DEVICEID4_SIZE ); *p++ =cpu_to_be32 (err->status);
Greg Kroah-Hartman
maintainerPATCH 66/159 NFSv4.1: Fix a protocol issue with CLOSE stateids
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Trond Myklebust <[email protected]>
commit 4a1e2feb9d246775dee0f78ed5b18826bae2b1c5 upstream.
According to RFC5661 Section 18.2.4, CLOSE is supposed to return the zero stateid. This means that nfs_clear_open_stateid_locked() cannot assume that the result stateid will always match the 'other' field of the existing open stateid when trying to determine a race with a parallel OPEN.
Instead, we look at the argument, and check for matches.
--- fs/nfs/nfs4proc.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-)
fs/nfs/nfs4proc.c +10 -5
@@ -1216,6 +1216,7 @@ static void nfs_resync_open_stateid_lock }static void nfs_clear_open_stateid_locked (struct nfs4_state *state, +nfs4_stateid *arg_stateid,nfs4_stateid *stateid,fmode_t fmode) {clear_bit (NFS_O_RDWR_STATE , &state->flags);
@@ -1234,8 +1235,9 @@ static void nfs_clear_open_stateid_lockeif (stateid ==NULL )return ;/* Handle races with OPEN */ -if (!nfs4_stateid_match_other (stateid, &state->open_stateid) || - !nfs4_stateid_is_newer (stateid, &state->open_stateid)) { +if (!nfs4_stateid_match_other (arg_stateid, &state->open_stateid) || + (nfs4_stateid_match_other (stateid, &state->open_stateid) && + !nfs4_stateid_is_newer (stateid, &state->open_stateid))) {nfs_resync_open_stateid_locked (state);return ; }
@@ -1244,10 +1246,12 @@ static void nfs_clear_open_stateid_lockenfs4_stateid_copy (&state->open_stateid, stateid); } -static void nfs_clear_open_stateid (struct nfs4_state *state,nfs4_stateid *stateid,fmode_t fmode) +static void nfs_clear_open_stateid (struct nfs4_state *state, +nfs4_stateid *arg_stateid, +nfs4_stateid *stateid,fmode_t fmode) {write_seqlock (&state->seqlock); -nfs_clear_open_stateid_locked (state, stateid, fmode); +nfs_clear_open_stateid_locked (state, arg_stateid, stateid, fmode);write_sequnlock (&state->seqlock);if (test_bit (NFS_STATE_RECLAIM_NOGRACE , &state->flags))nfs4_schedule_state_manager (state->owner->so_server->nfs_client);
@@ -2672,7 +2676,8 @@ static void nfs4_close_done(struct rpc_t goto out_release; } } -nfs_clear_open_stateid (state, res_stateid, calldata->arg.fmode); +nfs_clear_open_stateid (state, &calldata->arg.stateid, + res_stateid, calldata->arg.fmode); out_release:nfs_release_seqid (calldata->arg.seqid);nfs_refresh_inode (calldata->inode, calldata->res.fattr);
Greg Kroah-Hartman
maintainerPATCH 67/159 Revert "NFSv4: Remove incorrect check in can_open_delegated()"
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Trond Myklebust <[email protected]>
commit 36319608e28701c07cad80ae3be8b0fdfb1ab40f upstream.
This reverts commit 4e379d36c050b0117b5d10048be63a44f5036115.
This commit opens up a race between the recovery code and the open code.
--- fs/nfs/nfs4proc.c | 2 ++ 1 file changed, 2 insertions(+)
fs/nfs/nfs4proc.c +2 -0
@@ -1152,6 +1152,8 @@ static int can_open_delegated(struct nfsreturn 0 ;if ((delegation->type & fmode) != fmode)return 0 ; +if (test_bit (NFS_DELEGATION_NEED_RECLAIM , &delegation->flags)) +return 0 ;if (test_bit (NFS_DELEGATION_RETURNING , &delegation->flags))return 0 ;nfs_mark_delegation_referenced (delegation);
Greg Kroah-Hartman
maintainerPATCH 68/159 net: sunrpc: fix tracepoint Warning: unknown op ->
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Pratyush Anand <[email protected]>
commit 051ac3848a94f21cfdec899cc9c65ce7f9f116fa upstream.
perf stat -e sunrpc:svc_xprt_do_enqueue true results in
Warning: unknown op '->'
Warning: [sunrpc:svc_xprt_do_enqueue] unknown op '->'
Similar warning for svc_handle_xprt as well.
Actually TP_printk() should never dereference an address saved in the ring buffer that points somewhere in the kernel. There's no guarantee that that object still exists (with the exception of static strings).
Therefore change all the arguments for TP_printk(), so that it references values existing in the ring buffer only.
While doing that, also fix another possible bug when argument xprt could be NULL and TP_fast_assign() tries to access it's elements.
--- include/trace/events/sunrpc.h | 21 ++++++++++++++------- 1 file changed, 14 insertions(+), 7 deletions(-)
include/trace/events/sunrpc.h +14 -7
@@ -529,18 +529,21 @@ TRACE_EVENT(svc_xprt_do_enqueue,TP_STRUCT__entry (__field (struct svc_xprt *,xprt ) - __field(struct svc_rqst *, rqst) + __field_struct(struct sockaddr_storage, ss) +__field (int, pid) + __field(unsigned long, flags) ),TP_fast_assign ( __entry->xprt = xprt; - __entry->rqst = rqst; + xprt ?memcpy (&__entry->ss, &xprt->xpt_remote,sizeof (__entry->ss)) :memset (&__entry->ss,0 ,sizeof (__entry->ss)); + __entry->pid = rqst? rqst->rq_task->pid :0 ; + __entry->flags = xprt ? xprt->xpt_flags :0 ; ),TP_printk ("xprt=0x%p addr=%pIScp pid=%d flags=%s" , __entry->xprt, - (struct sockaddr *)&__entry->xprt->xpt_remote, - __entry->rqst ? __entry->rqst->rq_task->pid :0 , -show_svc_xprt_flags (__entry->xprt->xpt_flags)) + (struct sockaddr *)&__entry->ss, + __entry->pid,show_svc_xprt_flags (__entry->flags)) ); TRACE_EVENT(svc_xprt_dequeue,
@@ -589,16 +592,20 @@ TRACE_EVENT(svc_handle_xprt,TP_STRUCT__entry (__field (struct svc_xprt *,xprt )__field (int, len) + __field_struct(struct sockaddr_storage, ss) + __field(unsigned long, flags) ),TP_fast_assign ( __entry->xprt = xprt; + xprt ?memcpy (&__entry->ss, &xprt->xpt_remote,sizeof (__entry->ss)) :memset (&__entry->ss,0 ,sizeof (__entry->ss)); __entry->len = len; + __entry->flags = xprt ? xprt->xpt_flags :0 ; ),TP_printk ("xprt=0x%p addr=%pIScp len=%d flags=%s" , __entry->xprt, - (struct sockaddr *)&__entry->xprt->xpt_remote, __entry->len, -show_svc_xprt_flags (__entry->xprt->xpt_flags)) + (struct sockaddr *)&__entry->ss, + __entry->len,show_svc_xprt_flags (__entry->flags)) );#endif /* _TRACE_SUNRPC_H */
Greg Kroah-Hartman
maintainerPATCH 69/159 SUNRPC: Fix a thinko in xs_connect()
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Trond Myklebust <[email protected]>
commit 99b1a4c32ad22024ac6198a4337aaec5ea23168f upstream.
It is rather pointless to test the value of transport->inet after calling xs_reset_transport(), since it will always be zero, and so we will never see any exponential back off behaviour. Also don't force early connections for SOFTCONN tasks. If the server disconnects us, we should respect the exponential backoff.
--- net/sunrpc/xprtsock.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)
net/sunrpc/xprtsock.c +5 -4
@@ -2216,13 +2216,14 @@ static void xs_connect(struct rpc_xprt *WARN_ON_ONCE (!xprt_lock_connect (xprt, task, transport)); -/* Start by resetting any existing state */ -xs_reset_transport (transport); - -if (transport->sock !=NULL && !RPC_IS_SOFTCONN (task)) { +if (transport->sock !=NULL ) {dprintk ("RPC: xs_connect delayed xprt %p for %lu " "seconds\n" , xprt, xprt->reestablish_timeout /HZ ); + +/* Start by resetting any existing state */ +xs_reset_transport (transport); +queue_delayed_work (rpciod_workqueue, &transport->connect_worker, xprt->reestablish_timeout);
Greg Kroah-Hartman
maintainerPATCH 70/159 SUNRPC: xs_reset_transport must mark the connection as disconnected
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Trond Myklebust <[email protected]>
commit 0c78789e3a030615c6650fde89546cadf40ec2cc upstream.
In case the reconnection attempt fails.
--- net/sunrpc/xprtsock.c | 1 + 1 file changed, 1 insertion(+)
net/sunrpc/xprtsock.c +1 -0
@@ -834,6 +834,7 @@ static void xs_reset_transport(struct so sk->sk_user_data =NULL ;xs_restore_old_callbacks (transport, sk); +xprt_clear_connected (xprt);write_unlock_bh (&sk->sk_callback_lock);xs_sock_reset_connection_flags (xprt);
Greg Kroah-Hartman
maintainerPATCH 71/159 SUNRPC: Ensure that we wait for connections to complete before retrying
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Trond Myklebust <[email protected]>
commit 0fdea1e8a2853f79d39b8555cc9de16a7e0ab26f upstream.
Commit 718ba5b87343, moved the responsibility for unlocking the socket to xs_tcp_setup_socket, meaning that the socket will be unlocked before we know that it has finished trying to connect. The following patch is based on an initial patch by Russell King to ensure that we delay clearing the XPRT_CONNECTING flag until we either know that we failed to initiate a connection attempt, or the connection attempt itself failed.
--- include/linux/sunrpc/xprtsock.h | 3 +++ net/sunrpc/xprtsock.c | 11 ++++++++--- 2 files changed, 11 insertions(+), 3 deletions(-)
include/linux/sunrpc/xprtsock.h +3 -0
@@ -42,6 +42,7 @@ struct sock_xprt {/* * Connection of transports */ +unsigned long sock_state;struct delayed_work connect_worker;struct sockaddr_storage srcaddr;unsigned short srcport;
@@ -76,6 +77,8 @@ struct sock_xprt {*/ #define TCP_RPC_REPLY (1UL << 6) +#define XPRT_SOCK_CONNECTING 1U +#endif /* __KERNEL__ */ #endif /* _LINUX_SUNRPC_XPRTSOCK_H */
net/sunrpc/xprtsock.c +8 -3
@@ -1434,6 +1434,7 @@ out:static void xs_tcp_state_change (struct sock *sk) {struct rpc_xprt *xprt; +struct sock_xprt *transport;read_lock_bh (&sk->sk_callback_lock);if (!(xprt =xprt_from_sock (sk)))
@@ -1445,13 +1446,12 @@ static void xs_tcp_state_change(struct ssock_flag (sk,SOCK_ZAPPED ), sk->sk_shutdown); + transport = container_of(xprt, struct sock_xprt, xprt);trace_rpc_socket_state_change (xprt, sk->sk_socket);switch (sk->sk_state) {case TCP_ESTABLISHED :spin_lock (&xprt->transport_lock);if (!xprt_test_and_set_connected (xprt)) { -struct sock_xprt *transport =container_of (xprt, - struct sock_xprt, xprt);/* Reset TCP record info */ transport->tcp_offset =0 ;
@@ -1460,6 +1460,8 @@ static void xs_tcp_state_change(struct s transport->tcp_flags =TCP_RCV_COPY_FRAGHDR |TCP_RCV_COPY_XID ; xprt->connect_cookie++; +clear_bit (XPRT_SOCK_CONNECTING , &transport->sock_state); +xprt_clear_connecting (xprt);xprt_wake_pending_tasks (xprt, -EAGAIN ); }
@@ -1495,6 +1497,9 @@ static void xs_tcp_state_change(struct ssmp_mb__after_atomic ();break ;case TCP_CLOSE : +if (test_and_clear_bit (XPRT_SOCK_CONNECTING , + &transport->sock_state)) +xprt_clear_connecting (xprt);xs_sock_mark_closed (xprt); } out:
@@ -2111,6 +2116,7 @@ static int xs_tcp_finish_connecting(stru/* Tell the socket layer to start connecting... */ xprt->stat.connect_count++; xprt->stat.connect_start = jiffies; +set_bit (XPRT_SOCK_CONNECTING , &transport->sock_state); ret =kernel_connect (sock,xs_addr (xprt), xprt->addrlen,O_NONBLOCK );switch (ret) {case 0 :
@@ -2175,7 +2181,6 @@ static void xs_tcp_setup_socket(struct wcase -EINPROGRESS :case -EALREADY :xprt_unlock_connect (xprt, transport); -xprt_clear_connecting (xprt);return ;case -EINVAL :/* Happens, for instance, if the user specified a link
Greg Kroah-Hartman
maintainerPATCH 72/159 SUNRPC: Lock the transport layer on shutdown
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Trond Myklebust <[email protected]>
commit 79234c3db6842a3de03817211d891e0c2878f756 upstream.
Avoid all races with the connect/disconnect handlers by taking the transport lock.
--- net/sunrpc/xprt.c | 6 ++++++ 1 file changed, 6 insertions(+)
net/sunrpc/xprt.c +6 -0
@@ -611,6 +611,7 @@ static void xprt_autoclose(struct work_s xprt->ops->close (xprt);clear_bit (XPRT_CLOSE_WAIT , &xprt->state);xprt_release_write (xprt,NULL ); +wake_up_bit (&xprt->state,XPRT_LOCKED ); }/**
@@ -720,6 +721,7 @@ void xprt_unlock_connect(struct rpc_xprt xprt->ops->release_xprt (xprt,NULL ); out:spin_unlock_bh (&xprt->transport_lock); +wake_up_bit (&xprt->state,XPRT_LOCKED ); }/**
@@ -1389,6 +1391,10 @@ out:static void xprt_destroy (struct rpc_xprt *xprt) {dprintk ("RPC: destroying transport %p\n" , xprt); + +/* Exclude transport connect/disconnect handlers */ +wait_on_bit_lock (&xprt->state,XPRT_LOCKED ,TASK_UNINTERRUPTIBLE ); +del_timer_sync (&xprt->timer);rpc_xprt_debugfs_unregister (xprt);
Greg Kroah-Hartman
maintainerPATCH 73/159 rtc: s3c: fix disabled clocks for alarm
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Joonyoung Shim <[email protected]>
commit 1fb1c35f56bb6ab4a65920c648154b0f78f634a5 upstream.
The clock enable/disable codes for alarm have been removed from commit 24e1455493da ("drivers/rtc/rtc-s3c.c: delete duplicate clock control") and the clocks are disabled even if alarm is set, so alarm interrupt can't happen.
The s3c_rtc_setaie function can be called several times with 'enabled' argument having same value, so it needs to check whether clocks are enabled or not.
--- drivers/rtc/rtc-s3c.c | 24 ++++++++++++++++++------ 1 file changed, 18 insertions(+), 6 deletions(-)
drivers/rtc/rtc-s3c.c +18 -6
@@ -39,6 +39,7 @@ struct s3c_rtc {void __iomem *base;struct clk *rtc_clk;struct clk *rtc_src_clk; +bool clk_disabled;struct s3c_rtc_data *data;
@@ -71,9 +72,12 @@ static void s3c_rtc_enable_clk(struct s3unsigned long irq_flags;spin_lock_irqsave (&info->alarm_clk_lock, irq_flags); -clk_enable (info->rtc_clk); -if (info->data->needs_src_clk) -clk_enable (info->rtc_src_clk); +if (info->clk_disabled) { +clk_enable (info->rtc_clk); +if (info->data->needs_src_clk) +clk_enable (info->rtc_src_clk); + info->clk_disabled = false; + }spin_unlock_irqrestore (&info->alarm_clk_lock, irq_flags); }
@@ -82,9 +86,12 @@ static void s3c_rtc_disable_clk(struct sunsigned long irq_flags;spin_lock_irqsave (&info->alarm_clk_lock, irq_flags); -if (info->data->needs_src_clk) -clk_disable (info->rtc_src_clk); -clk_disable (info->rtc_clk); +if (!info->clk_disabled) { +if (info->data->needs_src_clk) +clk_disable (info->rtc_src_clk); +clk_disable (info->rtc_clk); + info->clk_disabled = true; + }spin_unlock_irqrestore (&info->alarm_clk_lock, irq_flags); }
@@ -128,6 +135,11 @@ static int s3c_rtc_setaie(struct devices3c_rtc_disable_clk (info); +if (enabled) +s3c_rtc_enable_clk (info); +else +s3c_rtc_disable_clk (info); +return 0 ; }
Greg Kroah-Hartman
maintainerPATCH 74/159 rtc: s5m: fix to update ctrl register
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Joonyoung Shim <[email protected]>
commit ff02c0444b83201ff76cc49deccac8cf2bffc7bc upstream.
According to datasheet, the S2MPS13X and S2MPS14X should update write buffer via setting WUDR bit to high after ctrl register is written.
If not, ALARM interrupt of rtc-s5m doesn't happen first time when i use tools/testing/selftests/timers/rtctest.c test program and hour format is used to 12 hour mode in Odroid-XU3 board.
One more issue is the RTC doesn't keep time on Odroid-XU3 board when i turn on board after power off even if RTC battery is connected. It can be solved as setting WUDR & RUDR bits to high at the same time after RTC_CTRL register is written. It's same with condition of only writing ALARM registers, so this is for only S2MPS14 and we should set WUDR & A_UDR bits to high on S2MPS13.
I can't find any reasonable description about this like fix from datasheet, but can find similar codes from rtc driver source of hardkernel kernel and vendor kernel.
--- drivers/rtc/rtc-s5m.c | 10 ++++++++++ 1 file changed, 10 insertions(+)
drivers/rtc/rtc-s5m.c +10 -0
@@ -635,6 +635,16 @@ static int s5m8767_rtc_init_reg(struct scase S2MPS13X : data[0 ] = (0 <<BCD_EN_SHIFT ) | (1 <<MODEL24_SHIFT ); ret =regmap_write (info->regmap, info->regs->ctrl, data[0 ]); +if (ret <0 ) +break ; + +/* +* Should set WUDR & (RUDR or AUDR) bits to high after writing +* RTC_CTRL register like writing Alarm registers. We can't find +* the description from datasheet but vendor code does that +* really. +*/ + ret =s5m8767_rtc_set_alarm_reg (info);break ;default :
Greg Kroah-Hartman
maintainerPATCH 75/159 rtc: abx80x: fix RTC write bit
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Mitja Spes <[email protected]>
commit 5f1b2f77646fc0ef2f36fc554f5722a1381d0892 upstream.
