5.4.24-stable review
Subsystems (5)
Greg Kroah-Hartman
maintainercover 5.4.24-stable review
This is the start of the stable review cycle for the 5.4.24 release. There are 152 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 Thu, 05 Mar 2020 17:42:10 +0000. Anything received after that time might be too late.
The whole patch series can be found in one patch at:
https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.4.24-rc1.gz
or in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-5.4.y
and the diffstat can be found below.
thanks,
greg k-h
-------------
Pseudo-Shortlog of commits:
Greg Kroah-Hartman <[email protected]>
Linux 5.4.24-rc1
Jim Mattson <[email protected]>
kvm: nVMX: VMWRITE checks unsupported field before read-only field
Jim Mattson <[email protected]>
kvm: nVMX: VMWRITE checks VMCS-link pointer before VMCS field
David Rientjes <[email protected]>
mm, thp: fix defrag setting if newline is not used
Wei Yang <[email protected]>
mm/huge_memory.c: use head to check huge zero page
John Hubbard <[email protected]>
mm/gup: allow FOLL_FORCE for get_user_pages_fast()
Vlastimil Babka <[email protected]>
mm/debug.c: always print flags in dump_page()
Waiman Long <[email protected]>
locking/lockdep: Fix lockdep_stats indentation problem
Daniel Jordan <[email protected]>
padata: always acquire cpu_hotplug_lock before pinst->lock
Christoph Hellwig <[email protected]>
xfs: clear kernel only flags in XFS_IOC_ATTRMULTI_BY_HANDLE
Sameer Pujar <[email protected]>
bus: tegra-aconnect: Remove PM_CLK dependency
Matteo Croce <[email protected]>
netfilter: nf_flowtable: fix documentation
Xin Long <[email protected]>
netfilter: nft_tunnel: no need to call htons() when dumping ports
Florian Fainelli <[email protected]>
thermal: brcmstb_thermal: Do not use DT coefficients
Linus Walleij <[email protected]>
thermal: db8500: Depromote debug print
Geert Uytterhoeven <[email protected]>
ubifs: Fix ino_t format warnings in orphan_delete()
Neeraj Upadhyay <[email protected]>
rcu: Allow only one expedited GP to run concurrently with wakeups
Sean Christopherson <[email protected]>
KVM: x86: Remove spurious clearing of async #PF MSR
Sean Christopherson <[email protected]>
KVM: x86: Remove spurious kvm_mmu_unload() from vcpu destruction path
Xiaochen Shen <[email protected]>
x86/resctrl: Check monitoring static key in the MBM overflow handler
Jiri Olsa <[email protected]>
perf ui gtk: Add missing zalloc object
Arnaldo Carvalho de Melo <[email protected]>
perf hists browser: Restore ESC as "Zoom out" of DSO/thread/etc
Uwe Kleine-König <[email protected]>
pwm: omap-dmtimer: put_device() after of_find_device_by_node()
Thomas Gleixner <[email protected]>
lib/vdso: Update coarse timekeeper unconditionally
Thomas Gleixner <[email protected]>
lib/vdso: Make __arch_update_vdso_data() logic understandable
Masami Hiramatsu <[email protected]>
kprobes: Set unoptimized flag after unoptimizing code
Janne Karhunen <[email protected]>
ima: ima/lsm policy rule loading logic bug fixes
Christophe JAILLET <[email protected]>
drivers: net: xgene: Fix the order of the arguments of 'alloc_etherdev_mqs()'
Lijun Ou <[email protected]>
RDMA/hns: Bugfix for posting a wqe with sge
Yixian Liu <[email protected]>
RDMA/hns: Simplify the calculation and usage of wqe idx for post verbs
Chao Yu <[email protected]>
f2fs: fix to add swap extent correctly
Cheng Jian <[email protected]>
sched/fair: Optimize select_idle_cpu
Sean Christopherson <[email protected]>
KVM: Check for a bad hva before dropping into the ghc slow path
Tom Lendacky <[email protected]>
KVM: SVM: Override default MMIO mask if memory encryption is enabled
Brian Norris <[email protected]>
mwifiex: delete unused mwifiex_get_intf_num()
Brian Norris <[email protected]>
mwifiex: drop most magic numbers from mwifiex_process_tdls_action_frame()
Aleksa Sarai <[email protected]>
namei: only return -ECHILD from follow_dotdot_rcu()
Masahiro Yamada <[email protected]>
kbuild: make single target builds even faster
Masahiro Yamada <[email protected]>
kbuild: remove unneeded variable, single-all
Masahiro Yamada <[email protected]>
kbuild: move headers_check rule to usr/include/Makefile
Masahiro Yamada <[email protected]>
kbuild: remove header compile test
Michael Ellerman <[email protected]>
selftests: Install settings files to fix TIMEOUT failures
Arthur Kiyanovski <[email protected]>
net: ena: make ena rxfh support ETH_RSS_HASH_NO_CHANGE
Ursula Braun <[email protected]>
net/smc: no peer ID in CLC decline for SMCD
Dmitry Bogdanov <[email protected]>
net: atlantic: fix out of range usage of active_vlans array
Pavel Belous <[email protected]>
net: atlantic: fix potential error handling
Pavel Belous <[email protected]>
net: atlantic: fix use after free kasan warn
Nikolay Aleksandrov <[email protected]>
net: netlink: cap max groups which will be considered in netlink_bind()
Alexandra Winter <[email protected]>
s390/qeth: vnicc Fix EOPNOTSUPP precedence
Bijan Mottahedeh <[email protected]>
nvme-pci: Hold cq_poll_lock while completing CQEs
Peter Chen <[email protected]>
usb: charger: assign specific number for enum value
Haiyang Zhang <[email protected]>
hv_netvsc: Fix unwanted wakeup in netvsc_attach()
Masahiro Yamada <[email protected]>
kbuild: fix DT binding schema rule to detect command line changes
Andrei Otcheretianski <[email protected]>
mac80211: Remove a redundant mutex unlock
Johannes Berg <[email protected]>
nl80211: fix potential leak in AP start
Tina Zhang <[email protected]>
drm/i915/gvt: Separate display reset from ALL_ENGINES reset
Tina Zhang <[email protected]>
drm/i915/gvt: Fix orphan vgpu dmabuf_objs' lifetime
Wolfram Sang <[email protected]>
i2c: jz4780: silence log flood on txabrt
Gustavo A. R. Silva <[email protected]>
i2c: altera: Fix potential integer overflow
Christophe JAILLET <[email protected]>
MIPS: VPE: Fix a double free and a memory leak in 'release_vpe()'
[email protected] <[email protected]>
HID: hiddev: Fix race in in hiddev_disconnect()
Christophe JAILLET <[email protected]>
HID: alps: Fix an error handling path in 'alps_input_configured()'
Cong Wang <[email protected]>
netfilter: xt_hashlimit: reduce hashlimit_mutex scope for htable_put()
Jozsef Kadlecsik <[email protected]>
netfilter: ipset: Fix forceadd evaluation path
Eugenio Pérez <[email protected]>
vhost: Check docket sk_family instead of call getname
Ursula Braun <[email protected]>
net/smc: transfer fasync_list in case of fallback
Jozsef Kadlecsik <[email protected]>
netfilter: ipset: Fix "INFO: rcu detected stall in hash_xxx" reports
Jens Axboe <[email protected]>
io_uring: fix 32-bit compatability with sendmsg/recvmsg
Rafael J. Wysocki <[email protected]>
cpufreq: Fix policy initialization for internal governor drivers
Shirish S <[email protected]>
amdgpu/gmc_v9: save/restore sdpif regs during S3
Orson Zhai <[email protected]>
Revert "PM / devfreq: Modify the device name as devfreq(X) for sysfs"
Steven Rostedt (VMware) <[email protected]>
tracing: Disable trace_printk() on post poned tests
Wolfram Sang <[email protected]>
macintosh: therm_windtunnel: fix regression when instantiating devices
Daniel Vetter <[email protected]>
drm/radeon: Inline drm_get_pci_dev
Daniel Vetter <[email protected]>
drm/amdgpu: Drop DRIVER_USE_AGP
Johan Korsnes <[email protected]>
HID: core: increase HID report buffer size to 8KiB
Johan Korsnes <[email protected]>
HID: core: fix off-by-one memset in hid_report_raw_event()
Hans de Goede <[email protected]>
HID: ite: Only bind to keyboard USB interface on Acer SW5-012 keyboard dock
Oliver Upton <[email protected]>
KVM: VMX: check descriptor table exits on instruction emulation
Mika Westerberg <[email protected]>
ACPI: watchdog: Fix gas->access_width usage
Mika Westerberg <[email protected]>
ACPICA: Introduce ACPI_ACCESS_BYTE_WIDTH() macro
Paul Moore <[email protected]>
audit: always check the netlink payload length in audit_receive_msg()
Paul Moore <[email protected]>
audit: fix error handling in audit_data_to_entry()
Dan Carpenter <[email protected]>
ext4: potential crash on allocation error in ext4_alloc_flex_bg_array()
Keith Busch <[email protected]>
nvme/pci: move cqe check after device shutdown
Nigel Kirkland <[email protected]>
nvme: prevent warning triggered by nvme_stop_keep_alive
Anton Eidelman <[email protected]>
nvme/tcp: fix bug on double requeue when send fails
Guangbin Huang <[email protected]>
net: hns3: fix a copying IPv6 address error in hclge_fd_get_flow_tuples()
Yufeng Mo <[email protected]>
net: hns3: add management table after IMP reset
Shay Bar <[email protected]>
mac80211: fix wrong 160/80+80 MHz setting
Sergey Matyukevich <[email protected]>
cfg80211: add missing policy for NL80211_ATTR_STATUS_CODE
Coly Li <[email protected]>
bcache: ignore pending signals when creating gc and allocator thread
Frank Sorenson <[email protected]>
cifs: Fix mode output in debugging statements
Bruce Allan <[email protected]>
ice: update Unit Load Status bitmask to check after reset
Arthur Kiyanovski <[email protected]>
net: ena: ena-com.c: prevent NULL pointer dereference
Sameeh Jubran <[email protected]>
net: ena: ethtool: use correct value for crc32 hash
Arthur Kiyanovski <[email protected]>
net: ena: fix corruption of dev_idx_to_host_tbl
Arthur Kiyanovski <[email protected]>
net: ena: fix incorrectly saving queue numbers when setting RSS indirection table
Arthur Kiyanovski <[email protected]>
net: ena: rss: store hash function as values and not bits
Sameeh Jubran <[email protected]>
net: ena: rss: fix failure to get indirection table
Sameeh Jubran <[email protected]>
net: ena: rss: do not allocate key when not supported
Arthur Kiyanovski <[email protected]>
net: ena: fix incorrect default RSS key
Arthur Kiyanovski <[email protected]>
net: ena: add missing ethtool TX timestamping indication
Arthur Kiyanovski <[email protected]>
net: ena: fix uses of round_jiffies()
Arthur Kiyanovski <[email protected]>
net: ena: fix potential crash when rxfh key is NULL
Brett Creeley <[email protected]>
i40e: Fix the conditional for i40e_vc_validate_vqs_bitmaps
Thierry Reding <[email protected]>
soc/tegra: fuse: Fix build with Tegra194 configuration
Daniel Kolesa <[email protected]>
amdgpu: Prevent build errors regarding soft/hard-float FP ABI tags
Isabel Zhang <[email protected]>
drm/amd/display: Add initialitions for PLL2 clock source
Yongqiang Sun <[email protected]>
drm/amd/display: Limit minimum DPPCLK to 100MHz.
Aric Cyr <[email protected]>
drm/amd/display: Check engine is not NULL before acquiring
Krishnamraju Eraparaju <[email protected]>
RDMA/siw: Remove unwanted WARN_ON in siw_cm_llp_data_ready()
Sung Lee <[email protected]>
drm/amd/display: Do not set optimized_require to false after plane disable
Kuninori Morimoto <[email protected]>
ARM: dts: sti: fixup sound frame-inversion for stihxxx-b2120.dtsi
Xiubo Li <[email protected]>
ceph: do not execute direct write in parallel if O_APPEND is specified
Kan Liang <[email protected]>
perf/x86/msr: Add Tremont support
Kan Liang <[email protected]>
perf/x86/cstate: Add Tremont support
Kan Liang <[email protected]>
perf/x86/intel: Add Elkhart Lake support
Peter Zijlstra <[email protected]>
arm/ftrace: Fix BE text poking
John Garry <[email protected]>
perf/smmuv3: Use platform_get_irq_optional() for wired interrupt
Trond Myklebust <[email protected]>
NFSv4: Fix races between open and dentry revalidation
Bjørn Mork <[email protected]>
qmi_wwan: unconditionally reject 2 ep interfaces
Bjørn Mork <[email protected]>
qmi_wwan: re-add DW5821e pre-production variant
Harald Freudenberger <[email protected]>
s390/zcrypt: fix card and queue total counter wrap
Sergey Matyukevich <[email protected]>
cfg80211: check wiphy driver existence for drvinfo report
Johannes Berg <[email protected]>
mac80211: consider more elements in parsing CRC
Jeff Moyer <[email protected]>
dax: pass NOWAIT flag to iomap_apply
Vincent Guittot <[email protected]>
sched/fair: Prevent unlimited runtime on throttled group
Peter Zijlstra (Intel) <[email protected]>
timers/nohz: Update NOHZ load in remote tick
Scott Wood <[email protected]>
sched/core: Don't skip remote tick for idle CPUs
Sean Paul <[email protected]>
drm/msm: Set dma maximum segment size for mdss
Corey Minyard <[email protected]>
ipmi:ssif: Handle a possible NULL pointer reference
Benjamin Poirier <[email protected]>
ipv6: Fix nlmsg_flags when splitting a multipath route
Benjamin Poirier <[email protected]>
ipv6: Fix route replacement with dev-only route
Taehee Yoo <[email protected]>
bonding: fix lockdep warning in bond_get_stats()
Taehee Yoo <[email protected]>
net: export netdev_next_lower_dev_rcu()
Taehee Yoo <[email protected]>
bonding: add missing netdev_update_lockdep_key()
Vasundhara Volam <[email protected]>
bnxt_en: Issue PCIe FLR in kdump kernel to cleanup pending DMAs.
Vasundhara Volam <[email protected]>
bnxt_en: Improve device shutdown method.
Xin Long <[email protected]>
sctp: move the format error check out of __sctp_sf_do_9_1_abort
Willem de Bruijn <[email protected]>
udp: rehash on disconnect
Paolo Abeni <[email protected]>
Revert "net: dev: introduce support for sch BYPASS for lockless qdisc"
Michal Kalderon <[email protected]>
qede: Fix race between rdma destroy workqueue and link change event
Dmitry Osipenko <[email protected]>
nfc: pn544: Fix occasional HW initialization failure
Rohit Maheshwari <[email protected]>
net/tls: Fix to avoid gettig invalid tls record
Jason Baron <[email protected]>
net: sched: correct flower port blocking
Arun Parameswaran <[email protected]>
net: phy: restore mdio regs in the iproc mdio driver
Horatiu Vultur <[email protected]>
net: mscc: fix in frame extraction
Alexandre Belloni <[email protected]>
net: macb: ensure interface is not suspended on at91rm9200
Jethro Beekman <[email protected]>
net: fib_rules: Correctly set table field when table number exceeds 8 bits
Florian Fainelli <[email protected]>
net: dsa: b53: Ensure the default VID is untagged
Aristeu Rozanski <[email protected]>
EDAC: skx_common: downgrade message importance on missing PCI device
Jens Axboe <[email protected]>
io_uring: grab ->fs as part of async offload
-------------
Diffstat:
Documentation/kbuild/makefiles.rst | 17 - Documentation/networking/nf_flowtable.txt | 2 +- Makefile | 108 +- arch/arm/boot/dts/stihxxx-b2120.dtsi | 2 +- arch/arm/kernel/ftrace.c | 7 +- arch/mips/kernel/vpe.c | 2 +- arch/x86/events/intel/core.c | 1 + arch/x86/events/intel/cstate.c | 22 +- arch/x86/events/msr.c | 3 +- arch/x86/kernel/cpu/resctrl/internal.h | 1 + arch/x86/kernel/cpu/resctrl/monitor.c | 4 +- arch/x86/kvm/svm.c | 43 + arch/x86/kvm/vmx/nested.c | 70 +- arch/x86/kvm/vmx/vmx.c | 15 + arch/x86/kvm/x86.c | 6 - drivers/acpi/acpi_watchdog.c | 3 +- drivers/bus/Kconfig | 1 - drivers/char/ipmi/ipmi_ssif.c | 10 +- drivers/cpufreq/cpufreq.c | 12 +- drivers/devfreq/devfreq.c | 4 +- drivers/edac/skx_common.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h | 1 + drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 37 +- drivers/gpu/drm/amd/display/dc/clk_mgr/Makefile | 6 + .../drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c | 6 + drivers/gpu/drm/amd/display/dc/dce/dce_aux.c | 2 +- drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c | 1 - .../gpu/drm/amd/display/dc/dcn21/dcn21_resource.c | 6 + .../drm/amd/include/asic_reg/dce/dce_12_0_offset.h | 2 + drivers/gpu/drm/i915/Kconfig.debug | 1 - drivers/gpu/drm/i915/gvt/dmabuf.c | 2 +- drivers/gpu/drm/i915/gvt/vgpu.c | 2 +- drivers/gpu/drm/msm/msm_drv.c | 8 + drivers/gpu/drm/radeon/radeon_drv.c | 43 +- drivers/gpu/drm/radeon/radeon_kms.c | 6 + drivers/hid/hid-alps.c | 2 +- drivers/hid/hid-core.c | 4 +- drivers/hid/hid-ite.c | 5 +- drivers/hid/usbhid/hiddev.c | 2 +- drivers/i2c/busses/i2c-altera.c | 2 +- drivers/i2c/busses/i2c-jz4780.c | 36 +- drivers/infiniband/hw/hns/hns_roce_device.h | 3 +- drivers/infiniband/hw/hns/hns_roce_hw_v1.c | 37 +- drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 80 +- drivers/infiniband/sw/siw/siw_cm.c | 5 +- drivers/macintosh/therm_windtunnel.c | 52 +- drivers/md/bcache/alloc.c | 18 +- drivers/md/bcache/btree.c | 13 + drivers/net/bonding/bond_main.c | 55 +- drivers/net/bonding/bond_options.c | 2 + drivers/net/dsa/b53/b53_common.c | 3 + drivers/net/ethernet/amazon/ena/ena_com.c | 96 +- drivers/net/ethernet/amazon/ena/ena_com.h | 9 + drivers/net/ethernet/amazon/ena/ena_ethtool.c | 46 +- drivers/net/ethernet/amazon/ena/ena_netdev.c | 6 +- drivers/net/ethernet/amazon/ena/ena_netdev.h | 2 + drivers/net/ethernet/apm/xgene/xgene_enet_main.c | 2 +- .../net/ethernet/aquantia/atlantic/aq_filters.c | 2 +- drivers/net/ethernet/aquantia/atlantic/aq_nic.c | 8 +- drivers/net/ethernet/aquantia/atlantic/aq_ring.c | 7 +- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 12 +- drivers/net/ethernet/cadence/macb_main.c | 6 +- .../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 22 +- drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c | 4 +- drivers/net/ethernet/intel/ice/ice_common.c | 17 +- drivers/net/ethernet/intel/ice/ice_hw_autogen.h | 6 + drivers/net/ethernet/mscc/ocelot_board.c | 8 + drivers/net/ethernet/qlogic/qede/qede.h | 2 + drivers/net/ethernet/qlogic/qede/qede_rdma.c | 29 +- drivers/net/hyperv/netvsc.c | 2 +- drivers/net/hyperv/netvsc_drv.c | 3 + drivers/net/phy/mdio-bcm-iproc.c | 20 + drivers/net/usb/qmi_wwan.c | 43 +- drivers/net/wireless/marvell/mwifiex/main.h | 13 - drivers/net/wireless/marvell/mwifiex/tdls.c | 75 +- drivers/nfc/pn544/i2c.c | 1 + drivers/nvme/host/core.c | 10 +- drivers/nvme/host/pci.c | 25 +- drivers/nvme/host/rdma.c | 2 +- drivers/nvme/host/tcp.c | 9 +- drivers/perf/arm_smmuv3_pmu.c | 2 +- drivers/pwm/pwm-omap-dmtimer.c | 21 +- drivers/s390/crypto/ap_bus.h | 4 +- drivers/s390/crypto/ap_card.c | 8 +- drivers/s390/crypto/ap_queue.c | 6 +- drivers/s390/crypto/zcrypt_api.c | 16 +- drivers/s390/net/qeth_l2_main.c | 29 +- drivers/soc/tegra/fuse/fuse-tegra30.c | 3 +- drivers/thermal/broadcom/brcmstb_thermal.c | 31 +- drivers/thermal/db8500_thermal.c | 4 +- drivers/vhost/net.c | 10 +- drivers/watchdog/wdat_wdt.c | 2 +- fs/ceph/file.c | 17 +- fs/cifs/cifsacl.c | 4 +- fs/cifs/connect.c | 2 +- fs/cifs/inode.c | 2 +- fs/dax.c | 3 + fs/ext4/super.c | 6 +- fs/f2fs/data.c | 32 +- fs/io_uring.c | 51 + fs/namei.c | 2 +- fs/nfs/nfs4file.c | 1 - fs/nfs/nfs4proc.c | 18 +- fs/ubifs/orphan.c | 4 +- fs/xfs/libxfs/xfs_attr.h | 7 +- fs/xfs/xfs_ioctl.c | 2 + fs/xfs/xfs_ioctl32.c | 2 + include/Kbuild | 1185 -------------------- include/acpi/actypes.h | 3 +- include/asm-generic/vdso/vsyscall.h | 4 +- include/linux/hid.h | 2 +- include/linux/netdevice.h | 7 +- include/linux/netfilter/ipset/ip_set.h | 11 +- include/linux/sched/nohz.h | 2 + include/net/flow_dissector.h | 9 + include/uapi/linux/usb/charger.h | 16 +- init/Kconfig | 22 +- kernel/audit.c | 40 +- kernel/auditfilter.c | 71 +- kernel/kprobes.c | 4 +- kernel/locking/lockdep_proc.c | 4 +- kernel/padata.c | 4 +- kernel/rcu/tree_exp.h | 11 +- kernel/sched/core.c | 31 +- kernel/sched/fair.c | 7 +- kernel/sched/loadavg.c | 33 +- kernel/time/vsyscall.c | 37 +- kernel/trace/trace.c | 2 + lib/Kconfig.debug | 11 - mm/debug.c | 8 +- mm/gup.c | 3 +- mm/huge_memory.c | 26 +- net/core/dev.c | 28 +- net/core/fib_rules.c | 2 +- net/ipv4/udp.c | 6 +- net/ipv6/ip6_fib.c | 7 +- net/ipv6/route.c | 1 + net/mac80211/mlme.c | 6 +- net/mac80211/util.c | 34 +- net/netfilter/ipset/ip_set_core.c | 34 +- net/netfilter/ipset/ip_set_hash_gen.h | 635 +++++++---- net/netfilter/nft_tunnel.c | 4 +- net/netfilter/xt_hashlimit.c | 12 +- net/netlink/af_netlink.c | 5 +- net/sched/cls_flower.c | 1 + net/sctp/sm_statefuns.c | 29 +- net/smc/af_smc.c | 2 + net/smc/smc_clc.c | 4 +- net/tls/tls_device.c | 20 +- net/wireless/ethtool.c | 8 +- net/wireless/nl80211.c | 5 +- scripts/Makefile.build | 9 - scripts/Makefile.headersinst | 18 - scripts/Makefile.lib | 18 +- security/integrity/ima/ima_policy.c | 44 +- tools/perf/ui/browsers/hists.c | 1 + tools/perf/ui/gtk/Build | 5 + tools/testing/selftests/ftrace/Makefile | 2 +- tools/testing/selftests/livepatch/Makefile | 2 + tools/testing/selftests/net/fib_tests.sh | 6 + tools/testing/selftests/rseq/Makefile | 2 + tools/testing/selftests/rtc/Makefile | 2 + usr/include/Makefile | 15 +- virt/kvm/kvm_main.c | 12 +- 165 files changed, 1820 insertions(+), 2267 deletions(-)
Greg Kroah-Hartman
maintainerPATCH 1/152 io_uring: grab ->fs as part of async offload
From: Jens Axboe <[email protected]>
[ Upstream commits 9392a27d88b9 and ff002b30181d ]
Ensure that the async work grabs ->fs from the queueing task if the punted commands needs to do lookups.
We don't have these two commits in 5.4-stable:
ff002b30181d30cdfbca316dadd099c3ca0d739c9392a27d88b9707145d713654eb26f0c29789e50
because they don't apply with the rework that was done in how io_uring handles offload. Since there's no io-wq in 5.4, it doesn't make sense to do two patches. I'm attaching my port of the two for 5.4-stable, it's been tested. Please queue it up for the next 5.4-stable, thanks!
--- fs/io_uring.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 46 insertions(+)
fs/io_uring.c +46 -0
@@ -71,6 +71,7 @@#include <linux/sizes.h> #include <linux/hugetlb.h> #include <linux/highmem.h> +#include <linux/fs_struct.h> #include <uapi/linux/io_uring.h>
@@ -334,6 +335,8 @@ struct io_kiocb {u32 result;u32 sequence; +struct fs_struct *fs; +struct work_struct work; };
@@ -651,6 +654,7 @@ static struct io_kiocb *io_get_req(struct io_ring_ctx *ctx,/* one is dropped after submission, the other at completion */ refcount_set (&req->refs,2 ); req->result =0 ; + req->fs =NULL ;return req; out:percpu_ref_put (&ctx->refs);
@@ -1663,6 +1667,16 @@ static int io_send_recvmsg(struct io_kiocb *req, const struct io_uring_sqe *sqe, ret = -EINTR ; } +if (req->fs) { +struct fs_struct *fs = req->fs; + +spin_lock (&req->fs->lock); +if (--fs->users) + fs =NULL ; +spin_unlock (&req->fs->lock); +if (fs) +free_fs_struct (fs); + }io_cqring_add_event (req->ctx, sqe->user_data, ret);io_put_req (req);return 0 ;
@@ -2159,6 +2173,7 @@ static inline bool io_sqe_needs_user(const struct io_uring_sqe *sqe)static void io_sq_wq_submit_work (struct work_struct *work) { struct io_kiocb *req = container_of(work, struct io_kiocb, work); +struct fs_struct *old_fs_struct = current->fs;struct io_ring_ctx *ctx = req->ctx;struct mm_struct *cur_mm =NULL ;struct async_list *async_list;
@@ -2178,6 +2193,15 @@ restart:/* Ensure we clear previously set non-block flag */ req->rw.ki_flags &= ~IOCB_NOWAIT ; +if (req->fs != current->fs && current->fs != old_fs_struct) { +task_lock (current); +if (req->fs) + current->fs = req->fs; +else + current->fs = old_fs_struct; +task_unlock (current); + } + ret =0 ;if (io_sqe_needs_user (sqe) && !cur_mm) {if (!mmget_not_zero (ctx->sqo_mm)) {
@@ -2276,6 +2300,11 @@ out:mmput (cur_mm); }revert_creds (old_cred); +if (old_fs_struct) { +task_lock (current); + current->fs = old_fs_struct; +task_unlock (current); + } }/*
@@ -2503,6 +2532,23 @@ err: req->user_data = s->sqe->user_data; +#if defined(CONFIG_NET ) +switch (READ_ONCE (s->sqe->opcode)) { +case IORING_OP_SENDMSG : +case IORING_OP_RECVMSG : +spin_lock (¤t->fs->lock); +if (!current->fs->in_exec) { + req->fs = current->fs; + req->fs->users++; + } +spin_unlock (¤t->fs->lock); +if (!req->fs) { + ret = -EAGAIN ; + goto err_req; + } + } +#endif +/* * If we already have a head request, queue this one for async * submittal once the head completes. If we don't have a head but
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 2/152 EDAC: skx_common: downgrade message importance on missing PCI device
In reply to Greg Kroah-HartmanFrom: Aristeu Rozanski <[email protected]>
[ Upstream commit 854bb48018d5da261d438b2232fa683bdb553979 ]
Both skx_edac and i10nm_edac drivers are loaded based on the matching CPU being available which leads the module to be automatically loaded in virtual machines as well. That will fail due the missing PCI devices. In both drivers the first function to make use of the PCI devices is skx_get_hi_lo() will simply print
EDAC skx: Can't get tolm/tohm
for each CPU core, which is noisy. This patch makes it a debug message.
--- drivers/edac/skx_common.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/edac/skx_common.c +1 -1
@@ -235,7 +235,7 @@ int skx_get_hi_lo(unsigned int did, int off[], u64 *tolm, u64 *tohm) pdev =pci_get_device (PCI_VENDOR_ID_INTEL , did,NULL );if (!pdev) { -skx_printk (KERN_ERR ,"Can't get tolm/tohm\n" ); +edac_dbg (2 ,"Can't get tolm/tohm\n" );return -ENODEV ; }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 3/152 net: dsa: b53: Ensure the default VID is untagged
In reply to Greg Kroah-HartmanFrom: Florian Fainelli <[email protected]>
[ Upstream commit d965a5432d4c3e6b9c3d2bc1d4a800013bbf76f6 ]
We need to ensure that the default VID is untagged otherwise the switch will be sending tagged frames and the results can be problematic. This is especially true with b53 switches that use VID 0 as their default VLAN since VID 0 has a special meaning.
--- drivers/net/dsa/b53/b53_common.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/net/dsa/b53/b53_common.c +3 -0
@@ -1353,6 +1353,9 @@ void b53_vlan_add(struct dsa_switch *ds,b53_get_vlan_entry (dev, vid, vl); +if (vid ==0 && vid ==b53_default_pvid (dev)) + untagged = true; + vl->members |=BIT (port);if (untagged && !dsa_is_cpu_port (ds, port)) vl->untag |=BIT (port);
Greg Kroah-Hartman
maintainerPATCH 4/152 net: fib_rules: Correctly set table field when table number exceeds 8 bits
In reply to Greg Kroah-HartmanFrom: Jethro Beekman <[email protected]>
[ Upstream commit 540e585a79e9d643ede077b73bcc7aa2d7b4d919 ]
In 709772e6e06564ed94ba740de70185ac3d792773, RT_TABLE_COMPAT was added to allow legacy software to deal with routing table numbers >= 256, but the same change to FIB rule queries was overlooked.
--- net/core/fib_rules.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
net/core/fib_rules.c +1 -1
@@ -967,7 +967,7 @@ static int fib_nl_fill_rule(struct sk_bu frh =nlmsg_data (nlh); frh->family = ops->family; - frh->table = rule->table; + frh->table = rule->table <256 ? rule->table :RT_TABLE_COMPAT ;if (nla_put_u32 (skb,FRA_TABLE , rule->table)) goto nla_put_failure;if (nla_put_u32 (skb,FRA_SUPPRESS_PREFIXLEN , rule->suppress_prefixlen))
Greg Kroah-Hartman
maintainerPATCH 5/152 net: macb: ensure interface is not suspended on at91rm9200
In reply to Greg Kroah-HartmanFrom: Alexandre Belloni <[email protected]>
[ Upstream commit e6a41c23df0d5da01540d2abef41591589c0b4be ]
Because of autosuspend, at91ether_start is called with clocks disabled. Ensure that pm_runtime doesn't suspend the interface as soon as it is opened as there is no pm_runtime support is the other relevant parts of the platform support for at91rm9200.
--- drivers/net/ethernet/cadence/macb_main.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/net/ethernet/cadence/macb_main.c +5 -1
@@ -3690,6 +3690,10 @@ static int at91ether_open(struct net_devu32 ctl;int ret; + ret =pm_runtime_get_sync (&lp->pdev->dev); +if (ret <0 ) +return ret; +/* Clear internal statistics */ ctl =macb_readl (lp,NCR );macb_writel (lp,NCR , ctl |MACB_BIT (CLRSTAT ));
@@ -3750,7 +3754,7 @@ static int at91ether_close(struct net_de q->rx_buffers, q->rx_buffers_dma); q->rx_buffers =NULL ; -return 0 ; +return pm_runtime_put (&lp->pdev->dev); }/* Transmit packet */
Greg Kroah-Hartman
maintainerPATCH 6/152 net: mscc: fix in frame extraction
In reply to Greg Kroah-HartmanFrom: Horatiu Vultur <[email protected]>
[ Upstream commit a81541041ceb55bcec9a8bb8ad3482263f0a205a ]
Each extracted frame on Ocelot has an IFH. The frame and IFH are extracted by reading chuncks of 4 bytes from a register.
In case the IFH and frames were read corretly it would try to read the next frame. In case there are no more frames in the queue, it checks if there were any previous errors and in that case clear the queue. But this check will always succeed also when there are no errors. Because when extracting the IFH the error is checked against 4(number of bytes read) and then the error is set only if the extraction of the frame failed. So in a happy case where there are no errors the err variable is still 4. So it could be a case where after the check that there are no more frames in the queue, a frame will arrive in the queue but because the error is not reseted, it would try to flush the queue. So the frame will be lost.
The fix consist in resetting the error after reading the IFH.
--- drivers/net/ethernet/mscc/ocelot_board.c | 8 ++++++++ 1 file changed, 8 insertions(+)
drivers/net/ethernet/mscc/ocelot_board.c +8 -0
@@ -112,6 +112,14 @@ static irqreturn_t ocelot_xtr_irq_handleif (err !=4 )break ; +/* At this point the IFH was read correctly, so it is safe to +* presume that there is no error. The err needs to be reset +* otherwise a frame could come in CPU queue between the while +* condition and the check for error later on. And in that case +* the new frame is just removed and not processed. +*/ + err =0 ; +ocelot_parse_ifh (ifh, &info); dev = ocelot->ports[info.port]->dev;
Greg Kroah-Hartman
maintainerPATCH 7/152 net: phy: restore mdio regs in the iproc mdio driver
In reply to Greg Kroah-HartmanFrom: Arun Parameswaran <[email protected]>
commit 6f08e98d62799e53c89dbf2c9a49d77e20ca648c upstream.
The mii management register in iproc mdio block
does not have a retention register so it is lost on suspend. Save and restore value of register while resuming from suspend.
--- drivers/net/phy/mdio-bcm-iproc.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+)
drivers/net/phy/mdio-bcm-iproc.c +20 -0
@@ -178,6 +178,23 @@ static int iproc_mdio_remove(struct platreturn 0 ; } +#ifdef CONFIG_PM_SLEEP +int iproc_mdio_resume (struct device *dev) +{ +struct platform_device *pdev =to_platform_device (dev); +struct iproc_mdio_priv *priv =platform_get_drvdata (pdev); + +/* restore the mii clock configuration */ +iproc_mdio_config_clk (priv->base); + +return 0 ; +} + +static const struct dev_pm_ops iproc_mdio_pm_ops = { + .resume = iproc_mdio_resume +}; +#endif /* CONFIG_PM_SLEEP */ +static const struct of_device_id iproc_mdio_of_match[] = { { .compatible ="brcm,iproc-mdio" , }, {/* sentinel */ },
@@ -188,6 +205,9 @@ static struct platform_driver iproc_mdio .driver = { .name = "iproc-mdio", .of_match_table = iproc_mdio_of_match, +#ifdef CONFIG_PM_SLEEP + .pm = &iproc_mdio_pm_ops, +#endif }, .probe = iproc_mdio_probe, .remove = iproc_mdio_remove,
Greg Kroah-Hartman
maintainerPATCH 8/152 net: sched: correct flower port blocking
In reply to Greg Kroah-HartmanFrom: Jason Baron <[email protected]>
[ Upstream commit 8a9093c79863b58cc2f9874d7ae788f0d622a596 ]
tc flower rules that are based on src or dst port blocking are sometimes ineffective due to uninitialized stack data. __skb_flow_dissect() extracts ports from the skb for tc flower to match against. However, the port dissection is not done when when the FLOW_DIS_IS_FRAGMENT bit is set in key_control->flags. All callers of __skb_flow_dissect(), zero-out the key_control field except for fl_classify() as used by the flower classifier. Thus, the FLOW_DIS_IS_FRAGMENT may be set on entry to __skb_flow_dissect(), since key_control is allocated on the stack and may not be initialized.
Since key_basic and key_control are present for all flow keys, let's make sure they are initialized.
--- include/net/flow_dissector.h | 9 +++++++++ net/sched/cls_flower.c | 1 + 2 files changed, 10 insertions(+)
include/net/flow_dissector.h +9 -0
@@ -5,6 +5,7 @@#include <linux/types.h> #include <linux/in6.h> #include <linux/siphash.h> +#include <linux/string.h> #include <uapi/linux/if_ether.h> /**
@@ -338,4 +339,12 @@ struct bpf_flow_dissector { void *data_end; }; +static inline void +flow_dissector_init_keys (struct flow_dissector_key_control *key_control, +struct flow_dissector_key_basic *key_basic) +{ +memset (key_control,0 ,sizeof (*key_control)); +memset (key_basic,0 ,sizeof (*key_basic)); +} +#endif
net/sched/cls_flower.c +1 -0
@@ -303,6 +303,7 @@ static int fl_classify(struct sk_buff *sstruct cls_fl_filter *f;list_for_each_entry_rcu (mask, &head->masks, list) { +flow_dissector_init_keys (&skb_key.control, &skb_key.basic);fl_clear_masked_range (&skb_key, mask);skb_flow_dissect_meta (skb, &mask->dissector, &skb_key);
Greg Kroah-Hartman
maintainerPATCH 9/152 net/tls: Fix to avoid gettig invalid tls record
In reply to Greg Kroah-HartmanFrom: Rohit Maheshwari <[email protected]>
[ Upstream commit 06f5201c6392f998a49ca9c9173e2930c8eb51d8 ]
Current code doesn't check if tcp sequence number is starting from (/after) 1st record's start sequnce number. It only checks if seq number is before 1st record's end sequnce number. This problem will always be a possibility in re-transmit case. If a record which belongs to a requested seq number is already deleted, tls_get_record will start looking into list and as per the check it will look if seq number is before the end seq of 1st record, which will always be true and will return 1st record always, it should in fact return NULL.
As part of the fix, start looking each record only if the sequence number lies in the list else return NULL.
There is one more check added, driver look for the start marker record to handle tcp packets which are before the tls offload start sequence number, hence return 1st record if the record is tls start marker and seq number is before the 1st record's starting sequence number.
--- net/tls/tls_device.c | 20 +++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-)
net/tls/tls_device.c +19 -1
@@ -581,7 +581,7 @@ struct tls_record_info *tls_get_record(s u32 seq, u64 *p_record_sn) { u64 record_sn = context->hint_record_sn; -struct tls_record_info *info; +struct tls_record_info *info, *last; info = context->retransmit_hint; if (!info ||
@@ -593,6 +593,24 @@ struct tls_record_info *tls_get_record(s struct tls_record_info, list);if (!info)return NULL ; +/* send the start_marker record if seq number is before the +* tls offload start marker sequence number. This record is +* required to handle TCP packets which are before TLS offload +* started. +* And if it's not start marker, look if this seq number +* belongs to the list. +*/ +if (likely (!tls_record_is_start_marker (info))) { +/* we have the first record, get the last record to see +* if this seq number belongs to the list. +*/ + last =list_last_entry (&context->records_list, + struct tls_record_info, list); + +if (!between (seq,tls_record_start_seq (info), + last->end_seq)) +return NULL ; + } record_sn = context->unacked_record_sn; }
Greg Kroah-Hartman
maintainerPATCH 10/152 nfc: pn544: Fix occasional HW initialization failure
In reply to Greg Kroah-HartmanFrom: Dmitry Osipenko <[email protected]>
[ Upstream commit c3331d2fe3fd4d5e321f2467d01f72de7edfb5d0 ]
The PN544 driver checks the "enable" polarity during of driver's probe and it's doing that by turning ON and OFF NFC with different polarities until enabling succeeds. It takes some time for the hardware to power-down, and thus, to deassert the IRQ that is raised by turning ON the hardware. Since the delay after last power-down of the polarity-checking process is missed in the code, the interrupt may trigger immediately after installing the IRQ handler (right after the checking is done), which results in IRQ handler trying to touch the disabled HW and ends with marking NFC as 'DEAD' during of the driver's probe:
pn544_hci_i2c 1-002a: NFC: nfc_en polarity : active high pn544_hci_i2c 1-002a: NFC: invalid len byte
shdlc: llc_shdlc_recv_frame: NULL Frame -> link is dead
This patch fixes the occasional NFC initialization failure on Nexus 7 device.
--- drivers/nfc/pn544/i2c.c | 1 + 1 file changed, 1 insertion(+)
drivers/nfc/pn544/i2c.c +1 -0
@@ -225,6 +225,7 @@ static void pn544_hci_i2c_platform_init( out:gpiod_set_value_cansleep (phy->gpiod_en, !phy->en_polarity); +usleep_range (10000 ,15000 ); }static void pn544_hci_i2c_enable_mode (struct pn544_i2c_phy *phy,int run_mode)
Greg Kroah-Hartman
maintainerPATCH 11/152 qede: Fix race between rdma destroy workqueue and link change event
In reply to Greg Kroah-HartmanFrom: Michal Kalderon <[email protected]>
[ Upstream commit af6565adb02d3129d3fae4d9d5da945abaf4417a ]
If an event is added while the rdma workqueue is being destroyed it could lead to several races, list corruption, null pointer dereference during queue_work or init_queue.
This fixes the race between the two flows which can occur during shutdown.
A kref object and a completion object are added to the rdma_dev structure, these are initialized before the workqueue is created. The refcnt is used to indicate work is being added to the workqueue and ensures the cleanup flow won't start while we're in the middle of adding the event.
Once the work is added, the refcnt is decreased and the cleanup flow is safe to run.
--- drivers/net/ethernet/qlogic/qede/qede.h | 2 + drivers/net/ethernet/qlogic/qede/qede_rdma.c | 29 ++++++++++++++++++++++++++- 2 files changed, 30 insertions(+), 1 deletion(-)
drivers/net/ethernet/qlogic/qede/qede.h +2 -0
@@ -163,6 +163,8 @@ struct qede_rdma_dev {struct list_head entry;struct list_head rdma_event_list;struct workqueue_struct *rdma_wq; +struct kref refcnt; +struct completion event_comp;bool exp_recovery; };
drivers/net/ethernet/qlogic/qede/qede_rdma.c +28 -1
@@ -59,6 +59,9 @@ static void _qede_rdma_dev_add(struct qestatic int qede_rdma_create_wq (struct qede_dev *edev) {INIT_LIST_HEAD (&edev->rdma_info.rdma_event_list); +kref_init (&edev->rdma_info.refcnt); +init_completion (&edev->rdma_info.event_comp); + edev->rdma_info.rdma_wq =create_singlethread_workqueue ("rdma_wq" );if (!edev->rdma_info.rdma_wq) {DP_NOTICE (edev,"qedr: Could not create workqueue\n" );
@@ -83,8 +86,23 @@ static void qede_rdma_cleanup_event(stru } } +static void qede_rdma_complete_event (struct kref *ref) +{ +struct qede_rdma_dev *rdma_dev = + container_of(ref, struct qede_rdma_dev, refcnt); + +/* no more events will be added after this */ +complete (&rdma_dev->event_comp); +} +static void qede_rdma_destroy_wq (struct qede_dev *edev) { +/* Avoid race with add_event flow, make sure it finishes before +* we start accessing the list and cleaning up the work +*/ +kref_put (&edev->rdma_info.refcnt, qede_rdma_complete_event); +wait_for_completion (&edev->rdma_info.event_comp); +qede_rdma_cleanup_event (edev);destroy_workqueue (edev->rdma_info.rdma_wq); }
@@ -310,15 +328,24 @@ static void qede_rdma_add_event(struct qif (!edev->rdma_info.qedr_dev)return ; +/* We don't want the cleanup flow to start while we're allocating and +* scheduling the work +*/ +if (!kref_get_unless_zero (&edev->rdma_info.refcnt)) +return ;/* already being destroyed */ + event_node =qede_rdma_get_free_event_node (edev);if (!event_node) -return ; + goto out; event_node->event = event; event_node->ptr = edev;INIT_WORK (&event_node->work, qede_rdma_handle_event);queue_work (edev->rdma_info.rdma_wq, &event_node->work); + +out: +kref_put (&edev->rdma_info.refcnt, qede_rdma_complete_event); }void qede_rdma_dev_event_open (struct qede_dev *edev)
Greg Kroah-Hartman
maintainerPATCH 12/152 Revert "net: dev: introduce support for sch BYPASS for lockless qdisc"
In reply to Greg Kroah-HartmanFrom: Paolo Abeni <[email protected]>
[ Upstream commit 379349e9bc3b42b8b2f8f7a03f64a97623fff323 ]
This reverts commit ba27b4cdaaa66561aaedb2101876e563738d36fe
Ahmed reported ouf-of-order issues bisected to commit ba27b4cdaaa6 ("net: dev: introduce support for sch BYPASS for lockless qdisc"). I can't find any working solution other than a plain revert.
This will introduce some minor performance regressions for pfifo_fast qdisc. I plan to address them in net-next with more indirect call wrapper boilerplate for qdiscs.
--- net/core/dev.c | 22 ++-------------------- 1 file changed, 2 insertions(+), 20 deletions(-)
net/core/dev.c +2 -20
@@ -3386,26 +3386,8 @@ static inline int __dev_xmit_skb(structqdisc_calculate_pkt_len (skb, q);if (q->flags &TCQ_F_NOLOCK ) { -if ((q->flags &TCQ_F_CAN_BYPASS ) &&READ_ONCE (q->empty) && -qdisc_run_begin (q)) { -if (unlikely (test_bit (__QDISC_STATE_DEACTIVATED, - &q->state))) { -__qdisc_drop (skb, &to_free); - rc =NET_XMIT_DROP ; - goto end_run; - } -qdisc_bstats_cpu_update (q, skb); - - rc =NET_XMIT_SUCCESS ; -if (sch_direct_xmit (skb, q, dev, txq,NULL , true)) -__qdisc_run (q); - -end_run: -qdisc_run_end (q); - }else { - rc = q->enqueue (skb, q, &to_free) &NET_XMIT_MASK ; -qdisc_run (q); - } + rc = q->enqueue (skb, q, &to_free) &NET_XMIT_MASK ; +qdisc_run (q);if (unlikely (to_free))kfree_skb_list (to_free);
Greg Kroah-Hartman
maintainerPATCH 13/152 udp: rehash on disconnect
In reply to Greg Kroah-HartmanFrom: Willem de Bruijn <[email protected]>
[ Upstream commit 303d0403b8c25e994e4a6e45389e173cf8706fb5 ]
As of the below commit, udp sockets bound to a specific address can coexist with one bound to the any addr for the same port.
The commit also phased out the use of socket hashing based only on port (hslot), in favor of always hashing on {addr, port} (hslot2).
The change broke the following behavior with disconnect (AF_UNSPEC):
server binds to 0.0.0.0:1337
server connects to 127.0.0.1:80
server disconnects
client connects to 127.0.0.1:1337
client sends "hello"
server reads "hello" // times out, packet did not find skOn connect the server acquires a specific source addr suitable for routing to its destination. On disconnect it reverts to the any addr.
The connect call triggers a rehash to a different hslot2. On disconnect, add the same to return to the original hslot2.
Skip this step if the socket is going to be unhashed completely.
--- net/ipv4/udp.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
net/ipv4/udp.c +5 -1
@@ -1856,8 +1856,12 @@ int __udp_disconnect(struct sock *sk, in inet->inet_dport =0 ;sock_rps_reset_rxhash (sk); sk->sk_bound_dev_if =0 ; -if (!(sk->sk_userlocks &SOCK_BINDADDR_LOCK )) +if (!(sk->sk_userlocks &SOCK_BINDADDR_LOCK )) {inet_reset_saddr (sk); +if (sk->sk_prot->rehash && + (sk->sk_userlocks &SOCK_BINDPORT_LOCK )) + sk->sk_prot->rehash (sk); + }if (!(sk->sk_userlocks &SOCK_BINDPORT_LOCK )) { sk->sk_prot->unhash (sk);
Greg Kroah-Hartman
maintainerPATCH 14/152 sctp: move the format error check out of __sctp_sf_do_9_1_abort
In reply to Greg Kroah-HartmanFrom: Xin Long <[email protected]>
[ Upstream commit 245709ec8be89af46ea7ef0444c9c80913999d99 ]
When T2 timer is to be stopped, the asoc should also be deleted, otherwise, there will be no chance to call sctp_association_free and the asoc could last in memory forever.
However, in sctp_sf_shutdown_sent_abort(), after adding the cmd SCTP_CMD_TIMER_STOP for T2 timer, it may return error due to the format error from __sctp_sf_do_9_1_abort() and miss adding SCTP_CMD_ASSOC_FAILED where the asoc will be deleted.
This patch is to fix it by moving the format error check out of __sctp_sf_do_9_1_abort(), and do it before adding the cmd SCTP_CMD_TIMER_STOP for T2 timer.
Thanks Hangbin for reporting this issue by the fuzz testing.
v1->v2:
- improve the comment in the code as Marcelo's suggestion.
--- net/sctp/sm_statefuns.c | 29 ++++++++++++++++++++--------- 1 file changed, 20 insertions(+), 9 deletions(-)
net/sctp/sm_statefuns.c +20 -9
@@ -170,6 +170,16 @@ static inline bool sctp_chunk_length_valreturn true; } +/* Check for format error in an ABORT chunk */ +static inline bool sctp_err_chunk_valid (struct sctp_chunk *chunk) +{ +struct sctp_errhdr *err; + +sctp_walk_errors (err, chunk->chunk_hdr); + +return (void *)err == (void *)chunk->chunk_end; +} +/********************************************************** * These are the state functions for handling chunk events. **********************************************************/
@@ -2255,6 +2265,9 @@ enum sctp_disposition sctp_sf_shutdown_p sctp_bind_addr_state(&asoc->base.bind_addr, &chunk->dest))return sctp_sf_discard_chunk (net, ep, asoc, type, arg, commands); +if (!sctp_err_chunk_valid (chunk)) +return sctp_sf_pdiscard (net, ep, asoc, type, arg, commands); +return __sctp_sf_do_9_1_abort (net, ep, asoc, type, arg, commands); }
@@ -2298,6 +2311,9 @@ enum sctp_disposition sctp_sf_shutdown_s sctp_bind_addr_state(&asoc->base.bind_addr, &chunk->dest))return sctp_sf_discard_chunk (net, ep, asoc, type, arg, commands); +if (!sctp_err_chunk_valid (chunk)) +return sctp_sf_pdiscard (net, ep, asoc, type, arg, commands); +/* Stop the T2-shutdown timer. */ sctp_add_cmd_sf (commands,SCTP_CMD_TIMER_STOP ,SCTP_TO (SCTP_EVENT_TIMEOUT_T2_SHUTDOWN ));
@@ -2565,6 +2581,9 @@ enum sctp_disposition sctp_sf_do_9_1_abo sctp_bind_addr_state(&asoc->base.bind_addr, &chunk->dest))return sctp_sf_discard_chunk (net, ep, asoc, type, arg, commands); +if (!sctp_err_chunk_valid (chunk)) +return sctp_sf_pdiscard (net, ep, asoc, type, arg, commands); +return __sctp_sf_do_9_1_abort (net, ep, asoc, type, arg, commands); }
@@ -2582,16 +2601,8 @@ static enum sctp_disposition __sctp_sf_d/* See if we have an error cause code in the chunk. */ len =ntohs (chunk->chunk_hdr->length); -if (len >=sizeof (struct sctp_chunkhdr ) +sizeof (struct sctp_errhdr )) { -struct sctp_errhdr *err; - -sctp_walk_errors (err, chunk->chunk_hdr); -if ((void *)err != (void *)chunk->chunk_end) -return sctp_sf_pdiscard (net, ep, asoc, type, arg, - commands); - +if (len >=sizeof (struct sctp_chunkhdr ) +sizeof (struct sctp_errhdr )) error = ((struct sctp_errhdr *)chunk->skb->data)->cause; - }sctp_add_cmd_sf (commands,SCTP_CMD_SET_SK_ERR ,SCTP_ERROR (ECONNRESET ));/* ASSOC_FAILED will DELETE_TCB. */
Greg Kroah-Hartman
maintainerPATCH 15/152 bnxt_en: Improve device shutdown method.
In reply to Greg Kroah-HartmanFrom: Vasundhara Volam <[email protected]>
[ Upstream commit 5567ae4a8d569d996d0d88d0eceb76205e4c7ce5 ]
Especially when bnxt_shutdown() is called during kexec, we need to disable MSIX and disable Bus Master to completely quiesce the device. Make these 2 calls unconditionally in the shutdown method.
--- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/broadcom/bnxt/bnxt.c +2 -2
@@ -11908,10 +11908,10 @@ static void bnxt_shutdown(struct pci_devdev_close (dev);bnxt_ulp_shutdown (bp); +bnxt_clear_int_mode (bp); +pci_disable_device (pdev);if (system_state ==SYSTEM_POWER_OFF ) { -bnxt_clear_int_mode (bp); -pci_disable_device (pdev);pci_wake_from_d3 (pdev, bp->wol);pci_set_power_state (pdev, PCI_D3hot); }
Greg Kroah-Hartman
maintainerPATCH 16/152 bnxt_en: Issue PCIe FLR in kdump kernel to cleanup pending DMAs.
In reply to Greg Kroah-HartmanFrom: Vasundhara Volam <[email protected]>
[ Upstream commit 8743db4a9acfd51f805ac0c87bcaae92c42d1061 ]
If crashed kernel does not shutdown the NIC properly, PCIe FLR is required in the kdump kernel in order to initialize all the functions properly.
--- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 8 ++++++++ 1 file changed, 8 insertions(+)
drivers/net/ethernet/broadcom/bnxt/bnxt.c +8 -0
@@ -11712,6 +11712,14 @@ static int bnxt_init_one(struct pci_devif (version_printed++ ==0 )pr_info ("%s" , version); +/* Clear any pending DMA transactions from crash kernel +* while loading driver in capture kernel. +*/ +if (is_kdump_kernel ()) { +pci_clear_master (pdev); +pcie_flr (pdev); + } + max_irqs =bnxt_get_max_irq (pdev); dev =alloc_etherdev_mq (sizeof (*bp), max_irqs);if (!dev)
Greg Kroah-Hartman
maintainerPATCH 17/152 bonding: add missing netdev_update_lockdep_key()
In reply to Greg Kroah-HartmanFrom: Taehee Yoo <[email protected]>
[ Upstream commit 064ff66e2bef84f1153087612032b5b9eab005bd ]
After bond_release(), netdev_update_lockdep_key() should be called. But both ioctl path and attribute path don't call netdev_update_lockdep_key().
This patch adds missing netdev_update_lockdep_key().