Fix RTC write bit as per application manual
--- drivers/rtc/rtc-abx80x.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/rtc/rtc-abx80x.c +1 -1
@@ -28,7 +28,7 @@#define ABX8XX_REG_WD 0x07#define ABX8XX_REG_CTRL1 0x10 -#define ABX8XX_CTRL_WRITE BIT(1) +#define ABX8XX_CTRL_WRITE BIT(0)#define ABX8XX_CTRL_12_24 BIT(6)#define ABX8XX_REG_CFG_KEY 0x1f
Greg Kroah-Hartman
maintainerPATCH 76/159 PCI,parisc: Enable 64-bit bus addresses on PA-RISC
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Helge Deller <[email protected]>
commit e02a653e15d8d32e9e768fd99a3271aafe5c5d77 upstream.
Commit 3a9ad0b ("PCI: Add pci_bus_addr_t") unconditionally introduced usage of 64-bit PCI bus addresses on all 64-bit platforms which broke PA-RISC.
It turned out that due to enabling the 64-bit addresses, the PCI logic decided to use the GMMIO instead of the LMMIO region. This commit simply disables registering the GMMIO and thus we fall back to use the LMMIO region as before.
Reverts commit 45ea2a5fed6dacb9bb0558d8b21eacc1c45d5bb4 ("PCI: Don't use 64-bit bus addresses on PA-RISC")
--- drivers/parisc/lba_pci.c | 7 +++++-- drivers/pci/Kconfig | 2 +- 2 files changed, 6 insertions(+), 3 deletions(-)
drivers/parisc/lba_pci.c +5 -2
@@ -1556,8 +1556,11 @@ lba_driver_probe(struct parisc_device *dif (lba_dev->hba.lmmio_space.flags)pci_add_resource_offset (&resources, &lba_dev->hba.lmmio_space, lba_dev->hba.lmmio_space_offset); -if (lba_dev->hba.gmmio_space.flags) -pci_add_resource (&resources, &lba_dev->hba.gmmio_space); +if (lba_dev->hba.gmmio_space.flags) { +/* pci_add_resource(&resources, &lba_dev->hba.gmmio_space); */ +pr_warn ("LBA: Not registering GMMIO space %pR\n" , + &lba_dev->hba.gmmio_space); + }pci_add_resource (&resources, &lba_dev->hba.bus_num);
drivers/pci/Kconfig +1 -1
@@ -2,7 +2,7 @@ # PCI configuration # config PCI_BUS_ADDR_T_64BIT - def_bool y if (ARCH_DMA_ADDR_T_64BIT || (64BIT && !PARISC)) + def_bool y if (ARCH_DMA_ADDR_T_64BIT || 64BIT) depends on PCI config PCI_MSI
Greg Kroah-Hartman
maintainerPATCH 77/159 parisc: Use double word condition in 64bit CAS operation
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: John David Anglin <[email protected]>
commit 1b59ddfcf1678de38a1f8ca9fb8ea5eebeff1843 upstream.
The attached change fixes the condition used in the "sub" instruction. A double word comparison is needed. This fixes the 64-bit LWS CAS operation on 64-bit kernels.
I can now enable 64-bit atomic support in GCC.
--- arch/parisc/kernel/syscall.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/parisc/kernel/syscall.S +1 -1
@@ -821,7 +821,7 @@ cas2_action: /* 64bit CAS */ #ifdef CONFIG_64BIT 19: ldd,ma 0(%sr3,%r26), %r29 - sub,= %r29, %r25, %r0 + sub,*= %r29, %r25, %r0 b,n cas2_end 20: std,ma %r24, 0(%sr3,%r26) copy %r0, %r28
Greg Kroah-Hartman
maintainerPATCH 78/159 parisc: Filter out spurious interrupts in PA-RISC irq handler
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Helge Deller <[email protected]>
commit b1b4e435e4ef7de77f07bf2a42c8380b960c2d44 upstream.
When detecting a serial port on newer PA-RISC machines (with iosapic) we have a long way to go to find the right IRQ line, registering it, then registering the serial port and the irq handler for the serial port. During this phase spurious interrupts for the serial port may happen which then crashes the kernel because the action handler might not have been set up yet.
So, basically it's a race condition between the serial port hardware and the CPU which sets up the necessary fields in the irq sructs. The main reason for this race is, that we unmask the serial port irqs too early without having set up everything properly before (which isn't easily possible because we need the IRQ number to register the serial ports).
This patch is a work-around for this problem. It adds checks to the CPU irq handler to verify if the IRQ action field has been initialized already. If not, we just skip this interrupt (which isn't critical for a serial port at bootup). The real fix would probably involve rewriting all PA-RISC specific IRQ code (for CPU, IOSAPIC, GSC and EISA) to use IRQ domains with proper parenting of the irq chips and proper irq enabling along this line.
This bug has been in the PA-RISC port since the beginning, but the crashes happened very rarely with currently used hardware. But on the latest machine which I bought (a C8000 workstation), which uses the fastest CPUs (4 x PA8900, 1GHz) and which has the largest possible L1 cache size (64MB each), the kernel crashed at every boot because of this race. So, without this patch the machine would currently be unuseable.
For the record, here is the flow logic:
1. serial_init_chip() in 8250_gsc.c calls iosapic_serial_irq().
2. iosapic_serial_irq() calls txn_alloc_irq() to find the irq.
3. iosapic_serial_irq() calls cpu_claim_irq() to register the CPU irq
4. cpu_claim_irq() unmasks the CPU irq (which it shouldn't!)
5. serial_init_chip() then registers the 8250 port.
Problems:
- In step 4 the CPU irq shouldn't have been registered yet, but after step 5
- If serial irq happens between 4 and 5 have finished, the kernel will crash
--- arch/parisc/kernel/irq.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
arch/parisc/kernel/irq.c +6 -2
@@ -507,8 +507,8 @@ void do_cpu_irq_mask(struct pt_regs *regstruct pt_regs *old_regs;unsigned long eirr_val;int irq, cpu =smp_processor_id (); -#ifdef CONFIG_SMP struct irq_desc *desc; +#ifdef CONFIG_SMP cpumask_t dest;#endif
@@ -521,8 +521,12 @@ void do_cpu_irq_mask(struct pt_regs *reg goto set_out; irq =eirr_to_irq (eirr_val); -#ifdef CONFIG_SMP +/* Filter out spurious interrupts, mostly from serial port at bootup */ desc =irq_to_desc (irq); +if (unlikely (!desc->action)) + goto set_out; + +#ifdef CONFIG_SMP cpumask_copy (&dest, desc->irq_data.affinity);if (irqd_is_per_cpu (&desc->irq_data) && !cpumask_test_cpu (smp_processor_id (), &dest)) {
Greg Kroah-Hartman
maintainerPATCH 79/159 vmscan: fix increasing nr_isolated incurred by putback unevictable pages
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jaewon Kim <[email protected]>
commit c54839a722a02818677bcabe57e957f0ce4f841d upstream.
reclaim_clean_pages_from_list() assumes that shrink_page_list() returns number of pages removed from the candidate list. But shrink_page_list() puts back mlocked pages without passing it to caller and without counting as nr_reclaimed. This increases nr_isolated.
To fix this, this patch changes shrink_page_list() to pass unevictable pages back to caller. Caller will take care those pages.
Minchan said:
It fixes two issues.
1. With unevictable page, cma_alloc will be successful.
Exactly speaking, cma_alloc of current kernel will fail due to unevictable pages.
2. fix leaking of NR_ISOLATED counter of vmstat
With it, too_many_isolated works. Otherwise, it could make hang until the process get SIGKILL.
--- mm/vmscan.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
mm/vmscan.c +1 -1
@@ -1153,7 +1153,7 @@ cull_mlocked:if (PageSwapCache (page))try_to_free_swap (page);unlock_page (page); -putback_lru_page (page); +list_add (&page->lru, &ret_pages);continue ; activate_locked:
Greg Kroah-Hartman
maintainerPATCH 80/159 fs: if a coredump already exists, unlink and recreate with O_EXCL
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jann Horn <[email protected]>
commit fbb1816942c04429e85dbf4c1a080accc534299e upstream.
It was possible for an attacking user to trick root (or another user) into writing his coredumps into an attacker-readable, pre-existing file using rename() or link(), causing the disclosure of secret data from the victim process' virtual memory. Depending on the configuration, it was also possible to trick root into overwriting system files with coredumps. Fix that issue by never writing coredumps into existing files.
Requirements for the attack:
- The attack only applies if the victim's process has a nonzero RLIMIT_CORE and is dumpable.
- The attacker can trick the victim into coredumping into an attacker-writable directory D, either because the core_pattern is relative and the victim's cwd is attacker-writable or because an absolute core_pattern pointing to a world-writable directory is used.
- The attacker has one of these:
A: on a system with protected_hardlinks=0:
execute access to a folder containing a victim-owned, attacker-readable file on the same partition as D, and the victim-owned file will be deleted before the main part of the attack takes place. (In practice, there are lots of files that fulfill this condition, e.g. entries in Debian's /var/lib/dpkg/info/.) This does not apply to most Linux systems because most distros set protected_hardlinks=1.
B: on a system with protected_hardlinks=1:
execute access to a folder containing a victim-owned, attacker-readable and attacker-writable file on the same partition as D, and the victim-owned file will be deleted before the main part of the attack takes place.
(This seems to be uncommon.)
C: on any system, independent of protected_hardlinks:
write access to a non-sticky folder containing a victim-owned, attacker-readable file on the same partition as D
(This seems to be uncommon.)
The basic idea is that the attacker moves the victim-owned file to where he expects the victim process to dump its core. The victim process dumps its core into the existing file, and the attacker reads the coredump from it.
If the attacker can't move the file because he does not have write access to the containing directory, he can instead link the file to a directory he controls, then wait for the original link to the file to be deleted (because the kernel checks that the link count of the corefile is 1).
A less reliable variant that requires D to be non-sticky works with link() and does not require deletion of the original link: link() the file into D, but then unlink() it directly before the kernel performs the link count check.
On systems with protected_hardlinks=0, this variant allows an attacker to not only gain information from coredumps, but also clobber existing, victim-writable files with coredumps. (This could theoretically lead to a privilege escalation.)
--- fs/coredump.c | 38 ++++++++++++++++++++++++++++++++------ 1 file changed, 32 insertions(+), 6 deletions(-)
fs/coredump.c +32 -6
@@ -506,10 +506,10 @@ void do_coredump(const siginfo_t *siginfconst struct cred *old_cred;struct cred *cred;int retval =0 ; -int flag =0 ;int ispipe;struct files_struct *displaced; -bool need_nonrelative = false; +/* require nonrelative corefile path and be extra careful */ +bool need_suid_safe = false;bool core_dumped = false;static atomic_t core_dump_count =ATOMIC_INIT (0 );struct coredump_params cprm = {
@@ -543,9 +543,8 @@ void do_coredump(const siginfo_t *siginf*/ if (__get_dumpable (cprm.mm_flags) ==SUID_DUMP_ROOT ) {/* Setuid core dump mode */ - flag =O_EXCL ;/* Stop rewrite attacks */ cred->fsuid =GLOBAL_ROOT_UID ;/* Dump root private */ - need_nonrelative = true; + need_suid_safe = true; } retval =coredump_wait (siginfo->si_signo, &core_state);
@@ -626,7 +625,7 @@ void do_coredump(const siginfo_t *siginfif (cprm.limit < binfmt->min_coredump) goto fail_unlock; -if (need_nonrelative && cn.corename[0 ] !='/' ) { +if (need_suid_safe && cn.corename[0 ] !='/' ) {printk (KERN_WARNING "Pid %d(%s) can only dump core " \"to fully qualified path!\n" ,task_tgid_vnr (current), current->comm);
@@ -634,8 +633,35 @@ void do_coredump(const siginfo_t *siginf goto fail_unlock; } +/* +* Unlink the file if it exists unless this is a SUID +* binary - in that case, we're running around with root +* privs and don't want to unlink another user's coredump. +*/ +if (!need_suid_safe) { +mm_segment_t old_fs; + + old_fs =get_fs (); +set_fs (KERNEL_DS ); +/* +* If it doesn't exist, that's fine. If there's some +* other problem, we'll catch it at the filp_open(). +*/ + (void )sys_unlink ((const char __user *)cn.corename); +set_fs (old_fs); + } + +/* +* There is a race between unlinking and creating the +* file, but if that causes an EEXIST here, that's +* fine - another process raced with us while creating +* the corefile, and the other process won. To userspace, +* what matters is that at least one of the two processes +* writes its coredump successfully, not which one. +*/ cprm.file =filp_open (cn.corename, -O_CREAT |2 |O_NOFOLLOW |O_LARGEFILE | flag, +O_CREAT |2 |O_NOFOLLOW | +O_LARGEFILE |O_EXCL ,0600 );if (IS_ERR (cprm.file)) goto fail_unlock;
Greg Kroah-Hartman
maintainerPATCH 81/159 fs: Dont dump core if the corefile would become world-readable.
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jann Horn <[email protected]>
commit 40f705a736eac10e7dca7ab5dd5ed675a6df031d upstream.
On a filesystem like vfat, all files are created with the same owner and mode independent of who created the file. When a vfat filesystem is mounted with root as owner of all files and read access for everyone, root's processes left world-readable coredumps on it (but other users' processes only left empty corefiles when given write access because of the uid mismatch).
Given that the old behavior was inconsistent and insecure, I don't see a problem with changing it. Now, all processes refuse to dump core unless the resulting corefile will only be readable by their owner.
--- fs/coredump.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
fs/coredump.c +6 -2
@@ -678,11 +678,15 @@ void do_coredump(const siginfo_t *siginfif (!S_ISREG (inode->i_mode)) goto close_fail;/* -* Dont allow local users get cute and trick others to coredump -* into their pre-created files. +* Don't dump core if the filesystem changed owner or mode +* of the file during file creation. This is an issue when +* a process dumps core while its cwd is e.g. on a vfat +* filesystem. */ if (!uid_eq (inode->i_uid,current_fsuid ())) goto close_fail; +if ((inode->i_mode &0677 ) !=0600 ) + goto close_fail;if (!(cprm.file->f_mode &FMODE_CAN_WRITE )) goto close_fail;if (do_truncate (cprm.file->f_path.dentry,0 ,0 , cprm.file))
Greg Kroah-Hartman
maintainerPATCH 82/159 mmc: sdhci-pci: set the clear transfer mode register quirk for O2Micro
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Adam Lee <[email protected]>
commit 143b648ddf1583905fa15d32be27a31442fc7933 upstream.
This patch fixes MMC not working issue on O2Micro/BayHub Host, which requires transfer mode register to be cleared when sending no DMA command.
--- drivers/mmc/host/sdhci-pci.c | 1 + 1 file changed, 1 insertion(+)
drivers/mmc/host/sdhci-pci.c +1 -0
@@ -549,6 +549,7 @@ static int jmicron_resume(struct sdhci_pstatic const struct sdhci_pci_fixes sdhci_o2 = { .probe = sdhci_pci_o2_probe, .quirks =SDHCI_QUIRK_NO_ENDATTR_IN_NOPDESC , + .quirks2 =SDHCI_QUIRK2_CLEAR_TRANSFERMODE_REG_BEFORE_CMD , .probe_slot = sdhci_pci_o2_probe_slot, .resume = sdhci_pci_o2_resume, };
Greg Kroah-Hartman
maintainerPATCH 83/159 mmc: sdhci: also get preset value and driver type for MMC_DDR52
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jisheng Zhang <[email protected]>
commit 0dafa60eb2506617e6968b97cc5a44914a7fb1a6 upstream.
commit bb8175a8aa42 ("mmc: sdhci: clarify DDR timing mode between SD-UHS and eMMC") added MMC_DDR52 as eMMC's DDR mode to be distinguished from SD-UHS, but it missed setting driver type for MMC_DDR52 timing mode.
So sometimes we get the following error on Marvell BG2Q DMP board:
[ 1.559598] mmcblk0: error -84 transferring data, sector 0, nr 8, cmd response 0x900, card status 0xb00 [ 1.569314] mmcblk0: retrying using single block read [ 1.575676] mmcblk0: error -84 transferring data, sector 2, nr 6, cmd response 0x900, card status 0x0 [ 1.585202] blk_update_request: I/O error, dev mmcblk0, sector 2 [ 1.591818] mmcblk0: error -84 transferring data, sector 3, nr 5, cmd response 0x900, card status 0x0 [ 1.601341] blk_update_request: I/O error, dev mmcblk0, sector 3
This patches fixes this by adding the missing driver type setting.
--- drivers/mmc/host/sdhci.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/mmc/host/sdhci.c +3 -1
@@ -1146,6 +1146,7 @@ static u16 sdhci_get_preset_value(struct preset =sdhci_readw (host,SDHCI_PRESET_FOR_SDR104 );break ;case MMC_TIMING_UHS_DDR50 : +case MMC_TIMING_MMC_DDR52 : preset =sdhci_readw (host,SDHCI_PRESET_FOR_DDR50 );break ;case MMC_TIMING_MMC_HS400 :
@@ -1598,7 +1599,8 @@ static void sdhci_do_set_ios(struct sdhc (ios->timing ==MMC_TIMING_UHS_SDR25 ) || (ios->timing ==MMC_TIMING_UHS_SDR50 ) || (ios->timing ==MMC_TIMING_UHS_SDR104 ) || - (ios->timing == MMC_TIMING_UHS_DDR50))) { + (ios->timing ==MMC_TIMING_UHS_DDR50 ) || + (ios->timing == MMC_TIMING_MMC_DDR52))) { u16 preset;sdhci_enable_preset_value (host, true);
Greg Kroah-Hartman
maintainerPATCH 84/159 mmc: core: fix race condition in mmc_wait_data_done
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jialing Fu <[email protected]>
commit 71f8a4b81d040b3d094424197ca2f1bf811b1245 upstream.
The following panic is captured in ker3.14, but the issue still exists in latest kernel.
---------------------------------------------------------------------
[ 20.738217] c0 3136 (Compiler) Unable to handle kernel NULL pointer dereference at virtual address 00000578 ...... [ 20.738499] c0 3136 (Compiler) PC is at _raw_spin_lock_irqsave+0x24/0x60 [ 20.738527] c0 3136 (Compiler) LR is at _raw_spin_lock_irqsave+0x20/0x60 [ 20.740134] c0 3136 (Compiler) Call trace: [ 20.740165] c0 3136 (Compiler) [<ffffffc0008ee900>] _raw_spin_lock_irqsave+0x24/0x60 [ 20.740200] c0 3136 (Compiler) [<ffffffc0000dd024>] __wake_up+0x1c/0x54 [ 20.740230] c0 3136 (Compiler) [<ffffffc000639414>] mmc_wait_data_done+0x28/0x34 [ 20.740262] c0 3136 (Compiler) [<ffffffc0006391a0>] mmc_request_done+0xa4/0x220 [ 20.740314] c0 3136 (Compiler) [<ffffffc000656894>] sdhci_tasklet_finish+0xac/0x264 [ 20.740352] c0 3136 (Compiler) [<ffffffc0000a2b58>] tasklet_action+0xa0/0x158 [ 20.740382] c0 3136 (Compiler) [<ffffffc0000a2078>] __do_softirq+0x10c/0x2e4 [ 20.740411] c0 3136 (Compiler) [<ffffffc0000a24bc>] irq_exit+0x8c/0xc0 [ 20.740439] c0 3136 (Compiler) [<ffffffc00008489c>] handle_IRQ+0x48/0xac [ 20.740469] c0 3136 (Compiler) [<ffffffc000081428>] gic_handle_irq+0x38/0x7c
----------------------------------------------------------------------
Because in SMP, "mrq" has race condition between below two paths:
path1: CPU0: <tasklet context>
static void mmc_wait_data_done(struct mmc_request *mrq)
{
mrq->host->context_info.is_done_rcv = true;
//
// If CPU0 has just finished "is_done_rcv = true" in path1, and at
// this moment, IRQ or ICache line missing happens in CPU0.
// What happens in CPU1 (path2)?
//
// If the mmcqd thread in CPU1(path2) hasn't entered to sleep mode:
// path2 would have chance to break from wait_event_interruptible
// in mmc_wait_for_data_req_done and continue to run for next
// mmc_request (mmc_blk_rw_rq_prep).
//
// Within mmc_blk_rq_prep, mrq is cleared to 0.
// If below line still gets host from "mrq" as the result of
// compiler, the panic happens as we traced.
wake_up_interruptible(&mrq->host->context_info.wait);}
path2: CPU1: <The mmcqd thread runs mmc_queue_thread>
static int mmc_wait_for_data_req_done(...
{
...
while (1) {
wait_event_interruptible(context_info->wait,
(context_info->is_done_rcv ||
context_info->is_new_req));
static void mmc_blk_rw_rq_prep(...
{
...
memset(brq, 0, sizeof(struct mmc_blk_request));This issue happens very coincidentally; however adding mdelay(1) in mmc_wait_data_done as below could duplicate it easily.
static void mmc_wait_data_done(struct mmc_request *mrq)
{
mrq->host->context_info.is_done_rcv = true;
+ mdelay(1);
wake_up_interruptible(&mrq->host->context_info.wait);
}At runtime, IRQ or ICache line missing may just happen at the same place of the mdelay(1).
This patch gets the mmc_context_info at the beginning of function, it can avoid this race condition.
--- drivers/mmc/core/core.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/mmc/core/core.c +4 -2
@@ -330,8 +330,10 @@ EXPORT_SYMBOL(mmc_start_bkops);*/ static void mmc_wait_data_done (struct mmc_request *mrq) { - mrq->host->context_info.is_done_rcv = true; -wake_up_interruptible (&mrq->host->context_info.wait); +struct mmc_context_info *context_info = &mrq->host->context_info; + + context_info->is_done_rcv = true; +wake_up_interruptible (&context_info->wait); }static void mmc_wait_done (struct mmc_request *mrq)
Greg Kroah-Hartman
maintainerPATCH 85/159 md/raid5: avoid races when changing cache size.
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: NeilBrown <[email protected]>
commit 2d5b569b665ea6d0b15c52529ff06300de81a7ce upstream.
Cache size can grow or shrink due to various pressures at any time. So when we resize the cache as part of a 'grow' operation (i.e. change the size to allow more devices) we need to blocks that automatic growing/shrinking.
So introduce a mutex. auto grow/shrink uses mutex_trylock() and just doesn't bother if there is a blockage. Resizing the whole cache holds the mutex to ensure that the correct number of new stripes is allocated.
This bug can result in some stripes not being freed when an array is stopped. This leads to the kmem_cache not being freed and a subsequent array can try to use the same kmem_cache and get confused.
--- drivers/md/raid5.c | 31 +++++++++++++++++++++++++------ drivers/md/raid5.h | 3 ++- 2 files changed, 27 insertions(+), 7 deletions(-)
drivers/md/raid5.c +25 -6
@@ -2151,6 +2151,9 @@ static int resize_stripes(struct r5confif (!sc)return -ENOMEM ; +/* Need to ensure auto-resizing doesn't interfere */ +mutex_lock (&conf->cache_size_mutex); +for (i = conf->max_nr_stripes; i; i--) { nsh =alloc_stripe (sc,GFP_KERNEL );if (!nsh)
@@ -2167,6 +2170,7 @@ static int resize_stripes(struct r5confkmem_cache_free (sc, nsh); }kmem_cache_destroy (sc); +mutex_unlock (&conf->cache_size_mutex);return -ENOMEM ; }/* Step 2 - Must use GFP_NOIO now.
@@ -2213,6 +2217,7 @@ static int resize_stripes(struct r5conf }else err = -ENOMEM ; +mutex_unlock (&conf->cache_size_mutex);/* Step 4, return new stripes to service */ while (!list_empty (&newstripes)) { nsh = list_entry(newstripes.next, struct stripe_head, lru);
@@ -5846,12 +5851,14 @@ static void raid5d(struct md_thread *thrpr_debug ("%d stripes handled\n" , handled);spin_unlock_irq (&conf->device_lock); -if (test_and_clear_bit (R5_ALLOC_MORE , &conf->cache_state)) { +if (test_and_clear_bit (R5_ALLOC_MORE , &conf->cache_state) && +mutex_trylock (&conf->cache_size_mutex)) {grow_one_stripe (conf, __GFP_NOWARN);/* Set flag even if allocation failed. This helps * slow down allocation requests when mem is short */ set_bit (R5_DID_ALLOC , &conf->cache_state); +mutex_unlock (&conf->cache_size_mutex); }async_tx_issue_pending_all ();
@@ -5883,18 +5890,22 @@ raid5_set_cache_size(struct mddev *mddevreturn -EINVAL ; conf->min_nr_stripes = size; +mutex_lock (&conf->cache_size_mutex);while (size < conf->max_nr_stripes &&drop_one_stripe (conf)) ; +mutex_unlock (&conf->cache_size_mutex); err =md_allow_write (mddev);if (err)return err; +mutex_lock (&conf->cache_size_mutex);while (size > conf->max_nr_stripes)if (!grow_one_stripe (conf,GFP_KERNEL ))break ; +mutex_unlock (&conf->cache_size_mutex);return 0 ; }
@@ -6360,11 +6371,18 @@ static unsigned long raid5_cache_scan(st struct shrink_control *sc) { struct r5conf *conf = container_of(shrink, struct r5conf, shrinker); -int ret =0 ; -while (ret < sc->nr_to_scan) { -if (drop_one_stripe (conf) ==0 ) -return SHRINK_STOP ; - ret++; +unsigned long ret =SHRINK_STOP ; + +if (mutex_trylock (&conf->cache_size_mutex)) { + ret=0 ; +while (ret < sc->nr_to_scan) { +if (drop_one_stripe (conf) ==0 ) { + ret =SHRINK_STOP ; +break ; + } + ret++; + } +mutex_unlock (&conf->cache_size_mutex); }return ret; }
@@ -6433,6 +6451,7 @@ static struct r5conf *setup_conf(struct goto abort;spin_lock_init (&conf->device_lock);seqcount_init (&conf->gen_lock); +mutex_init (&conf->cache_size_mutex);init_waitqueue_head (&conf->wait_for_stripe);init_waitqueue_head (&conf->wait_for_overlap);INIT_LIST_HEAD (&conf->handle_list);
drivers/md/raid5.h +2 -1
@@ -482,7 +482,8 @@ struct r5conf {*/ int active_name;char cache_name[2 ][32 ]; -struct kmem_cache *slab_cache;/* for allocating stripes */ +struct kmem_cache *slab_cache;/* for allocating stripes */ +struct mutex cache_size_mutex;/* Protect changes to cache size */ int seq_flush, seq_write;int quiesce;
Greg Kroah-Hartman
maintainerPATCH 86/159 md/raid5: dont let shrink_slab shrink too far.
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: NeilBrown <[email protected]>
commit 49895bcc7e566ba455eb2996607d6fbd3447ce16 upstream.
I have a report of drop_one_stripe() called from raid5_cache_scan() apparently finding ->max_nr_stripes == 0.
This should not be allowed.
So add a test to keep max_nr_stripes above min_nr_stripes.
Also use a 'mask' rather than a 'mod' in drop_one_stripe to ensure 'hash' is valid even if max_nr_stripes does reach zero.
--- drivers/md/raid5.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/md/raid5.c +3 -2
@@ -2245,7 +2245,7 @@ static int resize_stripes(struct r5confstatic int drop_one_stripe (struct r5conf *conf) {struct stripe_head *sh; -int hash = (conf->max_nr_stripes -1 ) %NR_STRIPE_HASH_LOCKS ; +int hash = (conf->max_nr_stripes -1 ) &STRIPE_HASH_LOCKS_MASK ;spin_lock_irq (conf->hash_locks + hash); sh =get_free_stripe (conf, hash);
@@ -6375,7 +6375,8 @@ static unsigned long raid5_cache_scan(stif (mutex_trylock (&conf->cache_size_mutex)) { ret=0 ; -while (ret < sc->nr_to_scan) { +while (ret < sc->nr_to_scan && + conf->max_nr_stripes > conf->min_nr_stripes) {if (drop_one_stripe (conf) ==0 ) { ret =SHRINK_STOP ;break ;
Greg Kroah-Hartman
maintainerPATCH 87/159 md/raid10: always set reshape_safe when initializing reshape_position.
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: NeilBrown <[email protected]>
commit 299b0685e31c9f3dcc2d58ee3beca761a40b44b3 upstream.
'reshape_position' tracks where in the reshape we have reached. 'reshape_safe' tracks where in the reshape we have safely recorded in the metadata.
These are compared to determine when to update the metadata. So it is important that reshape_safe is initialised properly. Currently it isn't. When starting a reshape from the beginning it usually has the correct value by luck. But when reducing the number of devices in a RAID10, it has the wrong value and this leads to the metadata not being updated correctly.
This can lead to corruption if the reshape is not allowed to complete.
This patch is suitable for any -stable kernel which supports RAID10 reshape, which is 3.5 and later.
--- drivers/md/raid10.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/md/raid10.c +4 -1
@@ -3566,6 +3566,7 @@ static struct r10conf *setup_conf(struct/* far_copies must be 1 */ conf->prev.stride = conf->dev_sectors; } + conf->reshape_safe = conf->reshape_progress;spin_lock_init (&conf->device_lock);INIT_LIST_HEAD (&conf->retry_list);
@@ -3770,7 +3771,6 @@ static int run(struct mddev *mddev) } conf->offset_diff = min_offset_diff; - conf->reshape_safe = conf->reshape_progress;clear_bit (MD_RECOVERY_SYNC , &mddev->recovery);clear_bit (MD_RECOVERY_CHECK , &mddev->recovery);set_bit (MD_RECOVERY_RESHAPE , &mddev->recovery);
@@ -4113,6 +4113,7 @@ static int raid10_start_reshape(struct m conf->reshape_progress = size; }else conf->reshape_progress =0 ; + conf->reshape_safe = conf->reshape_progress;spin_unlock_irq (&conf->device_lock);if (mddev->delta_disks && mddev->bitmap) {
@@ -4180,6 +4181,7 @@ abort: rdev->new_data_offset = rdev->data_offset;smp_wmb (); conf->reshape_progress = MaxSector; + conf->reshape_safe = MaxSector; mddev->reshape_position = MaxSector;spin_unlock_irq (&conf->device_lock);return ret;
@@ -4534,6 +4536,7 @@ static void end_reshape(struct r10conf *md_finish_reshape (conf->mddev);smp_wmb (); conf->reshape_progress = MaxSector; + conf->reshape_safe = MaxSector;spin_unlock_irq (&conf->device_lock);/* read-ahead size must cover two whole stripes, which is
Greg Kroah-Hartman
maintainerPATCH 88/159 md: flush ->event_work before stopping array.
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: NeilBrown <[email protected]>
commit ee5d004fd0591536a061451eba2b187092e9127c upstream.
The 'event_work' worker used by dm-raid may still be running when the array is stopped. This can result in an oops.
So flush the workqueue on which it is run after detaching and before destroying the device.
--- drivers/md/md.c | 2 ++ 1 file changed, 2 insertions(+)
drivers/md/md.c +2 -0
@@ -5365,6 +5365,8 @@ static void __md_stop(struct mddev *mdde {struct md_personality *pers = mddev->pers;mddev_detach (mddev); +/* Ensure ->event_work is done */ +flush_workqueue (md_misc_wq);spin_lock (&mddev->lock); mddev->ready =0 ; mddev->pers =NULL ;
Greg Kroah-Hartman
maintainerPATCH 89/159 iommu/fsl: Really fix init section(s) content
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Emil Medve <[email protected]>
commit 57fb907da89977640ef183556a621336c1348fa0 upstream.
'0f1fb99 iommu/fsl: Fix section mismatch' was intended to address the modpost warning and the potential crash. Crash which is actually easy to trigger with a 'unbind' followed by a 'bind' sequence. The fix is wrong as fsl_of_pamu_driver.driver gets added by bus_add_driver() to a couple of klist(s) which become invalid/corrupted as soon as the init sections are freed. Depending on when/how the init sections storage is reused various/random errors and crashes will happen
'cd70d46 iommu/fsl: Various cleanups' contains annotations that go further down the wrong path laid by '0f1fb99 iommu/fsl: Fix section mismatch'
Now remove all the incorrect annotations from the above mentioned patches (not exactly a revert) and those previously existing in the code, This fixes the modpost warning(s), the unbind/bind sequence crashes and the random errors/crashes
--- drivers/iommu/fsl_pamu.c | 26 +++++++++++++------------- 1 file changed, 13 insertions(+), 13 deletions(-)
drivers/iommu/fsl_pamu.c +13 -13
@@ -41,7 +41,6 @@ struct pamu_isr_data {static struct paace *ppaact;static struct paace *spaact; -static struct ome *omt__initdata ;/* * Table for matching compatible strings, for device tree
@@ -50,7 +49,7 @@ static struct ome *omt __initdata;* SOCs. For the older SOCs "fsl,qoriq-device-config-1.0" * string would be used. */ -static const struct of_device_id guts_device_ids[]__initconst = { +static const struct of_device_id guts_device_ids[] = { { .compatible ="fsl,qoriq-device-config-1.0" , }, { .compatible ="fsl,qoriq-device-config-2.0" , }, {}
@@ -599,7 +598,7 @@ found_cpu_node:* Memory accesses to QMAN and BMAN private memory need not be coherent, so * clear the PAACE entry coherency attribute for them. */ -static void __init setup_qbman_paace (struct paace *ppaace,int paace_type) +static void setup_qbman_paace (struct paace *ppaace,int paace_type) {switch (paace_type) {case QMAN_PAACE :
@@ -629,7 +628,7 @@ static void __init setup_qbman_paace(str* this table to translate device transaction to appropriate corenet * transaction. */ -static void __init setup_omt (struct ome *omt) +static void setup_omt (struct ome *omt) {struct ome *ome;
@@ -666,7 +665,7 @@ static void __init setup_omt(struct ome* Get the maximum number of PAACT table entries * and subwindows supported by PAMU */ -static void __init get_pamu_cap_values (unsigned long pamu_reg_base) +static void get_pamu_cap_values (unsigned long pamu_reg_base) {u32 pc_val;
@@ -676,9 +675,9 @@ static void __init get_pamu_cap_values(u }/* Setup PAMU registers pointing to PAACT, SPAACT and OMT */ -static int __init setup_one_pamu (unsigned long pamu_reg_base,unsigned long pamu_reg_size, -phys_addr_t ppaact_phys,phys_addr_t spaact_phys, -phys_addr_t omt_phys) +static int setup_one_pamu (unsigned long pamu_reg_base,unsigned long pamu_reg_size, +phys_addr_t ppaact_phys,phys_addr_t spaact_phys, +phys_addr_t omt_phys) {u32 *pc;struct pamu_mmap_regs *pamu_regs;
@@ -720,7 +719,7 @@ static int __init setup_one_pamu(unsigne }/* Enable all device LIODNS */ -static void __init setup_liodns (void ) +static void setup_liodns (void ) {int i, len;struct paace *ppaace;
@@ -846,7 +845,7 @@ struct ccsr_law {/* * Create a coherence subdomain for a given memory block. */ -static int __init create_csd (phys_addr_t phys,size_t size,u32 csd_port_id) +static int create_csd (phys_addr_t phys,size_t size,u32 csd_port_id) {struct device_node *np;const __be32 *iprop;
@@ -988,7 +987,7 @@ error:static const struct {u32 svr;u32 port_id; -} port_id_map[]__initconst = { +} port_id_map[] = { {(SVR_P2040 <<8 ) |0x10 ,0xFF000000 },/* P2040 1.0 */ {(SVR_P2040 <<8 ) |0x11 ,0xFF000000 },/* P2040 1.1 */ {(SVR_P2041 <<8 ) |0x10 ,0xFF000000 },/* P2041 1.0 */
@@ -1006,7 +1005,7 @@ static const struct {#define SVR_SECURITY 0x80000/* The Security (E) bit */ -static int __init fsl_pamu_probe (struct platform_device *pdev) +static int fsl_pamu_probe (struct platform_device *pdev) {struct device *dev = &pdev->dev;void __iomem *pamu_regs =NULL ;
@@ -1022,6 +1021,7 @@ static int __init fsl_pamu_probe(structint irq;phys_addr_t ppaact_phys;phys_addr_t spaact_phys; +struct ome *omt;phys_addr_t omt_phys;size_t mem_size =0 ;unsigned int order =0 ;
@@ -1200,7 +1200,7 @@ error:return ret; } -static struct platform_driver fsl_of_pamu_driver__initdata = { +static struct platform_driver fsl_of_pamu_driver = { .driver = { .name ="fsl-of-pamu" , },
Greg Kroah-Hartman
maintainerPATCH 90/159 iommu/io-pgtable-arm: Unmap and free table when overwriting with block
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Will Deacon <[email protected]>
commit cf27ec930be906e142c752f9161197d69ca534d7 upstream.
When installing a block mapping, we unconditionally overwrite a non-leaf PTE if we find one. However, this can cause a problem if the following sequence of events occur:
(1) iommu_map called for a 4k (i.e. PAGE_SIZE) mapping at some address
- We initialise the page table all the way down to a leaf entry
- No TLB maintenance is required, because we're going from invalid to valid.