Test commands:
ip link add bond0 type bond
ip link add bond1 type bond
ifenslave bond0 bond1
ifenslave -d bond0 bond1
ifenslave bond1 bond0
Splat looks like:
[ 29.501182][ T1046] WARNING: possible circular locking dependency detected
[ 29.501945][ T1039] hardirqs last disabled at (1962): [<ffffffffac6c807f>] handle_mm_fault+0x13f/0x700
[ 29.503442][ T1046] 5.5.0+ #322 Not tainted
[ 29.503447][ T1046] ------------------------------------------------------
[ 29.504277][ T1039] softirqs last enabled at (1180): [<ffffffffade00678>] __do_softirq+0x678/0x981
[ 29.505443][ T1046] ifenslave/1046 is trying to acquire lock:
[ 29.505886][ T1039] softirqs last disabled at (1169): [<ffffffffac19c18a>] irq_exit+0x17a/0x1a0
[ 29.509997][ T1046] ffff88805d5da280 (&dev->addr_list_lock_key#3){+...}, at: dev_mc_sync_multiple+0x95/0x120
[ 29.511243][ T1046]
[ 29.511243][ T1046] but task is already holding lock:
[ 29.512192][ T1046] ffff8880460f2280 (&dev->addr_list_lock_key#4){+...}, at: bond_enslave+0x4482/0x47b0 [bonding]
[ 29.514124][ T1046]
[ 29.514124][ T1046] which lock already depends on the new lock.
[ 29.514124][ T1046]
[ 29.517297][ T1046]
[ 29.517297][ T1046] the existing dependency chain (in reverse order) is:
[ 29.518231][ T1046]
[ 29.518231][ T1046] -> #1 (&dev->addr_list_lock_key#4){+...}:
[ 29.519076][ T1046] _raw_spin_lock+0x30/0x70
[ 29.519588][ T1046] dev_mc_sync_multiple+0x95/0x120
[ 29.520208][ T1046] bond_enslave+0x448d/0x47b0 [bonding]
[ 29.520862][ T1046] bond_option_slaves_set+0x1a3/0x370 [bonding]
[ 29.521640][ T1046] __bond_opt_set+0x1ff/0xbb0 [bonding]
[ 29.522438][ T1046] __bond_opt_set_notify+0x2b/0xf0 [bonding]
[ 29.523251][ T1046] bond_opt_tryset_rtnl+0x92/0xf0 [bonding]
[ 29.524082][ T1046] bonding_sysfs_store_option+0x8a/0xf0 [bonding]
[ 29.524959][ T1046] kernfs_fop_write+0x276/0x410
[ 29.525620][ T1046] vfs_write+0x197/0x4a0
[ 29.526218][ T1046] ksys_write+0x141/0x1d0
[ 29.526818][ T1046] do_syscall_64+0x99/0x4f0
[ 29.527430][ T1046] entry_SYSCALL_64_after_hwframe+0x49/0xbe
[ 29.528265][ T1046]
[ 29.528265][ T1046] -> #0 (&dev->addr_list_lock_key#3){+...}:
[ 29.529272][ T1046] __lock_acquire+0x2d8d/0x3de0
[ 29.529935][ T1046] lock_acquire+0x164/0x3b0
[ 29.530638][ T1046] _raw_spin_lock+0x30/0x70
[ 29.531187][ T1046] dev_mc_sync_multiple+0x95/0x120
[ 29.531790][ T1046] bond_enslave+0x448d/0x47b0 [bonding]
[ 29.532451][ T1046] bond_option_slaves_set+0x1a3/0x370 [bonding]
[ 29.533163][ T1046] __bond_opt_set+0x1ff/0xbb0 [bonding]
[ 29.533789][ T1046] __bond_opt_set_notify+0x2b/0xf0 [bonding]
[ 29.534595][ T1046] bond_opt_tryset_rtnl+0x92/0xf0 [bonding]
[ 29.535500][ T1046] bonding_sysfs_store_option+0x8a/0xf0 [bonding]
[ 29.536379][ T1046] kernfs_fop_write+0x276/0x410
[ 29.537057][ T1046] vfs_write+0x197/0x4a0
[ 29.537640][ T1046] ksys_write+0x141/0x1d0
[ 29.538251][ T1046] do_syscall_64+0x99/0x4f0
[ 29.538870][ T1046] entry_SYSCALL_64_after_hwframe+0x49/0xbe
[ 29.539659][ T1046]
[ 29.539659][ T1046] other info that might help us debug this:
[ 29.539659][ T1046]
[ 29.540953][ T1046] Possible unsafe locking scenario:
[ 29.540953][ T1046]
[ 29.541883][ T1046] CPU0 CPU1
[ 29.542540][ T1046] ---- ----
[ 29.543209][ T1046] lock(&dev->addr_list_lock_key#4);
[ 29.543880][ T1046] lock(&dev->addr_list_lock_key#3);
[ 29.544873][ T1046] lock(&dev->addr_list_lock_key#4);
[ 29.545863][ T1046] lock(&dev->addr_list_lock_key#3);
[ 29.546525][ T1046]
[ 29.546525][ T1046] *** DEADLOCK ***
[ 29.546525][ T1046]
[ 29.547542][ T1046] 5 locks held by ifenslave/1046:
[ 29.548196][ T1046] #0: ffff88806044c478 (sb_writers#5){.+.+}, at: vfs_write+0x3bb/0x4a0
[ 29.549248][ T1046] #1: ffff88805af00890 (&of->mutex){+.+.}, at: kernfs_fop_write+0x1cf/0x410
[ 29.550343][ T1046] #2: ffff88805b8b54b0 (kn->count#157){.+.+}, at: kernfs_fop_write+0x1f2/0x410
[ 29.551575][ T1046] #3: ffffffffaecf4cf0 (rtnl_mutex){+.+.}, at: bond_opt_tryset_rtnl+0x5f/0xf0 [bonding]
[ 29.552819][ T1046] #4: ffff8880460f2280 (&dev->addr_list_lock_key#4){+...}, at: bond_enslave+0x4482/0x47b0 [bonding]
[ 29.554175][ T1046]
[ 29.554175][ T1046] stack backtrace:
[ 29.554907][ T1046] CPU: 0 PID: 1046 Comm: ifenslave Not tainted 5.5.0+ #322
[ 29.555854][ T1046] Hardware name: innotek GmbH VirtualBox/VirtualBox, BIOS VirtualBox 12/01/2006
[ 29.557064][ T1046] Call Trace:
[ 29.557504][ T1046] dump_stack+0x96/0xdb
[ 29.558054][ T1046] check_noncircular+0x371/0x450
[ 29.558723][ T1046] ? print_circular_bug.isra.35+0x310/0x310
[ 29.559486][ T1046] ? hlock_class+0x130/0x130
[ 29.560100][ T1046] ? __lock_acquire+0x2d8d/0x3de0
[ 29.560761][ T1046] __lock_acquire+0x2d8d/0x3de0
[ 29.561366][ T1046] ? register_lock_class+0x14d0/0x14d0
[ 29.562045][ T1046] ? find_held_lock+0x39/0x1d0
[ 29.562641][ T1046] lock_acquire+0x164/0x3b0
[ 29.563199][ T1046] ? dev_mc_sync_multiple+0x95/0x120
[ 29.563872][ T1046] _raw_spin_lock+0x30/0x70
[ 29.564464][ T1046] ? dev_mc_sync_multiple+0x95/0x120
[ 29.565146][ T1046] dev_mc_sync_multiple+0x95/0x120
[ 29.565793][ T1046] bond_enslave+0x448d/0x47b0 [bonding]
[ 29.566487][ T1046] ? bond_update_slave_arr+0x940/0x940 [bonding]
[ 29.567279][ T1046] ? bstr_printf+0xc20/0xc20
[ 29.567857][ T1046] ? stack_trace_consume_entry+0x160/0x160
[ 29.568614][ T1046] ? deactivate_slab.isra.77+0x2c5/0x800
[ 29.569320][ T1046] ? check_chain_key+0x236/0x5d0
[ 29.569939][ T1046] ? sscanf+0x93/0xc0
[ 29.570442][ T1046] ? vsscanf+0x1e20/0x1e20
[ 29.571003][ T1046] bond_option_slaves_set+0x1a3/0x370 [bonding][ ... ]
--- drivers/net/bonding/bond_main.c | 2 ++ drivers/net/bonding/bond_options.c | 2 ++ 2 files changed, 4 insertions(+)
drivers/net/bonding/bond_main.c +2 -0
@@ -3550,6 +3550,8 @@ static int bond_do_ioctl(struct net_devicase BOND_RELEASE_OLD :case SIOCBONDRELEASE : res =bond_release (bond_dev, slave_dev); +if (!res) +netdev_update_lockdep_key (slave_dev);break ;case BOND_SETHWADDR_OLD :case SIOCBONDSETHWADDR :
drivers/net/bonding/bond_options.c +2 -0
@@ -1398,6 +1398,8 @@ static int bond_option_slaves_set(structcase '-' :slave_dbg (bond->dev, dev,"Releasing interface\n" ); ret =bond_release (bond->dev, dev); +if (!ret) +netdev_update_lockdep_key (dev);break ;default :
Greg Kroah-Hartman
maintainerPATCH 18/152 net: export netdev_next_lower_dev_rcu()
In reply to Greg Kroah-HartmanFrom: Taehee Yoo <[email protected]>
[ Upstream commit 7151affeef8d527f50b4b68a871fd28bd660023f ]
netdev_next_lower_dev_rcu() will be used to implement a function, which is to walk all lower interfaces.
There are already functions that they walk their lower interface. (netdev_walk_all_lower_dev_rcu, netdev_walk_all_lower_dev()). But, there would be cases that couldn't be covered by given netdev_walk_all_lower_dev_{rcu}() function.
So, some modules would want to implement own function, which is to walk all lower interfaces.
In the next patch, netdev_next_lower_dev_rcu() will be used. In addition, this patch removes two unused prototypes in netdevice.h.
--- include/linux/netdevice.h | 7 +++---- net/core/dev.c | 6 +++--- 2 files changed, 6 insertions(+), 7 deletions(-)
include/linux/netdevice.h +3 -4
@@ -72,6 +72,8 @@ void netdev_set_default_ethtool_ops(stru#define NET_RX_SUCCESS 0/* keep 'em coming, baby */ #define NET_RX_DROP 1/* packet dropped */ +#define MAX_NEST_DEV 8 +/* * Transmit return codes: transmit return codes originate from three different * namespaces:
@@ -4294,11 +4296,8 @@ void *netdev_lower_get_next(struct net_dldev ; \ ldev = netdev_lower_get_next(dev, &(iter))) -struct net_device *netdev_all_lower_get_next (struct net_device *dev, +struct net_device *netdev_next_lower_dev_rcu (struct net_device *dev,struct list_head **iter); -struct net_device *netdev_all_lower_get_next_rcu (struct net_device *dev, -struct list_head **iter); - int netdev_walk_all_lower_dev(struct net_device *dev, int (*fn)(struct net_device *lower_dev, void *data),
net/core/dev.c +3 -3
@@ -146,7 +146,6 @@#include "net-sysfs.h" #define MAX_GRO_SKBS 8 -#define MAX_NEST_DEV 8/* This should be increased if a protocol with a bigger head is added. */ #define GRO_MAX_HEAD (MAX_HEADER + 128)
@@ -6914,8 +6913,8 @@ static int __netdev_walk_all_lower_dev(sreturn 0 ; } -static struct net_device *netdev_next_lower_dev_rcu (struct net_device *dev, -struct list_head **iter) +struct net_device *netdev_next_lower_dev_rcu (struct net_device *dev, +struct list_head **iter) {struct netdev_adjacent *lower;
@@ -6927,6 +6926,7 @@ static struct net_device *netdev_next_loreturn lower->dev; } +EXPORT_SYMBOL (netdev_next_lower_dev_rcu);static u8 __netdev_upper_depth (struct net_device *dev) {
Greg Kroah-Hartman
maintainerPATCH 19/152 bonding: fix lockdep warning in bond_get_stats()
In reply to Greg Kroah-HartmanFrom: Taehee Yoo <[email protected]>
[ Upstream commit b3e80d44f5b1b470dd9e2dbc6816e63a5c519709 ]
In the "struct bonding", there is stats_lock.
This lock protects "bond_stats" in the "struct bonding". bond_stats is updated in the bond_get_stats() and this function would be executed concurrently. So, the lock is needed.
Bonding interfaces would be nested.
So, either stats_lock should use dynamic lockdep class key or stats_lock should be used by spin_lock_nested(). In the current code, stats_lock is using a dynamic lockdep class key.
But there is no updating stats_lock_key routine So, lockdep warning will occur.
Test commands:
ip link add bond0 type bond
ip link add bond1 type bond
ip link set bond0 master bond1
ip link set bond0 nomaster
ip link set bond1 master bond0
Splat looks like:
[ 38.420603][ T957] 5.5.0+ #394 Not tainted
[ 38.421074][ T957] ------------------------------------------------------
[ 38.421837][ T957] ip/957 is trying to acquire lock:
[ 38.422399][ T957] ffff888063262cd8 (&bond->stats_lock_key#2){+.+.}, at: bond_get_stats+0x90/0x4d0 [bonding]
[ 38.423528][ T957]
[ 38.423528][ T957] but task is already holding lock:
[ 38.424526][ T957] ffff888065fd2cd8 (&bond->stats_lock_key){+.+.}, at: bond_get_stats+0x90/0x4d0 [bonding]
[ 38.426075][ T957]
[ 38.426075][ T957] which lock already depends on the new lock.
[ 38.426075][ T957]
[ 38.428536][ T957]
[ 38.428536][ T957] the existing dependency chain (in reverse order) is:
[ 38.429475][ T957]
[ 38.429475][ T957] -> #1 (&bond->stats_lock_key){+.+.}:
[ 38.430273][ T957] _raw_spin_lock+0x30/0x70
[ 38.430812][ T957] bond_get_stats+0x90/0x4d0 [bonding]
[ 38.431451][ T957] dev_get_stats+0x1ec/0x270
[ 38.432088][ T957] bond_get_stats+0x1a5/0x4d0 [bonding]
[ 38.432767][ T957] dev_get_stats+0x1ec/0x270
[ 38.433322][ T957] rtnl_fill_stats+0x44/0xbe0
[ 38.433866][ T957] rtnl_fill_ifinfo+0xeb2/0x3720
[ 38.434474][ T957] rtmsg_ifinfo_build_skb+0xca/0x170
[ 38.435081][ T957] rtmsg_ifinfo_event.part.33+0x1b/0xb0
[ 38.436848][ T957] rtnetlink_event+0xcd/0x120
[ 38.437455][ T957] notifier_call_chain+0x90/0x160
[ 38.438067][ T957] netdev_change_features+0x74/0xa0
[ 38.438708][ T957] bond_compute_features.isra.45+0x4e6/0x6f0 [bonding]
[ 38.439522][ T957] bond_enslave+0x3639/0x47b0 [bonding]
[ 38.440225][ T957] do_setlink+0xaab/0x2ef0
[ 38.440786][ T957] __rtnl_newlink+0x9c5/0x1270
[ 38.441463][ T957] rtnl_newlink+0x65/0x90
[ 38.442075][ T957] rtnetlink_rcv_msg+0x4a8/0x890
[ 38.442774][ T957] netlink_rcv_skb+0x121/0x350
[ 38.443451][ T957] netlink_unicast+0x42e/0x610
[ 38.444282][ T957] netlink_sendmsg+0x65a/0xb90
[ 38.444992][ T957] ____sys_sendmsg+0x5ce/0x7a0
[ 38.445679][ T957] ___sys_sendmsg+0x10f/0x1b0
[ 38.446365][ T957] __sys_sendmsg+0xc6/0x150
[ 38.447007][ T957] do_syscall_64+0x99/0x4f0
[ 38.447668][ T957] entry_SYSCALL_64_after_hwframe+0x49/0xbe
[ 38.448538][ T957]
[ 38.448538][ T957] -> #0 (&bond->stats_lock_key#2){+.+.}:
[ 38.449554][ T957] __lock_acquire+0x2d8d/0x3de0
[ 38.450148][ T957] lock_acquire+0x164/0x3b0
[ 38.450711][ T957] _raw_spin_lock+0x30/0x70
[ 38.451292][ T957] bond_get_stats+0x90/0x4d0 [bonding]
[ 38.451950][ T957] dev_get_stats+0x1ec/0x270
[ 38.452425][ T957] bond_get_stats+0x1a5/0x4d0 [bonding]
[ 38.453362][ T957] dev_get_stats+0x1ec/0x270
[ 38.453825][ T957] rtnl_fill_stats+0x44/0xbe0
[ 38.454390][ T957] rtnl_fill_ifinfo+0xeb2/0x3720
[ 38.456257][ T957] rtmsg_ifinfo_build_skb+0xca/0x170
[ 38.456998][ T957] rtmsg_ifinfo_event.part.33+0x1b/0xb0
[ 38.459351][ T957] rtnetlink_event+0xcd/0x120
[ 38.460086][ T957] notifier_call_chain+0x90/0x160
[ 38.460829][ T957] netdev_change_features+0x74/0xa0
[ 38.461752][ T957] bond_compute_features.isra.45+0x4e6/0x6f0 [bonding]
[ 38.462705][ T957] bond_enslave+0x3639/0x47b0 [bonding]
[ 38.463476][ T957] do_setlink+0xaab/0x2ef0
[ 38.464141][ T957] __rtnl_newlink+0x9c5/0x1270
[ 38.464897][ T957] rtnl_newlink+0x65/0x90
[ 38.465522][ T957] rtnetlink_rcv_msg+0x4a8/0x890
[ 38.466215][ T957] netlink_rcv_skb+0x121/0x350
[ 38.466895][ T957] netlink_unicast+0x42e/0x610
[ 38.467583][ T957] netlink_sendmsg+0x65a/0xb90
[ 38.468285][ T957] ____sys_sendmsg+0x5ce/0x7a0
[ 38.469202][ T957] ___sys_sendmsg+0x10f/0x1b0
[ 38.469884][ T957] __sys_sendmsg+0xc6/0x150
[ 38.470587][ T957] do_syscall_64+0x99/0x4f0
[ 38.471245][ T957] entry_SYSCALL_64_after_hwframe+0x49/0xbe
[ 38.472093][ T957]
[ 38.472093][ T957] other info that might help us debug this:
[ 38.472093][ T957]
[ 38.473438][ T957] Possible unsafe locking scenario:
[ 38.473438][ T957]
[ 38.474898][ T957] CPU0 CPU1
[ 38.476234][ T957] ---- ----
[ 38.480171][ T957] lock(&bond->stats_lock_key);
[ 38.480808][ T957] lock(&bond->stats_lock_key#2);
[ 38.481791][ T957] lock(&bond->stats_lock_key);
[ 38.482754][ T957] lock(&bond->stats_lock_key#2);
[ 38.483416][ T957]
[ 38.483416][ T957] *** DEADLOCK ***
[ 38.483416][ T957]
[ 38.484505][ T957] 3 locks held by ip/957:
[ 38.485048][ T957] #0: ffffffffbccf6230 (rtnl_mutex){+.+.}, at: rtnetlink_rcv_msg+0x457/0x890
[ 38.486198][ T957] #1: ffff888065fd2cd8 (&bond->stats_lock_key){+.+.}, at: bond_get_stats+0x90/0x4d0 [bonding]
[ 38.487625][ T957] #2: ffffffffbc9254c0 (rcu_read_lock){....}, at: bond_get_stats+0x5/0x4d0 [bonding]
[ 38.488897][ T957]
[ 38.488897][ T957] stack backtrace:
[ 38.489646][ T957] CPU: 1 PID: 957 Comm: ip Not tainted 5.5.0+ #394
[ 38.490497][ T957] Hardware name: innotek GmbH VirtualBox/VirtualBox, BIOS VirtualBox 12/01/2006
[ 38.492810][ T957] Call Trace:
[ 38.493219][ T957] dump_stack+0x96/0xdb
[ 38.493709][ T957] check_noncircular+0x371/0x450
[ 38.494344][ T957] ? lookup_address+0x60/0x60
[ 38.494923][ T957] ? print_circular_bug.isra.35+0x310/0x310
[ 38.495699][ T957] ? hlock_class+0x130/0x130
[ 38.496334][ T957] ? __lock_acquire+0x2d8d/0x3de0
[ 38.496979][ T957] __lock_acquire+0x2d8d/0x3de0
[ 38.497607][ T957] ? register_lock_class+0x14d0/0x14d0
[ 38.498333][ T957] ? check_chain_key+0x236/0x5d0
[ 38.499003][ T957] lock_acquire+0x164/0x3b0
[ 38.499800][ T957] ? bond_get_stats+0x90/0x4d0 [bonding]
[ 38.500706][ T957] _raw_spin_lock+0x30/0x70
[ 38.501435][ T957] ? bond_get_stats+0x90/0x4d0 [bonding]
[ 38.502311][ T957] bond_get_stats+0x90/0x4d0 [bonding][ ... ]
But, there is another problem.
The dynamic lockdep class key is protected by RTNL, but bond_get_stats() would be called outside of RTNL.
So, it would use an invalid dynamic lockdep class key.
In order to fix this issue, stats_lock uses spin_lock_nested() instead of a dynamic lockdep key.
The bond_get_stats() calls bond_get_lowest_level_rcu() to get the correct nest level value, which will be used by spin_lock_nested(). The "dev->lower_level" indicates lower nest level value, but this value is invalid outside of RTNL.
So, bond_get_lowest_level_rcu() returns valid lower nest level value in the RCU critical section.bond_get_lowest_level_rcu() will be work only when LOCKDEP is enabled.
--- drivers/net/bonding/bond_main.c | 53 +++++++++++++++++++++++++++++++++++++--- 1 file changed, 50 insertions(+), 3 deletions(-)
drivers/net/bonding/bond_main.c +50 -3
@@ -3436,6 +3436,47 @@ static void bond_fold_stats(struct rtnl_ } } +#ifdef CONFIG_LOCKDEP +static int bond_get_lowest_level_rcu (struct net_device *dev) +{ +struct net_device *ldev, *next, *now, *dev_stack[MAX_NEST_DEV +1 ]; +struct list_head *niter, *iter, *iter_stack[MAX_NEST_DEV +1 ]; +int cur =0 , max =0 ; + + now = dev; + iter = &dev->adj_list.lower; + +while (1 ) { + next =NULL ; +while (1 ) { + ldev =netdev_next_lower_dev_rcu (now, &iter); +if (!ldev) +break ; + + next = ldev; + niter = &ldev->adj_list.lower; + dev_stack[cur] = now; + iter_stack[cur++] = iter; +if (max <= cur) + max = cur; +break ; + } + +if (!next) { +if (!cur) +return max; + next = dev_stack[--cur]; + niter = iter_stack[cur]; + } + + now = next; + iter = niter; + } + +return max; +} +#endif +static void bond_get_stats (struct net_device *bond_dev,struct rtnl_link_stats64 *stats) {
@@ -3443,11 +3484,17 @@ static void bond_get_stats(struct net_destruct rtnl_link_stats64 temp;struct list_head *iter;struct slave *slave; +int nest_level =0 ; -spin_lock (&bond->stats_lock); -memcpy (stats, &bond->bond_stats,sizeof (*stats));rcu_read_lock (); +#ifdef CONFIG_LOCKDEP + nest_level =bond_get_lowest_level_rcu (bond_dev); +#endif + +spin_lock_nested (&bond->stats_lock, nest_level); +memcpy (stats, &bond->bond_stats,sizeof (*stats)); +bond_for_each_slave_rcu (bond, slave, iter) {const struct rtnl_link_stats64 *new =dev_get_stats (slave->dev, &temp);
@@ -3457,10 +3504,10 @@ static void bond_get_stats(struct net_de/* save off the slave stats for the next run */ memcpy (&slave->slave_stats, new,sizeof (*new)); } -rcu_read_unlock ();memcpy (&bond->bond_stats, stats,sizeof (*stats));spin_unlock (&bond->stats_lock); +rcu_read_unlock (); }static int bond_do_ioctl (struct net_device *bond_dev,struct ifreq *ifr,int cmd)
Greg Kroah-Hartman
maintainerPATCH 20/152 ipv6: Fix route replacement with dev-only route
In reply to Greg Kroah-HartmanFrom: Benjamin Poirier <[email protected]>
[ Upstream commit e404b8c7cfb31654c9024d497cec58a501501692 ]
After commit 27596472473a ("ipv6: fix ECMP route replacement") it is no longer possible to replace an ECMP-able route by a non ECMP-able route. For example,
ip route add 2001:db8::1/128 via fe80::1 dev dummy0 ip route replace 2001:db8::1/128 dev dummy0
does not work as expected.
Tweak the replacement logic so that point 3 in the log of the above commit becomes:
3. If the new route is not ECMP-able, and no matching non-ECMP-able route exists, replace matching ECMP-able route (if any) or add the new route.
We can now summarize the entire replace semantics to:
When doing a replace, prefer replacing a matching route of the same "ECMP-able-ness" as the replace argument. If there is no such candidate, fallback to the first route found.
--- net/ipv6/ip6_fib.c | 7 ++++--- tools/testing/selftests/net/fib_tests.sh | 6 ++++++ 2 files changed, 10 insertions(+), 3 deletions(-)
net/ipv6/ip6_fib.c +4 -3
@@ -1050,8 +1050,7 @@ static int fib6_add_rt2node(struct fib6_ found++;break ; } -if (rt_can_ecmp) - fallback_ins = fallback_ins ?: ins; + fallback_ins = fallback_ins ?: ins; goto next_iter; }
@@ -1094,7 +1093,9 @@ next_iter: }if (fallback_ins && !found) { -/* No ECMP-able route found, replace first non-ECMP one */ +/* No matching route with same ecmp-able-ness found, replace +* first matching route +*/ ins = fallback_ins; iter =rcu_dereference_protected (*ins,lockdep_is_held (&rt->fib6_table->tb6_lock));
tools/testing/selftests/net/fib_tests.sh +6 -0
@@ -910,6 +910,12 @@ ipv6_rt_replace_mpath() check_route6 "2001:db8:104::/64 via 2001:db8:101::3 dev veth1 metric 1024" log_test $? 0 "Multipath with single path via multipath attribute" + # multipath with dev-only + add_initial_route6 "nexthop via 2001:db8:101::2 nexthop via 2001:db8:103::2" + run_cmd "$IP -6 ro replace 2001:db8:104::/64 dev veth1" + check_route6 "2001:db8:104::/64 dev veth1 metric 1024" + log_test $? 0 "Multipath with dev-only" + # route replace fails - invalid nexthop 1 add_initial_route6 "nexthop via 2001:db8:101::2 nexthop via 2001:db8:103::2" run_cmd "$IP -6 ro replace 2001:db8:104::/64 nexthop via 2001:db8:111::3 nexthop via 2001:db8:103::3"
Greg Kroah-Hartman
maintainerPATCH 21/152 ipv6: Fix nlmsg_flags when splitting a multipath route
In reply to Greg Kroah-HartmanFrom: Benjamin Poirier <[email protected]>
[ Upstream commit afecdb376bd81d7e16578f0cfe82a1aec7ae18f3 ]
When splitting an RTA_MULTIPATH request into multiple routes and adding the second and later components, we must not simply remove NLM_F_REPLACE but instead replace it by NLM_F_CREATE. Otherwise, it may look like the netlink message was malformed.
For example,
ip route add 2001:db8::1/128 dev dummy0 ip route change 2001:db8::1/128 nexthop via fe80::30:1 dev dummy0 \ nexthop via fe80::30:2 dev dummy0 results in the following warnings: [ 1035.057019] IPv6: RTM_NEWROUTE with no NLM_F_CREATE or NLM_F_REPLACE [ 1035.057517] IPv6: NLM_F_CREATE should be set when creating new route
This patch makes the nlmsg sequence look equivalent for __ip6_ins_rt() to what it would get if the multipath route had been added in multiple netlink operations:
ip route add 2001:db8::1/128 dev dummy0 ip route change 2001:db8::1/128 nexthop via fe80::30:1 dev dummy0 ip route append 2001:db8::1/128 nexthop via fe80::30:2 dev dummy0
--- net/ipv6/route.c | 1 + 1 file changed, 1 insertion(+)
net/ipv6/route.c +1 -0
@@ -5155,6 +5155,7 @@ static int ip6_route_multipath_add(struc*/ cfg->fc_nlinfo.nlh->nlmsg_flags &= ~(NLM_F_EXCL |NLM_F_REPLACE ); + cfg->fc_nlinfo.nlh->nlmsg_flags |=NLM_F_CREATE ; nhn++; }
Greg Kroah-Hartman
maintainerPATCH 22/152 ipmi:ssif: Handle a possible NULL pointer reference
In reply to Greg Kroah-HartmanFrom: Corey Minyard <[email protected]>
[ Upstream commit 6b8526d3abc02c08a2f888e8c20b7ac9e5776dfe ]
In error cases a NULL can be passed to memcpy. The length will always be zero, so it doesn't really matter, but go ahead and check for NULL, anyway, to be more precise and avoid static analysis errors.
--- drivers/char/ipmi/ipmi_ssif.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-)
drivers/char/ipmi/ipmi_ssif.c +7 -3
@@ -775,10 +775,14 @@ static void msg_done_handler(struct ssif_info *ssif_info, int result, flags =ipmi_ssif_lock_cond (ssif_info, &oflags); msg = ssif_info->curr_msg;if (msg) { +if (data) { +if (len >IPMI_MAX_MSG_LENGTH ) + len =IPMI_MAX_MSG_LENGTH ; +memcpy (msg->rsp, data, len); + }else { + len =0 ; + } msg->rsp_size = len; -if (msg->rsp_size >IPMI_MAX_MSG_LENGTH ) - msg->rsp_size =IPMI_MAX_MSG_LENGTH ; -memcpy (msg->rsp, data, msg->rsp_size); ssif_info->curr_msg =NULL ; }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 23/152 drm/msm: Set dma maximum segment size for mdss
In reply to Greg Kroah-HartmanFrom: Sean Paul <[email protected]>
[ Upstream commit db735fc4036bbe1fbe606819b5f0ff26cc76cdff ]
Turning on CONFIG_DMA_API_DEBUG_SG results in the following error:
[ 12.078665] msm ae00000.mdss: DMA-API: mapping sg segment longer than device claims to support [len=3526656] [max=65536] [ 12.089870] WARNING: CPU: 6 PID: 334 at /mnt/host/source/src/third_party/kernel/v4.19/kernel/dma/debug.c:1301 debug_dma_map_sg+0x1dc/0x318 [ 12.102655] Modules linked in: joydev [ 12.106442] CPU: 6 PID: 334 Comm: frecon Not tainted 4.19.0 #2 [ 12.112450] Hardware name: Google Cheza (rev3+) (DT) [ 12.117566] pstate: 60400009 (nZCv daif +PAN -UAO) [ 12.122506] pc : debug_dma_map_sg+0x1dc/0x318 [ 12.126995] lr : debug_dma_map_sg+0x1dc/0x318 [ 12.131487] sp : ffffff800cc3ba80 [ 12.134913] x29: ffffff800cc3ba80 x28: 0000000000000000 [ 12.140395] x27: 0000000000000004 x26: 0000000000000004 [ 12.145868] x25: ffffff8008e55b18 x24: 0000000000000000 [ 12.151337] x23: 00000000ffffffff x22: ffffff800921c000 [ 12.156809] x21: ffffffc0fa75b080 x20: ffffffc0f7195090 [ 12.162280] x19: ffffffc0f1c53280 x18: 0000000000000000 [ 12.167749] x17: 0000000000000000 x16: 0000000000000000 [ 12.173218] x15: 0000000000000000 x14: 0720072007200720 [ 12.178689] x13: 0720072007200720 x12: 0720072007200720 [ 12.184161] x11: 0720072007200720 x10: 0720072007200720 [ 12.189641] x9 : ffffffc0f1fc6b60 x8 : 0000000000000000 [ 12.195110] x7 : ffffff8008132ce0 x6 : 0000000000000000 [ 12.200585] x5 : 0000000000000000 x4 : ffffff8008134734 [ 12.206058] x3 : ffffff800cc3b830 x2 : ffffffc0f1fc6240 [ 12.211532] x1 : 25045a74f48a7400 x0 : 25045a74f48a7400 [ 12.217006] Call trace: [ 12.219535] debug_dma_map_sg+0x1dc/0x318 [ 12.223671] get_pages+0x19c/0x20c [ 12.227177] msm_gem_fault+0x64/0xfc [ 12.230874] __do_fault+0x3c/0x140 [ 12.234383] __handle_mm_fault+0x70c/0xdb8 [ 12.238603] handle_mm_fault+0xac/0xc4 [ 12.242473] do_page_fault+0x1bc/0x3d4 [ 12.246342] do_translation_fault+0x54/0x88 [ 12.250652] do_mem_abort+0x60/0xf0 [ 12.254250] el0_da+0x20/0x24 [ 12.257317] irq event stamp: 67260 [ 12.260828] hardirqs last enabled at (67259): [<ffffff8008132d0c>] console_unlock+0x214/0x608 [ 12.269693] hardirqs last disabled at (67260): [<ffffff8008080e0c>] do_debug_exception+0x5c/0x178 [ 12.278820] softirqs last enabled at (67256): [<ffffff8008081664>] __do_softirq+0x4d4/0x520 [ 12.287510] softirqs last disabled at (67249): [<ffffff80080be574>] irq_exit+0xa8/0x100 [ 12.295742] ---[ end trace e63cfc40c313ffab ]---
The root of the problem is that the default segment size for sgt is (UINT_MAX & PAGE_MASK), and the default segment size for device dma is 64K. As such, if you compare the 2, you would deduce that the sg segment will overflow the device's capacity. In reality, the hardware can accommodate the larger sg segments, it's just not initializing its max segment properly. This patch initializes the max segment size for the mdss device, which gets rid of that pesky warning.
--- drivers/gpu/drm/msm/msm_drv.c | 8 ++++++++ 1 file changed, 8 insertions(+)
drivers/gpu/drm/msm/msm_drv.c +8 -0
@@ -441,6 +441,14 @@ static int msm_drm_init(struct device *dev, struct drm_driver *drv)if (ret) goto err_msm_uninit; +if (!dev->dma_parms) { + dev->dma_parms =devm_kzalloc (dev,sizeof (*dev->dma_parms), +GFP_KERNEL ); +if (!dev->dma_parms) +return -ENOMEM ; + } +dma_set_max_seg_size (dev,DMA_BIT_MASK (32 )); +msm_gem_shrinker_init (ddev);switch (get_mdp_ver (pdev)) {
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 24/152 sched/core: Dont skip remote tick for idle CPUs
In reply to Greg Kroah-HartmanFrom: Scott Wood <[email protected]>
[ Upstream commit 488603b815a7514c7009e6fc339d74ed4a30f343 ]
This will be used in the next patch to get a loadavg update from nohz cpus. The delta check is skipped because idle_sched_class doesn't update se.exec_start.
--- kernel/sched/core.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-)
kernel/sched/core.c +10 -8
@@ -3668,22 +3668,24 @@ static void sched_tick_remote(struct wor* statistics and checks timeslices in a time-independent way, regardless * of when exactly it is running. */ -if (idle_cpu (cpu) || !tick_nohz_tick_stopped_cpu (cpu)) +if (!tick_nohz_tick_stopped_cpu (cpu)) goto out_requeue;rq_lock_irq (rq, &rf); curr = rq->curr; -if (is_idle_task (curr) ||cpu_is_offline (cpu)) +if (cpu_is_offline (cpu)) goto out_unlock;update_rq_clock (rq); - delta =rq_clock_task (rq) - curr->se.exec_start; -/* -* Make sure the next tick runs within a reasonable -* amount of time. -*/ -WARN_ON_ONCE (delta > (u64 )NSEC_PER_SEC *3 ); +if (!is_idle_task (curr)) { +/* +* Make sure the next tick runs within a reasonable +* amount of time. +*/ + delta =rq_clock_task (rq) - curr->se.exec_start; +WARN_ON_ONCE (delta > (u64 )NSEC_PER_SEC *3 ); + } curr->sched_class->task_tick (rq, curr,0 ); out_unlock:
Greg Kroah-Hartman
maintainerPATCH 25/152 timers/nohz: Update NOHZ load in remote tick
In reply to Greg Kroah-HartmanFrom: Peter Zijlstra (Intel) <[email protected]>
[ Upstream commit ebc0f83c78a2d26384401ecf2d2fa48063c0ee27 ]
The way loadavg is tracked during nohz only pays attention to the load upon entering nohz. This can be particularly noticeable if full nohz is entered while non-idle, and then the cpu goes idle and stays that way for a long time.
Use the remote tick to ensure that full nohz cpus report their deltas within a reasonable time.
[ swood: Added changelog and removed recheck of stopped tick. ]
--- include/linux/sched/nohz.h | 2 ++ kernel/sched/core.c | 4 +++- kernel/sched/loadavg.c | 33 +++++++++++++++++++++++---------- 3 files changed, 28 insertions(+), 11 deletions(-)
include/linux/sched/nohz.h +2 -0
@@ -15,9 +15,11 @@ static inline void nohz_balance_enter_idle(int cpu) { }#ifdef CONFIG_NO_HZ_COMMON void calc_load_nohz_start (void ); +void calc_load_nohz_remote (struct rq *rq);void calc_load_nohz_stop (void );#else static inline void calc_load_nohz_start (void ) { } +static inline void calc_load_nohz_remote (struct rq *rq) { }static inline void calc_load_nohz_stop (void ) { }#endif /* CONFIG_NO_HZ_COMMON */
kernel/sched/core.c +3 -1
@@ -3676,6 +3676,7 @@ static void sched_tick_remote(struct work_struct *work)if (cpu_is_offline (cpu)) goto out_unlock; + curr = rq->curr;update_rq_clock (rq);if (!is_idle_task (curr)) {
@@ -3688,10 +3689,11 @@ static void sched_tick_remote(struct work_struct *work) } curr->sched_class->task_tick (rq, curr,0 ); +calc_load_nohz_remote (rq); out_unlock:rq_unlock_irq (rq, &rf); - out_requeue: +/* * Run the remote tick once per second (1Hz). This arbitrary * frequency is large enough to avoid overload but short enough
kernel/sched/loadavg.c +23 -10
@@ -231,16 +231,11 @@ static inline int calc_load_read_idx(void)return calc_load_idx &1 ; } -void calc_load_nohz_start (void ) +static void calc_load_nohz_fold (struct rq *rq) { -struct rq *this_rq =this_rq ();long delta; -/* -* We're going into NO_HZ mode, if there's any pending delta, fold it -* into the pending NO_HZ delta. -*/ - delta =calc_load_fold_active (this_rq,0 ); + delta =calc_load_fold_active (rq,0 );if (delta) {int idx =calc_load_write_idx ();
@@ -248,6 +243,24 @@ void calc_load_nohz_start(void) } } +void calc_load_nohz_start (void ) +{ +/* +* We're going into NO_HZ mode, if there's any pending delta, fold it +* into the pending NO_HZ delta. +*/ +calc_load_nohz_fold (this_rq ()); +} + +/* +* Keep track of the load for NOHZ_FULL, must be called between +* calc_load_nohz_{start,stop}(). +*/ +void calc_load_nohz_remote (struct rq *rq) +{ +calc_load_nohz_fold (rq); +} +void calc_load_nohz_stop (void ) {struct rq *this_rq =this_rq ();
@@ -268,7 +281,7 @@ void calc_load_nohz_stop(void) this_rq->calc_load_update +=LOAD_FREQ ; } -static long calc_load_nohz_fold (void ) +static long calc_load_nohz_read (void ) {int idx =calc_load_read_idx ();long delta =0 ;
@@ -323,7 +336,7 @@ static void calc_global_nohz(void) }#else /* !CONFIG_NO_HZ_COMMON */ -static inline long calc_load_nohz_fold (void ) {return 0 ; } +static inline long calc_load_nohz_read (void ) {return 0 ; }static inline void calc_global_nohz (void ) { }#endif /* CONFIG_NO_HZ_COMMON */
@@ -346,7 +359,7 @@ void calc_global_load(unsigned long ticks)/* * Fold the 'old' NO_HZ-delta to include all NO_HZ CPUs. */ - delta =calc_load_nohz_fold (); + delta =calc_load_nohz_read ();if (delta)atomic_long_add (delta, &calc_load_tasks);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 26/152 sched/fair: Prevent unlimited runtime on throttled group
In reply to Greg Kroah-HartmanFrom: Vincent Guittot <[email protected]>
[ Upstream commit 2a4b03ffc69f2dedc6388e9a6438b5f4c133a40d ]
When a running task is moved on a throttled task group and there is no other task enqueued on the CPU, the task can keep running using 100% CPU whatever the allocated bandwidth for the group and although its cfs rq is throttled. Furthermore, the group entity of the cfs_rq and its parents are not enqueued but only set as curr on their respective cfs_rqs.
We have the following sequence:
sched_move_task
-dequeue_task: dequeue task and group_entities. -put_prev_task: put task and group entities. -sched_change_group: move task to new group.
-enqueue_task: enqueue only task but not group entities because cfs_rq is
throttled.
-set_next_task : set task and group_entities as current sched_entity of
their cfs_rq.
Another impact is that the root cfs_rq runnable_load_avg at root rq stays null because the group_entities are not enqueued. This situation will stay the same until an "external" event triggers a reschedule. Let trigger it immediately instead.
--- kernel/sched/core.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)
kernel/sched/core.c +8 -1
@@ -7058,8 +7058,15 @@ void sched_move_task(struct task_struct *tsk)if (queued)enqueue_task (rq, tsk, queue_flags); -if (running) +if (running) {set_next_task (rq, tsk); +/* +* After changing group, the running task may have joined a +* throttled one but it's still the running task. Trigger a +* resched to make sure that task can still run. +*/ +resched_curr (rq); + }task_rq_unlock (rq, tsk, &rf); }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 27/152 dax: pass NOWAIT flag to iomap_apply
In reply to Greg Kroah-HartmanFrom: Jeff Moyer <[email protected]>
[ Upstream commit 96222d53842dfe54869ec4e1b9d4856daf9105a2 ]
fstests generic/471 reports a failure when run with MOUNT_OPTIONS="-o dax". The reason is that the initial pwrite to an empty file with the RWF_NOWAIT flag set does not return -EAGAIN. It turns out that dax_iomap_rw doesn't pass that flag through to iomap_apply.
With this patch applied, generic/471 passes for me.
--- fs/dax.c | 3 +++ 1 file changed, 3 insertions(+)
fs/dax.c +3 -0
@@ -1207,6 +1207,9 @@ dax_iomap_rw(struct kiocb *iocb, struct iov_iter *iter,lockdep_assert_held (&inode->i_rwsem); } +if (iocb->ki_flags &IOCB_NOWAIT ) + flags |=IOMAP_NOWAIT ; +while (iov_iter_count (iter)) { ret =iomap_apply (inode, pos,iov_iter_count (iter), flags, ops, iter, dax_iomap_actor);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 28/152 mac80211: consider more elements in parsing CRC
In reply to Greg Kroah-HartmanFrom: Johannes Berg <[email protected]>
[ Upstream commit a04564c99bb4a92f805a58e56b2d22cc4978f152 ]
We only use the parsing CRC for checking if a beacon changed, and elements with an ID > 63 cannot be represented in the filter. Thus, like we did before with WMM and Cisco vendor elements, just statically add these forgotten items to the CRC:
- WLAN_EID_VHT_OPERATION
- WLAN_EID_OPMODE_NOTIF
I guess that in most cases when VHT/HE operation change, the HT operation also changed, and so the change was picked up, but we did notice that pure operating mode notification changes were ignored.
--- net/mac80211/util.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-)
net/mac80211/util.c +12 -6
@@ -1063,16 +1063,22 @@ _ieee802_11_parse_elems_crc(const u8 *start, size_t len, bool action, elem_parse_failed = true;break ;case WLAN_EID_VHT_OPERATION : -if (elen >=sizeof (struct ieee80211_vht_operation )) +if (elen >=sizeof (struct ieee80211_vht_operation )) { elems->vht_operation = (void *)pos; -else - elem_parse_failed = true; +if (calc_crc) + crc =crc32_be (crc, pos -2 , elen +2 ); +break ; + } + elem_parse_failed = true;break ;case WLAN_EID_OPMODE_NOTIF : -if (elen >0 ) +if (elen >0 ) { elems->opmode_notif = pos; -else - elem_parse_failed = true; +if (calc_crc) + crc =crc32_be (crc, pos -2 , elen +2 ); +break ; + } + elem_parse_failed = true;break ;case WLAN_EID_MESH_ID : elems->mesh_id = pos;
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 29/152 cfg80211: check wiphy driver existence for drvinfo report
In reply to Greg Kroah-HartmanFrom: Sergey Matyukevich <[email protected]>
[ Upstream commit bfb7bac3a8f47100ebe7961bd14e924c96e21ca7 ]
When preparing ethtool drvinfo, check if wiphy driver is defined before dereferencing it. Driver may not exist, e.g. if wiphy is attached to a virtual platform device.
--- net/wireless/ethtool.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
net/wireless/ethtool.c +6 -2
@@ -7,9 +7,13 @@void cfg80211_get_drvinfo (struct net_device *dev,struct ethtool_drvinfo *info) {struct wireless_dev *wdev = dev->ieee80211_ptr; +struct device *pdev =wiphy_dev (wdev->wiphy); -strlcpy (info->driver,wiphy_dev (wdev->wiphy)->driver->name, -sizeof (info->driver)); +if (pdev->driver) +strlcpy (info->driver, pdev->driver->name, +sizeof (info->driver)); +else +strlcpy (info->driver,"N/A" ,sizeof (info->driver));strlcpy (info->version,init_utsname ()->release,sizeof (info->version));
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 30/152 s390/zcrypt: fix card and queue total counter wrap
In reply to Greg Kroah-HartmanFrom: Harald Freudenberger <[email protected]>
[ Upstream commit fcd98d4002539f1e381916fc1b6648938c1eac76 ]
The internal statistic counters for the total number of requests processed per card and per queue used integers. So they do wrap after a rather huge amount of crypto requests processed. This patch introduces uint64 counters which should hold much longer but still may wrap. The sysfs attributes request_count for card and queue also used only %ld and now display the counter value with %llu.
This is not a security relevant fix. The int overflow which happened is not in any way exploitable as a security breach.
--- drivers/s390/crypto/ap_bus.h | 4 ++-- drivers/s390/crypto/ap_card.c | 8 ++++---- drivers/s390/crypto/ap_queue.c | 6 +++--- drivers/s390/crypto/zcrypt_api.c | 16 +++++++++------- 4 files changed, 18 insertions(+), 16 deletions(-)
drivers/s390/crypto/ap_bus.h +2 -2
@@ -162,7 +162,7 @@ struct ap_card {unsigned int functions;/* AP device function bitfield. */ int queue_depth;/* AP queue depth.*/ int id;/* AP card number. */ -atomic_t total_request_count;/* # requests ever for this AP device.*/ +atomic64_t total_request_count;/* # requests ever for this AP device.*/ };#define to_ap_card (x) container_of((x), struct ap_card, ap_dev.device)
@@ -179,7 +179,7 @@ struct ap_queue {enum ap_state state;/* State of the AP device. */ int pendingq_count;/* # requests on pendingq list. */ int requestq_count;/* # requests on requestq list. */ -int total_request_count;/* # requests ever for this AP device.*/ +u64 total_request_count;/* # requests ever for this AP device.*/ int request_timeout;/* Request timeout in jiffies. */ struct timer_list timeout;/* Timer for request timeouts. */ struct list_head pendingq;/* List of message sent to AP queue. */
drivers/s390/crypto/ap_card.c +4 -4
@@ -63,13 +63,13 @@ static ssize_t request_count_show(struct device *dev, char *buf) {struct ap_card *ac =to_ap_card (dev); -unsigned int req_cnt; +u64 req_cnt; req_cnt =0 ;spin_lock_bh (&ap_list_lock); - req_cnt =atomic_read (&ac->total_request_count); + req_cnt =atomic64_read (&ac->total_request_count);spin_unlock_bh (&ap_list_lock); -return snprintf (buf,PAGE_SIZE ,"%d\n" , req_cnt); +return snprintf (buf,PAGE_SIZE ,"%llu\n" , req_cnt); } static ssize_t request_count_store(struct device *dev,
@@ -83,7 +83,7 @@ static ssize_t request_count_store(struct device *dev,for_each_ap_queue (aq, ac) aq->total_request_count =0 ;spin_unlock_bh (&ap_list_lock); -atomic_set (&ac->total_request_count,0 ); +atomic64_set (&ac->total_request_count,0 );return count; }
drivers/s390/crypto/ap_queue.c +3 -3
@@ -479,12 +479,12 @@ static ssize_t request_count_show(struct device *dev, char *buf) {struct ap_queue *aq =to_ap_queue (dev); -unsigned int req_cnt; +u64 req_cnt;spin_lock_bh (&aq->lock); req_cnt = aq->total_request_count;spin_unlock_bh (&aq->lock); -return snprintf (buf,PAGE_SIZE ,"%d\n" , req_cnt); +return snprintf (buf,PAGE_SIZE ,"%llu\n" , req_cnt); } static ssize_t request_count_store(struct device *dev,
@@ -676,7 +676,7 @@ void ap_queue_message(struct ap_queue *aq, struct ap_message *ap_msg)list_add_tail (&ap_msg->list, &aq->requestq); aq->requestq_count++; aq->total_request_count++; -atomic_inc (&aq->card->total_request_count); +atomic64_inc (&aq->card->total_request_count);/* Send/receive as many request from the queue as possible. */ ap_wait (ap_sm_event_loop (aq,AP_EVENT_POLL ));spin_unlock_bh (&aq->lock);
drivers/s390/crypto/zcrypt_api.c +9 -7
@@ -605,8 +605,8 @@ static inline bool zcrypt_card_compare(struct zcrypt_card *zc, weight +=atomic_read (&zc->load); pref_weight +=atomic_read (&pref_zc->load);if (weight == pref_weight) -return atomic_read (&zc->card->total_request_count) > -atomic_read (&pref_zc->card->total_request_count); +return atomic64_read (&zc->card->total_request_count) > +atomic64_read (&pref_zc->card->total_request_count);return weight > pref_weight; }
@@ -1216,11 +1216,12 @@ static void zcrypt_qdepth_mask(char qdepth[], size_t max_adapters)spin_unlock (&zcrypt_list_lock); } -static void zcrypt_perdev_reqcnt (int reqcnt[],size_t max_adapters) +static void zcrypt_perdev_reqcnt (u32 reqcnt[],size_t max_adapters) {struct zcrypt_card *zc;struct zcrypt_queue *zq;int card; +u64 cnt;memset (reqcnt,0 ,sizeof (int ) * max_adapters);spin_lock (&zcrypt_list_lock);
@@ -1232,8 +1233,9 @@ static void zcrypt_perdev_reqcnt(int reqcnt[], size_t max_adapters) || card >= max_adapters)continue ;spin_lock (&zq->queue->lock); - reqcnt[card] = zq->queue->total_request_count; + cnt = zq->queue->total_request_count;spin_unlock (&zq->queue->lock); + reqcnt[card] = (cnt <UINT_MAX ) ? (u32 ) cnt :UINT_MAX ; } }local_bh_enable ();
@@ -1411,9 +1413,9 @@ static long zcrypt_unlocked_ioctl(struct file *filp, unsigned int cmd,return 0 ; }case ZCRYPT_PERDEV_REQCNT : { -int *reqcnt; +u32 *reqcnt; - reqcnt =kcalloc (AP_DEVICES ,sizeof (int ),GFP_KERNEL ); + reqcnt =kcalloc (AP_DEVICES ,sizeof (u32),GFP_KERNEL );if (!reqcnt)return -ENOMEM ;zcrypt_perdev_reqcnt (reqcnt,AP_DEVICES );
@@ -1470,7 +1472,7 @@ static long zcrypt_unlocked_ioctl(struct file *filp, unsigned int cmd, }case Z90STAT_PERDEV_REQCNT : {/* the old ioctl supports only 64 adapters */ -int reqcnt[MAX_ZDEV_CARDIDS ]; +u32 reqcnt[MAX_ZDEV_CARDIDS ];zcrypt_perdev_reqcnt (reqcnt,MAX_ZDEV_CARDIDS );if (copy_to_user ((int __user *) arg, reqcnt,sizeof (reqcnt)))
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 31/152 qmi_wwan: re-add DW5821e pre-production variant
In reply to Greg Kroah-HartmanFrom: Bjørn Mork <[email protected]>
[ Upstream commit 88bf54603f6f2c137dfee1abf6436ceac3528d2d ]
Commit f25e1392fdb5 removed the support for the pre-production variant of the Dell DW5821e to avoid probing another USB interface unnecessarily. However, the pre-production samples are found in the wild, and this lack of support is causing problems for users of such samples. It is therefore necessary to support both variants.
Matching on both interfaces 0 and 1 is not expected to cause any problem with either variant, as only the QMI function will be probed successfully on either. Interface 1 will be rejected based on the HID class for the production variant:
T: Bus=01 Lev=03 Prnt=04 Port=00 Cnt=01 Dev#= 16 Spd=480 MxCh= 0 D: Ver= 2.10 Cls=ef(misc ) Sub=02 Prot=01 MxPS=64 #Cfgs= 2 P: Vendor=413c ProdID=81d7 Rev=03.18 S: Manufacturer=DELL S: Product=DW5821e Snapdragon X20 LTE S: SerialNumber=0123456789ABCDEF C: #Ifs= 6 Cfg#= 1 Atr=a0 MxPwr=500mA I: If#= 0 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=ff Driver=qmi_wwan I: If#= 1 Alt= 0 #EPs= 1 Cls=03(HID ) Sub=00 Prot=00 Driver=usbhid I: If#= 2 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option I: If#= 3 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option I: If#= 4 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option I: If#= 5 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=ff Driver=option
And interface 0 will be rejected based on too few endpoints for the pre-production variant:
T: Bus=01 Lev=02 Prnt=02 Port=03 Cnt=03 Dev#= 7 Spd=480 MxCh= 0
D: Ver= 2.10 Cls=ef(misc ) Sub=02 Prot=01 MxPS=64 #Cfgs= 2
P: Vendor=413c ProdID=81d7 Rev= 3.18
S: Manufacturer=DELL
S: Product=DW5821e Snapdragon X20 LTE
S: SerialNumber=0123456789ABCDEF
C: #Ifs= 5 Cfg#= 1 Atr=a0 MxPwr=500mA
I: If#= 0 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=ff Driver=
I: If#= 1 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=ff Driver=qmi_wwan
I: If#= 2 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option
I: If#= 3 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option
I: If#= 4 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option
--- drivers/net/usb/qmi_wwan.c | 1 + 1 file changed, 1 insertion(+)
drivers/net/usb/qmi_wwan.c +1 -0
@@ -1363,6 +1363,7 @@ static const struct usb_device_id products[] = { {QMI_FIXED_INTF(0x413c, 0x81b6, 8)}, /* Dell Wireless 5811e */ {QMI_FIXED_INTF(0x413c, 0x81b6, 10)}, /* Dell Wireless 5811e */ {QMI_FIXED_INTF(0x413c, 0x81d7, 0)}, /* Dell Wireless 5821e */ + {QMI_FIXED_INTF(0x413c, 0x81d7, 1)}, /* Dell Wireless 5821e preproduction config */ {QMI_FIXED_INTF(0x413c, 0x81e0, 0)}, /* Dell Wireless 5821e with eSIM support*/ {QMI_FIXED_INTF(0x03f0, 0x4e1d, 8)}, /* HP lt4111 LTE/EV-DO/HSPA+ Gobi 4G Module */ {QMI_FIXED_INTF(0x03f0, 0x9d1d, 1)}, /* HP lt4120 Snapdragon X5 LTE */
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 32/152 qmi_wwan: unconditionally reject 2 ep interfaces
In reply to Greg Kroah-HartmanFrom: Bjørn Mork <[email protected]>
[ Upstream commit 00516d13d4cfa56ce39da144db2dbf08b09b9357 ]
We have been using the fact that the QMI and DIAG functions usually are the only ones with class/subclass/protocol being ff/ff/ff on Quectel modems. This has allowed us to match the QMI function without knowing the exact interface number, which can vary depending on firmware configuration.