(2) iommu_unmap is called on the mapping installed in (1)
- We walk the page table to the final (leaf) entry and zero it
- We only changed a valid leaf entry, so we invalidate leaf-only
(3) iommu_map is called on the same address as (1), but this time for a 2MB (i.e. BLOCK_SIZE) mapping)
- We walk the page table down to the penultimate level, where we find a table entry
- We overwrite the table entry with a block mapping and return without any TLB maintenance and without freeing the memory used by the now-orphaned table.
This last step can lead to a walk-cache caching the overwritten table entry, causing unexpected faults when the new mapping is accessed by a device. One way to fix this would be to collapse the page table when freeing the last page at a given level, but this would require expensive iteration on every map call. Instead, this patch detects the case when we are overwriting a table entry and explicitly unmaps the table first, which takes care of both freeing and TLB invalidation.
--- drivers/iommu/io-pgtable-arm.c | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-)
drivers/iommu/io-pgtable-arm.c +16 -1
@@ -200,6 +200,10 @@ typedef u64 arm_lpae_iopte;static bool selftest_running = false; +static int __arm_lpae_unmap (struct arm_lpae_io_pgtable *data, +unsigned long iova,size_t size,int lvl, +arm_lpae_iopte *ptep); + static int arm_lpae_init_pte(struct arm_lpae_io_pgtable *data, unsigned long iova, phys_addr_t paddr, arm_lpae_iopte prot, int lvl,
@@ -207,10 +211,21 @@ static int arm_lpae_init_pte(struct arm_ {arm_lpae_iopte pte = prot; -/* We require an unmap first */ if (iopte_leaf (*ptep, lvl)) { +/* We require an unmap first */ WARN_ON (!selftest_running);return -EEXIST ; + }else if (iopte_type (*ptep, lvl) ==ARM_LPAE_PTE_TYPE_TABLE ) { +/* +* We need to unmap and free the old table before +* overwriting it with a block entry. +*/ +arm_lpae_iopte *tblp; +size_t sz =ARM_LPAE_BLOCK_SIZE (lvl, data); + + tblp = ptep -ARM_LPAE_LVL_IDX (iova, lvl, data); +if (WARN_ON (__arm_lpae_unmap (data, iova, sz, lvl, tblp) != sz)) +return -EINVAL ; }if (data->iop.cfg.quirks &IO_PGTABLE_QUIRK_ARM_NS )
Greg Kroah-Hartman
maintainerPATCH 91/159 iommu/tegra-smmu: Parameterize number of TLB lines
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Thierry Reding <[email protected]>
commit 11cec15bf3fb498206ef63b1fa26c27689e02d0e upstream.
The number of TLB lines was increased from 16 on Tegra30 to 32 on Tegra114 and later. Parameterize the value so that the initial default can be set accordingly.
On Tegra30, initializing the value to 32 would effectively disable the TLB and hence cause massive latencies for memory accesses translated through the SMMU. This is especially noticeable for isochronuous clients such as display, whose FIFOs would continuously underrun.
--- drivers/iommu/tegra-smmu.c | 9 +++++++-- drivers/memory/tegra/tegra114.c | 1 + drivers/memory/tegra/tegra124.c | 1 + drivers/memory/tegra/tegra30.c | 1 + include/soc/tegra/mc.h | 1 + 5 files changed, 11 insertions(+), 2 deletions(-)
drivers/iommu/tegra-smmu.c +7 -2
@@ -26,6 +26,7 @@ struct tegra_smmu {const struct tegra_smmu_soc *soc;unsigned long pfn_mask; +unsigned long tlb_mask;unsigned long *asids;struct mutex lock;
@@ -65,7 +66,8 @@ static inline u32 smmu_readl(struct tegr#define SMMU_TLB_CONFIG 0x14#define SMMU_TLB_CONFIG_HIT_UNDER_MISS (1 << 29)#define SMMU_TLB_CONFIG_ROUND_ROBIN_ARBITRATION (1 << 28) -#define SMMU_TLB_CONFIG_ACTIVE_LINES (x) ((x) & 0x3f) +#define SMMU_TLB_CONFIG_ACTIVE_LINES (smmu) \ + ((smmu)->soc->num_tlb_lines & (smmu)->tlb_mask)#define SMMU_PTC_CONFIG 0x18#define SMMU_PTC_CONFIG_ENABLE (1 << 29)
@@ -716,6 +718,9 @@ struct tegra_smmu *tegra_smmu_probe(stru smmu->pfn_mask =BIT_MASK (mc->soc->num_address_bits -PAGE_SHIFT ) -1 ;dev_dbg (dev,"address bits: %u, PFN mask: %#lx\n" , mc->soc->num_address_bits, smmu->pfn_mask); + smmu->tlb_mask = (smmu->soc->num_tlb_lines <<1 ) -1 ; +dev_dbg (dev,"TLB lines: %u, mask: %#lx\n" , smmu->soc->num_tlb_lines, + smmu->tlb_mask); value =SMMU_PTC_CONFIG_ENABLE |SMMU_PTC_CONFIG_INDEX_MAP (0x3f );
@@ -725,7 +730,7 @@ struct tegra_smmu *tegra_smmu_probe(strusmmu_writel (smmu, value,SMMU_PTC_CONFIG ); value =SMMU_TLB_CONFIG_HIT_UNDER_MISS | -SMMU_TLB_CONFIG_ACTIVE_LINES (0x20 ); +SMMU_TLB_CONFIG_ACTIVE_LINES (smmu);if (soc->supports_round_robin_arbitration) value |=SMMU_TLB_CONFIG_ROUND_ROBIN_ARBITRATION ;
drivers/memory/tegra/tegra114.c +1 -0
@@ -935,6 +935,7 @@ static const struct tegra_smmu_soc tegra .num_swgroups = ARRAY_SIZE(tegra114_swgroups), .supports_round_robin_arbitration = false, .supports_request_limit = false, + .num_tlb_lines = 32, .num_asids = 4, .ops = &tegra114_smmu_ops, };
drivers/memory/tegra/tegra124.c +1 -0
@@ -981,6 +981,7 @@ static const struct tegra_smmu_soc tegra .num_swgroups = ARRAY_SIZE(tegra124_swgroups), .supports_round_robin_arbitration = true, .supports_request_limit = true, + .num_tlb_lines = 32, .num_asids = 128, .ops = &tegra124_smmu_ops, };
drivers/memory/tegra/tegra30.c +1 -0
@@ -957,6 +957,7 @@ static const struct tegra_smmu_soc tegra .num_swgroups = ARRAY_SIZE(tegra30_swgroups), .supports_round_robin_arbitration = false, .supports_request_limit = false, + .num_tlb_lines = 16, .num_asids = 4, .ops = &tegra30_smmu_ops, };
include/soc/tegra/mc.h +1 -0
@@ -59,6 +59,7 @@ struct tegra_smmu_soc {bool supports_round_robin_arbitration;bool supports_request_limit; +unsigned int num_tlb_lines;unsigned int num_asids;const struct tegra_smmu_ops *ops;
Greg Kroah-Hartman
maintainerPATCH 92/159 iommu/vt-d: Really use upper context table when necessary
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Joerg Roedel <[email protected]>
commit 4df4eab168c1c4058603be55a3169d4a45779cc0 upstream.
There is a bug in iommu_context_addr() which will always use the lower context table, even when the upper context table needs to be used. Fix this issue.
--- drivers/iommu/intel-iommu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iommu/intel-iommu.c +1 -1
@@ -681,6 +681,7 @@ static inline struct context_entry *iommstruct context_entry *context;u64 *entry; + entry = &root->lo;if (ecs_enabled (iommu)) {if (devfn >=0x80 ) { devfn -=0x80 ;
@@ -688,7 +689,6 @@ static inline struct context_entry *iomm } devfn *=2 ; } - entry = &root->lo;if (*entry &1 ) context =phys_to_virt (*entry &VTD_PAGE_MASK );else {
Greg Kroah-Hartman
maintainerPATCH 93/159 eCryptfs: Invalidate dcache entries when lower i_nlink is zero
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Tyler Hicks <[email protected]>
commit 5556e7e6d30e8e9b5ee51b0e5edd526ee80e5e36 upstream.
Consider eCryptfs dcache entries to be stale when the corresponding lower inode's i_nlink count is zero. This solves a problem caused by the lower inode being directly modified, without going through the eCryptfs mount, leaving stale eCryptfs dentries cached and the eCryptfs inode's i_nlink count not being cleared.
--- fs/ecryptfs/dentry.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-)
fs/ecryptfs/dentry.c +8 -8
@@ -45,20 +45,20 @@static int ecryptfs_d_revalidate (struct dentry *dentry,unsigned int flags) {struct dentry *lower_dentry =ecryptfs_dentry_to_lower (dentry); -int rc; - -if (!(lower_dentry->d_flags &DCACHE_OP_REVALIDATE )) -return 1 ; +int rc =1 ;if (flags &LOOKUP_RCU )return -ECHILD ; - rc = lower_dentry->d_op->d_revalidate (lower_dentry, flags); +if (lower_dentry->d_flags &DCACHE_OP_REVALIDATE ) + rc = lower_dentry->d_op->d_revalidate (lower_dentry, flags); +if (d_really_is_positive (dentry)) { -struct inode *lower_inode = -ecryptfs_inode_to_lower (d_inode (dentry)); +struct inode *inode =d_inode (dentry); -fsstack_copy_attr_all (d_inode (dentry), lower_inode); +fsstack_copy_attr_all (inode,ecryptfs_inode_to_lower (inode)); +if (!inode->i_nlink) +return 0 ; }return rc; }
Greg Kroah-Hartman
maintainerPATCH 94/159 hfs: fix B-tree corruption after insertion at position 0
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Hin-Tak Leung <[email protected]>
commit b4cc0efea4f0bfa2477c56af406cfcf3d3e58680 upstream.
Fix B-tree corruption when a new record is inserted at position 0 in the node in hfs_brec_insert().
This is an identical change to the corresponding hfs b-tree code to Sergei Antonov's "hfsplus: fix B-tree corruption after insertion at position 0", to keep similar code paths in the hfs and hfsplus drivers in sync, where appropriate.
--- fs/hfs/brec.c | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-)
fs/hfs/brec.c +11 -9
@@ -131,13 +131,16 @@ skip:hfs_bnode_write (node, entry, data_off + key_len, entry_len);hfs_bnode_dump (node); -if (new_node) { -/* update parent key if we inserted a key -* at the start of the first node -*/ -if (!rec && new_node != node) -hfs_brec_update_parent (fd); +/* +* update parent key if we inserted a key +* at the start of the node and it is not the new node +*/ +if (!rec && new_node != node) { +hfs_bnode_read_key (node, fd->search_key, data_off + size); +hfs_brec_update_parent (fd); + } +if (new_node) {hfs_bnode_put (fd->bnode);if (!new_node->parent) {hfs_btree_inc_height (tree);
@@ -166,9 +169,6 @@ skip: goto again; } -if (!rec) -hfs_brec_update_parent (fd); -return 0 ; }
@@ -366,6 +366,8 @@ again:if (IS_ERR (parent))return PTR_ERR (parent);__hfs_brec_find (parent, fd); +if (fd->record <0 ) +return -ENOENT ;hfs_bnode_dump (parent); rec = fd->record;
Greg Kroah-Hartman
maintainerPATCH 95/159 ideapad-laptop: Add Lenovo Yoga 3 14 to no_hw_rfkill dmi list
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Hans de Goede <[email protected]>
commit fa92a31b3335478c545cdc8e79e1e9b788184e6b upstream.
Like some of the other Yoga models the Lenovo Yoga 3 14 does not have a hw rfkill switch, and trying to read the hw rfkill switch through the ideapad module causes it to always reported blocking breaking wifi.
This commit adds the Lenovo Yoga 3 14 to the no_hw_rfkill dmi list, fixing the wifi breakage.
--- drivers/platform/x86/ideapad-laptop.c | 7 +++++++ 1 file changed, 7 insertions(+)
drivers/platform/x86/ideapad-laptop.c +7 -0
@@ -852,6 +852,13 @@ static const struct dmi_system_id no_hw_ }, }, { + .ident = "Lenovo Yoga 3 14", + .matches = { +DMI_MATCH (DMI_SYS_VENDOR ,"LENOVO" ), + DMI_MATCH(DMI_PRODUCT_VERSION, "Lenovo Yoga 3 14"), + }, + }, + { .ident ="Lenovo Yoga 3 Pro 1370" , .matches = { DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
Greg Kroah-Hartman
maintainerPATCH 96/159 IB/srp: Handle partial connection success correctly
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Bart Van Assche <[email protected]>
commit c257ea6f9f9aed0b173e0c2932bb8dac5612cdc6 upstream.
Avoid that the following kernel warning is reported if the SRP target system accepts fewer channels per connection than what was requested by the initiator system:
WARNING: at drivers/infiniband/ulp/srp/ib_srp.c:617 srp_destroy_qp+0xb1/0x120 [ib_srp]() Call Trace: [<ffffffff8105d67f>] warn_slowpath_common+0x7f/0xc0 [<ffffffff8105d6da>] warn_slowpath_null+0x1a/0x20 [<ffffffffa05419e1>] srp_destroy_qp+0xb1/0x120 [ib_srp] [<ffffffffa05445fb>] srp_create_ch_ib+0x19b/0x420 [ib_srp] [<ffffffffa0545257>] srp_create_target+0x7d7/0xa94 [ib_srp] [<ffffffff8138dac0>] dev_attr_store+0x20/0x30 [<ffffffff812079ef>] sysfs_write_file+0xef/0x170 [<ffffffff81191fc4>] vfs_write+0xb4/0x130 [<ffffffff8119276f>] sys_write+0x5f/0xa0 [<ffffffff815a0a59>] system_call_fastpath+0x16/0x1b
--- drivers/infiniband/ulp/srp/ib_srp.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/infiniband/ulp/srp/ib_srp.c +2 -1
@@ -3266,7 +3266,7 @@ static ssize_t srp_create_target(structsrp_free_ch_ib (target, ch);srp_free_req_data (target, ch); target->ch_count = ch - target->ch; -break ; + goto connected; } }
@@ -3276,6 +3276,7 @@ static ssize_t srp_create_target(struct node_idx++; } +connected: target->scsi_host->nr_hw_queues = target->ch_count; ret =srp_add_target (host, target);
Greg Kroah-Hartman
maintainerPATCH 97/159 IB/srp: Stop the scsi_eh_<n> and scsi_tmf_<n> threads if login fails
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Bart Van Assche <[email protected]>
commit bc44bd1d864664f3658352c6aaaa02557d49165d upstream.
scsi_host_alloc() not only allocates memory for a SCSI host but also creates the scsi_eh_<n> kernel thread and the scsi_tmf_<n> workqueue. Stop these threads if login fails by calling scsi_host_put().
--- drivers/infiniband/ulp/srp/ib_srp.c | 9 +++++++++ 1 file changed, 9 insertions(+)
drivers/infiniband/ulp/srp/ib_srp.c +9 -0
@@ -2761,6 +2761,13 @@ static int srp_sdev_count(struct Scsi_Horeturn c; } +/* +* Return values: +* < 0 upon failure. Caller is responsible for SRP target port cleanup. +* 0 and target->state == SRP_TARGET_REMOVED if asynchronous target port +* removal has been scheduled. +* 0 and target->state != SRP_TARGET_REMOVED upon success. +*/ static int srp_add_target (struct srp_host *host,struct srp_target_port *target) {struct srp_rport_identifiers ids;
@@ -3299,6 +3306,8 @@ out:mutex_unlock (&host->add_target_mutex);scsi_host_put (target->scsi_host); +if (ret <0 ) +scsi_host_put (target->scsi_host);return ret;
Greg Kroah-Hartman
maintainerPATCH 98/159 IB/qib: Change lkey table allocation to support more MRs
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Mike Marciniszyn <[email protected]>
commit d6f1c17e162b2a11e708f28fa93f2f79c164b442 upstream.
The lkey table is allocated with with a get_user_pages() with an order based on a number of index bits from a module parameter.
The underlying kernel code cannot allocate that many contiguous pages.
There is no reason the underlying memory needs to be physically contiguous.
This patch:
- switches the allocation/deallocation to vmalloc/vfree
- caps the number of bits to 23 to insure at least 1 generation bit
o this matches the module parameter description
--- drivers/infiniband/hw/qib/qib_keys.c | 4 ++++ drivers/infiniband/hw/qib/qib_verbs.c | 14 ++++++++++---- drivers/infiniband/hw/qib/qib_verbs.h | 2 ++ 3 files changed, 16 insertions(+), 4 deletions(-)
drivers/infiniband/hw/qib/qib_keys.c +4 -0
@@ -86,6 +86,10 @@ int qib_alloc_lkey(struct qib_mregion *m* unrestricted LKEY. */ rkt->gen++; +/* +* bits are capped in qib_verbs.c to insure enough bits +* for generation number +*/ mr->lkey = (r << (32 - ib_qib_lkey_table_size)) | ((((1 << (24 - ib_qib_lkey_table_size)) -1 ) & rkt->gen) <<8 );
drivers/infiniband/hw/qib/qib_verbs.c +10 -4
@@ -40,6 +40,7 @@#include <linux/rculist.h> #include <linux/mm.h> #include <linux/random.h> +#include <linux/vmalloc.h> #include "qib.h" #include "qib_common.h"
@@ -2089,10 +2090,16 @@ int qib_register_ib_device(struct qib_de* the LKEY). The remaining bits act as a generation number or tag. */ spin_lock_init (&dev->lk_table.lock); +/* insure generation is at least 4 bits see keys.c */ +if (ib_qib_lkey_table_size >MAX_LKEY_TABLE_BITS ) { +qib_dev_warn (dd,"lkey bits %u too large, reduced to %u\n" , + ib_qib_lkey_table_size,MAX_LKEY_TABLE_BITS ); + ib_qib_lkey_table_size =MAX_LKEY_TABLE_BITS ; + } dev->lk_table.max =1 << ib_qib_lkey_table_size; lk_tab_size = dev->lk_table.max *sizeof (*dev->lk_table.table); dev->lk_table.table = (struct qib_mregion __rcu **) -__get_free_pages (GFP_KERNEL ,get_order (lk_tab_size)); +vmalloc (lk_tab_size);if (dev->lk_table.table ==NULL ) { ret = -ENOMEM ; goto err_lk;
@@ -2265,7 +2272,7 @@ err_tx:sizeof (struct qib_pio_header ), dev->pio_hdrs, dev->pio_hdrs_phys); err_hdrs: -free_pages ((unsigned long ) dev->lk_table.table,get_order (lk_tab_size)); +vfree (dev->lk_table.table); err_lk:kfree (dev->qp_table); err_qpt:
@@ -2319,8 +2326,7 @@ void qib_unregister_ib_device(struct qibsizeof (struct qib_pio_header ), dev->pio_hdrs, dev->pio_hdrs_phys); lk_tab_size = dev->lk_table.max *sizeof (*dev->lk_table.table); -free_pages ((unsigned long ) dev->lk_table.table, -get_order (lk_tab_size)); +vfree (dev->lk_table.table);kfree (dev->qp_table); }
drivers/infiniband/hw/qib/qib_verbs.h +2 -0
@@ -647,6 +647,8 @@ struct qib_qpn_table {struct qpn_map map[QPNMAP_ENTRIES ]; }; +#define MAX_LKEY_TABLE_BITS 23 +struct qib_lkey_table {spinlock_t lock;/* protect changes in this struct */ u32 next;/* next unused index (speeds search) */
Greg Kroah-Hartman
maintainerPATCH 99/159 IB/uverbs: reject invalid or unknown opcodes
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Christoph Hellwig <[email protected]>
commit b632ffa7cee439ba5dce3b3bc4a5cbe2b3e20133 upstream.