The ability to silently reject the DIAG function, which is usually handled by the option driver, is important for this method to work. This is done based on the knowledge that it has exactly 2 bulk endpoints. QMI function control interfaces will have either 3 or 1 endpoint. This rule is universal so the quirk condition can be removed.
The fixed layouts known from the Gobi1k and Gobi2k modems have been gradually replaced by more dynamic layouts, and many vendors now use configurable layouts without changing device IDs. Renaming the class/subclass/protocol matching macro makes it more obvious that this is now not Quectel specific anymore.
--- drivers/net/usb/qmi_wwan.c | 42 ++++++++++++++------------------------ 1 file changed, 15 insertions(+), 27 deletions(-)
drivers/net/usb/qmi_wwan.c +15 -27
@@ -61,7 +61,6 @@ enum qmi_wwan_flags {enum qmi_wwan_quirks {QMI_WWAN_QUIRK_DTR =1 <<0 ,/* needs "set DTR" request */ -QMI_WWAN_QUIRK_QUECTEL_DYNCFG =1 <<1 ,/* check num. endpoints */ };struct qmimux_hdr {
@@ -916,16 +915,6 @@ static const struct driver_info qmi_wwan_info_quirk_dtr = { .data = QMI_WWAN_QUIRK_DTR, }; -static const struct driver_info qmi_wwan_info_quirk_quectel_dyncfg = { - .description ="WWAN/QMI device" , - .flags =FLAG_WWAN |FLAG_SEND_ZLP , - .bind = qmi_wwan_bind, - .unbind = qmi_wwan_unbind, - .manage_power = qmi_wwan_manage_power, - .rx_fixup = qmi_wwan_rx_fixup, - .data =QMI_WWAN_QUIRK_DTR |QMI_WWAN_QUIRK_QUECTEL_DYNCFG , -}; -#define HUAWEI_VENDOR_ID 0x12D1/* map QMI/wwan function by a fixed interface number */
@@ -946,14 +935,18 @@ static const struct driver_info qmi_wwan_info_quirk_quectel_dyncfg = {#define QMI_GOBI_DEVICE (vend, prod) \ QMI_FIXED_INTF(vend, prod, 0) -/* Quectel does not use fixed interface numbers on at least some of their -* devices. We need to check the number of endpoints to ensure that we bind to -* the correct interface. +/* Many devices have QMI and DIAG functions which are distinguishable +* from other vendor specific functions by class, subclass and +* protocol all being 0xff. The DIAG function has exactly 2 endpoints +* and is silently rejected when probed. +* +* This makes it possible to match dynamically numbered QMI functions +* as seen on e.g. many Quectel modems. */ -#define QMI_QUIRK_QUECTEL_DYNCFG (vend, prod) \ +#define QMI_MATCH_FF_FF_FF (vend, prod) \ USB_DEVICE_AND_INTERFACE_INFO(vend, prod, USB_CLASS_VENDOR_SPEC, \ USB_SUBCLASS_VENDOR_SPEC, 0xff), \ - .driver_info = (unsigned long)&qmi_wwan_info_quirk_quectel_dyncfg + .driver_info = (unsigned long)&qmi_wwan_info_quirk_dtrstatic const struct usb_device_id products[] = {/* 1. CDC ECM like devices match on the control interface */
@@ -1059,10 +1052,10 @@ static const struct usb_device_id products[] = { USB_DEVICE_AND_INTERFACE_INFO(0x03f0, 0x581d, USB_CLASS_VENDOR_SPEC, 1, 7), .driver_info = (unsigned long )&qmi_wwan_info, }, - {QMI_QUIRK_QUECTEL_DYNCFG(0x2c7c, 0x0125)}, /* Quectel EC25, EC20 R2.0 Mini PCIe */ - {QMI_QUIRK_QUECTEL_DYNCFG(0x2c7c, 0x0306)}, /* Quectel EP06/EG06/EM06 */ - {QMI_QUIRK_QUECTEL_DYNCFG(0x2c7c, 0x0512)}, /* Quectel EG12/EM12 */ - {QMI_QUIRK_QUECTEL_DYNCFG(0x2c7c, 0x0800)}, /* Quectel RM500Q-GL */ + {QMI_MATCH_FF_FF_FF(0x2c7c, 0x0125)}, /* Quectel EC25, EC20 R2.0 Mini PCIe */ + {QMI_MATCH_FF_FF_FF(0x2c7c, 0x0306)}, /* Quectel EP06/EG06/EM06 */ + {QMI_MATCH_FF_FF_FF(0x2c7c, 0x0512)}, /* Quectel EG12/EM12 */ + {QMI_MATCH_FF_FF_FF(0x2c7c, 0x0800)}, /* Quectel RM500Q-GL *//* 3. Combined interface devices matching on interface number */ {QMI_FIXED_INTF(0x0408, 0xea42, 4)}, /* Yota / Megafon M100-1 */
@@ -1455,7 +1448,6 @@ static int qmi_wwan_probe(struct usb_interface *intf, {struct usb_device_id *id = (struct usb_device_id *)prod;struct usb_interface_descriptor *desc = &intf->cur_altsetting->desc; -const struct driver_info *info;/* Workaround to enable dynamic IDs. This disables usbnet * blacklisting functionality. Which, if required, can be
@@ -1491,12 +1483,8 @@ static int qmi_wwan_probe(struct usb_interface *intf,* different. Ignore the current interface if the number of endpoints * equals the number for the diag interface (two). */ - info = (void *)id->driver_info; - -if (info->data &QMI_WWAN_QUIRK_QUECTEL_DYNCFG ) { -if (desc->bNumEndpoints ==2 ) -return -ENODEV ; - } +if (desc->bNumEndpoints ==2 ) +return -ENODEV ;return usbnet_probe (intf, id); }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 33/152 NFSv4: Fix races between open and dentry revalidation
In reply to Greg Kroah-HartmanFrom: Trond Myklebust <[email protected]>
[ Upstream commit cf5b4059ba7197d6cef9c0e024979d178ed8c8ec ]
We want to make sure that we revalidate the dentry if and only if we've done an OPEN by filename.
In order to avoid races with remote changes to the directory on the server, we want to save the verifier before calling OPEN. The exception is if the server returned a delegation with our OPEN, as we then know that the filename can't have changed on the server.
--- fs/nfs/nfs4file.c | 1 - fs/nfs/nfs4proc.c | 18 ++++++++++++++++-- 2 files changed, 16 insertions(+), 3 deletions(-)
fs/nfs/nfs4file.c +0 -1
@@ -86,7 +86,6 @@ nfs4_file_open(struct inode *inode, struct file *filp)if (inode !=d_inode (dentry)) goto out_drop; -nfs_set_verifier (dentry,nfs_save_change_attribute (dir));nfs_file_set_open_context (filp, ctx);nfs_fscache_open_file (inode, filp); err =0 ;
fs/nfs/nfs4proc.c +16 -2
@@ -2962,10 +2962,13 @@ static int _nfs4_open_and_get_state(struct nfs4_opendata *opendata,struct dentry *dentry;struct nfs4_state *state;fmode_t acc_mode =_nfs4_ctx_to_accessmode (ctx); +struct inode *dir =d_inode (opendata->dir); +unsigned long dir_verifier;unsigned int seq;int ret; seq =raw_seqcount_begin (&sp->so_reclaim_seqcount); + dir_verifier =nfs_save_change_attribute (dir); ret =_nfs4_proc_open (opendata, ctx);if (ret !=0 )
@@ -2993,8 +2996,19 @@ static int _nfs4_open_and_get_state(struct nfs4_opendata *opendata,dput (ctx->dentry); ctx->dentry = dentry = alias; } -nfs_set_verifier (dentry, -nfs_save_change_attribute (d_inode (opendata->dir))); + } + +switch (opendata->o_arg.claim) { +default : +break ; +case NFS4_OPEN_CLAIM_NULL : +case NFS4_OPEN_CLAIM_DELEGATE_CUR : +case NFS4_OPEN_CLAIM_DELEGATE_PREV : +if (!opendata->rpc_done) +break ; +if (opendata->o_res.delegation_type !=0 ) + dir_verifier =nfs_save_change_attribute (dir); +nfs_set_verifier (dentry, dir_verifier); }/* Parse layoutget results before we check for access */
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 34/152 perf/smmuv3: Use platform_get_irq_optional() for wired interrupt
In reply to Greg Kroah-HartmanFrom: John Garry <[email protected]>
[ Upstream commit 0ca2c0319a7bce0e152b51b866979d62dc261e48 ]
Even though a SMMUv3 PMCG implementation may use an MSI as the form of interrupt source, the kernel would still complain that it does not find the wired (GSIV) interrupt in this case:
root@(none)$ dmesg | grep arm-smmu-v3-pmcg | grep "not found" [ 59.237219] arm-smmu-v3-pmcg arm-smmu-v3-pmcg.8.auto: IRQ index 0 not found [ 59.322841] arm-smmu-v3-pmcg arm-smmu-v3-pmcg.9.auto: IRQ index 0 not found [ 59.422155] arm-smmu-v3-pmcg arm-smmu-v3-pmcg.10.auto: IRQ index 0 not found [ 59.539014] arm-smmu-v3-pmcg arm-smmu-v3-pmcg.11.auto: IRQ index 0 not found [ 59.640329] arm-smmu-v3-pmcg arm-smmu-v3-pmcg.12.auto: IRQ index 0 not found [ 59.743112] arm-smmu-v3-pmcg arm-smmu-v3-pmcg.13.auto: IRQ index 0 not found [ 59.880577] arm-smmu-v3-pmcg arm-smmu-v3-pmcg.14.auto: IRQ index 0 not found [ 60.017528] arm-smmu-v3-pmcg arm-smmu-v3-pmcg.15.auto: IRQ index 0 not found
Use platform_get_irq_optional() to silence the warning.
If neither interrupt source is found, then the driver will still warn that IRQ setup errored and the probe will fail.
--- drivers/perf/arm_smmuv3_pmu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/perf/arm_smmuv3_pmu.c +1 -1
@@ -772,7 +772,7 @@ static int smmu_pmu_probe(struct platform_device *pdev) smmu_pmu->reloc_base = smmu_pmu->reg_base; } - irq =platform_get_irq (pdev,0 ); + irq =platform_get_irq_optional (pdev,0 );if (irq >0 ) smmu_pmu->irq = irq;
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 35/152 arm/ftrace: Fix BE text poking
In reply to Greg Kroah-HartmanFrom: Peter Zijlstra <[email protected]>
[ Upstream commit be993e44badc448add6a18d6f12b20615692c4c3 ]
The __patch_text() function already applies __opcode_to_mem_*(), so when __opcode_to_mem_*() is not the identity (BE*), it is applied twice, wrecking the instruction.
--- arch/arm/kernel/ftrace.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-)
arch/arm/kernel/ftrace.c +2 -5
@@ -81,13 +81,10 @@ static int ftrace_modify_code(unsigned long pc, unsigned long old, {unsigned long replaced; -if (IS_ENABLED (CONFIG_THUMB2_KERNEL )) { +if (IS_ENABLED (CONFIG_THUMB2_KERNEL )) old =__opcode_to_mem_thumb32 (old); - new =__opcode_to_mem_thumb32 (new); - }else { +else old =__opcode_to_mem_arm (old); - new =__opcode_to_mem_arm (new); - }if (validate) {if (probe_kernel_read (&replaced, (void *)pc,MCOUNT_INSN_SIZE ))
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 36/152 perf/x86/intel: Add Elkhart Lake support
In reply to Greg Kroah-HartmanFrom: Kan Liang <[email protected]>
[ Upstream commit eda23b387f6c4bb2971ac7e874a09913f533b22c ]
Elkhart Lake also uses Tremont CPU. From the perspective of Intel PMU, there is nothing changed compared with Jacobsville.
Share the perf code with Jacobsville.
--- arch/x86/events/intel/core.c | 1 + 1 file changed, 1 insertion(+)
arch/x86/events/intel/core.c +1 -0
@@ -4746,6 +4746,7 @@ __init int intel_pmu_init(void)break ;case INTEL_FAM6_ATOM_TREMONT_D : +case INTEL_FAM6_ATOM_TREMONT : x86_pmu.late_ack = true;memcpy (hw_cache_event_ids, glp_hw_cache_event_ids,sizeof (hw_cache_event_ids));
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 37/152 perf/x86/cstate: Add Tremont support
In reply to Greg Kroah-HartmanFrom: Kan Liang <[email protected]>
[ Upstream commit ecf71fbccb9ac5cb964eb7de59bb9da3755b7885 ]
Tremont is Intel's successor to Goldmont Plus. From the perspective of Intel cstate residency counters, there is nothing changed compared with Goldmont Plus and Goldmont.
Share glm_cstates with Goldmont Plus and Goldmont. Update the comments for Tremont.
--- arch/x86/events/intel/cstate.c | 22 +++++++++++++--------- 1 file changed, 13 insertions(+), 9 deletions(-)
arch/x86/events/intel/cstate.c +13 -9
@@ -40,17 +40,18 @@* Model specific counters: * MSR_CORE_C1_RES: CORE C1 Residency Counter * perf code: 0x00 -* Available model: SLM,AMT,GLM,CNL +* Available model: SLM,AMT,GLM,CNL,TNT * Scope: Core (each processor core has a MSR) * MSR_CORE_C3_RESIDENCY: CORE C3 Residency Counter * perf code: 0x01 * Available model: NHM,WSM,SNB,IVB,HSW,BDW,SKL,GLM, -* CNL,KBL,CML +* CNL,KBL,CML,TNT * Scope: Core * MSR_CORE_C6_RESIDENCY: CORE C6 Residency Counter * perf code: 0x02 * Available model: SLM,AMT,NHM,WSM,SNB,IVB,HSW,BDW, -* SKL,KNL,GLM,CNL,KBL,CML,ICL,TGL +* SKL,KNL,GLM,CNL,KBL,CML,ICL,TGL, +* TNT * Scope: Core * MSR_CORE_C7_RESIDENCY: CORE C7 Residency Counter * perf code: 0x03
@@ -60,17 +61,18 @@* MSR_PKG_C2_RESIDENCY: Package C2 Residency Counter. * perf code: 0x00 * Available model: SNB,IVB,HSW,BDW,SKL,KNL,GLM,CNL, -* KBL,CML,ICL,TGL +* KBL,CML,ICL,TGL,TNT * Scope: Package (physical package) * MSR_PKG_C3_RESIDENCY: Package C3 Residency Counter. * perf code: 0x01 * Available model: NHM,WSM,SNB,IVB,HSW,BDW,SKL,KNL, -* GLM,CNL,KBL,CML,ICL,TGL +* GLM,CNL,KBL,CML,ICL,TGL,TNT * Scope: Package (physical package) * MSR_PKG_C6_RESIDENCY: Package C6 Residency Counter. * perf code: 0x02 -* Available model: SLM,AMT,NHM,WSM,SNB,IVB,HSW,BDW -* SKL,KNL,GLM,CNL,KBL,CML,ICL,TGL +* Available model: SLM,AMT,NHM,WSM,SNB,IVB,HSW,BDW, +* SKL,KNL,GLM,CNL,KBL,CML,ICL,TGL, +* TNT * Scope: Package (physical package) * MSR_PKG_C7_RESIDENCY: Package C7 Residency Counter. * perf code: 0x03
@@ -87,7 +89,8 @@* Scope: Package (physical package) * MSR_PKG_C10_RESIDENCY: Package C10 Residency Counter. * perf code: 0x06 -* Available model: HSW ULT,KBL,GLM,CNL,CML,ICL,TGL +* Available model: HSW ULT,KBL,GLM,CNL,CML,ICL,TGL, +* TNT * Scope: Package (physical package) * */
@@ -640,8 +643,9 @@ static const struct x86_cpu_id intel_cstates_match[] __initconst = {X86_CSTATES_MODEL (INTEL_FAM6_ATOM_GOLDMONT , glm_cstates),X86_CSTATES_MODEL (INTEL_FAM6_ATOM_GOLDMONT_D , glm_cstates), -X86_CSTATES_MODEL (INTEL_FAM6_ATOM_GOLDMONT_PLUS , glm_cstates), +X86_CSTATES_MODEL (INTEL_FAM6_ATOM_TREMONT_D , glm_cstates), +X86_CSTATES_MODEL (INTEL_FAM6_ATOM_TREMONT , glm_cstates),X86_CSTATES_MODEL (INTEL_FAM6_ICELAKE_L , icl_cstates), X86_CSTATES_MODEL(INTEL_FAM6_ICELAKE, icl_cstates),
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 38/152 perf/x86/msr: Add Tremont support
In reply to Greg Kroah-HartmanFrom: Kan Liang <[email protected]>
[ Upstream commit 0aa0e0d6b34b89649e6b5882a7e025a0eb9bd832 ]
Tremont is Intel's successor to Goldmont Plus. SMI_COUNT MSR is also supported.
--- arch/x86/events/msr.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
arch/x86/events/msr.c +2 -1
@@ -75,8 +75,9 @@ static bool test_intel(int idx, void *data)case INTEL_FAM6_ATOM_GOLDMONT :case INTEL_FAM6_ATOM_GOLDMONT_D : -case INTEL_FAM6_ATOM_GOLDMONT_PLUS : +case INTEL_FAM6_ATOM_TREMONT_D : +case INTEL_FAM6_ATOM_TREMONT :case INTEL_FAM6_XEON_PHI_KNL :case INTEL_FAM6_XEON_PHI_KNM :
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 39/152 ceph: do not execute direct write in parallel if O_APPEND is specified
In reply to Greg Kroah-HartmanFrom: Xiubo Li <[email protected]>
[ Upstream commit 8e4473bb50a1796c9c32b244e5dbc5ee24ead937 ]
In O_APPEND & O_DIRECT mode, the data from different writers will be possibly overlapping each other since they take the shared lock.
For example, both Writer1 and Writer2 are in O_APPEND and O_DIRECT mode:
Writer1 Writer2
shared_lock() shared_lock()
getattr(CAP_SIZE) getattr(CAP_SIZE)
iocb->ki_pos = EOF iocb->ki_pos = EOF
write(data1)
write(data2)
shared_unlock() shared_unlock()The data2 will overlap the data1 from the same file offset, the old EOF.
Switch to exclusive lock instead when O_APPEND is specified.
--- fs/ceph/file.c | 17 +++++++++++------ 1 file changed, 11 insertions(+), 6 deletions(-)
fs/ceph/file.c +11 -6
@@ -1418,6 +1418,7 @@ static ssize_t ceph_write_iter(struct kiocb *iocb, struct iov_iter *from)struct ceph_cap_flush *prealloc_cf;ssize_t count, written =0 ;int err, want, got; +bool direct_lock = false;loff_t pos;loff_t limit =max (i_size_read (inode), fsc->max_file_size);
@@ -1428,8 +1429,11 @@ static ssize_t ceph_write_iter(struct kiocb *iocb, struct iov_iter *from)if (!prealloc_cf)return -ENOMEM ; +if ((iocb->ki_flags & (IOCB_DIRECT |IOCB_APPEND )) ==IOCB_DIRECT ) + direct_lock = true; + retry_snap: -if (iocb->ki_flags &IOCB_DIRECT ) +if (direct_lock)ceph_start_io_direct (inode);else ceph_start_io_write (inode);
@@ -1519,14 +1523,15 @@ static ssize_t ceph_write_iter(struct kiocb *iocb, struct iov_iter *from)/* we might need to revert back to that point */ data = *from; -if (iocb->ki_flags &IOCB_DIRECT ) { +if (iocb->ki_flags &IOCB_DIRECT ) written =ceph_direct_read_write (iocb, &data, snapc, &prealloc_cf); -ceph_end_io_direct (inode); - }else { +else written =ceph_sync_write (iocb, &data, pos, snapc); +if (direct_lock) +ceph_end_io_direct (inode); +else ceph_end_io_write (inode); - }if (written >0 )iov_iter_advance (from, written);ceph_put_snap_context (snapc);
@@ -1577,7 +1582,7 @@ static ssize_t ceph_write_iter(struct kiocb *iocb, struct iov_iter *from) goto out_unlocked; out: -if (iocb->ki_flags &IOCB_DIRECT ) +if (direct_lock)ceph_end_io_direct (inode);else ceph_end_io_write (inode);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 40/152 ARM: dts: sti: fixup sound frame-inversion for stihxxx-b2120.dtsi
In reply to Greg Kroah-HartmanFrom: Kuninori Morimoto <[email protected]>
[ Upstream commit f24667779b5348279e5e4328312a141a730a1fc7 ]
frame-inversion is "flag" not "uint32".
This patch fixup it.
--- arch/arm/boot/dts/stihxxx-b2120.dtsi | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/boot/dts/stihxxx-b2120.dtsi +1 -1
@@ -46,7 +46,7 @@ /* DAC */ format = "i2s"; mclk-fs = <256>; - frame-inversion = <1>; + frame-inversion; cpu { sound-dai = <&sti_uni_player2>; };
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 41/152 drm/amd/display: Do not set optimized_require to false after plane disable
In reply to Greg Kroah-HartmanFrom: Sung Lee <[email protected]>
[ Upstream commit df36f6cf23ada812930afa8ee76681d4ad307c61 ]
[WHY]
The optimized_require flag is needed to set watermarks and clocks lower in certain conditions. This flag is set to true and then set to false while programming front end in dcn20.
[HOW]
Do not set the flag to false while disabling plane.
--- drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c | 1 - 1 file changed, 1 deletion(-)
drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c +0 -1
@@ -493,7 +493,6 @@ static void dcn20_plane_atomic_disable(struct dc *dc, struct pipe_ctx *pipe_ctx) dpp->funcs->dpp_dppclk_control (dpp, false, false); hubp->power_gated = true; - dc->optimized_required = false;/* We're powering off, no need to optimize */ dc->hwss.plane_atomic_power_down(dc, pipe_ctx->plane_res.dpp,
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 42/152 RDMA/siw: Remove unwanted WARN_ON in siw_cm_llp_data_ready()
In reply to Greg Kroah-HartmanFrom: Krishnamraju Eraparaju <[email protected]>
[ Upstream commit 663218a3e715fd9339d143a3e10088316b180f4f ]
Warnings like below can fill up the dmesg while disconnecting RDMA connections.
Hence, remove the unwanted WARN_ON.
WARNING: CPU: 6 PID: 0 at drivers/infiniband/sw/siw/siw_cm.c:1229 siw_cm_llp_data_ready+0xc1/0xd0 [siw] RIP: 0010:siw_cm_llp_data_ready+0xc1/0xd0 [siw] Call Trace: <IRQ> tcp_data_queue+0x226/0xb40 tcp_rcv_established+0x220/0x620 tcp_v4_do_rcv+0x12a/0x1e0 tcp_v4_rcv+0xb05/0xc00 ip_local_deliver_finish+0x69/0x210 ip_local_deliver+0x6b/0xe0 ip_rcv+0x273/0x362 __netif_receive_skb_core+0xb35/0xc30 netif_receive_skb_internal+0x3d/0xb0 napi_gro_frags+0x13b/0x200 t4_ethrx_handler+0x433/0x7d0 [cxgb4] process_responses+0x318/0x580 [cxgb4] napi_rx_handler+0x14/0x100 [cxgb4] net_rx_action+0x149/0x3b0 __do_softirq+0xe3/0x30a irq_exit+0x100/0x110 do_IRQ+0x7f/0xe0 common_interrupt+0xf/0xf </IRQ>
--- drivers/infiniband/sw/siw/siw_cm.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)
drivers/infiniband/sw/siw/siw_cm.c +2 -3
@@ -1225,10 +1225,9 @@ static void siw_cm_llp_data_ready(struct sock *sk)read_lock (&sk->sk_callback_lock); cep =sk_to_cep (sk); -if (!cep) { -WARN_ON (1 ); +if (!cep) goto out; - } +siw_dbg_cep (cep,"state: %d\n" , cep->state);switch (cep->state) {
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 43/152 drm/amd/display: Check engine is not NULL before acquiring
In reply to Greg Kroah-HartmanFrom: Aric Cyr <[email protected]>
[ Upstream commit 2b63d0ec0daf79ba503fa8bfa25e07dc3da274f3 ]
[Why]
Engine can be NULL in some cases, so we must not acquire it.
[How]
Check for NULL engine before acquiring.
--- drivers/gpu/drm/amd/display/dc/dce/dce_aux.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/amd/display/dc/dce/dce_aux.c +1 -1
@@ -386,7 +386,7 @@ static bool acquire( {enum gpio_result result; -if (!is_engine_available (engine)) +if ((engine ==NULL ) || !is_engine_available (engine))return false; result = dal_ddc_open(ddc, GPIO_MODE_HARDWARE,
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 44/152 drm/amd/display: Limit minimum DPPCLK to 100MHz.
In reply to Greg Kroah-HartmanFrom: Yongqiang Sun <[email protected]>
[ Upstream commit 6c81917a0485ee2a1be0dc23321ac10ecfd9578b ]
[Why]
Underflow is observed when plug in a 4K@60 monitor with 1366x768 eDP due to DPPCLK is too low.
[How]
Limit minimum DPPCLK to 100MHz.
--- drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c | 6 ++++++ 1 file changed, 6 insertions(+)
drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c +6 -0
@@ -91,6 +91,12 @@ void rn_update_clocks(struct clk_mgr *clk_mgr_base,rn_vbios_smu_set_min_deep_sleep_dcfclk (clk_mgr, clk_mgr_base->clks.dcfclk_deep_sleep_khz); } +// workaround: Limit dppclk to 100Mhz to avoid lower eDP panel switch to plus 4K monitor underflow. +if (!IS_DIAG_DC (dc->ctx->dce_environment)) { +if (new_clocks->dppclk_khz <100000 ) + new_clocks->dppclk_khz =100000 ; + } +if (should_set_clock (safe_to_lower, new_clocks->dppclk_khz, clk_mgr->base.clks.dppclk_khz)) {if (clk_mgr->base.clks.dppclk_khz > new_clocks->dppclk_khz) dpp_clock_lowered = true;
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 45/152 drm/amd/display: Add initialitions for PLL2 clock source
In reply to Greg Kroah-HartmanFrom: Isabel Zhang <[email protected]>
[ Upstream commit c134c3cabae46a56ab2e1f5e5fa49405e1758838 ]
[Why]
Starting from 14nm, the PLL is built into the PHY and the PLL is mapped to PHY on 1 to 1 basis. In the code, the DP port is mapped to a PLL that was not initialized. This causes DP to HDMI dongle to not light up the display.