We have many WR opcodes that are only supported in kernel space and/or require optional information to be copied into the WR structure. Reject all those not explicitly handled so that we can't pass invalid information to drivers.
--- drivers/infiniband/core/uverbs_cmd.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-)
drivers/infiniband/core/uverbs_cmd.c +9 -1
@@ -2244,6 +2244,12 @@ ssize_t ib_uverbs_post_send(struct ib_uv next->send_flags = user_wr->send_flags;if (is_ud) { +if (next->opcode !=IB_WR_SEND && + next->opcode !=IB_WR_SEND_WITH_IMM ) { + ret = -EINVAL ; + goto out_put; + } + next->wr.ud.ah =idr_read_ah (user_wr->wr.ud.ah, file->ucontext);if (!next->wr.ud.ah) {
@@ -2283,9 +2289,11 @@ ssize_t ib_uverbs_post_send(struct ib_uv user_wr->wr.atomic.compare_add; next->wr.atomic.swap = user_wr->wr.atomic.swap; next->wr.atomic.rkey = user_wr->wr.atomic.rkey; +case IB_WR_SEND :break ;default : -break ; + ret = -EINVAL ; + goto out_put; } }
Greg Kroah-Hartman
maintainerPATCH 100/159 IB/uverbs: Fix race between ib_uverbs_open and remove_one
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Yishai Hadas <[email protected]>
commit 35d4a0b63dc0c6d1177d4f532a9deae958f0662c upstream.
Before this commit there was a device look-up table that was protected by a spin_lock used by ib_uverbs_open and by ib_uverbs_remove_one. When it was dropped and container_of was used instead, it enabled the race with remove_one as dev might be freed just after:
dev = container_of(inode->i_cdev, struct ib_uverbs_device, cdev) but before the kref_get.
In addition, this buggy patch added some dead code as container_of(x,y,z) can never be NULL and so dev can never be NULL. As a result the comment above ib_uverbs_open saying "the open method will either immediately run -ENXIO" is wrong as it can never happen.
The solution follows Jason Gunthorpe suggestion from below URL:
https://www.mail-archive.com/[email protected]/msg25692.html
cdev will hold a kref on the parent (the containing structure, ib_uverbs_device) and only when that kref is released it is guaranteed that open will never be called again.
In addition, fixes the active count scheme to use an atomic not a kref to prevent WARN_ON as pointed by above comment from Jason.
--- drivers/infiniband/core/uverbs.h | 3 +- drivers/infiniband/core/uverbs_main.c | 43 +++++++++++++++++++++++----------- 2 files changed, 32 insertions(+), 14 deletions(-)
drivers/infiniband/core/uverbs.h +2 -1
@@ -85,7 +85,7 @@*/ struct ib_uverbs_device { -struct kref ref; +atomic_t refcount;int num_comp_vectors;struct completion comp;struct device *dev;
@@ -94,6 +94,7 @@ struct ib_uverbs_device {struct cdev cdev;struct rb_root xrcd_tree;struct mutex xrcd_tree_mutex; +struct kobject kobj; };struct ib_uverbs_event_file {
drivers/infiniband/core/uverbs_main.c +30 -13
125 lines not on this page · raw message
Greg Kroah-Hartman
maintainerPATCH 101/159 IB/iser: Fix missing return status check in iser_send_data_out
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit d16739055bd1f562ae4d83e69f7f7f1cefcfbe16 upstream.
Since commit "IB/iser: Fix race between iser connection teardown..." iser_initialize_task_headers() might fail, so we need to check that.
--- drivers/infiniband/ulp/iser/iser_initiator.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)
drivers/infiniband/ulp/iser/iser_initiator.c +5 -3
@@ -454,7 +454,7 @@ int iser_send_data_out(struct iscsi_connunsigned long buf_offset;unsigned long data_seg_len;uint32_t itt; -int err =0 ; +int err;struct ib_sge *tx_dsg; itt = (__force uint32_t )hdr->itt;
@@ -475,7 +475,9 @@ int iser_send_data_out(struct iscsi_connmemcpy (&tx_desc->iscsi_header, hdr,sizeof (struct iscsi_hdr ));/* build the tx desc */ -iser_initialize_task_headers (task, tx_desc); + err =iser_initialize_task_headers (task, tx_desc); +if (err) + goto send_data_out_error; mem_reg = &iser_task->rdma_reg[ISER_DIR_OUT ]; tx_dsg = &tx_desc->tx_sg[1 ];
@@ -502,7 +504,7 @@ int iser_send_data_out(struct iscsi_conn send_data_out_error:kmem_cache_free (ig.desc_cache, tx_desc); -iser_err ("conn %p failed err %d\n" ,conn, err); +iser_err ("conn %p failed err %d\n" , conn, err);return err; }
Greg Kroah-Hartman
maintainerPATCH 102/159 IB/iser: Fix possible bogus DMA unmapping
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit 8d5944d80359e645feb2ebd069a6f4caf7825e40 upstream.
If iser_initialize_task_headers() routine failed before dma mapping, we should not attempt to unmap in cleanup_task().
--- drivers/infiniband/ulp/iser/iscsi_iser.c | 12 ++++++++---- drivers/infiniband/ulp/iser/iscsi_iser.h | 2 ++ 2 files changed, 10 insertions(+), 4 deletions(-)
drivers/infiniband/ulp/iser/iscsi_iser.c +8 -4
@@ -201,6 +201,7 @@ iser_initialize_task_headers(struct iscs goto out; } + tx_desc->mapped = true; tx_desc->dma_addr = dma_addr; tx_desc->tx_sg[0 ].addr = tx_desc->dma_addr; tx_desc->tx_sg[0 ].length =ISER_HEADERS_LEN ;
@@ -360,16 +361,19 @@ iscsi_iser_task_xmit(struct iscsi_task *static void iscsi_iser_cleanup_task (struct iscsi_task *task) {struct iscsi_iser_task *iser_task = task->dd_data; -struct iser_tx_desc *tx_desc = &iser_task->desc; -struct iser_conn *iser_conn = task->conn->dd_data; +struct iser_tx_desc *tx_desc = &iser_task->desc; +struct iser_conn *iser_conn = task->conn->dd_data;struct iser_device *device = iser_conn->ib_conn.device;/* DEVICE_REMOVAL event might have already released the device */ if (!device)return ; -ib_dma_unmap_single (device->ib_device, - tx_desc->dma_addr,ISER_HEADERS_LEN ,DMA_TO_DEVICE ); +if (likely (tx_desc->mapped)) { +ib_dma_unmap_single (device->ib_device, tx_desc->dma_addr, +ISER_HEADERS_LEN ,DMA_TO_DEVICE ); + tx_desc->mapped = false; + }/* mgmt tasks do not need special cleanup */ if (!task->sc)
drivers/infiniband/ulp/iser/iscsi_iser.h +2 -0
@@ -270,6 +270,7 @@ enum iser_desc_type {* sg[1] optionally points to either of immediate data * unsolicited data-out or control * @num_sge: number sges used on this TX task +* @mapped: Is the task header mapped */ struct iser_tx_desc {struct iser_hdr iser_header;
@@ -278,6 +279,7 @@ struct iser_tx_desc {u64 dma_addr;struct ib_sge tx_sg[2 ];int num_sge; +bool mapped; };#define ISER_RX_PAD_SIZE (256 - (ISER_RX_PAYLOAD_SIZE + \
Greg Kroah-Hartman
maintainerPATCH 103/159 IB/mlx5: avoid destroying a NULL mr in reg_user_mr error flow
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Haggai Eran <[email protected]>
commit 11d748045c6dadb279d1acdb6d2ea8f3f2ede85b upstream.
The mlx5_ib_reg_user_mr() function will attempt to call clean_mr() in its error flow even though there is never a case where the error flow occurs with a valid MR pointer to destroy.
Remove the clean_mr() call and the incorrect comment above it.
support for MMU notifiers")
--- drivers/infiniband/hw/mlx5/mr.c | 12 ------------ 1 file changed, 12 deletions(-)
drivers/infiniband/hw/mlx5/mr.c +0 -12
@@ -1119,19 +1119,7 @@ struct ib_mr *mlx5_ib_reg_user_mr(structreturn &mr->ibmr; error: -/* -* Destroy the umem *before* destroying the MR, to ensure we -* will not have any in-flight notifiers when destroying the -* MR. -* -* As the MR is completely invalid to begin with, and this -* error path is only taken if we can't push the mr entry into -* the pagefault tree, this is safe. -*/ -ib_umem_release (umem); -/* Kill the MR, and return an error code. */ -clean_mr (mr);return ERR_PTR (err); }
Greg Kroah-Hartman
maintainerPATCH 104/159 IB/mlx4: Fix potential deadlock when sending mad to wire
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jack Morgenstein <[email protected]>
commit 90c1d8b6350cca9d8a234f03c77a317a7613bcee upstream.
send_mad_to_wire takes the same spinlock that is taken in the interrupt context. Therefore, it needs irqsave/restore.
--- drivers/infiniband/hw/mlx4/mcg.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-)
drivers/infiniband/hw/mlx4/mcg.c +4 -3
@@ -206,15 +206,16 @@ static int send_mad_to_wire(struct mlx4_ {struct mlx4_ib_dev *dev = ctx->dev;struct ib_ah_attr ah_attr; +unsigned long flags; -spin_lock (&dev->sm_lock); +spin_lock_irqsave (&dev->sm_lock, flags);if (!dev->sm_ah[ctx->port -1 ]) {/* port is not yet Active, sm_ah not ready */ -spin_unlock (&dev->sm_lock); +spin_unlock_irqrestore (&dev->sm_lock, flags);return -EAGAIN ; }mlx4_ib_query_ah (dev->sm_ah[ctx->port -1 ], &ah_attr); -spin_unlock (&dev->sm_lock); +spin_unlock_irqrestore (&dev->sm_lock, flags);return mlx4_ib_send_to_wire (dev,mlx4_master_func_num (dev->dev), ctx->port,IB_QPT_GSI ,0 ,1 ,IB_QP1_QKEY , &ah_attr,NULL , mad);
Greg Kroah-Hartman
maintainerPATCH 105/159 IB/mlx4: Forbid using sysfs to change RoCE pkeys
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jack Morgenstein <[email protected]>
commit 2b135db3e81301d0452e6aa107349abe67b097d6 upstream.
The pkey mapping for RoCE must remain the default mapping:
VFs:
virtual index 0 = mapped to real index 0 (0xFFFF) All others indices: mapped to a real pkey index containing an
invalid pkey.
PF:
virtual index i = real index i.
Don't allow users to change these mappings using files found in sysfs.
--- drivers/infiniband/hw/mlx4/sysfs.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/infiniband/hw/mlx4/sysfs.c +4 -1
@@ -640,6 +640,8 @@ static int add_port(struct mlx4_ib_dev *struct mlx4_port *p;int i;int ret; +int is_eth =rdma_port_get_link_layer (&dev->ib_dev, port_num) == +IB_LINK_LAYER_ETHERNET ; p =kzalloc (sizeof *p,GFP_KERNEL );if (!p)
@@ -657,7 +659,8 @@ static int add_port(struct mlx4_ib_dev * p->pkey_group.name ="pkey_idx" ; p->pkey_group.attrs = -alloc_group_attrs (show_port_pkey, store_port_pkey, +alloc_group_attrs (show_port_pkey, + is_eth ?NULL : store_port_pkey, dev->dev->caps.pkey_table_len[port_num]);if (!p->pkey_group.attrs) { ret = -ENOMEM ;
Greg Kroah-Hartman
maintainerPATCH 106/159 IB/mlx4: Use correct SL on AH query under RoCE
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Noa Osherovich <[email protected]>
commit 5e99b139f1b68acd65e36515ca347b03856dfb5a upstream.
The mlx4 IB driver implementation for ib_query_ah used a wrong offset (28 instead of 29) when link type is Ethernet. Fixed to use the correct one.
--- drivers/infiniband/hw/mlx4/ah.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/infiniband/hw/mlx4/ah.c +5 -1
@@ -148,9 +148,13 @@ int mlx4_ib_query_ah(struct ib_ah *ibah,enum rdma_link_layer ll;memset (ah_attr,0 ,sizeof *ah_attr); - ah_attr->sl =be32_to_cpu (ah->av.ib.sl_tclass_flowlabel) >>28 ; ah_attr->port_num =be32_to_cpu (ah->av.ib.port_pd) >>24 ; ll =rdma_port_get_link_layer (ibah->device, ah_attr->port_num); +if (ll ==IB_LINK_LAYER_ETHERNET ) + ah_attr->sl =be32_to_cpu (ah->av.eth.sl_tclass_flowlabel) >>29 ; +else + ah_attr->sl =be32_to_cpu (ah->av.ib.sl_tclass_flowlabel) >>28 ; + ah_attr->dlid = ll ==IB_LINK_LAYER_INFINIBAND ?be16_to_cpu (ah->av.ib.dlid) :0 ;if (ah->av.ib.stat_rate) ah_attr->static_rate = ah->av.ib.stat_rate -MLX4_STAT_RATE_OFFSET ;
Greg Kroah-Hartman
maintainerPATCH 107/159 IB/mlx4: Fix incorrect cq flushing in error state
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Ariel Nahum <[email protected]>
commit 799cdaf8a98f13d4fba3162e21e1e63f21045010 upstream.
When handling a device internal error, the driver is responsible to drain the completion queue with flush errors.
In case a completion queue was assigned to multiple send queues, the driver iterates over the send queues and generates flush errors of inflight wqes. The driver must correctly pass the wc array with an offset as a result of the previous send queue iteration. Not doing so will overwrite previously set completions and return a wrong number of polled completions which includes ones which were not correctly set.
--- drivers/infiniband/hw/mlx4/cq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/infiniband/hw/mlx4/cq.c +1 -1
@@ -629,7 +629,7 @@ static void mlx4_ib_poll_sw_comp(struct* simulated FLUSH_ERR completions */ list_for_each_entry (qp, &cq->send_qp_list, cq_send_list) { -mlx4_ib_qp_sw_comp (qp, num_entries, wc, npolled,1 ); +mlx4_ib_qp_sw_comp (qp, num_entries, wc + *npolled, npolled,1 );if (*npolled >= num_entries) goto out; }
Greg Kroah-Hartman
maintainerPATCH 108/159 stmmac: fix check for phydev being open
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Alexey Brodkin <[email protected]>
commit dfc50fcaad574e5c8c85cbc83eca1426b2413fa4 upstream.
Current check of phydev with IS_ERR(phydev) may make not much sense because of_phy_connect() returns NULL on failure instead of error value.
Still for checking result of phy_connect() IS_ERR() makes perfect sense.
So let's use combined check IS_ERR_OR_NULL() that covers both cases.
--- drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +4 -1
@@ -829,8 +829,11 @@ static int stmmac_init_phy(struct net_de phydev =phy_connect (dev, phy_id_fmt, &stmmac_adjust_link, interface); -if (IS_ERR (phydev)) { +if (IS_ERR_OR_NULL (phydev)) {pr_err ("%s: Could not attach to PHY\n" , dev->name); +if (!phydev) +return -ENODEV ; +return PTR_ERR (phydev); }
Re: [PATCH 4.1 108/159] stmmac: fix check for phydev being open
Hello.
Quoted 1 line · Greg Kroah-Hartman
4.1-stable review patch. If anyone has any objections, please let me know.
NAK on the same grounds as the 3.x patches. of_phy_connect() isn't called
yet. If the patch has been committed already, it's OK -- it should do no harm.
Quoted 30 lines
------------------
From: Alexey Brodkin <[email protected]>
commit
dfc50fcaad574e5c8c85cbc83eca1426b2413fa4upstream.Current check of phydev with
IS_ERR(phydev)may make not much sense becauseof_phy_connect()returns NULL on failure instead of error value.Still for checking result of
phy_connect()IS_ERR()makes perfect sense.So let's use combined check
IS_ERR_OR_NULL()that covers both cases.Cc Sergei Shtylyov <[email protected]>Cc Giuseppe Cavallaro <[email protected]>Cc David Miller <[email protected]>Signed-off-by Alexey Brodkin <[email protected]>Signed-off-by David S. Miller <[email protected]>Signed-off-by Greg Kroah-Hartman <[email protected]>--- drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +0
-0@@ -829,8 +829,11 @@ static int stmmac_init_phy(struct net_de phydev =phy_connect (dev, phy_id_fmt, &stmmac_adjust_link, interface);
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c static int stmmac_init_phy(struct net_de · line 839
Quoted 10 lines
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +4
-1-if (IS_ERR (phydev)) { +if (IS_ERR_OR_NULL (phydev)) {pr_err ("%s: Could not attach to PHY\n" , dev->name); +if (!phydev) +return -ENODEV ; +return PTR_ERR (phydev); }
MBR, Sergei
Greg Kroah-Hartman
maintainerRe: [PATCH 4.1 108/159] stmmac: fix check for phydev being open
Quoted 9 lines · Sergei Shtylyov
Hello.
Quoted 1 line · Greg Kroah-Hartman
4.1-stable review patch. If anyone has any objections, please let me know.
NAK on the same grounds as the 3.x patches.
of_phy_connect()isn't called
yet. If the patch has been committed already, it's OK -- it should do no harm.
Yeah, it's already merged, sorry.
greg k-h
Re: [PATCH 4.1 108/159] stmmac: fix check for phydev being open
Quoted 7 lines · Greg Kroah-Hartman
Quoted 5 lines
Quoted 1 line
4.1-stable review patch. If anyone has any objections, please let me know.