[How]
Initializations added for PLL2 when creating resources.
--- drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c | 6 ++++++ 1 file changed, 6 insertions(+)
drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c +6 -0
@@ -57,6 +57,7 @@#include "dcn20/dcn20_dccg.h" #include "dcn21_hubbub.h" #include "dcn10/dcn10_resource.h" +#include "dce110/dce110_resource.h" #include "dcn20/dcn20_dwb.h" #include "dcn20/dcn20_mmhubbub.h"
@@ -824,6 +825,7 @@ static const struct dc_debug_options debug_defaults_diags = {enum dcn20_clk_src_array_id {DCN20_CLK_SRC_PLL0 ,DCN20_CLK_SRC_PLL1 , +DCN20_CLK_SRC_PLL2 ,DCN20_CLK_SRC_TOTAL_DCN21 };
@@ -1492,6 +1494,10 @@ static bool construct(dcn21_clock_source_create (ctx, ctx->dc_bios,CLOCK_SOURCE_COMBO_PHY_PLL1 , &clk_src_regs[1 ], false); + pool->base.clock_sources[DCN20_CLK_SRC_PLL2 ] = +dcn21_clock_source_create (ctx, ctx->dc_bios, +CLOCK_SOURCE_COMBO_PHY_PLL2 , + &clk_src_regs[2 ], false); pool->base.clk_src_count =DCN20_CLK_SRC_TOTAL_DCN21 ;
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 46/152 amdgpu: Prevent build errors regarding soft/hard-float FP ABI tags
In reply to Greg Kroah-HartmanFrom: Daniel Kolesa <[email protected]>
[ Upstream commit 416611d9b6eebaeae58ed26cc7d23131c69126b1 ]
On PowerPC, the compiler will tag object files with whether they use hard or soft float FP ABI and whether they use 64 or 128-bit long double ABI. On systems with 64-bit long double ABI, a tag will get emitted whenever a double is used, as on those systems a long double is the same as a double. This will prevent linkage as other files are being compiled with hard-float.
On ppc64, this code will never actually get used for the time being, as the only currently existing hardware using it are the Renoir APUs. Therefore, until this is testable and can be fixed properly, at least make sure the build will not fail.
--- drivers/gpu/drm/amd/display/dc/clk_mgr/Makefile | 6 ++++++ 1 file changed, 6 insertions(+)
drivers/gpu/drm/amd/display/dc/clk_mgr/Makefile +6 -0
@@ -91,6 +91,12 @@ ifdef CONFIG_DRM_AMD_DC_DCN2_1 ############################################################################### CLK_MGR_DCN21 = rn_clk_mgr.o rn_clk_mgr_vbios_smu.o +# prevent build errors regarding soft-float vs hard-float FP ABI tags +# this code is currently unused on ppc64, as it applies to Renoir APUs only +ifdef CONFIG_PPC64 +CFLAGS_$(AMDDALPATH)/dc/clk_mgr/dcn21/rn_clk_mgr.o := $(call cc-option,-mno-gnu-attribute) +endif + AMD_DAL_CLK_MGR_DCN21 = $(addprefix $(AMDDALPATH)/dc/clk_mgr/dcn21/,$(CLK_MGR_DCN21)) AMD_DISPLAY_FILES += $(AMD_DAL_CLK_MGR_DCN21)
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 47/152 soc/tegra: fuse: Fix build with Tegra194 configuration
In reply to Greg Kroah-HartmanFrom: Thierry Reding <[email protected]>
[ Upstream commit 6f4ecbe284df5f22e386a640d9a4b32cede62030 ]
If only Tegra194 support is enabled, the tegra30_fuse_read() and tegra30_fuse_init() function are not declared and cause a build failure. Add Tegra194 to the preprocessor guard to make sure these functions are available for Tegra194-only builds as well.
--- drivers/soc/tegra/fuse/fuse-tegra30.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/soc/tegra/fuse/fuse-tegra30.c +2 -1
@@ -35,7 +35,8 @@defined (CONFIG_ARCH_TEGRA_124_SOC ) || \defined (CONFIG_ARCH_TEGRA_132_SOC ) || \defined (CONFIG_ARCH_TEGRA_210_SOC ) || \ - defined(CONFIG_ARCH_TEGRA_186_SOC) +defined (CONFIG_ARCH_TEGRA_186_SOC ) || \ + defined(CONFIG_ARCH_TEGRA_194_SOC)static u32 tegra30_fuse_read_early (struct tegra_fuse *fuse,unsigned int offset) {if (WARN_ON (!fuse->base))
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 48/152 i40e: Fix the conditional for i40e_vc_validate_vqs_bitmaps
In reply to Greg Kroah-HartmanFrom: Brett Creeley <[email protected]>
[ Upstream commit f27f37a04a69890ac85d9155f03ee2d23b678d8f ]
Commit d9d6a9aed3f6 ("i40e: Fix virtchnl_queue_select bitmap validation") introduced a necessary change for verifying how queue bitmaps from the iavf driver get validated. Unfortunately, the conditional was reversed. Fix this.
--- drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c +2 -2
@@ -2363,7 +2363,7 @@ static int i40e_vc_enable_queues_msg(struct i40e_vf *vf, u8 *msg) goto error_param; } -if (i40e_vc_validate_vqs_bitmaps (vqs)) { +if (!i40e_vc_validate_vqs_bitmaps (vqs)) { aq_ret =I40E_ERR_PARAM ; goto error_param; }
@@ -2425,7 +2425,7 @@ static int i40e_vc_disable_queues_msg(struct i40e_vf *vf, u8 *msg) goto error_param; } -if (i40e_vc_validate_vqs_bitmaps (vqs)) { +if (!i40e_vc_validate_vqs_bitmaps (vqs)) { aq_ret =I40E_ERR_PARAM ; goto error_param; }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 49/152 net: ena: fix potential crash when rxfh key is NULL
In reply to Greg Kroah-HartmanFrom: Arthur Kiyanovski <[email protected]>
[ Upstream commit 91a65b7d3ed8450f31ab717a65dcb5f9ceb5ab02 ]
When ethtool -X is called without an hkey, ena_com_fill_hash_function() is called with key=NULL, which is passed to memcpy causing a crash.
This commit fixes this issue by checking key is not NULL.
--- drivers/net/ethernet/amazon/ena/ena_com.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-)
drivers/net/ethernet/amazon/ena/ena_com.c +9 -8
@@ -2297,15 +2297,16 @@ int ena_com_fill_hash_function(struct ena_com_dev *ena_dev,switch (func) {case ENA_ADMIN_TOEPLITZ : -if (key_len >sizeof (hash_key->key)) { -pr_err ("key len (%hu) is bigger than the max supported (%zu)\n" , - key_len,sizeof (hash_key->key)); -return -EINVAL ; +if (key) { +if (key_len !=sizeof (hash_key->key)) { +pr_err ("key len (%hu) doesn't equal the supported size (%zu)\n" , + key_len,sizeof (hash_key->key)); +return -EINVAL ; + } +memcpy (hash_key->key, key, key_len); + rss->hash_init_val = init_val; + hash_key->keys_num = key_len >>2 ; } - -memcpy (hash_key->key, key, key_len); - rss->hash_init_val = init_val; - hash_key->keys_num = key_len >>2 ;break ;case ENA_ADMIN_CRC32 : rss->hash_init_val = init_val;
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 50/152 net: ena: fix uses of round_jiffies()
In reply to Greg Kroah-HartmanFrom: Arthur Kiyanovski <[email protected]>
[ Upstream commit 2a6e5fa2f4c25b66c763428a3e65363214946931 ]
Quoted 1 line
From the documentation of
round_jiffies():
"Rounds a time delta in the future (in jiffies) up or down to (approximately) full seconds. This is useful for timers for which the exact time they fire does not matter too much, as long as they fire approximately every X seconds.
By rounding these timers to whole seconds, all such timers will fire at the same time, rather than at various times spread out. The goal of this is to have the CPU wake up less, which saves power."
There are 2 parts to this patch:
================================
Part 1:
-------
In our case we need timer_service to be called approximately every X=1 seconds, and the exact time does not matter, so using round_jiffies() is the right way to go.
Therefore we add round_jiffies() to the mod_timer() in ena_timer_service().
Part 2:
-------round_jiffies() is used in check_for_missing_keep_alive() when getting the jiffies of the expiration of the keep_alive timeout. Here it is actually a mistake to use round_jiffies() because we want the exact time when keep_alive should expire and not an approximate rounded time, which can cause early, false positive, timeouts.
Therefore we remove round_jiffies() in the calculation of keep_alive_expired() in check_for_missing_keep_alive().
--- drivers/net/ethernet/amazon/ena/ena_netdev.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/net/ethernet/amazon/ena/ena_netdev.c +3 -3
@@ -3035,8 +3035,8 @@ static void check_for_missing_keep_alive(struct ena_adapter *adapter)if (adapter->keep_alive_timeout ==ENA_HW_HINTS_NO_TIMEOUT )return ; - keep_alive_expired =round_jiffies (adapter->last_keep_alive_jiffies + - adapter->keep_alive_timeout); + keep_alive_expired = adapter->last_keep_alive_jiffies + + adapter->keep_alive_timeout;if (unlikely (time_is_before_jiffies (keep_alive_expired))) {netif_err (adapter, drv, adapter->netdev,"Keep alive watchdog timeout.\n" );
@@ -3138,7 +3138,7 @@ static void ena_timer_service(struct timer_list *t) }/* Reset the timer */ -mod_timer (&adapter->timer_service, jiffies +HZ ); +mod_timer (&adapter->timer_service,round_jiffies (jiffies +HZ )); } static int ena_calc_io_queue_num(struct pci_dev *pdev,
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 51/152 net: ena: add missing ethtool TX timestamping indication
In reply to Greg Kroah-HartmanFrom: Arthur Kiyanovski <[email protected]>
[ Upstream commit cf6d17fde93bdda23c9b02dd5906a12bf8c55209 ]
Current implementation of the driver calls skb_tx_timestamp()to add a software tx timestamp to the skb, however the software-transmit capability is not reported in ethtool -T.
This commit updates the ethtool structure to report the software-transmit capability in ethtool -T using the standard ethtool_op_get_ts_info(). This function reports all software timestamping capabilities (tx and rx), as well as setting phc_index = -1. phc_index is the index of the PTP hardware clock device that will be used for hardware timestamps. Since we don't have such a device in ENA, using the default -1 value is the correct setting.
--- drivers/net/ethernet/amazon/ena/ena_ethtool.c | 1 + 1 file changed, 1 insertion(+)
drivers/net/ethernet/amazon/ena/ena_ethtool.c +1 -0
@@ -805,6 +805,7 @@ static const struct ethtool_ops ena_ethtool_ops = { .get_channels = ena_get_channels, .get_tunable = ena_get_tunable, .set_tunable = ena_set_tunable, + .get_ts_info = ethtool_op_get_ts_info, };void ena_set_ethtool_ops (struct net_device *netdev)
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 52/152 net: ena: fix incorrect default RSS key
In reply to Greg Kroah-HartmanFrom: Arthur Kiyanovski <[email protected]>
[ Upstream commit 0d1c3de7b8c78a5e44b74b62ede4a63629f5d811 ]
Bug description:
When running "ethtool -x <if_name>" the key shows up as all zeros.
When we use "ethtool -X <if_name> hfunc toeplitz hkey <some:random:key>" to set the key and then try to retrieve it using "ethtool -x <if_name>" then we return the correct key because we return the one we saved.
Bug cause:
We don't fetch the key from the device but instead return the key that we have saved internally which is by default set to zero upon allocation.
Fix:
This commit fixes the issue by initializing the key to a random value using netdev_rss_key_fill().
--- drivers/net/ethernet/amazon/ena/ena_com.c | 15 +++++++++++++++ drivers/net/ethernet/amazon/ena/ena_com.h | 1 + 2 files changed, 16 insertions(+)
drivers/net/ethernet/amazon/ena/ena_com.c +15 -0
@@ -1041,6 +1041,19 @@ static int ena_com_get_feature(struct ena_com_dev *ena_dev, feature_ver); } +static void ena_com_hash_key_fill_default_key (struct ena_com_dev *ena_dev) +{ +struct ena_admin_feature_rss_flow_hash_control *hash_key = + (ena_dev->rss).hash_key; + +netdev_rss_key_fill (&hash_key->key,sizeof (hash_key->key)); +/* The key is stored in the device in u32 array +* as well as the API requires the key to be passed in this +* format. Thus the size of our array should be divided by 4 +*/ + hash_key->keys_num =sizeof (hash_key->key) /sizeof (u32); +} +static int ena_com_hash_key_allocate (struct ena_com_dev *ena_dev) {struct ena_rss *rss = &ena_dev->rss;
@@ -2631,6 +2644,8 @@ int ena_com_rss_init(struct ena_com_dev *ena_dev, u16 indr_tbl_log_size)if (unlikely (rc)) goto err_hash_key; +ena_com_hash_key_fill_default_key (ena_dev); + rc =ena_com_hash_ctrl_init (ena_dev);if (unlikely (rc)) goto err_hash_ctrl;
drivers/net/ethernet/amazon/ena/ena_com.h +1 -0
@@ -44,6 +44,7 @@#include <linux/spinlock.h> #include <linux/types.h> #include <linux/wait.h> +#include <linux/netdevice.h> #include "ena_common_defs.h" #include "ena_admin_defs.h"
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 53/152 net: ena: rss: do not allocate key when not supported
In reply to Greg Kroah-HartmanFrom: Sameeh Jubran <[email protected]>
[ Upstream commit 6a4f7dc82d1e3abd3feb0c60b5041056fcd9880c ]
Currently we allocate the key whether the device supports setting the key or not. This commit adds a check to the allocation function and handles the error accordingly.
--- drivers/net/ethernet/amazon/ena/ena_com.c | 24 ++++++++++++++++++++--- 1 file changed, 21 insertions(+), 3 deletions(-)
drivers/net/ethernet/amazon/ena/ena_com.c +21 -3
@@ -1057,6 +1057,20 @@ static void ena_com_hash_key_fill_default_key(struct ena_com_dev *ena_dev)static int ena_com_hash_key_allocate (struct ena_com_dev *ena_dev) {struct ena_rss *rss = &ena_dev->rss; +struct ena_admin_feature_rss_flow_hash_control *hash_key; +struct ena_admin_get_feat_resp get_resp; +int rc; + + hash_key = (ena_dev->rss).hash_key; + + rc =ena_com_get_feature_ex (ena_dev, &get_resp, +ENA_ADMIN_RSS_HASH_FUNCTION , + ena_dev->rss.hash_key_dma_addr, +sizeof (ena_dev->rss.hash_key),0 ); +if (unlikely (rc)) { + hash_key =NULL ; +return -EOPNOTSUPP ; + } rss->hash_key = dma_alloc_coherent(ena_dev->dmadev, sizeof(*rss->hash_key),
@@ -2640,11 +2654,15 @@ int ena_com_rss_init(struct ena_com_dev *ena_dev, u16 indr_tbl_log_size)if (unlikely (rc)) goto err_indr_tbl; +/* The following function might return unsupported in case the +* device doesn't support setting the key / hash function. We can safely +* ignore this error and have indirection table support only. +*/ rc =ena_com_hash_key_allocate (ena_dev); -if (unlikely (rc)) +if (unlikely (rc) && rc != -EOPNOTSUPP ) goto err_hash_key; - -ena_com_hash_key_fill_default_key (ena_dev); +else if (rc != -EOPNOTSUPP ) +ena_com_hash_key_fill_default_key (ena_dev); rc =ena_com_hash_ctrl_init (ena_dev);if (unlikely (rc))
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 54/152 net: ena: rss: fix failure to get indirection table
In reply to Greg Kroah-HartmanFrom: Sameeh Jubran <[email protected]>
[ Upstream commit 0c8923c0a64fb5d14bebb9a9065d2dc25ac5e600 ]
On old hardware, getting / setting the hash function is not supported while gettting / setting the indirection table is.
This commit enables us to still show the indirection table on older hardwares by setting the hash function and key to NULL.
--- drivers/net/ethernet/amazon/ena/ena_ethtool.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+)
drivers/net/ethernet/amazon/ena/ena_ethtool.c +14 -0
@@ -648,7 +648,21 @@ static int ena_get_rxfh(struct net_device *netdev, u32 *indir, u8 *key,if (rc)return rc; +/* We call this function in order to check if the device +* supports getting/setting the hash function. +*/ rc =ena_com_get_hash_function (adapter->ena_dev, &ena_func, key); + +if (rc) { +if (rc == -EOPNOTSUPP ) { + key =NULL ; + hfunc =NULL ; + rc =0 ; + } + +return rc; + } +if (rc)return rc;
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 55/152 net: ena: rss: store hash function as values and not bits
In reply to Greg Kroah-HartmanFrom: Arthur Kiyanovski <[email protected]>
[ Upstream commit 4844470d472d660c26149ad764da2406adb13423 ]
The device receives, stores and retrieves the hash function value as bits and not as their enum value.
The bug:
* In ena_com_set_hash_function() we set cmd.u.flow_hash_func.selected_func to the bit value of rss->hash_func. (1 << rss->hash_func)
* In ena_com_get_hash_function() we retrieve the hash function and store it's bit value in rss->hash_func. (Now the bit value of rss->hash_func is stored in rss->hash_func instead of it's enum value)
The fix:
This commit fixes the issue by converting the retrieved hash function values from the device to the matching enum value of the set bit using ffs(). ffs() finds the first set bit's index in a word. Since the function returns 1 for the LSB's index, we need to subtract 1 from the returned value (note that BIT(0) is 1).
--- drivers/net/ethernet/amazon/ena/ena_com.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/net/ethernet/amazon/ena/ena_com.c +5 -1
@@ -2370,7 +2370,11 @@ int ena_com_get_hash_function(struct ena_com_dev *ena_dev,if (unlikely (rc))return rc; - rss->hash_func = get_resp.u.flow_hash_func.selected_func; +/* ffs() returns 1 in case the lsb is set */ + rss->hash_func =ffs (get_resp.u.flow_hash_func.selected_func); +if (rss->hash_func) + rss->hash_func--; +if (func) *func = rss->hash_func;
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 56/152 net: ena: fix incorrectly saving queue numbers when setting RSS indirection table
In reply to Greg Kroah-HartmanFrom: Arthur Kiyanovski <[email protected]>
[ Upstream commit 92569fd27f5cb0ccbdf7c7d70044b690e89a0277 ]
The indirection table has the indices of the Rx queues. When we store it during set indirection operation, we convert the indices to our internal representation of the indices.
Our internal representation of the indices is: even indices for Tx and uneven indices for Rx, where every Tx/Rx pair are in a consecutive order starting from 0. For example if the driver has 3 queues (3 for Tx and 3 for Rx) then the indices are as follows:
0 1 2 3 4 5
Tx Rx Tx Rx Tx Rx
The BUG:
The issue is that when we satisfy a get request for the indirection table, we don't convert the indices back to the original representation.
The FIX:
Simply apply the inverse function for the indices of the indirection table after we set it.
--- drivers/net/ethernet/amazon/ena/ena_ethtool.c | 24 ++++++++++++++++++- drivers/net/ethernet/amazon/ena/ena_netdev.h | 2 ++ 2 files changed, 25 insertions(+), 1 deletion(-)
drivers/net/ethernet/amazon/ena/ena_ethtool.c +23 -1
@@ -636,6 +636,28 @@ static u32 ena_get_rxfh_key_size(struct net_device *netdev)return ENA_HASH_KEY_SIZE ; } +static int ena_indirection_table_get (struct ena_adapter *adapter,u32 *indir) +{ +struct ena_com_dev *ena_dev = adapter->ena_dev; +int i, rc; + +if (!indir) +return 0 ; + + rc =ena_com_indirect_table_get (ena_dev, indir); +if (rc) +return rc; + +/* Our internal representation of the indices is: even indices +* for Tx and uneven indices for Rx. We need to convert the Rx +* indices to be consecutive +*/ +for (i =0 ; i <ENA_RX_RSS_TABLE_SIZE ; i++) + indir[i] =ENA_IO_RXQ_IDX_TO_COMBINED_IDX (indir[i]); + +return rc; +} +static int ena_get_rxfh (struct net_device *netdev,u32 *indir,u8 *key,u8 *hfunc) {
@@ -644,7 +666,7 @@ static int ena_get_rxfh(struct net_device *netdev, u32 *indir, u8 *key,u8 func;int rc; - rc =ena_com_indirect_table_get (adapter->ena_dev, indir); + rc =ena_indirection_table_get (adapter, indir);if (rc)return rc;
drivers/net/ethernet/amazon/ena/ena_netdev.h +2 -0
@@ -127,6 +127,8 @@#define ENA_IO_TXQ_IDX (q) (2 * (q))#define ENA_IO_RXQ_IDX (q) (2 * (q) + 1) +#define ENA_IO_TXQ_IDX_TO_COMBINED_IDX (q) ((q) / 2) +#define ENA_IO_RXQ_IDX_TO_COMBINED_IDX (q) (((q) - 1) / 2)#define ENA_MGMNT_IRQ_IDX 0#define ENA_IO_IRQ_FIRST_IDX 1
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 57/152 net: ena: fix corruption of dev_idx_to_host_tbl
In reply to Greg Kroah-HartmanFrom: Arthur Kiyanovski <[email protected]>
[ Upstream commit e3f89f91e98ce07dc0f121a3b70d21aca749ba39 ]
The function ena_com_ind_tbl_convert_from_device() has an overflow bug as explained below. Either way, this function is not needed at all since we don't retrieve the indirection table from the device at any point which means that this conversion is not needed.
The bug:
The for loop iterates over all io_sq_queues, when passing the actual number of used queues the io_sq_queues[i].idx equals 0 since they are uninitialized which results in the following code to be executed till the end of the loop:
dev_idx_to_host_tbl[0] = i;
This results dev_idx_to_host_tbl[0] in being equal to ENA_TOTAL_NUM_QUEUES - 1.
--- drivers/net/ethernet/amazon/ena/ena_com.c | 28 ----------------------- 1 file changed, 28 deletions(-)
drivers/net/ethernet/amazon/ena/ena_com.c +0 -28
@@ -1281,30 +1281,6 @@ static int ena_com_ind_tbl_convert_to_device(struct ena_com_dev *ena_dev)return 0 ; } -static int ena_com_ind_tbl_convert_from_device (struct ena_com_dev *ena_dev) -{ -u16 dev_idx_to_host_tbl[ENA_TOTAL_NUM_QUEUES ] = { (u16)-1 }; -struct ena_rss *rss = &ena_dev->rss; -u8 idx; -u16 i; - -for (i =0 ; i <ENA_TOTAL_NUM_QUEUES ; i++) - dev_idx_to_host_tbl[ena_dev->io_sq_queues[i].idx] = i; - -for (i =0 ; i <1 << rss->tbl_log_size; i++) { -if (rss->rss_ind_tbl[i].cq_idx >ENA_TOTAL_NUM_QUEUES ) -return -EINVAL ; - idx = (u8 )rss->rss_ind_tbl[i].cq_idx; - -if (dev_idx_to_host_tbl[idx] >ENA_TOTAL_NUM_QUEUES ) -return -EINVAL ; - - rss->host_rss_ind_tbl[i] = dev_idx_to_host_tbl[idx]; - } - -return 0 ; -} -static void ena_com_update_intr_delay_resolution (struct ena_com_dev *ena_dev,u16 intr_delay_resolution) {
@@ -2638,10 +2614,6 @@ int ena_com_indirect_table_get(struct ena_com_dev *ena_dev, u32 *ind_tbl)if (!ind_tbl)return 0 ; - rc =ena_com_ind_tbl_convert_from_device (ena_dev); -if (unlikely (rc)) -return rc; -for (i =0 ; i < (1 << rss->tbl_log_size); i++) ind_tbl[i] = rss->host_rss_ind_tbl[i];
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 58/152 net: ena: ethtool: use correct value for crc32 hash
In reply to Greg Kroah-HartmanFrom: Sameeh Jubran <[email protected]>
[ Upstream commit 886d2089276e40d460731765083a741c5c762461 ]
Up till kernel 4.11 there was no enum defined for crc32 hash in ethtool, thus the xor enum was used for supporting crc32.
--- drivers/net/ethernet/amazon/ena/ena_ethtool.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/amazon/ena/ena_ethtool.c +2 -2
@@ -693,7 +693,7 @@ static int ena_get_rxfh(struct net_device *netdev, u32 *indir, u8 *key, func =ETH_RSS_HASH_TOP ;break ;case ENA_ADMIN_CRC32 : - func =ETH_RSS_HASH_XOR ; + func =ETH_RSS_HASH_CRC32 ;break ;default : netif_err(adapter, drv, netdev,
@@ -739,7 +739,7 @@ static int ena_set_rxfh(struct net_device *netdev, const u32 *indir,case ETH_RSS_HASH_TOP : func =ENA_ADMIN_TOEPLITZ ;break ; -case ETH_RSS_HASH_XOR : +case ETH_RSS_HASH_CRC32 : func =ENA_ADMIN_CRC32 ;break ;default :
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 59/152 net: ena: ena-com.c: prevent NULL pointer dereference
In reply to Greg Kroah-HartmanFrom: Arthur Kiyanovski <[email protected]>
[ Upstream commit c207979f5ae10ed70aff1bb13f39f0736973de99 ]
comp_ctx can be NULL in a very rare case when an admin command is executed during the execution of ena_remove().
The bug scenario is as follows:
* ena_destroy_device() sets the comp_ctx to be NULL
* An admin command is executed before executing unregister_netdev(), this can still happen because our device can still receive callbacks from the netdev infrastructure such as ethtool commands.
* When attempting to access the comp_ctx, the bug occurs since it's set to NULL
Fix:
Added a check that comp_ctx is not NULL
--- drivers/net/ethernet/amazon/ena/ena_com.c | 5 +++++ 1 file changed, 5 insertions(+)
drivers/net/ethernet/amazon/ena/ena_com.c +5 -0
@@ -200,6 +200,11 @@ static void comp_ctxt_release(struct ena_com_admin_queue *queue,static struct ena_comp_ctx *get_comp_ctxt (struct ena_com_admin_queue *queue,u16 command_id,bool capture) { +if (unlikely (!queue->comp_ctx)) { +pr_err ("Completion context is NULL\n" ); +return NULL ; + } +if (unlikely (command_id >= queue->q_depth)) {pr_err ("command id is larger than the queue size. cmd_id: %u queue size %d\n" , command_id, queue->q_depth);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 60/152 ice: update Unit Load Status bitmask to check after reset
In reply to Greg Kroah-HartmanFrom: Bruce Allan <[email protected]>
[ Upstream commit cf8fc2a0863f9ff27ebd2efcdb1f7d378b9fb8a6 ]
After a reset the Unit Load Status bits in the GLNVM_ULD register to check for completion should be 0x7FF before continuing. Update the mask to check (minus the three reserved bits that are always set).
--- drivers/net/ethernet/intel/ice/ice_common.c | 17 ++++++++++++----- drivers/net/ethernet/intel/ice/ice_hw_autogen.h | 6 ++++++ 2 files changed, 18 insertions(+), 5 deletions(-)
drivers/net/ethernet/intel/ice/ice_common.c +12 -5
@@ -934,7 +934,7 @@ void ice_deinit_hw(struct ice_hw *hw)*/ enum ice_status ice_check_reset (struct ice_hw *hw) { -u32 cnt, reg =0 , grst_delay; +u32 cnt, reg =0 , grst_delay, uld_mask;/* Poll for Device Active state in case a recent CORER, GLOBR, * or EMPR has occurred. The grst delay value is in 100ms units.