NAK on the same grounds as the 3.x patches.
of_phy_connect()isn't called
yet. If the patch has been committed already, it's OK -- it should do no harm.Yeah, it's already merged, sorry.
Sorry for not letting you know in due time.
Quoted 1 line
greg k-h
MBR, Sergei
Greg Kroah-Hartman
maintainerPATCH 109/159 stmmac: troubleshoot unexpected bits in des0 & des1
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Alexey Brodkin <[email protected]>
commit f1590670ce069eefeb93916391a67643e6ad1630 upstream.
Current implementation of descriptor init procedure only takes care about setting/clearing ownership flag in "des0"/"des1" fields while it is perfectly possible to get unexpected bits set because of the following factors:
[1] On driver probe underlying memory allocated with dma_alloc_coherent() might not be zeroed and so it will be filled with garbage.
[2] During driver operation some bits could be set by SD/MMC controller (for example error flags etc).
And unexpected and/or randomly set flags in "des0"/"des1" fields may lead to unpredictable behavior of GMAC DMA block.
This change addresses both items above with:
[1] Use of dma_zalloc_coherent() instead of simple dma_alloc_coherent() to make sure allocated memory is zeroed. That shouldn't affect performance because this allocation only happens once on driver probe.
[2] Do explicit zeroing of both "des0" and "des1" fields of all buffer descriptors during initialization of DMA transfer.
And while at it fixed identation of dma_free_coherent() counterpart as well.
--- drivers/net/ethernet/stmicro/stmmac/descs.h | 2 + drivers/net/ethernet/stmicro/stmmac/enh_desc.c | 3 + drivers/net/ethernet/stmicro/stmmac/norm_desc.c | 3 + drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 44 +++++++++++----------- 4 files changed, 28 insertions(+), 24 deletions(-)
drivers/net/ethernet/stmicro/stmmac/descs.h +2 -0
@@ -158,6 +158,8 @@ struct dma_desc { u32 buffer2_size:13; u32 reserved4:3; } etx;/* -- enhanced -- */ + +u64 all_flags; } des01;unsigned int des2;unsigned int des3;
drivers/net/ethernet/stmicro/stmmac/enh_desc.c +2 -1
@@ -240,6 +240,7 @@ static int enh_desc_get_rx_status(void *static void enh_desc_init_rx_desc (struct dma_desc *p,int disable_rx_ic,int mode,int end) { + p->des01.all_flags =0 ; p->des01.erx.own =1 ; p->des01.erx.buffer1_size = BUF_SIZE_8KiB -1 ;
@@ -254,7 +255,7 @@ static void enh_desc_init_rx_desc(structstatic void enh_desc_init_tx_desc (struct dma_desc *p,int mode,int end) { - p->des01.etx.own =0 ; + p->des01.all_flags =0 ;if (mode ==STMMAC_CHAIN_MODE )ehn_desc_tx_set_on_chain (p, end);else
drivers/net/ethernet/stmicro/stmmac/norm_desc.c +2 -1
@@ -123,6 +123,7 @@ static int ndesc_get_rx_status(void *datstatic void ndesc_init_rx_desc (struct dma_desc *p,int disable_rx_ic,int mode,int end) { + p->des01.all_flags =0 ; p->des01.rx.own =1 ; p->des01.rx.buffer1_size = BUF_SIZE_2KiB -1 ;
@@ -137,7 +138,7 @@ static void ndesc_init_rx_desc(struct dmstatic void ndesc_init_tx_desc (struct dma_desc *p,int mode,int end) { - p->des01.tx.own =0 ; + p->des01.all_flags =0 ;if (mode ==STMMAC_CHAIN_MODE )ndesc_tx_set_on_chain (p, end);else
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +22 -22
63 lines not on this page · raw message
Greg Kroah-Hartman
maintainerPATCH 110/159 net: stmmac: dwmac-rk: Fix clk rate when provided by soc
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: =?UTF-8?q?Heiko=20St=C3=BCbner?= <[email protected]>
commit c48fa33c1fb2ccdb4bcc863a7b841f11efe0f8b0 upstream.
The first iteration of the dwmac-rk support did access an intermediate clock directly below the pll selector. This was removed in a subsequent revision, but the clock and one invocation remained. This results in the driver trying to set the rate of a non-existent clock when the soc and not some external source provides the phy clock for RMII phys.
So set the rate of the correct clock and remove the remaining now completely unused definition.
--- drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c +1 -2
@@ -38,7 +38,6 @@ struct rk_priv_data {bool clock_input;struct clk *clk_mac; -struct clk *clk_mac_pll;struct clk *gmac_clkin;struct clk *mac_clk_rx;struct clk *mac_clk_tx;
@@ -208,7 +207,7 @@ static int gmac_clk_init(struct rk_priv_dev_info (dev,"%s: clock input from PHY\n" , __func__); }else {if (bsp_priv->phy_iface ==PHY_INTERFACE_MODE_RMII ) -clk_set_rate (bsp_priv->clk_mac_pll,50000000 ); +clk_set_rate (bsp_priv->clk_mac,50000000 ); }return 0 ;
Greg Kroah-Hartman
maintainerPATCH 111/159 hfs,hfsplus: cache pages correctly between bnode_create and bnode_free
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Hin-Tak Leung <[email protected]>
commit 7cb74be6fd827e314f81df3c5889b87e4c87c569 upstream.
Pages looked up by __hfs_bnode_create() (called by hfs_bnode_create() and hfs_bnode_find() for finding or creating pages corresponding to an inode) are immediately kmap()'ed and used (both read and write) and kunmap()'ed, and should not be page_cache_release()'ed until hfs_bnode_free().
This patch fixes a problem I first saw in July 2012: merely running "du" on a large hfsplus-mounted directory a few times on a reasonably loaded system would get the hfsplus driver all confused and complaining about B-tree inconsistencies, and generates a "BUG: Bad page state". Most recently, I can generate this problem on up-to-date Fedora 22 with shipped kernel 4.0.5, by running "du /" (="/" + "/home" + "/mnt" + other smaller mounts) and "du /mnt" simultaneously on two windows, where /mnt is a lightly-used QEMU VM image of the full Mac OS X 10.9:
$ df -i / /home /mnt Filesystem Inodes IUsed IFree IUse% Mounted on /dev/mapper/fedora-root 3276800 551665 2725135 17% / /dev/mapper/fedora-home 52879360 716221 52163139 2% /home /dev/nbd0p2 4294967295 1387818 4293579477 1% /mnt
After applying the patch, I was able to run "du /" (60+ times) and "du /mnt" (150+ times) continuously and simultaneously for 6+ hours.
There are many reports of the hfsplus driver getting confused under load and generating "BUG: Bad page state" or other similar issues over the years. [1]
The unpatched code [2] has always been wrong since it entered the kernel tree. The only reason why it gets away with it is that the kmap/memcpy/kunmap follow very quickly after the page_cache_release() so the kernel has not had a chance to reuse the memory for something else, most of the time.
The current RW driver appears to have followed the design and development of the earlier read-only hfsplus driver [3], where-by version 0.1 (Dec 2001) had a B-tree node-centric approach to read_cache_page()/page_cache_release() per bnode_get()/bnode_put(), migrating towards version 0.2 (June 2002) of caching and releasing pages per inode extents. When the current RW code first entered the kernel [2] in 2005, there was an REF_PAGES conditional (and "//" commented out code) to switch between B-node centric paging to inode-centric paging. There was a mistake with the direction of one of the REF_PAGES conditionals in __hfs_bnode_create(). In a subsequent "remove debug code" commit [4], the read_cache_page()/page_cache_release() per bnode_get()/bnode_put() were removed, but a page_cache_release() was mistakenly left in (propagating the "REF_PAGES <-> !REF_PAGE" mistake), and the commented-out page_cache_release() in bnode_release() (which should be spanned by !REF_PAGES) was never enabled.
References:
[1]:
Michael Fox, Apr 2013
http://www.spinics.net/lists/linux-fsdevel/msg63807.html
("hfsplus volume suddenly inaccessable after 'hfs: recoff %d too large'")
Sasha Levin, Feb 2015
http://lkml.org/lkml/2015/2/20/85 ("use after free")
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/740814
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1027887
https://bugzilla.kernel.org/show_bug.cgi?id=42342
https://bugzilla.kernel.org/show_bug.cgi?id=63841
https://bugzilla.kernel.org/show_bug.cgi?id=78761
[2]:
http://git.kernel.org/cgit/linux/kernel/git/tglx/history.git/commit/\
fs/hfs/bnode.c?id=d1081202f1d0ee35ab0beb490da4b65d4bc763db commit d1081202f1d0ee35ab0beb490da4b65d4bc763db
Author: Andrew Morton <[email protected]> Date: Wed Feb 25 16:17:36 2004 -0800
[PATCH] HFS rewrite
http://git.kernel.org/cgit/linux/kernel/git/tglx/history.git/commit/\
fs/hfsplus/bnode.c?id=91556682e0bf004d98a529bf829d339abb98bbbd
commit 91556682e0bf004d98a529bf829d339abb98bbbd
Author: Andrew Morton <[email protected]> Date: Wed Feb 25 16:17:48 2004 -0800
[PATCH] HFS+ support
[3]:
http://sourceforge.net/projects/linux-hfsplus/
http://sourceforge.net/projects/linux-hfsplus/files/Linux%202.4.x%20patch/hfsplus%200.1/
http://sourceforge.net/projects/linux-hfsplus/files/Linux%202.4.x%20patch/hfsplus%200.2/
http://linux-hfsplus.cvs.sourceforge.net/viewvc/linux-hfsplus/linux/\
fs/hfsplus/bnode.c?r1=1.4&r2=1.5
Date: Thu Jun 6 09:45:14 2002 +0000
Use buffer cache instead of page cache in bnode.c. Cache inode extents.
[4]:
http://git.kernel.org/cgit/linux/kernel/git/\
stable/linux-stable.git/commit/?id=a5e3985fa014029eb6795664c704953720cc7f7d
commit a5e3985fa014029eb6795664c704953720cc7f7d
Author: Roman Zippel <[email protected]> Date: Tue Sep 6 15:18:47 2005 -0700
[PATCH] hfs: remove debug code
--- fs/hfs/bnode.c | 9 ++++----- fs/hfsplus/bnode.c | 3 --- 2 files changed, 4 insertions(+), 8 deletions(-)
fs/hfs/bnode.c +4 -5
@@ -288,7 +288,6 @@ static struct hfs_bnode *__hfs_bnode_crepage_cache_release (page); goto fail; } -page_cache_release (page); node->page[i] = page; }
@@ -398,11 +397,11 @@ node_error:void hfs_bnode_free (struct hfs_bnode *node) { -//int i; +int i; -//for (i = 0; i < node->tree->pages_per_bnode; i++) -// if (node->page[i]) -// page_cache_release(node->page[i]); +for (i =0 ; i < node->tree->pages_per_bnode; i++) +if (node->page[i]) +page_cache_release (node->page[i]);kfree (node); }
fs/hfsplus/bnode.c +0 -3
@@ -454,7 +454,6 @@ static struct hfs_bnode *__hfs_bnode_crepage_cache_release (page); goto fail; } -page_cache_release (page); node->page[i] = page; }
@@ -566,13 +565,11 @@ node_error:void hfs_bnode_free (struct hfs_bnode *node) { -#if 0 int i;for (i =0 ; i < node->tree->pages_per_bnode; i++)if (node->page[i])page_cache_release (node->page[i]); -#endif kfree (node); }
Greg Kroah-Hartman
maintainerPATCH 112/159 lib/decompressors: use real out buf size for gunzip with kernel
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Yinghai Lu <[email protected]>
commit 2d3862d26e67a59340ba1cf1748196c76c5787de upstream.
When loading x86 64bit kernel above 4GiB with patched grub2, got kernel gunzip error.
| early console in decompress_kernel | decompress_kernel: | input: [0x807f2143b4-0x807ff61aee] | output: [0x807cc00000-0x807f3ea29b] 0x027ea29c: output_len | boot via startup_64 | KASLR using RDTSC... | new output: [0x46fe000000-0x470138cfff] 0x0338d000: output_run_size | decompress: [0x46fe000000-0x47007ea29b] <=== [0x807f2143b4-0x807ff61aee] | | Decompressing Linux... gz... | | uncompression error | | -- System halted
the new buffer is at 0x46fe000000ULL, decompressor_gzip is using 0xffffffb901ffffff as out_len. gunzip in lib/zlib_inflate/inflate.c cap that len to 0x01ffffff and decompress fails later.
We could hit this problem with crashkernel booting that uses kexec loading kernel above 4GiB.
We have decompress_* support:
1. inbuf[]/outbuf[] for kernel preboot.
2. inbuf[]/flush() for initramfs
3. fill()/flush() for initrd.
This bug only affect kernel preboot path that use outbuf[].
Add __decompress and take real out_buf_len for gunzip instead of guessing wrong buf size.
--- arch/arm/boot/compressed/decompress.c | 2 +- arch/m32r/boot/compressed/misc.c | 3 ++- arch/mips/boot/compressed/decompress.c | 4 ++-- arch/s390/boot/compressed/misc.c | 2 +- arch/sh/boot/compressed/misc.c | 2 +- arch/unicore32/boot/compressed/misc.c | 4 ++-- arch/x86/boot/compressed/misc.c | 3 ++- lib/decompress_bunzip2.c | 6 +++--- lib/decompress_inflate.c | 31 ++++++++++++++++++++++++++----- lib/decompress_unlz4.c | 6 +++--- lib/decompress_unlzma.c | 7 +++---- lib/decompress_unlzo.c | 13 ++++++++++++- lib/decompress_unxz.c | 12 +++++++++++- 13 files changed, 69 insertions(+), 26 deletions(-)
arch/arm/boot/compressed/decompress.c +1 -1
@@ -57,5 +57,5 @@ extern char * strstr(const char * s1, coint do_decompress (u8 *input,int len,u8 *output,void (*error)(char *x)) { -return decompress (input, len,NULL ,NULL , output,NULL , error); +return __decompress (input, len,NULL ,NULL , output,0 ,NULL , error); }
arch/m32r/boot/compressed/misc.c +2 -1
@@ -86,6 +86,7 @@ decompress_kernel(int mmu_on, unsigned c free_mem_end_ptr = free_mem_ptr +BOOT_HEAP_SIZE ;puts ("\nDecompressing Linux... " ); -decompress (input_data, input_len,NULL ,NULL , output_data,NULL , error); +__decompress (input_data, input_len,NULL ,NULL , output_data,0 , +NULL , error);puts ("done.\nBooting the kernel.\n" ); }
arch/mips/boot/compressed/decompress.c +2 -2
@@ -111,8 +111,8 @@ void decompress_kernel(unsigned long booputs ("\n" );/* Decompress the kernel with according algorithm */ -decompress ((char *)zimage_start, zimage_size,0 ,0 , - (void *)VMLINUX_LOAD_ADDRESS_ULL ,0 , error); +__decompress ((char *)zimage_start, zimage_size,0 ,0 , + (void *)VMLINUX_LOAD_ADDRESS_ULL ,0 ,0 , error);/* FIXME: should we flush cache here? */ puts ("Now, booting the kernel...\n" );
arch/s390/boot/compressed/misc.c +1 -1
@@ -167,7 +167,7 @@ unsigned long decompress_kernel(void)#endif puts ("Uncompressing Linux... " ); -decompress (input_data, input_len,NULL ,NULL , output,NULL , error); +__decompress (input_data, input_len,NULL ,NULL , output,0 ,NULL , error);puts ("Ok, booting the kernel.\n" );return (unsigned long ) output; }
arch/sh/boot/compressed/misc.c +1 -1
@@ -132,7 +132,7 @@ void decompress_kernel(void)puts ("Uncompressing Linux... " );cache_control (CACHE_ENABLE ); -decompress (input_data, input_len,NULL ,NULL , output,NULL , error); +__decompress (input_data, input_len,NULL ,NULL , output,0 ,NULL , error);cache_control (CACHE_DISABLE );puts ("Ok, booting the kernel.\n" ); }
arch/unicore32/boot/compressed/misc.c +2 -2
@@ -119,8 +119,8 @@ unsigned long decompress_kernel(unsigned output_ptr =get_unaligned_le32 (tmp);arch_decomp_puts ("Uncompressing Linux..." ); -decompress (input_data, input_data_end - input_data,NULL ,NULL , - output_data,NULL , error); +__decompress (input_data, input_data_end - input_data,NULL ,NULL , + output_data,0 ,NULL , error);arch_decomp_puts (" done, booting the kernel.\n" );return output_ptr; }
arch/x86/boot/compressed/misc.c +2 -1
@@ -424,7 +424,8 @@ asmlinkage __visible void *decompress_ke#endif debug_putstr ("\nDecompressing Linux... " ); -decompress (input_data, input_len,NULL ,NULL , output,NULL , error); +__decompress (input_data, input_len,NULL ,NULL , output, output_len, +NULL , error);parse_elf (output);/* * 32-bit always performs relocations. 64-bit relocations are only
lib/decompress_bunzip2.c +3 -3
@@ -743,12 +743,12 @@ exit_0: }#ifdef PREBOOT -STATIC int INIT decompress(unsigned char *buf, long len, +STATIC int INIT __decompress(unsigned char *buf, long len,long (*fill)(void *,unsigned long ),long (*flush)(void *,unsigned long ), -unsigned char *outbuf, +unsigned char *outbuf,long olen,long *pos, -void (*error)(char *x)) +void (*error)(char *x)) {return bunzip2 (buf, len -4 , fill, flush, outbuf, pos, error); }
lib/decompress_inflate.c +26 -5
@@ -1,4 +1,5 @@#ifdef STATIC +#define PREBOOT /* Pre-boot environment: included */ /* prevent inclusion of _LINUX_KERNEL_H in pre-boot environment: lots
@@ -33,23 +34,23 @@ static long INIT nofill(void *buffer, un }/* Included from initramfs et al code */ -STATIC int INIT gunzip(unsigned char *buf, long len, +STATIC int INIT __gunzip(unsigned char *buf, long len,long (*fill)(void *,unsigned long ),long (*flush)(void *,unsigned long ), -unsigned char *out_buf, +unsigned char *out_buf,long out_len,long *pos,void (*error)(char *x)) {u8 *zbuf;struct z_stream_s *strm;int rc; -size_t out_len; rc =-1 ;if (flush) { out_len =0x8000 ;/* 32 K */ out_buf =malloc (out_len); }else { - out_len = ((size_t )~0 ) - (size_t )out_buf;/* no limit */ +if (!out_len) + out_len = ((size_t )~0 ) - (size_t )out_buf;/* no limit */ }if (!out_buf) {error ("Out of memory while allocating output buffer" );
@@ -181,4 +182,24 @@ gunzip_nomem1:return rc;/* returns Z_OK (0) if successful */ } -#define decompress gunzip +#ifndef PREBOOT +STATIC int INIT gunzip(unsigned char *buf, long len, +long (*fill)(void *,unsigned long ), +long (*flush)(void *,unsigned long ), +unsigned char *out_buf, +long *pos, +void (*error)(char *x)) +{ +return __gunzip (buf, len, fill, flush, out_buf,0 , pos, error); +} +#else +STATIC int INIT __decompress(unsigned char *buf, long len, +long (*fill)(void *,unsigned long ), +long (*flush)(void *,unsigned long ), +unsigned char *out_buf,long out_len, +long *pos, +void (*error)(char *x)) +{ +return __gunzip (buf, len, fill, flush, out_buf, out_len, pos, error); +} +#endif
lib/decompress_unlz4.c +3 -3
@@ -196,12 +196,12 @@ exit_0: }#ifdef PREBOOT -STATIC int INIT decompress(unsigned char *buf, long in_len, +STATIC int INIT __decompress(unsigned char *buf, long in_len,long (*fill)(void *,unsigned long ),long (*flush)(void *,unsigned long ), -unsigned char *output, +unsigned char *output,long out_len,long *posp, -void (*error)(char *x) +void (*error)(char *x) ) {return unlz4 (buf, in_len -4 , fill, flush, output, posp, error);
lib/decompress_unlzma.c +3 -4
@@ -667,13 +667,12 @@ exit_0: }#ifdef PREBOOT -STATIC int INIT decompress(unsigned char *buf, long in_len, +STATIC int INIT __decompress(unsigned char *buf, long in_len,long (*fill)(void *,unsigned long ),long (*flush)(void *,unsigned long ), -unsigned char *output, +unsigned char *output,long out_len,long *posp, -void (*error)(char *x) - ) +void (*error)(char *x)) {return unlzma (buf, in_len -4 , fill, flush, output, posp, error); }
lib/decompress_unlzo.c +12 -1
@@ -31,6 +31,7 @@*/ #ifdef STATIC +#define PREBOOT #include "lzo/lzo1x_decompress_safe.c" #else #include <linux/decompress/unlzo.h>
@@ -287,4 +288,14 @@ exit:return ret; } -#define decompress unlzo +#ifdef PREBOOT +STATIC int INIT __decompress(unsigned char *buf, long len, +long (*fill)(void *,unsigned long ), +long (*flush)(void *,unsigned long ), +unsigned char *out_buf,long olen, +long *pos, +void (*error)(char *x)) +{ +return unlzo (buf, len, fill, flush, out_buf, pos, error); +} +#endif
lib/decompress_unxz.c +11 -1
@@ -394,4 +394,14 @@ error_alloc_state:* This macro is used by architecture-specific files to decompress * the kernel image. */ -#define decompress unxz +#ifdef XZ_PREBOOT +STATIC int INIT __decompress(unsigned char *buf, long len, +long (*fill)(void *,unsigned long ), +long (*flush)(void *,unsigned long ), +unsigned char *out_buf,long olen, +long *pos, +void (*error)(char *x)) +{ +return unxz (buf, len, fill, flush, out_buf, pos, error); +} +#endif
Greg Kroah-Hartman
maintainerPATCH 113/159 jbd2: avoid infinite loop when destroying aborted journal
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jan Kara <[email protected]>
commit 841df7df196237ea63233f0f9eaa41db53afd70f upstream.