@@ -956,13 +956,20 @@ enum ice_status ice_check_reset(struct ice_hw *hw)return ICE_ERR_RESET_FAILED ; } -#define ICE_RESET_DONE_MASK (GLNVM_ULD_CORER_DONE_M | \ - GLNVM_ULD_GLOBR_DONE_M) +#define ICE_RESET_DONE_MASK (GLNVM_ULD_PCIER_DONE_M |\ + GLNVM_ULD_PCIER_DONE_1_M |\ + GLNVM_ULD_CORER_DONE_M |\ + GLNVM_ULD_GLOBR_DONE_M |\ + GLNVM_ULD_POR_DONE_M |\ + GLNVM_ULD_POR_DONE_1_M |\ + GLNVM_ULD_PCIER_DONE_2_M) + + uld_mask =ICE_RESET_DONE_MASK ;/* Device is Active; check Global Reset processes are done */ for (cnt =0 ; cnt <ICE_PF_RESET_WAIT_COUNT ; cnt++) { - reg =rd32 (hw,GLNVM_ULD ) &ICE_RESET_DONE_MASK ; -if (reg ==ICE_RESET_DONE_MASK ) { + reg =rd32 (hw,GLNVM_ULD ) & uld_mask; +if (reg == uld_mask) {ice_debug (hw,ICE_DBG_INIT ,"Global reset processes done. %d\n" , cnt);break ;
drivers/net/ethernet/intel/ice/ice_hw_autogen.h +6 -0
@@ -273,8 +273,14 @@#define GLNVM_GENS_SR_SIZE_S 5#define GLNVM_GENS_SR_SIZE_M ICE_M(0x7, 5)#define GLNVM_ULD 0x000B6008 +#define GLNVM_ULD_PCIER_DONE_M BIT(0) +#define GLNVM_ULD_PCIER_DONE_1_M BIT(1)#define GLNVM_ULD_CORER_DONE_M BIT(3)#define GLNVM_ULD_GLOBR_DONE_M BIT(4) +#define GLNVM_ULD_POR_DONE_M BIT(5) +#define GLNVM_ULD_POR_DONE_1_M BIT(8) +#define GLNVM_ULD_PCIER_DONE_2_M BIT(9) +#define GLNVM_ULD_PE_DONE_M BIT(10)#define GLPCI_CNF2 0x000BE004#define GLPCI_CNF2_CACHELINE_SIZE_M BIT(1)#define PF_FUNC_RID 0x0009E880
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 61/152 cifs: Fix mode output in debugging statements
In reply to Greg Kroah-HartmanFrom: Frank Sorenson <[email protected]>
[ Upstream commit f52aa79df43c4509146140de0241bc21a4a3b4c7 ]
A number of the debug statements output file or directory mode in hex. Change these to print using octal.
--- fs/cifs/cifsacl.c | 4 ++-- fs/cifs/connect.c | 2 +- fs/cifs/inode.c | 2 +- 3 files changed, 4 insertions(+), 4 deletions(-)
fs/cifs/cifsacl.c +2 -2
@@ -603,7 +603,7 @@ static void access_flags_to_mode(__le32 ace_flags, int type, umode_t *pmode, ((flags & FILE_EXEC_RIGHTS) == FILE_EXEC_RIGHTS)) *pmode |= (S_IXUGO & (*pbits_to_set)); -cifs_dbg (NOISY ,"access flags 0x%x mode now 0x%x\n" , flags, *pmode); +cifs_dbg (NOISY ,"access flags 0x%x mode now %04o\n" , flags, *pmode);return ; }
@@ -632,7 +632,7 @@ static void mode_to_access_flags(umode_t mode, umode_t bits_to_use,if (mode &S_IXUGO ) *pace_flags |=SET_FILE_EXEC_RIGHTS ; -cifs_dbg (NOISY ,"mode: 0x%x, access flags now 0x%x\n" , +cifs_dbg (NOISY ,"mode: %04o, access flags now 0x%x\n" , mode, *pace_flags);return ; }
fs/cifs/connect.c +1 -1
@@ -4094,7 +4094,7 @@ int cifs_setup_cifs_sb(struct smb_vol *pvolume_info, cifs_sb->mnt_gid = pvolume_info->linux_gid; cifs_sb->mnt_file_mode = pvolume_info->file_mode; cifs_sb->mnt_dir_mode = pvolume_info->dir_mode; -cifs_dbg (FYI ,"file mode: 0x%hx dir mode: 0x%hx\n" , +cifs_dbg (FYI ,"file mode: %04ho dir mode: %04ho\n" , cifs_sb->mnt_file_mode, cifs_sb->mnt_dir_mode); cifs_sb->actimeo = pvolume_info->actimeo;
fs/cifs/inode.c +1 -1
@@ -1586,7 +1586,7 @@ int cifs_mkdir(struct inode *inode, struct dentry *direntry, umode_t mode)struct TCP_Server_Info *server;char *full_path; -cifs_dbg (FYI ,"In cifs_mkdir, mode = 0x%hx inode = 0x%p\n" , +cifs_dbg (FYI ,"In cifs_mkdir, mode = %04ho inode = 0x%p\n" , mode, inode); cifs_sb =CIFS_SB (inode->i_sb);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 62/152 bcache: ignore pending signals when creating gc and allocator thread
In reply to Greg Kroah-HartmanFrom: Coly Li <[email protected]>
[ Upstream commit 0b96da639a4874311e9b5156405f69ef9fc3bef8 ]
When run a cache set, all the bcache btree node of this cache set will be checked by bch_btree_check(). If the bcache btree is very large, iterating all the btree nodes will occupy too much system memory and the bcache registering process might be selected and killed by system OOM killer. kthread_run() will fail if current process has pending signal, therefore the kthread creating in run_cache_set() for gc and allocator kernel threads are very probably failed for a very large bcache btree.
Indeed such OOM is safe and the registering process will exit after the registration done. Therefore this patch flushes pending signals during the cache set start up, specificly in bch_cache_allocator_start() and bch_gc_thread_start(), to make sure run_cache_set() won't fail for large cahced data set.
--- drivers/md/bcache/alloc.c | 18 ++++++++++++++++-- drivers/md/bcache/btree.c | 13 +++++++++++++ 2 files changed, 29 insertions(+), 2 deletions(-)
drivers/md/bcache/alloc.c +16 -2
@@ -67,6 +67,7 @@#include <linux/blkdev.h> #include <linux/kthread.h> #include <linux/random.h> +#include <linux/sched/signal.h> #include <trace/events/bcache.h> #define MAX_OPEN_BUCKETS 128
@@ -733,8 +734,21 @@ int bch_open_buckets_alloc(struct cache_set *c)int bch_cache_allocator_start (struct cache *ca) { -struct task_struct *k =kthread_run (bch_allocator_thread, - ca,"bcache_allocator" ); +struct task_struct *k; + +/* +* In case previous btree check operation occupies too many +* system memory for bcache btree node cache, and the +* registering process is selected by OOM killer. Here just +* ignore the SIGKILL sent by OOM killer if there is, to +* avoid kthread_run() being failed by pending signals. The +* bcache registering process will exit after the registration +* done. +*/ +if (signal_pending (current)) +flush_signals (current); + + k =kthread_run (bch_allocator_thread, ca,"bcache_allocator" );if (IS_ERR (k))return PTR_ERR (k);
drivers/md/bcache/btree.c +13 -0
@@ -34,6 +34,7 @@#include <linux/random.h> #include <linux/rcupdate.h> #include <linux/sched/clock.h> +#include <linux/sched/signal.h> #include <linux/rculist.h> #include <linux/delay.h> #include <trace/events/bcache.h>
@@ -1908,6 +1909,18 @@ static int bch_gc_thread(void *arg)int bch_gc_thread_start (struct cache_set *c) { +/* +* In case previous btree check operation occupies too many +* system memory for bcache btree node cache, and the +* registering process is selected by OOM killer. Here just +* ignore the SIGKILL sent by OOM killer if there is, to +* avoid kthread_run() being failed by pending signals. The +* bcache registering process will exit after the registration +* done. +*/ +if (signal_pending (current)) +flush_signals (current); + c->gc_thread =kthread_run (bch_gc_thread, c,"bcache_gc" );return PTR_ERR_OR_ZERO (c->gc_thread); }
Signature · 1 line
2.20.1
Re: [PATCH 5.4 062/152] bcache: ignore pending signals when creating gc and allocator thread
Quoted 13 lines · Greg Kroah-Hartman
From: Coly Li <[email protected]>
[ Upstream commit
0b96da639a4874311e9b5156405f69ef9fc3bef8]When run a cache set, all the bcache btree node of this cache set will be checked by
bch_btree_check(). If the bcache btree is very large, iterating all the btree nodes will occupy too much system memory and the bcache registering process might be selected and killed by system OOM killer.kthread_run()will fail if current process has pending signal, therefore the kthread creating inrun_cache_set()for gc and allocator kernel threads are very probably failed for a very large bcache btree.
Quoted 5 lines
Indeed such OOM is safe and the registering process will exit after the registration done. Therefore this patch flushes pending signals during the cache set start up, specificly in
bch_cache_allocator_start()andbch_gc_thread_start(), to make surerun_cache_set()won't fail for large cahced data set.
Please drop this one, it's being reverted in mainline.
Signature · 1 line
Jens Axboe
Greg Kroah-Hartman
maintainerRe: [PATCH 5.4 062/152] bcache: ignore pending signals when creating gc and allocator thread
Quoted 14 lines · Jens Axboe
Quoted 13 lines · Greg Kroah-Hartman
From: Coly Li <[email protected]>
[ Upstream commit
0b96da639a4874311e9b5156405f69ef9fc3bef8]When run a cache set, all the bcache btree node of this cache set will be checked by
bch_btree_check(). If the bcache btree is very large, iterating all the btree nodes will occupy too much system memory and the bcache registering process might be selected and killed by system OOM killer.kthread_run()will fail if current process has pending signal, therefore the kthread creating inrun_cache_set()for gc and allocator kernel threads are very probably failed for a very large bcache btree.
Quoted 7 lines
Quoted 5 lines
Indeed such OOM is safe and the registering process will exit after the registration done. Therefore this patch flushes pending signals during the cache set start up, specificly in
bch_cache_allocator_start()andbch_gc_thread_start(), to make surerun_cache_set()won't fail for large cahced data set.Please drop this one, it's being reverted in mainline.
Dropped from all trees now, thanks.
greg k-h
Re: [PATCH 5.4 062/152] bcache: ignore pending signals when creating gc and allocator thread
Quoted 21 lines · Greg Kroah-Hartman
Quoted 20 lines · Jens Axboe
Quoted 18 lines · Greg Kroah-Hartman
From: Coly Li <[email protected]>
[ Upstream commit
0b96da639a4874311e9b5156405f69ef9fc3bef8]When run a cache set, all the bcache btree node of this cache set will be checked by
bch_btree_check(). If the bcache btree is very large, iterating all the btree nodes will occupy too much system memory and the bcache registering process might be selected and killed by system OOM killer.kthread_run()will fail if current process has pending signal, therefore the kthread creating inrun_cache_set()for gc and allocator kernel threads are very probably failed for a very large bcache btree.Indeed such OOM is safe and the registering process will exit after the registration done. Therefore this patch flushes pending signals during the cache set start up, specificly in
bch_cache_allocator_start()andbch_gc_thread_start(), to make surerun_cache_set()won't fail for large cahced data set.
Quoted 3 lines
Quoted 1 line
Please drop this one, it's being reverted in mainline.
Dropped from all trees now, thanks.
Thanks!
Signature · 1 line
Jens Axboe
Greg Kroah-Hartman
maintainerPATCH 63/152 cfg80211: add missing policy for NL80211_ATTR_STATUS_CODE
In reply to Greg Kroah-HartmanFrom: Sergey Matyukevich <[email protected]>
[ Upstream commit ea75080110a4c1fa011b0a73cb8f42227143ee3e ]
The nl80211_policy is missing for NL80211_ATTR_STATUS_CODE attribute. As a result, for strictly validated commands, it's assumed to not be supported.
--- net/wireless/nl80211.c | 1 + 1 file changed, 1 insertion(+)
net/wireless/nl80211.c +1 -0
@@ -437,6 +437,7 @@ const struct nla_policy nl80211_policy[NUM_NL80211_ATTR] = { [NL80211_ATTR_CONTROL_PORT_NO_ENCRYPT] = { .type = NLA_FLAG }, [NL80211_ATTR_CONTROL_PORT_OVER_NL80211] = { .type = NLA_FLAG }, [NL80211_ATTR_PRIVACY] = { .type = NLA_FLAG }, + [NL80211_ATTR_STATUS_CODE] = { .type = NLA_U16 }, [NL80211_ATTR_CIPHER_SUITE_GROUP] = { .type = NLA_U32 }, [NL80211_ATTR_WPA_VERSIONS] = { .type = NLA_U32 }, [NL80211_ATTR_PID] = { .type = NLA_U32 },
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 64/152 mac80211: fix wrong 160/80+80 MHz setting
In reply to Greg Kroah-HartmanFrom: Shay Bar <[email protected]>
[ Upstream commit 33181ea7f5a62a17fbe55f0f73428ecb5e686be8 ]
Before this patch, STA's would set new width of 160/80+80 MHz based on AP capability only. This is wrong because STA may not support > 80MHz BW.
Fix is to verify STA has 160/80+80 MHz capability before increasing its width to > 80MHz.
The "support_80_80" and "support_160" setting is based on:
"Table 9-272 — Setting of the Supported Channel Width Set subfield and Extended NSS BW Support subfield at a STA transmitting the VHT Capabilities Information field"
Quoted 1 line
From "Draft P802.11REVmd_D3.0.pdf"
--- net/mac80211/util.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-)
net/mac80211/util.c +14 -2
@@ -2993,10 +2993,22 @@ bool ieee80211_chandef_vht_oper(struct ieee80211_hw *hw,int cf0, cf1;int ccfs0, ccfs1, ccfs2;int ccf0, ccf1; +u32 vht_cap; +bool support_80_80 = false; +bool support_160 = false;if (!oper || !htop)return false; + vht_cap = hw->wiphy->bands[chandef->chan->band]->vht_cap.cap; + support_160 = (vht_cap & (IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_MASK | +IEEE80211_VHT_CAP_EXT_NSS_BW_MASK )); + support_80_80 = ((vht_cap & +IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_160_80PLUS80MHZ ) || + (vht_cap &IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_160MHZ && + vht_cap &IEEE80211_VHT_CAP_EXT_NSS_BW_MASK ) || + ((vht_cap &IEEE80211_VHT_CAP_EXT_NSS_BW_MASK ) >> +IEEE80211_VHT_CAP_EXT_NSS_BW_SHIFT >1 )); ccfs0 = oper->center_freq_seg0_idx; ccfs1 = oper->center_freq_seg1_idx; ccfs2 = (le16_to_cpu(htop->operation_mode) &
@@ -3024,10 +3036,10 @@ bool ieee80211_chandef_vht_oper(struct ieee80211_hw *hw,unsigned int diff; diff =abs (ccf1 - ccf0); -if (diff ==8 ) { +if ((diff ==8 ) && support_160) { new.width =NL80211_CHAN_WIDTH_160 ; new.center_freq1 = cf1; - }else if (diff >8 ) { + }else if ((diff >8 ) && support_80_80) { new.width =NL80211_CHAN_WIDTH_80P80 ; new.center_freq2 = cf1; }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 65/152 net: hns3: add management table after IMP reset
In reply to Greg Kroah-HartmanFrom: Yufeng Mo <[email protected]>
[ Upstream commit d0db7ed397517c8b2be24a0d1abfa15df776908e ]
In the current process, the management table is missing after the IMP reset. This patch adds the management table to the reset process.
--- drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 7 +++++++ 1 file changed, 7 insertions(+)
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +7 -0
@@ -9437,6 +9437,13 @@ static int hclge_reset_ae_dev(struct hnae3_ae_dev *ae_dev)return ret; } + ret =init_mgr_tbl (hdev); +if (ret) { +dev_err (&pdev->dev, +"failed to reinit manager table, ret = %d\n" , ret); +return ret; + } + ret =hclge_init_fd_config (hdev);if (ret) {dev_err (&pdev->dev,"fd table init fail, ret=%d\n" , ret);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 66/152 net: hns3: fix a copying IPv6 address error in hclge_fd_get_flow_tuples()
In reply to Greg Kroah-HartmanFrom: Guangbin Huang <[email protected]>
[ Upstream commit 47327c9315b2f3ae4ab659457977a26669631f20 ]
The IPv6 address defined in struct in6_addr is specified as big endian, but there is no specified endian in struct hclge_fd_rule_tuples, so it will cause a problem if directly use memcpy() to copy ipv6 address between these two structures since this field in struct hclge_fd_rule_tuples is little endian.
This patch fixes this problem by using be32_to_cpu() to convert endian of IPv6 address of struct in6_addr before copying.
--- .../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-)
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +9 -6
@@ -6003,6 +6003,9 @@ static int hclge_get_all_rules(struct hnae3_handle *handle,static void hclge_fd_get_flow_tuples (const struct flow_keys *fkeys,struct hclge_fd_rule_tuples *tuples) { +#define flow_ip6_src fkeys->addrs.v6addrs.src.in6_u.u6_addr32 +#define flow_ip6_dst fkeys->addrs.v6addrs.dst.in6_u.u6_addr32 + tuples->ether_proto =be16_to_cpu (fkeys->basic.n_proto); tuples->ip_proto = fkeys->basic.ip_proto; tuples->dst_port =be16_to_cpu (fkeys->ports.dst);
@@ -6011,12 +6014,12 @@ static void hclge_fd_get_flow_tuples(const struct flow_keys *fkeys, tuples->src_ip[3 ] =be32_to_cpu (fkeys->addrs.v4addrs.src); tuples->dst_ip[3 ] =be32_to_cpu (fkeys->addrs.v4addrs.dst); }else { -memcpy (tuples->src_ip, - fkeys->addrs.v6addrs.src.in6_u.u6_addr32, -sizeof (tuples->src_ip)); -memcpy (tuples->dst_ip, - fkeys->addrs.v6addrs.dst.in6_u.u6_addr32, -sizeof (tuples->dst_ip)); +int i; + +for (i =0 ; i <IPV6_SIZE ; i++) { + tuples->src_ip[i] =be32_to_cpu (flow_ip6_src[i]); + tuples->dst_ip[i] =be32_to_cpu (flow_ip6_dst[i]); + } } }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 67/152 nvme/tcp: fix bug on double requeue when send fails
In reply to Greg Kroah-HartmanFrom: Anton Eidelman <[email protected]>
[ Upstream commit 2d570a7c0251c594489a2c16b82b14ae30345c03 ]
When nvme_tcp_io_work() fails to send to socket due to connection close/reset, error_recovery work is triggered from nvme_tcp_state_change() socket callback. This cancels all the active requests in the tagset, which requeues them.
The failed request, however, was ended and thus requeued individually as well unless send returned -EPIPE. Another return code to be treated the same way is -ECONNRESET.
Double requeue caused BUG_ON(blk_queued_rq(rq))
in blk_mq_requeue_request() from either the individual requeue of the failed request or the bulk requeue from blk_mq_tagset_busy_iter(, nvme_cancel_request, );
--- drivers/nvme/host/tcp.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/nvme/host/tcp.c +6 -1
@@ -1054,7 +1054,12 @@ static void nvme_tcp_io_work(struct work_struct *w) }else if (unlikely (result <0 )) {dev_err (queue->ctrl->ctrl.device,"failed to send request %d\n" , result); -if (result != -EPIPE ) + +/* +* Fail the request unless peer closed the connection, +* in which case error recovery flow will complete all. +*/ +if ((result != -EPIPE ) && (result != -ECONNRESET ))nvme_tcp_fail_request (queue->request);nvme_tcp_done_send_req (queue);return ;
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 68/152 nvme: prevent warning triggered by nvme_stop_keep_alive
In reply to Greg Kroah-HartmanFrom: Nigel Kirkland <[email protected]>
[ Upstream commit 97b2512ad000a409b4073dd1a71e4157d76675cb ]
Delayed keep alive work is queued on system workqueue and may be cancelled via nvme_stop_keep_alive from nvme_reset_wq, nvme_fc_wq or nvme_wq.
Check_flush_dependency detects mismatched attributes between the work-queue context used to cancel the keep alive work and system-wq. Specifically system-wq does not have the WQ_MEM_RECLAIM flag, whereas the contexts used to cancel keep alive work have WQ_MEM_RECLAIM flag.
Example warning:
workqueue: WQ_MEM_RECLAIM nvme-reset-wq:nvme_fc_reset_ctrl_work [nvme_fc]
is flushing !WQ_MEM_RECLAIM events:nvme_keep_alive_work [nvme_core]
To avoid the flags mismatch, delayed keep alive work is queued on nvme_wq.
However this creates a secondary concern where work and a request to cancel that work may be in the same work queue - namely err_work in the rdma and tcp transports, which will want to flush/cancel the keep alive work which will now be on nvme_wq.
After reviewing the transports, it looks like err_work can be moved to nvme_reset_wq. In fact that aligns them better with transition into RESETTING and performing related reset work in nvme_reset_wq.
Change nvme-rdma and nvme-tcp to perform err_work in nvme_reset_wq.
--- drivers/nvme/host/core.c | 10 +++++----- drivers/nvme/host/rdma.c | 2 +- drivers/nvme/host/tcp.c | 2 +- 3 files changed, 7 insertions(+), 7 deletions(-)
drivers/nvme/host/core.c +5 -5
@@ -66,8 +66,8 @@ MODULE_PARM_DESC(streams, "turn on support for Streams write directives");* nvme_reset_wq - hosts nvme reset works * nvme_delete_wq - hosts nvme delete works * -* nvme_wq will host works such are scan, aen handling, fw activation, -* keep-alive error recovery, periodic reconnects etc. nvme_reset_wq +* nvme_wq will host works such as scan, aen handling, fw activation, +* keep-alive, periodic reconnects etc. nvme_reset_wq * runs reset works which also flush works hosted on nvme_wq for * serialization purposes. nvme_delete_wq host controller deletion * works which flush reset works for serialization.
@@ -972,7 +972,7 @@ static void nvme_keep_alive_end_io(struct request *rq, blk_status_t status) startka = true;spin_unlock_irqrestore (&ctrl->lock, flags);if (startka) -schedule_delayed_work (&ctrl->ka_work, ctrl->kato *HZ ); +queue_delayed_work (nvme_wq, &ctrl->ka_work, ctrl->kato *HZ ); }static int nvme_keep_alive (struct nvme_ctrl *ctrl)
@@ -1002,7 +1002,7 @@ static void nvme_keep_alive_work(struct work_struct *work)dev_dbg (ctrl->device,"reschedule traffic based keep-alive timer\n" ); ctrl->comp_seen = false; -schedule_delayed_work (&ctrl->ka_work, ctrl->kato *HZ ); +queue_delayed_work (nvme_wq, &ctrl->ka_work, ctrl->kato *HZ );return ; }
@@ -1019,7 +1019,7 @@ static void nvme_start_keep_alive(struct nvme_ctrl *ctrl)if (unlikely (ctrl->kato ==0 ))return ; -schedule_delayed_work (&ctrl->ka_work, ctrl->kato *HZ ); +queue_delayed_work (nvme_wq, &ctrl->ka_work, ctrl->kato *HZ ); }void nvme_stop_keep_alive (struct nvme_ctrl *ctrl)
drivers/nvme/host/rdma.c +1 -1
@@ -1088,7 +1088,7 @@ static void nvme_rdma_error_recovery(struct nvme_rdma_ctrl *ctrl)if (!nvme_change_ctrl_state (&ctrl->ctrl,NVME_CTRL_RESETTING ))return ; -queue_work (nvme_wq, &ctrl->err_work); +queue_work (nvme_reset_wq, &ctrl->err_work); } static void nvme_rdma_wr_error(struct ib_cq *cq, struct ib_wc *wc,
drivers/nvme/host/tcp.c +1 -1
@@ -422,7 +422,7 @@ static void nvme_tcp_error_recovery(struct nvme_ctrl *ctrl)if (!nvme_change_ctrl_state (ctrl,NVME_CTRL_RESETTING ))return ; -queue_work (nvme_wq, &to_tcp_ctrl (ctrl)->err_work); +queue_work (nvme_reset_wq, &to_tcp_ctrl (ctrl)->err_work); } static int nvme_tcp_process_nvme_cqe(struct nvme_tcp_queue *queue,
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 69/152 nvme/pci: move cqe check after device shutdown
In reply to Greg Kroah-HartmanFrom: Keith Busch <[email protected]>
[ Upstream commit fa46c6fb5d61b1f17b06d7c6ef75478b576304c7 ]
Many users have reported nvme triggered irq_startup() warnings during shutdown. The driver uses the nvme queue's irq to synchronize scanning for completions, and enabling an interrupt affined to only offline CPUs triggers the alarming warning.
Move the final CQE check to after disabling the device and all registered interrupts have been torn down so that we do not have any IRQ to synchronize.
--- drivers/nvme/host/pci.c | 23 ++++++++++++++++++----- 1 file changed, 18 insertions(+), 5 deletions(-)
drivers/nvme/host/pci.c +18 -5
@@ -1407,6 +1407,23 @@ static void nvme_disable_admin_queue(struct nvme_dev *dev, bool shutdown)nvme_poll_irqdisable (nvmeq,-1 ); } +/* +* Called only on a device that has been disabled and after all other threads +* that can check this device's completion queues have synced. This is the +* last chance for the driver to see a natural completion before +* nvme_cancel_request() terminates all incomplete requests. +*/ +static void nvme_reap_pending_cqes (struct nvme_dev *dev) +{ +u16 start, end; +int i; + +for (i = dev->ctrl.queue_count -1 ; i >0 ; i--) { +nvme_process_cq (&dev->queues[i], &start, &end,-1 ); +nvme_complete_cqes (&dev->queues[i], start, end); + } +} +static int nvme_cmb_qdepth (struct nvme_dev *dev,int nr_io_queues,int entry_size) {
@@ -2241,11 +2258,6 @@ static bool __nvme_disable_io_queues(struct nvme_dev *dev, u8 opcode)if (timeout ==0 )return false; -/* handle any remaining CQEs */ -if (opcode == nvme_admin_delete_cq && - !test_bit (NVMEQ_DELETE_ERROR , &nvmeq->flags)) -nvme_poll_irqdisable (nvmeq,-1 ); - sent--;if (nr_queues) goto retry;
@@ -2434,6 +2446,7 @@ static void nvme_dev_disable(struct nvme_dev *dev, bool shutdown)nvme_suspend_io_queues (dev);nvme_suspend_queue (&dev->queues[0 ]);nvme_pci_disable (dev); +nvme_reap_pending_cqes (dev);blk_mq_tagset_busy_iter (&dev->tagset, nvme_cancel_request, &dev->ctrl);blk_mq_tagset_busy_iter (&dev->admin_tagset, nvme_cancel_request, &dev->ctrl);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 70/152 ext4: potential crash on allocation error in ext4_alloc_flex_bg_array()
In reply to Greg Kroah-HartmanFrom: Dan Carpenter <[email protected]>
commit 37b0b6b8b99c0e1c1f11abbe7cf49b6d03795b3f upstream.