Commit 6f6a6fda2945 "jbd2: fix ocfs2 corrupt when updating journal superblock fails" changed jbd2_cleanup_journal_tail() to return EIO when the journal is aborted. That makes logic in jbd2_log_do_checkpoint() bail out which is fine, except that jbd2_journal_destroy() expects jbd2_log_do_checkpoint() to always make a progress in cleaning the journal. Without it jbd2_journal_destroy() just loops in an infinite loop.
Fix jbd2_journal_destroy() to cleanup journal checkpoint lists of jbd2_log_do_checkpoint() fails with error.
--- fs/jbd2/checkpoint.c | 39 +++++++++++++++++++++++++++++++++------ fs/jbd2/commit.c | 2 +- fs/jbd2/journal.c | 11 ++++++++++- include/linux/jbd2.h | 3 ++- 4 files changed, 46 insertions(+), 9 deletions(-)
fs/jbd2/checkpoint.c +33 -6
@@ -417,12 +417,12 @@ int jbd2_cleanup_journal_tail(journal_t* journal_clean_one_cp_list * * Find all the written-back checkpoint buffers in the given list and -* release them. +* release them. If 'destroy' is set, clean all buffers unconditionally. * * Called with j_list_lock held. * Returns 1 if we freed the transaction, 0 otherwise. */ -static int journal_clean_one_cp_list (struct journal_head *jh) +static int journal_clean_one_cp_list (struct journal_head *jh,bool destroy) {struct journal_head *last_jh;struct journal_head *next_jh = jh;
@@ -436,7 +436,10 @@ static int journal_clean_one_cp_list(str do { jh = next_jh; next_jh = jh->b_cpnext; - ret = __try_to_free_cp_buf(jh); + if (!destroy) + ret = __try_to_free_cp_buf(jh); + else + ret = __jbd2_journal_remove_checkpoint(jh) + 1; if (!ret) return freed; if (ret == 2)
@@ -459,10 +462,11 @@ static int journal_clean_one_cp_list(str* journal_clean_checkpoint_list * * Find all the written-back checkpoint buffers in the journal and release them. +* If 'destroy' is set, release all buffers unconditionally. * * Called with j_list_lock held. */ -void __jbd2_journal_clean_checkpoint_list (journal_t *journal) +void __jbd2_journal_clean_checkpoint_list (journal_t *journal,bool destroy) {transaction_t *transaction, *last_transaction, *next_transaction;int ret;
@@ -476,7 +480,8 @@ void __jbd2_journal_clean_checkpoint_lis do { transaction = next_transaction; next_transaction = transaction->t_cpnext; - ret = journal_clean_one_cp_list(transaction->t_checkpoint_list); + ret = journal_clean_one_cp_list(transaction->t_checkpoint_list, + destroy); /* * This function only frees up some memory if possible so we * dont have an obligation to finish processing. Bail out if
@@ -492,7 +497,7 @@ void __jbd2_journal_clean_checkpoint_lis* we can possibly see not yet submitted buffers on io_list */ ret =journal_clean_one_cp_list (transaction-> - t_checkpoint_io_list); + t_checkpoint_io_list, destroy);if (need_resched ())return ;/*
@@ -506,6 +511,28 @@ void __jbd2_journal_clean_checkpoint_lis }/* +* Remove buffers from all checkpoint lists as journal is aborted and we just +* need to free memory +*/ +void jbd2_journal_destroy_checkpoint (journal_t *journal) +{ +/* +* We loop because __jbd2_journal_clean_checkpoint_list() may abort +* early due to a need of rescheduling. +*/ +while (1 ) { +spin_lock (&journal->j_list_lock); +if (!journal->j_checkpoint_transactions) { +spin_unlock (&journal->j_list_lock); +break ; + } +__jbd2_journal_clean_checkpoint_list (journal, true); +spin_unlock (&journal->j_list_lock); +cond_resched (); + } +} + +/* * journal_remove_checkpoint: called after a buffer has been committed * to disk (either by being write-back flushed to disk, or being * committed to the log).
fs/jbd2/commit.c +1 -1
@@ -510,7 +510,7 @@ void jbd2_journal_commit_transaction(jou* frees some memory */ spin_lock (&journal->j_list_lock); -__jbd2_journal_clean_checkpoint_list (journal); +__jbd2_journal_clean_checkpoint_list (journal, false);spin_unlock (&journal->j_list_lock);jbd_debug (3 ,"JBD2: commit phase 1\n" );
fs/jbd2/journal.c +10 -1
@@ -1708,8 +1708,17 @@ int jbd2_journal_destroy(journal_t *jourwhile (journal->j_checkpoint_transactions !=NULL ) {spin_unlock (&journal->j_list_lock);mutex_lock (&journal->j_checkpoint_mutex); -jbd2_log_do_checkpoint (journal); + err =jbd2_log_do_checkpoint (journal);mutex_unlock (&journal->j_checkpoint_mutex); +/* +* If checkpointing failed, just free the buffers to avoid +* looping forever +*/ +if (err) { +jbd2_journal_destroy_checkpoint (journal); +spin_lock (&journal->j_list_lock); +break ; + }spin_lock (&journal->j_list_lock); }
include/linux/jbd2.h +2 -1
@@ -1042,8 +1042,9 @@ void jbd2_update_log_tail(journal_t *jouextern void jbd2_journal_commit_transaction (journal_t *);/* Checkpoint list management */ -void __jbd2_journal_clean_checkpoint_list (journal_t *journal); +void __jbd2_journal_clean_checkpoint_list (journal_t *journal,bool destroy);int __jbd2_journal_remove_checkpoint (struct journal_head *); +void jbd2_journal_destroy_checkpoint (journal_t *journal);void __jbd2_journal_insert_checkpoint (struct journal_head *,transaction_t *);
Greg Kroah-Hartman
maintainerPATCH 114/159 ipv6: Make MLD packets to only be processed locally
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Angga <[email protected]>
[ Upstream commit 4c938d22c88a9ddccc8c55a85e0430e9c62b1ac5 ]
Before commit daad151263cf ("ipv6: Make ipv6_is_mld() inline and use it from ip6_mc_input().") MLD packets were only processed locally. After the change, a copy of MLD packet goes through ip6_mr_input, causing MRT6MSG_NOCACHE message to be generated to user space.
Make MLD packet only processed locally.
--- net/ipv6/ip6_input.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
net/ipv6/ip6_input.c +3 -3
@@ -331,10 +331,10 @@ int ip6_mc_input(struct sk_buff *skb)if (offset <0 ) goto out; -if (!ipv6_is_mld (skb, nexthdr, offset)) - goto out; +if (ipv6_is_mld (skb, nexthdr, offset)) + deliver = true; - deliver = true; + goto out; }/* unknown RA - process it normally */ }
Greg Kroah-Hartman
maintainerPATCH 115/159 rhashtable: fix for resize events during table walk
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Phil Sutter <[email protected]>
[ Upstream commit 142b942a75cb10ede1b42bf85368d41449ab4e3b ]
If rhashtable_walk_next detects a resize operation in progress, it jumps to the new table and continues walking that one. But it misses to drop the reference to it's current item, leading it to continue traversing the new table's bucket in which the current item is sorted into, and after reaching that bucket's end continues traversing the new table's second bucket instead of the first one, thereby potentially missing items.
This fixes the rhashtable runtime test for me. Bug probably introduced by Herbert Xu's patch eddee5ba ("rhashtable: Fix walker behaviour during rehash") although not explicitly tested.
--- lib/rhashtable.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
lib/rhashtable.c +2 -2
@@ -612,6 +612,8 @@ next: iter->skip =0 ; } + iter->p =NULL ; +/* Ensure we see any new tables. */ smp_rmb ();
@@ -622,8 +624,6 @@ next:return ERR_PTR (-EAGAIN ); } - iter->p =NULL ; - out:return obj;
Greg Kroah-Hartman
maintainerPATCH 116/159 net: graceful exit from netif_alloc_netdev_queues()
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Eric Dumazet <[email protected]>
[ Upstream commit d339727c2b1a10f25e6636670ab6e1841170e328 ]
User space can crash kernel with
ip link add ifb10 numtxqueues 100000 type ifb
We must replace a BUG_ON() by proper test and return -EINVAL for crazy values.
--- net/core/dev.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
net/core/dev.c +2 -1
@@ -6297,7 +6297,8 @@ static int netif_alloc_netdev_queues(strstruct netdev_queue *tx;size_t sz = count *sizeof (*tx); -BUG_ON (count <1 || count >0xffff ); +if (count <1 || count >0xffff ) +return -EINVAL ; tx =kzalloc (sz,GFP_KERNEL | __GFP_NOWARN | __GFP_REPEAT);if (!tx) {
Greg Kroah-Hartman
maintainerPATCH 117/159 Revert "dev: set iflink to 0 for virtual interfaces"
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Nicolas Dichtel <[email protected]>
[ Upstream commit 95ec655bc465ccb2a3329d4aff9a45e3c8188db5 ]
This reverts commit e1622baf54df8cc958bf29d71de5ad545ea7d93c.
The side effect of this commit is to add a '@NONE' after each virtual interface name with a 'ip link'. It may break existing scripts.
--- net/core/dev.c | 4 ---- 1 file changed, 4 deletions(-)
net/core/dev.c +0 -4
@@ -672,10 +672,6 @@ int dev_get_iflink(const struct net_deviif (dev->netdev_ops && dev->netdev_ops->ndo_get_iflink)return dev->netdev_ops->ndo_get_iflink (dev); -/* If dev->rtnl_link_ops is set, it's a virtual interface. */ -if (dev->rtnl_link_ops) -return 0 ; -return dev->ifindex; }EXPORT_SYMBOL (dev_get_iflink);
Greg Kroah-Hartman
maintainerPATCH 118/159 rtnetlink: verify IFLA_VF_INFO attributes before passing them to driver
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Daniel Borkmann <[email protected]>
[ Upstream commit 4f7d2cdfdde71ffe962399b7020c674050329423 ]
Jason Gunthorpe reported that since commit c02db8c6290b ("rtnetlink: make SR-IOV VF interface symmetric"), we don't verify IFLA_VF_INFO attributes anymore with respect to their policy, that is, ifla_vfinfo_policy[].
Before, they were part of ifla_policy[], but they have been nested since placed under IFLA_VFINFO_LIST, that contains the attribute IFLA_VF_INFO, which is another nested attribute for the actual VF attributes such as IFLA_VF_MAC, IFLA_VF_VLAN, etc.
Despite the policy being split out from ifla_policy[] in this commit, it's never applied anywhere. nla_for_each_nested() only does basic nla_ok() testing for struct nlattr, but it doesn't know about the data context and their requirements.
Fix, on top of Jason's initial work, does 1) parsing of the attributes with the right policy, and 2) using the resulting parsed attribute table from 1) instead of the nla_for_each_nested() loop (just like we used to do when still part of ifla_policy[]).
Reference: http://thread.gmane.org/gmane.linux.network/368913
--- net/core/rtnetlink.c | 187 ++++++++++++++++++++++++++------------------------- 1 file changed, 96 insertions(+), 91 deletions(-)
net/core/rtnetlink.c +96 -91
@@ -1287,10 +1287,6 @@ static const struct nla_policy ifla_info [IFLA_INFO_SLAVE_DATA] = { .type = NLA_NESTED }, }; -static const struct nla_policy ifla_vfinfo_policy[IFLA_VF_INFO_MAX +1 ] = { - [IFLA_VF_INFO ] = { .type =NLA_NESTED }, -}; -static const struct nla_policy ifla_vf_policy[IFLA_VF_MAX +1 ] = { [IFLA_VF_MAC ] = { .len =sizeof (struct ifla_vf_mac ) }, [IFLA_VF_VLAN ] = { .len =sizeof (struct ifla_vf_vlan ) },
@@ -1437,96 +1433,98 @@ static int validate_linkmsg(struct net_dreturn 0 ; } -static int do_setvfinfo (struct net_device *dev,struct nlattr *attr) +static int do_setvfinfo (struct net_device *dev,struct nlattr **tb) { -int rem, err = -EINVAL ; -struct nlattr *vf;const struct net_device_ops *ops = dev->netdev_ops; +int err = -EINVAL ; -nla_for_each_nested (vf, attr, rem) { -switch (nla_type (vf)) { -case IFLA_VF_MAC : { -struct ifla_vf_mac *ivm; - ivm =nla_data (vf); - err = -EOPNOTSUPP ; -if (ops->ndo_set_vf_mac) - err = ops->ndo_set_vf_mac (dev, ivm->vf, - ivm->mac); -break ; - } -case IFLA_VF_VLAN : { -struct ifla_vf_vlan *ivv; - ivv =nla_data (vf); - err = -EOPNOTSUPP ; -if (ops->ndo_set_vf_vlan) - err = ops->ndo_set_vf_vlan (dev, ivv->vf, - ivv->vlan, - ivv->qos); -break ; - } -case IFLA_VF_TX_RATE : { -struct ifla_vf_tx_rate *ivt; -struct ifla_vf_info ivf; - ivt =nla_data (vf); - err = -EOPNOTSUPP ; -if (ops->ndo_get_vf_config) - err = ops->ndo_get_vf_config (dev, ivt->vf, - &ivf); -if (err) -break ; - err = -EOPNOTSUPP ; -if (ops->ndo_set_vf_rate) - err = ops->ndo_set_vf_rate (dev, ivt->vf, - ivf.min_tx_rate, - ivt->rate); -break ; - } -case IFLA_VF_RATE : { -struct ifla_vf_rate *ivt; - ivt =nla_data (vf); - err = -EOPNOTSUPP ; -if (ops->ndo_set_vf_rate) - err = ops->ndo_set_vf_rate (dev, ivt->vf, - ivt->min_tx_rate, - ivt->max_tx_rate); -break ; - } -case IFLA_VF_SPOOFCHK : { -struct ifla_vf_spoofchk *ivs; - ivs =nla_data (vf); - err = -EOPNOTSUPP ; -if (ops->ndo_set_vf_spoofchk) - err = ops->ndo_set_vf_spoofchk (dev, ivs->vf, - ivs->setting); -break ; - } -case IFLA_VF_LINK_STATE : { -struct ifla_vf_link_state *ivl; - ivl =nla_data (vf); - err = -EOPNOTSUPP ; -if (ops->ndo_set_vf_link_state) - err = ops->ndo_set_vf_link_state (dev, ivl->vf, - ivl->link_state); -break ; - } -case IFLA_VF_RSS_QUERY_EN : { -struct ifla_vf_rss_query_en *ivrssq_en; +if (tb[IFLA_VF_MAC ]) { +struct ifla_vf_mac *ivm =nla_data (tb[IFLA_VF_MAC ]); - ivrssq_en =nla_data (vf); - err = -EOPNOTSUPP ; -if (ops->ndo_set_vf_rss_query_en) - err = ops->ndo_set_vf_rss_query_en (dev, - ivrssq_en->vf, - ivrssq_en->setting); -break ; - } -default : - err = -EINVAL ; -break ; - } -if (err) -break ; + err = -EOPNOTSUPP ; +if (ops->ndo_set_vf_mac) + err = ops->ndo_set_vf_mac (dev, ivm->vf, + ivm->mac); +if (err <0 ) +return err; + } + +if (tb[IFLA_VF_VLAN ]) { +struct ifla_vf_vlan *ivv =nla_data (tb[IFLA_VF_VLAN ]); + + err = -EOPNOTSUPP ; +if (ops->ndo_set_vf_vlan) + err = ops->ndo_set_vf_vlan (dev, ivv->vf, ivv->vlan, + ivv->qos); +if (err <0 ) +return err; } + +if (tb[IFLA_VF_TX_RATE ]) { +struct ifla_vf_tx_rate *ivt =nla_data (tb[IFLA_VF_TX_RATE ]); +struct ifla_vf_info ivf; + + err = -EOPNOTSUPP ; +if (ops->ndo_get_vf_config) + err = ops->ndo_get_vf_config (dev, ivt->vf, &ivf); +if (err <0 ) +return err; + + err = -EOPNOTSUPP ; +if (ops->ndo_set_vf_rate) + err = ops->ndo_set_vf_rate (dev, ivt->vf, + ivf.min_tx_rate, + ivt->rate); +if (err <0 ) +return err; + } + +if (tb[IFLA_VF_RATE ]) { +struct ifla_vf_rate *ivt =nla_data (tb[IFLA_VF_RATE ]); + + err = -EOPNOTSUPP ; +if (ops->ndo_set_vf_rate) + err = ops->ndo_set_vf_rate (dev, ivt->vf, + ivt->min_tx_rate, + ivt->max_tx_rate); +if (err <0 ) +return err; + } + +if (tb[IFLA_VF_SPOOFCHK ]) { +struct ifla_vf_spoofchk *ivs =nla_data (tb[IFLA_VF_SPOOFCHK ]); + + err = -EOPNOTSUPP ; +if (ops->ndo_set_vf_spoofchk) + err = ops->ndo_set_vf_spoofchk (dev, ivs->vf, + ivs->setting); +if (err <0 ) +return err; + } + +if (tb[IFLA_VF_LINK_STATE ]) { +struct ifla_vf_link_state *ivl =nla_data (tb[IFLA_VF_LINK_STATE ]); + + err = -EOPNOTSUPP ; +if (ops->ndo_set_vf_link_state) + err = ops->ndo_set_vf_link_state (dev, ivl->vf, + ivl->link_state); +if (err <0 ) +return err; + } + +if (tb[IFLA_VF_RSS_QUERY_EN ]) { +struct ifla_vf_rss_query_en *ivrssq_en; + + err = -EOPNOTSUPP ; + ivrssq_en =nla_data (tb[IFLA_VF_RSS_QUERY_EN ]); +if (ops->ndo_set_vf_rss_query_en) + err = ops->ndo_set_vf_rss_query_en (dev, ivrssq_en->vf, + ivrssq_en->setting); +if (err <0 ) +return err; + } +return err; }
@@ -1722,14 +1720,21 @@ static int do_setlink(const struct sk_bu }if (tb[IFLA_VFINFO_LIST ]) { +struct nlattr *vfinfo[IFLA_VF_MAX +1 ];struct nlattr *attr;int rem; +nla_for_each_nested (attr, tb[IFLA_VFINFO_LIST ], rem) { -if (nla_type (attr) !=IFLA_VF_INFO ) { +if (nla_type (attr) !=IFLA_VF_INFO || +nla_len (attr) <NLA_HDRLEN ) { err = -EINVAL ; goto errout; } - err =do_setvfinfo (dev, attr); + err =nla_parse_nested (vfinfo,IFLA_VF_MAX , attr, + ifla_vf_policy); +if (err <0 ) + goto errout; + err =do_setvfinfo (dev, vfinfo);if (err <0 ) goto errout; status |=DO_SETLINK_NOTIFY ;
Greg Kroah-Hartman
maintainerPATCH 120/159 net/tipc: initialize security state for new connection socket
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Stephen Smalley <[email protected]>
[ Upstream commit fdd75ea8df370f206a8163786e7470c1277a5064 ]
Calling connect() with an AF_TIPC socket would trigger a series of error messages from SELinux along the lines of:
SELinux: Invalid class 0
type=AVC msg=audit(1434126658.487:34500): avc: denied { <unprintable> }
for pid=292 comm="kworker/u16:5" scontext=system_u:system_r:kernel_t:s0 tcontext=system_u:object_r:unlabeled_t:s0 tclass=<unprintable> permissive=0
This was due to a failure to initialize the security state of the new connection sock by the tipc code, leaving it with junk in the security class field and an unlabeled secid. Add a call to security_sk_clone() to inherit the security state from the parent socket.
--- net/tipc/socket.c | 1 + 1 file changed, 1 insertion(+)
net/tipc/socket.c +1 -0
@@ -2009,6 +2009,7 @@ static int tipc_accept(struct socket *so res =tipc_sk_create (sock_net (sock->sk), new_sock,0 ,1 );if (res) goto exit; +security_sk_clone (sock->sk, new_sock->sk); new_sk = new_sock->sk; new_tsock =tipc_sk (new_sk);
Greg Kroah-Hartman
maintainerPATCH 121/159 bridge: mdb: zero out the local br_ip variable before use
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Nikolay Aleksandrov <[email protected]>
[ Upstream commit f1158b74e54f2e2462ba5e2f45a118246d9d5b43 ]
Since commit b0e9a30dd669 ("bridge: Add vlan id to multicast groups") there's a check in br_ip_equal() for a matching vlan id, but the mdb functions were not modified to use (or at least zero it) so when an entry was added it would have a garbage vlan id (from the local br_ip variable in __br_mdb_add/del) and this would prevent it from being matched and also deleted. So zero out the whole local ip var to protect ourselves from future changes and also to fix the current bug, since there's no vlan id support in the mdb uapi - use always vlan id 0. Example before patch:
root@debian:~# bridge mdb add dev br0 port eth1 grp 239.0.0.1 permanent root@debian:~# bridge mdb dev br0 port eth1 grp 239.0.0.1 permanent root@debian:~# bridge mdb del dev br0 port eth1 grp 239.0.0.1 permanent RTNETLINK answers: Invalid argument
After patch:
root@debian:~# bridge mdb add dev br0 port eth1 grp 239.0.0.1 permanent root@debian:~# bridge mdb dev br0 port eth1 grp 239.0.0.1 permanent root@debian:~# bridge mdb del dev br0 port eth1 grp 239.0.0.1 permanent root@debian:~# bridge mdb
--- net/bridge/br_mdb.c | 2 ++ 1 file changed, 2 insertions(+)
net/bridge/br_mdb.c +2 -0
@@ -371,6 +371,7 @@ static int __br_mdb_add(struct net *net,if (!p || p->br != br || p->state ==BR_STATE_DISABLED )return -EINVAL ; +memset (&ip,0 ,sizeof (ip)); ip.proto = entry->addr.proto;if (ip.proto ==htons (ETH_P_IP )) ip.u.ip4 = entry->addr.u.ip4;
@@ -417,6 +418,7 @@ static int __br_mdb_del(struct net_bridgif (!netif_running (br->dev) || br->multicast_disabled)return -EINVAL ; +memset (&ip,0 ,sizeof (ip)); ip.proto = entry->addr.proto;if (ip.proto ==htons (ETH_P_IP )) {if (timer_pending (&br->ip4_other_query.timer))
Greg Kroah-Hartman
maintainerPATCH 122/159 net: pktgen: fix race between pktgen_thread_worker() and kthread_stop()
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Oleg Nesterov <[email protected]>
[ Upstream commit fecdf8be2d91e04b0a9a4f79ff06499a36f5d14f ]
pktgen_thread_worker() is obviously racy, kthread_stop() can come between the kthread_should_stop() check and set_current_state().
--- net/core/pktgen.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
net/core/pktgen.c +3 -1
@@ -3490,8 +3490,10 @@ static int pktgen_thread_worker(void *arpktgen_rem_thread (t);/* Wait for kthread_stop */ -while (!kthread_should_stop ()) { +for (;;) {set_current_state (TASK_INTERRUPTIBLE ); +if (kthread_should_stop ()) +break ;schedule (); }__set_current_state (TASK_RUNNING );
Greg Kroah-Hartman
maintainerPATCH 123/159 bridge: fix potential crash in __netdev_pick_tx()
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Eric Dumazet <[email protected]>
[ Upstream commit a7d35f9d73e9ffa74a02304b817e579eec632f67 ]
Commit c29390c6dfee ("xps: must clear sender_cpu before forwarding") fixed an issue in normal forward path, caused by sender_cpu & napi_id skb fields being an union.
Bridge is another point where skb can be forwarded, so we need the same cure.
Bug triggers if packet was received on a NIC using skb_mark_napi_id()
--- net/bridge/br_forward.c | 1 + 1 file changed, 1 insertion(+)
net/bridge/br_forward.c +1 -0
@@ -42,6 +42,7 @@ int br_dev_queue_push_xmit(struct sock * }else {skb_push (skb,ETH_HLEN );br_drop_fake_rtable (skb); +skb_sender_cpu_clear (skb);dev_queue_xmit (skb); }
Greg Kroah-Hartman
maintainerPATCH 124/159 net: do not process device backlog during unregistration
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Julian Anastasov <[email protected]>
[ Upstream commit e9e4dd3267d0c5234c5c0f47440456b10875dec9 ]
commit 381c759d9916 ("ipv4: Avoid crashing in ip_error") fixes a problem where processed packet comes from device with destroyed inetdev (dev->ip_ptr). This is not expected because inetdev_destroy is called in NETDEV_UNREGISTER phase and packets should not be processed after dev_close_many() and synchronize_net(). Above fix is still required because inetdev_destroy can be called for other reasons. But it shows the real problem: backlog can keep packets for long time and they do not hold reference to device. Such packets are then delivered to upper levels at the same time when device is unregistered.
Calling flush_backlog after NETDEV_UNREGISTER_FINAL still accounts all packets from backlog but before that some packets continue to be delivered to upper levels long after the synchronize_net call which is supposed to wait the last ones. Also, as Eric pointed out, processed packets, mostly from other devices, can continue to add new packets to backlog.
Fix the problem by moving flush_backlog early, after the device driver is stopped and before the synchronize_net() call. Then use netif_running check to make sure we do not add more packets to backlog. We have to do it in enqueue_to_backlog context when the local IRQ is disabled. As result, after the flush_backlog and synchronize_net sequence all packets should be accounted.
Thanks to Eric W. Biederman for the test script and his valuable feedback!
--- net/core/dev.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
net/core/dev.c +4 -2
@@ -3337,6 +3337,8 @@ static int enqueue_to_backlog(struct sk_local_irq_save (flags);rps_lock (sd); +if (!netif_running (skb->dev)) + goto drop; qlen =skb_queue_len (&sd->input_pkt_queue);if (qlen <= netdev_max_backlog && !skb_flow_limit (skb, qlen)) {if (qlen) {
@@ -3358,6 +3360,7 @@ enqueue: goto enqueue; } +drop: sd->dropped++;rps_unlock (sd);
@@ -6023,6 +6026,7 @@ static void rollback_registered_many(strunlist_netdevice (dev); dev->reg_state =NETREG_UNREGISTERING ; +on_each_cpu (flush_backlog, dev,1 ); }synchronize_net ();
@@ -6647,8 +6651,6 @@ void netdev_run_todo(void) dev->reg_state =NETREG_UNREGISTERED ; -on_each_cpu (flush_backlog, dev,1 ); -netdev_wait_allrefs (dev);/* paranoia */
Greg Kroah-Hartman
maintainerPATCH 125/159 net: call rcu_read_lock early in process_backlog
In reply to Greg Kroah-Hartman4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Julian Anastasov <[email protected]>
[ Upstream commit 2c17d27c36dcce2b6bf689f41a46b9e909877c21 ]
Incoming packet should be either in backlog queue or in RCU read-side section. Otherwise, the final sequence of flush_backlog() and synchronize_net() may miss packets that can run without device reference:
CPU 1 CPU 2
skb->dev: no reference
process_backlog:__skb_dequeue
process_backlog:local_irq_enableon_each_cpu forflush_backlog => IPI(hardirq): flush_backlog
- packet not found in backlog
CPU delayed ...synchronize_net
- no ongoing RCU
read-side sections
netdev_run_todo,rcu_barrier: no
ongoing callbacks
__netif_receive_skb_core:rcu_read_lock
- too late
free dev
process packet for freed dev
--- net/core/dev.c | 30 +++++++++++++++--------------- 1 file changed, 15 insertions(+), 15 deletions(-)
net/core/dev.c +15 -15
@@ -3666,8 +3666,6 @@ static int __netif_receive_skb_core(stru pt_prev =NULL ; -rcu_read_lock (); - another_round: skb->skb_iif = skb->dev->ifindex;
@@ -3677,7 +3675,7 @@ another_round: skb->protocol == cpu_to_be16(ETH_P_8021AD)) { skb =skb_vlan_untag (skb);if (unlikely (!skb)) - goto unlock; + goto out; }#ifdef CONFIG_NET_CLS_ACT
@@ -3707,7 +3705,7 @@ skip_taps:if (static_key_false (&ingress_needed)) { skb =handle_ing (skb, &pt_prev, &ret, orig_dev);if (!skb) - goto unlock; + goto out; } skb->tc_verd =0 ;
@@ -3724,7 +3722,7 @@ ncls:if (vlan_do_receive (&skb)) goto another_round;else if (unlikely (!skb)) - goto unlock; + goto out; } rx_handler =rcu_dereference (skb->dev->rx_handler);
@@ -3736,7 +3734,7 @@ ncls:switch (rx_handler (&skb)) {case RX_HANDLER_CONSUMED : ret =NET_RX_SUCCESS ; - goto unlock; + goto out;case RX_HANDLER_ANOTHER : goto another_round;case RX_HANDLER_EXACT :
@@ -3790,8 +3788,7 @@ drop: ret =NET_RX_DROP ; } -unlock: -rcu_read_unlock (); +out:return ret; }
@@ -3822,29 +3819,30 @@ static int __netif_receive_skb(struct skstatic int netif_receive_skb_internal (struct sk_buff *skb) { +int ret; +net_timestamp_check (netdev_tstamp_prequeue, skb);if (skb_defer_rx_timestamp (skb))return NET_RX_SUCCESS ; +rcu_read_lock (); +#ifdef CONFIG_RPS if (static_key_false (&rps_needed)) {struct rps_dev_flow voidflow, *rflow = &voidflow; -int cpu, ret; - -rcu_read_lock (); - - cpu =get_rps_cpu (skb->dev, skb, &rflow); +int cpu =get_rps_cpu (skb->dev, skb, &rflow);if (cpu >=0 ) { ret =enqueue_to_backlog (skb, cpu, &rflow->last_qtail);rcu_read_unlock ();return ret; } -rcu_read_unlock (); }#endif -return __netif_receive_skb (skb); + ret =__netif_receive_skb (skb); +rcu_read_unlock (); +return ret; }/**
@@ -4389,8 +4387,10 @@ static int process_backlog(struct napi_sstruct sk_buff *skb;while ((skb =__skb_dequeue (&sd->process_queue))) { +rcu_read_lock ();local_irq_enable ();__netif_receive_skb (skb); +rcu_read_unlock ();local_irq_disable ();input_queue_head_incr (sd);if (++work >= quota) {
Re: [PATCH 4.1 125/159] net: call rcu_read_lock early in process_backlog
Quoted 4 lines · Greg Kroah-Hartman
4.1-stable review patch. If anyone has any objections, please let me know.
------------------
Quoted 8 lines
From: Julian Anastasov <[email protected]>
[ Upstream commit
2c17d27c36dcce2b6bf689f41a46b9e909877c21]Incoming packet should be either in backlog queue or in RCU read-side section. Otherwise, the final sequence of
flush_backlog()andsynchronize_net()may miss packets that can run without device reference:
<snip>
Several of our 4.1.9-rc1 running systems are experiencing hangs requiring hardware/sysrq reset with this patch applied. Reverting it fixes the hangs completely.
4.2 includes this patch as well but I have no such problems there. 4.2.2-rc1 works fine as well.
For now I think this patch should be reverted in 4.1.9.
The hangs have occured so far on Xen PV and KVM x86_64 virtual machines, they will hang completely within minutes or hours depending on the type of workload. The workloads are all fairly light, one running low traffic email/antispam, another running monitoring and metrics of ~5 hosts and one running a single terminal IRC client. All but the IRC one will hang within a few minutes of booting.
When they lock up they only respond to sysrq, with ttyS0/hvc0 not echoing anything typed in back, and are completely dead on the network. One system managed to report rcu stalls but no backtraces (I'll look over the debug config, if there is any interest).
My bare metal desktop has yet to be able to hit it, but it might be entirely down to a different type of workload.
Something missing in 4.1?
Re: [PATCH 4.1 125/159] net: call rcu_read_lock early in process_backlog
Hello,
Quoted 13 lines · Andre Tomt (LKML)
Quoted 4 lines · Greg Kroah-Hartman
4.1-stable review patch. If anyone has any objections, please let me know.
------------------
Quoted 8 lines
From: Julian Anastasov <[email protected]>
[ Upstream commit
2c17d27c36dcce2b6bf689f41a46b9e909877c21]Incoming packet should be either in backlog queue or in RCU read-side section. Otherwise, the final sequence of
flush_backlog()andsynchronize_net()may miss packets that can run without device reference:
Quoted 5 lines
<snip>
Several of our 4.1.9-rc1 running systems are experiencing hangs requiring hardware/sysrq reset with this patch applied. Reverting it fixes the hangs completely.
Bad to hear that...
Quoted 2 lines
4.2 includes this patch as well but I have no such problems there. 4.2.2-rc1 works fine as well.
No problem with 4.2+? Same setup/config?
Quoted 14 lines
For now I think this patch should be reverted in 4.1.9.
The hangs have occured so far on Xen PV and KVM
x86_64virtual machines, they will hang completely within minutes or hours depending on the type of workload. The workloads are all fairly light, one running low traffic email/antispam, another running monitoring and metrics of ~5 hosts and one running a single terminal IRC client. All but the IRC one will hang within a few minutes of booting.When they lock up they only respond to sysrq, with ttyS0/hvc0 not echoing anything typed in back, and are completely dead on the network. One system managed to report rcu stalls but no backtraces (I'll look over the debug config, if there is any interest).
Quoted 4 lines
My bare metal desktop has yet to be able to hit it, but it might be entirely down to a different type of workload.
Something missing in 4.1?
They are 2 related patches, the first one is
[PATCH 4.1 124/159] net: do not process device backlog during unregistration
But the problematic patch calls rcu_read_lock while
local IRQ is disabled (in process_backlog), this is something that should be noted for the patch. I'll try to see what Xen does. It would be useful to see .config and any kind of backtraces/stalls, it will help also to other developers to catch the problem...
Regards
--
Julian Anastasov <[email protected]>