If sbi->s_flex_groups_allocated is zero and the first allocation fails then this code will crash. The problem is that "i--" will set "i" to -1 but when we compare "i >= sbi->s_flex_groups_allocated" then the -1 is type promoted to unsigned and becomes UINT_MAX. Since UINT_MAX is more than zero, the condition is true so we call kvfree(new_groups[-1]). The loop will carry on freeing invalid memory until it crashes.
--- fs/ext4/super.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
fs/ext4/super.c +3 -3
@@ -2343,7 +2343,7 @@ int ext4_alloc_flex_bg_array(struct supe {struct ext4_sb_info *sbi =EXT4_SB (sb);struct flex_groups **old_groups, **new_groups; -int size, i; +int size, i, j;if (!sbi->s_log_groups_per_flex)return 0 ;
@@ -2364,8 +2364,8 @@ int ext4_alloc_flex_bg_array(struct supe sizeof(struct flex_groups)), GFP_KERNEL);if (!new_groups[i]) { -for (i--; i >= sbi->s_flex_groups_allocated; i--) -kvfree (new_groups[i]); +for (j = sbi->s_flex_groups_allocated; j < i; j++) +kvfree (new_groups[j]);kvfree (new_groups);ext4_msg (sb,KERN_ERR ,"not enough memory for %d flex groups" , size);
Greg Kroah-Hartman
maintainerPATCH 71/152 audit: fix error handling in audit_data_to_entry()
In reply to Greg Kroah-HartmanFrom: Paul Moore <[email protected]>
commit 2ad3e17ebf94b7b7f3f64c050ff168f9915345eb upstream.
Commit 219ca39427bf ("audit: use union for audit_field values since they are mutually exclusive") combined a number of separate fields in the audit_field struct into a single union. Generally this worked just fine because they are generally mutually exclusive. Unfortunately in audit_data_to_entry() the overlap can be a problem when a specific error case is triggered that causes the error path code to attempt to cleanup an audit_field struct and the cleanup involves attempting to free a stored LSM string (the lsm_str field). Currently the code always has a non-NULL value in the audit_field.lsm_str field as the top of the for-loop transfers a value into audit_field.val (both .lsm_str and .val are part of the same union); if audit_data_to_entry() fails and the audit_field struct is specified to contain a LSM string, but the audit_field.lsm_str has not yet been properly set, the error handling code will attempt to free the bogus audit_field.lsm_str value that was set with audit_field.val at the top of the for-loop.
This patch corrects this by ensuring that the audit_field.val is only set when needed (it is cleared when the audit_field struct is allocated with kcalloc()). It also corrects a few other issues to ensure that in case of error the proper error code is returned.
--- kernel/auditfilter.c | 71 ++++++++++++++++++++++++++++----------------------- 1 file changed, 39 insertions(+), 32 deletions(-)
kernel/auditfilter.c +39 -32
@@ -456,6 +456,7 @@ static struct audit_entry *audit_data_to bufp = data->buf;for (i =0 ; i < data->field_count; i++) {struct audit_field *f = &entry->rule.fields[i]; +u32 f_val; err = -EINVAL ;
@@ -464,12 +465,12 @@ static struct audit_entry *audit_data_to goto exit_free; f->type = data->fields[i]; - f->val = data->values[i]; + f_val = data->values[i];/* Support legacy tests for a valid loginuid */ -if ((f->type ==AUDIT_LOGINUID ) && (f->val ==AUDIT_UID_UNSET )) { +if ((f->type ==AUDIT_LOGINUID ) && (f_val ==AUDIT_UID_UNSET )) { f->type =AUDIT_LOGINUID_SET ; - f->val =0 ; + f_val =0 ; entry->rule.pflags |=AUDIT_LOGINUID_LEGACY ; }
@@ -485,7 +486,7 @@ static struct audit_entry *audit_data_tocase AUDIT_SUID :case AUDIT_FSUID :case AUDIT_OBJ_UID : - f->uid =make_kuid (current_user_ns (), f->val); + f->uid =make_kuid (current_user_ns (), f_val);if (!uid_valid (f->uid)) goto exit_free;break ;
@@ -494,11 +495,12 @@ static struct audit_entry *audit_data_tocase AUDIT_SGID :case AUDIT_FSGID :case AUDIT_OBJ_GID : - f->gid =make_kgid (current_user_ns (), f->val); + f->gid =make_kgid (current_user_ns (), f_val);if (!gid_valid (f->gid)) goto exit_free;break ;case AUDIT_ARCH : + f->val = f_val; entry->rule.arch_f = f;break ;case AUDIT_SUBJ_USER :
@@ -511,11 +513,13 @@ static struct audit_entry *audit_data_tocase AUDIT_OBJ_TYPE :case AUDIT_OBJ_LEV_LOW :case AUDIT_OBJ_LEV_HIGH : - str =audit_unpack_string (&bufp, &remain, f->val); -if (IS_ERR (str)) + str =audit_unpack_string (&bufp, &remain, f_val); +if (IS_ERR (str)) { + err =PTR_ERR (str); goto exit_free; - entry->rule.buflen += f->val; - + } + entry->rule.buflen += f_val; + f->lsm_str = str; err =security_audit_rule_init (f->type, f->op, str, (void **)&f->lsm_rule);/* Keep currently invalid fields around in case they
@@ -524,68 +528,71 @@ static struct audit_entry *audit_data_topr_warn ("audit rule for LSM \'%s\' is invalid\n" , str); err =0 ; - } -if (err) { -kfree (str); + }else if (err) goto exit_free; - }else - f->lsm_str = str;break ;case AUDIT_WATCH : - str =audit_unpack_string (&bufp, &remain, f->val); -if (IS_ERR (str)) + str =audit_unpack_string (&bufp, &remain, f_val); +if (IS_ERR (str)) { + err =PTR_ERR (str); goto exit_free; - entry->rule.buflen += f->val; - - err =audit_to_watch (&entry->rule, str, f->val, f->op); + } + err =audit_to_watch (&entry->rule, str, f_val, f->op);if (err) {kfree (str); goto exit_free; } + entry->rule.buflen += f_val;break ;case AUDIT_DIR : - str =audit_unpack_string (&bufp, &remain, f->val); -if (IS_ERR (str)) + str =audit_unpack_string (&bufp, &remain, f_val); +if (IS_ERR (str)) { + err =PTR_ERR (str); goto exit_free; - entry->rule.buflen += f->val; - + } err =audit_make_tree (&entry->rule, str, f->op);kfree (str);if (err) goto exit_free; + entry->rule.buflen += f_val;break ;case AUDIT_INODE : + f->val = f_val; err =audit_to_inode (&entry->rule, f);if (err) goto exit_free;break ;case AUDIT_FILTERKEY : -if (entry->rule.filterkey || f->val >AUDIT_MAX_KEY_LEN ) +if (entry->rule.filterkey || f_val >AUDIT_MAX_KEY_LEN ) goto exit_free; - str =audit_unpack_string (&bufp, &remain, f->val); -if (IS_ERR (str)) + str =audit_unpack_string (&bufp, &remain, f_val); +if (IS_ERR (str)) { + err =PTR_ERR (str); goto exit_free; - entry->rule.buflen += f->val; + } + entry->rule.buflen += f_val; entry->rule.filterkey = str;break ;case AUDIT_EXE : -if (entry->rule.exe || f->val >PATH_MAX ) +if (entry->rule.exe || f_val >PATH_MAX ) goto exit_free; - str =audit_unpack_string (&bufp, &remain, f->val); + str =audit_unpack_string (&bufp, &remain, f_val);if (IS_ERR (str)) { err =PTR_ERR (str); goto exit_free; } - entry->rule.buflen += f->val; - - audit_mark =audit_alloc_mark (&entry->rule, str, f->val); + audit_mark =audit_alloc_mark (&entry->rule, str, f_val);if (IS_ERR (audit_mark)) {kfree (str); err =PTR_ERR (audit_mark); goto exit_free; } + entry->rule.buflen += f_val; entry->rule.exe = audit_mark;break ; +default : + f->val = f_val; +break ; } }
Greg Kroah-Hartman
maintainerPATCH 72/152 audit: always check the netlink payload length in audit_receive_msg()
In reply to Greg Kroah-HartmanFrom: Paul Moore <[email protected]>
commit 756125289285f6e55a03861bf4b6257aa3d19a93 upstream.
This patch ensures that we always check the netlink payload length in audit_receive_msg() before we take any action on the payload itself.
--- kernel/audit.c | 40 +++++++++++++++++++++------------------- 1 file changed, 21 insertions(+), 19 deletions(-)
kernel/audit.c +21 -19
@@ -1100,13 +1100,11 @@ static void audit_log_feature_change(intaudit_log_end (ab); } -static int audit_set_feature (struct sk_buff *skb) +static int audit_set_feature (struct audit_features *uaf) { -struct audit_features *uaf;int i;BUILD_BUG_ON (AUDIT_LAST_FEATURE +1 >ARRAY_SIZE (audit_feature_names)); - uaf =nlmsg_data (nlmsg_hdr (skb));/* if there is ever a version 2 we should handle that here */
@@ -1174,6 +1172,7 @@ static int audit_receive_msg(struct sk_b {u32 seq;void *data; +int data_len;int err;struct audit_buffer *ab;u16 msg_type = nlh->nlmsg_type;
@@ -1187,6 +1186,7 @@ static int audit_receive_msg(struct sk_b seq = nlh->nlmsg_seq; data =nlmsg_data (nlh); + data_len =nlmsg_len (nlh);switch (msg_type) {case AUDIT_GET : {
@@ -1210,7 +1210,7 @@ static int audit_receive_msg(struct sk_bstruct audit_status s;memset (&s,0 ,sizeof (s));/* guard against past and future API changes */ -memcpy (&s, data,min_t (size_t,sizeof (s),nlmsg_len (nlh))); +memcpy (&s, data,min_t (size_t,sizeof (s), data_len));if (s.mask &AUDIT_STATUS_ENABLED ) { err =audit_set_enabled (s.enabled);if (err <0 )
@@ -1314,7 +1314,9 @@ static int audit_receive_msg(struct sk_breturn err;break ;case AUDIT_SET_FEATURE : - err =audit_set_feature (skb); +if (data_len <sizeof (struct audit_features )) +return -EINVAL ; + err =audit_set_feature (data);if (err)return err;break ;
@@ -1326,6 +1328,8 @@ static int audit_receive_msg(struct sk_b err =audit_filter (msg_type,AUDIT_FILTER_USER );if (err ==1 ) {/* match or error */ +char *str = data; + err =0 ;if (msg_type ==AUDIT_USER_TTY ) { err =tty_audit_push ();
@@ -1333,26 +1337,24 @@ static int audit_receive_msg(struct sk_bbreak ; }audit_log_user_recv_msg (&ab, msg_type); -if (msg_type !=AUDIT_USER_TTY ) +if (msg_type !=AUDIT_USER_TTY ) { +/* ensure NULL termination */ + str[data_len -1 ] ='\0' ;audit_log_format (ab," msg='%.*s'" ,AUDIT_MESSAGE_TEXT_MAX , - (char *)data); -else { -int size; - + str); + }else {audit_log_format (ab," data=" ); - size =nlmsg_len (nlh); -if (size >0 && - ((unsigned char *)data)[size -1 ] =='\0' ) - size--; -audit_log_n_untrustedstring (ab, data, size); +if (data_len >0 && str[data_len -1 ] =='\0' ) + data_len--; +audit_log_n_untrustedstring (ab, str, data_len); }audit_log_end (ab); }break ;case AUDIT_ADD_RULE :case AUDIT_DEL_RULE : -if (nlmsg_len (nlh) <sizeof (struct audit_rule_data )) +if (data_len <sizeof (struct audit_rule_data ))return -EINVAL ;if (audit_enabled ==AUDIT_LOCKED ) { audit_log_common_recv_msg(audit_context(), &ab,
@@ -1364,7 +1366,7 @@ static int audit_receive_msg(struct sk_baudit_log_end (ab);return -EPERM ; } - err =audit_rule_change (msg_type, seq, data,nlmsg_len (nlh)); + err =audit_rule_change (msg_type, seq, data, data_len);break ;case AUDIT_LIST_RULES : err =audit_list_rules_send (skb, seq);
@@ -1379,7 +1381,7 @@ static int audit_receive_msg(struct sk_bcase AUDIT_MAKE_EQUIV : {void *bufp = data;u32 sizes[2 ]; -size_t msglen =nlmsg_len (nlh); +size_t msglen = data_len;char *old, *new; err = -EINVAL ;
@@ -1455,7 +1457,7 @@ static int audit_receive_msg(struct sk_bmemset (&s,0 ,sizeof (s));/* guard against past and future API changes */ -memcpy (&s, data,min_t (size_t,sizeof (s),nlmsg_len (nlh))); +memcpy (&s, data,min_t (size_t,sizeof (s), data_len));/* check if new data is valid */ if ((s.enabled !=0 && s.enabled !=1 ) || (s.log_passwd !=0 && s.log_passwd !=1 ))
Greg Kroah-Hartman
maintainerPATCH 73/152 ACPICA: Introduce ACPI_ACCESS_BYTE_WIDTH() macro
In reply to Greg Kroah-HartmanFrom: Mika Westerberg <[email protected]>
commit 1dade3a7048ccfc675650cd2cf13d578b095e5fb upstream.
Sometimes it is useful to find the access_width field value in bytes and not in bits so add a helper that can be used for this purpose.
--- include/acpi/actypes.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
include/acpi/actypes.h +2 -1
@@ -532,11 +532,12 @@ typedef u64 acpi_integer; strnlen (a, ACPI_NAMESEG_SIZE) == ACPI_NAMESEG_SIZE)/* -* Algorithm to obtain access bit width. +* Algorithm to obtain access bit or byte width. * Can be used with access_width of struct acpi_generic_address and access_size of * struct acpi_resource_generic_register. */ #define ACPI_ACCESS_BIT_WIDTH (size) (1 << ((size) + 2)) +#define ACPI_ACCESS_BYTE_WIDTH (size) (1 << ((size) - 1))/******************************************************************************* *
Greg Kroah-Hartman
maintainerPATCH 74/152 ACPI: watchdog: Fix gas->access_width usage
In reply to Greg Kroah-HartmanFrom: Mika Westerberg <[email protected]>
commit 2ba33a4e9e22ac4dda928d3e9b5978a3a2ded4e0 upstream.
ACPI Generic Address Structure (GAS) access_width field is not in bytes as the driver seems to expect in few places so fix this by using the newly introduced macro ACPI_ACCESS_BYTE_WIDTH().
--- drivers/acpi/acpi_watchdog.c | 3 +-- drivers/watchdog/wdat_wdt.c | 2 +- 2 files changed, 2 insertions(+), 3 deletions(-)
drivers/acpi/acpi_watchdog.c +1 -2
@@ -126,12 +126,11 @@ void __init acpi_watchdog_init(void) gas = &entries[i].register_region; res.start = gas->address; + res.end = res.start +ACPI_ACCESS_BYTE_WIDTH (gas->access_width) -1 ;if (gas->space_id ==ACPI_ADR_SPACE_SYSTEM_MEMORY ) { res.flags =IORESOURCE_MEM ; - res.end = res.start +ALIGN (gas->access_width,4 ) -1 ; }else if (gas->space_id ==ACPI_ADR_SPACE_SYSTEM_IO ) { res.flags =IORESOURCE_IO ; - res.end = res.start + gas->access_width -1 ; }else {pr_warn ("Unsupported address space: %u\n" , gas->space_id);
drivers/watchdog/wdat_wdt.c +1 -1
@@ -389,7 +389,7 @@ static int wdat_wdt_probe(struct platformemset (&r,0 ,sizeof (r)); r.start = gas->address; - r.end = r.start + gas->access_width -1 ; + r.end = r.start +ACPI_ACCESS_BYTE_WIDTH (gas->access_width) -1 ;if (gas->space_id ==ACPI_ADR_SPACE_SYSTEM_MEMORY ) { r.flags =IORESOURCE_MEM ; }else if (gas->space_id ==ACPI_ADR_SPACE_SYSTEM_IO ) {
Greg Kroah-Hartman
maintainerPATCH 75/152 KVM: VMX: check descriptor table exits on instruction emulation
In reply to Greg Kroah-HartmanFrom: Oliver Upton <[email protected]>
commit 86f7e90ce840aa1db407d3ea6e9b3a52b2ce923c upstream.
KVM emulates UMIP on hardware that doesn't support it by setting the 'descriptor table exiting' VM-execution control and performing instruction emulation. When running nested, this emulation is broken as KVM refuses to emulate L2 instructions by default.
Correct this regression by allowing the emulation of descriptor table instructions if L1 hasn't requested 'descriptor table exiting'.
--- arch/x86/kvm/vmx/vmx.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+)
arch/x86/kvm/vmx/vmx.c +15 -0
@@ -7165,6 +7165,7 @@ static int vmx_check_intercept_io(structelse intercept =nested_vmx_check_io_bitmaps (vcpu, port, size); +/* FIXME: produce nested vmexit and return X86EMUL_INTERCEPTED. */ return intercept ?X86EMUL_UNHANDLEABLE :X86EMUL_CONTINUE ; }
@@ -7194,6 +7195,20 @@ static int vmx_check_intercept(struct kvcase x86_intercept_outs:return vmx_check_intercept_io (vcpu, info); +case x86_intercept_lgdt: +case x86_intercept_lidt: +case x86_intercept_lldt: +case x86_intercept_ltr: +case x86_intercept_sgdt: +case x86_intercept_sidt: +case x86_intercept_sldt: +case x86_intercept_str: +if (!nested_cpu_has2 (vmcs12,SECONDARY_EXEC_DESC )) +return X86EMUL_CONTINUE ; + +/* FIXME: produce nested vmexit and return X86EMUL_INTERCEPTED. */ +break ; +/* TODO: check more intercepts... */ default :break ;
Greg Kroah-Hartman
maintainerPATCH 76/152 HID: ite: Only bind to keyboard USB interface on Acer SW5-012 keyboard dock
In reply to Greg Kroah-HartmanFrom: Hans de Goede <[email protected]>
commit beae56192a2570578ae45050e73c5ff9254f63e6 upstream.
Commit 8f18eca9ebc5 ("HID: ite: Add USB id match for Acer SW5-012 keyboard dock") added the USB id for the Acer SW5-012's keyboard dock to the hid-ite driver to fix the rfkill driver not working.
Most keyboard docks with an ITE 8595 keyboard/touchpad controller have the "Wireless Radio Control" bits which need the special hid-ite driver on the second USB interface (the mouse interface) and their touchpad only supports mouse emulation, so using generic hid-input handling for anything but the "Wireless Radio Control" bits is fine. On these devices we simply bind to all USB interfaces.
But unlike other ITE8595 using keyboard docks, the Acer Aspire Switch 10 (SW5-012)'s touchpad not only does mouse emulation it also supports HID-multitouch and all the keys including the "Wireless Radio Control" bits have been moved to the first USB interface (the keyboard intf).
So we need hid-ite to handle the first (keyboard) USB interface and have it NOT bind to the second (mouse) USB interface so that that can be handled by hid-multitouch.c and we get proper multi-touch support.
This commit changes the hid_device_id for the SW5-012 keyboard dock to only match on hid devices from the HID_GROUP_GENERIC group, this way hid-ite will not bind the the mouse/multi-touch interface which has HID_GROUP_MULTITOUCH_WIN_8 as group.
This fixes the regression to mouse-emulation mode introduced by adding the keyboard dock USB id.
--- drivers/hid/hid-ite.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/hid/hid-ite.c +3 -2
@@ -41,8 +41,9 @@ static const struct hid_device_id ite_de { HID_USB_DEVICE(USB_VENDOR_ID_ITE, USB_DEVICE_ID_ITE8595) }, { HID_USB_DEVICE(USB_VENDOR_ID_258A, USB_DEVICE_ID_258A_6A88) },/* ITE8595 USB kbd ctlr, with Synaptics touchpad connected to it. */ - { HID_USB_DEVICE(USB_VENDOR_ID_SYNAPTICS, - USB_DEVICE_ID_SYNAPTICS_ACER_SWITCH5_012) }, + { HID_DEVICE(BUS_USB, HID_GROUP_GENERIC, +USB_VENDOR_ID_SYNAPTICS , + USB_DEVICE_ID_SYNAPTICS_ACER_SWITCH5_012) }, { } };MODULE_DEVICE_TABLE (hid, ite_devices);
Greg Kroah-Hartman
maintainerPATCH 77/152 HID: core: fix off-by-one memset in hid_report_raw_event()
In reply to Greg Kroah-HartmanFrom: Johan Korsnes <[email protected]>
commit 5ebdffd25098898aff1249ae2f7dbfddd76d8f8f upstream.
In case a report is greater than HID_MAX_BUFFER_SIZE, it is truncated, but the report-number byte is not correctly handled. This results in a off-by-one in the following memset, causing a kernel Oops and ensuing system crash.
Note: With commit 8ec321e96e05 ("HID: Fix slab-out-of-bounds read in hid_field_extract") I no longer hit the kernel Oops as we instead fail "controlled" at probe if there is a report too long in the HID report-descriptor. hid_report_raw_event() is an exported symbol, so presumabely we cannot always rely on this being the case.
function.")
--- drivers/hid/hid-core.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/hid/hid-core.c +3 -1
@@ -1741,7 +1741,9 @@ int hid_report_raw_event(struct hid_devi rsize = ((report->size -1 ) >>3 ) +1 ; -if (rsize >HID_MAX_BUFFER_SIZE ) +if (report_enum->numbered && rsize >=HID_MAX_BUFFER_SIZE ) + rsize =HID_MAX_BUFFER_SIZE -1 ; +else if (rsize >HID_MAX_BUFFER_SIZE ) rsize =HID_MAX_BUFFER_SIZE ;if (csize < rsize) {
Greg Kroah-Hartman
maintainerPATCH 78/152 HID: core: increase HID report buffer size to 8KiB
In reply to Greg Kroah-HartmanFrom: Johan Korsnes <[email protected]>
commit 84a4062632462c4320704fcdf8e99e89e94c0aba upstream.
We have a HID touch device that reports its opens and shorts test results in HID buffers of size 8184 bytes. The maximum size of the HID buffer is currently set to 4096 bytes, causing probe of this device to fail. With this patch we increase the maximum size of the HID buffer to 8192 bytes, making device probe and acquisition of said buffers succeed.
--- include/linux/hid.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
include/linux/hid.h +1 -1
@@ -492,7 +492,7 @@ struct hid_report_enum { };#define HID_MIN_BUFFER_SIZE 64/* make sure there is at least a packet size of space */ -#define HID_MAX_BUFFER_SIZE 4096/* 4kb */ +#define HID_MAX_BUFFER_SIZE 8192/* 8kb */ #define HID_CONTROL_FIFO_SIZE 256/* to init devices with >100 reports */ #define HID_OUTPUT_FIFO_SIZE 64
Greg Kroah-Hartman
maintainerPATCH 79/152 drm/amdgpu: Drop DRIVER_USE_AGP
In reply to Greg Kroah-HartmanFrom: Daniel Vetter <[email protected]>
commit 8a3bddf67ce88b96531fb22c5a75d7f4dc41d155 upstream.
This doesn't do anything except auto-init drm_agp support when you call drm_get_pci_dev(). Which amdgpu stopped doing with
commit b58c11314a1706bf094c489ef5cb28f76478c704
Author: Alex Deucher <[email protected]> Date: Fri Jun 2 17:16:31 2017 -0400
drm/amdgpu: drop deprecated drm_get_pci_dev and drm_put_dev
No idea whether this was intentional or accidental breakage, but I guess anyone who manages to boot a this modern gpu behind an agp bridge deserves a price. A price I never expect anyone to ever collect :-)
--- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +1 -1
@@ -1421,7 +1421,7 @@ amdgpu_get_crtc_scanout_position(structstatic struct drm_driver kms_driver = { .driver_features = -DRIVER_USE_AGP |DRIVER_ATOMIC | +DRIVER_ATOMIC |DRIVER_GEM |DRIVER_RENDER |DRIVER_MODESET |DRIVER_SYNCOBJ , .load = amdgpu_driver_load_kms,
Greg Kroah-Hartman
maintainerPATCH 80/152 drm/radeon: Inline drm_get_pci_dev
In reply to Greg Kroah-HartmanFrom: Daniel Vetter <[email protected]>
commit eb12c957735b582607e5842a06d1f4c62e185c1d upstream.
It's the last user, and more importantly, it's the last non-legacy user of anything in drm_pci.c.
The only tricky bit is the agp initialization. But a close look shows that radeon does not use the drm_agp midlayer (the main use of that is drm_bufs for legacy drivers), and instead could use the agp subsystem directly (like nouveau does already). Hence we can just pull this in too.
A further step would be to entirely drop the use of drm_device->agp, but feels like too much churn just for this patch.
--- drivers/gpu/drm/radeon/radeon_drv.c | 43 ++++++++++++++++++++++++++++++++++-- drivers/gpu/drm/radeon/radeon_kms.c | 6 +++++ 2 files changed, 47 insertions(+), 2 deletions(-)
drivers/gpu/drm/radeon/radeon_drv.c +41 -2
@@ -37,6 +37,7 @@#include <linux/vga_switcheroo.h> #include <linux/mmu_notifier.h> +#include <drm/drm_agpsupport.h> #include <drm/drm_crtc_helper.h> #include <drm/drm_drv.h> #include <drm/drm_fb_helper.h>
@@ -325,6 +326,7 @@ static int radeon_pci_probe(struct pci_d const struct pci_device_id *ent) {unsigned long flags =0 ; +struct drm_device *dev;int ret;if (!ent)
@@ -365,7 +367,44 @@ static int radeon_pci_probe(struct pci_dif (ret)return ret; -return drm_get_pci_dev (pdev, ent, &kms_driver); + dev =drm_dev_alloc (&kms_driver, &pdev->dev); +if (IS_ERR (dev)) +return PTR_ERR (dev); + + ret =pci_enable_device (pdev); +if (ret) + goto err_free; + + dev->pdev = pdev; +#ifdef __alpha__ + dev->hose = pdev->sysdata; +#endif + +pci_set_drvdata (pdev, dev); + +if (pci_find_capability (dev->pdev,PCI_CAP_ID_AGP )) + dev->agp =drm_agp_init (dev); +if (dev->agp) { + dev->agp->agp_mtrr =arch_phys_wc_add ( + dev->agp->agp_info.aper_base, + dev->agp->agp_info.aper_size * +1024 *1024 ); + } + + ret =drm_dev_register (dev, ent->driver_data); +if (ret) + goto err_agp; + +return 0 ; + +err_agp: +if (dev->agp) +arch_phys_wc_del (dev->agp->agp_mtrr); +kfree (dev->agp); +pci_disable_device (pdev); +err_free: +drm_dev_put (dev); +return ret; } static void
@@ -578,7 +617,7 @@ radeon_get_crtc_scanout_position(structstatic struct drm_driver kms_driver = { .driver_features = -DRIVER_USE_AGP |DRIVER_GEM |DRIVER_RENDER , +DRIVER_GEM |DRIVER_RENDER , .load = radeon_driver_load_kms, .open = radeon_driver_open_kms, .postclose = radeon_driver_postclose_kms,
drivers/gpu/drm/radeon/radeon_kms.c +6 -0
@@ -31,6 +31,7 @@#include <linux/uaccess.h> #include <linux/vga_switcheroo.h> +#include <drm/drm_agpsupport.h> #include <drm/drm_fb_helper.h> #include <drm/drm_file.h> #include <drm/drm_ioctl.h>
@@ -77,6 +78,11 @@ void radeon_driver_unload_kms(struct drmradeon_modeset_fini (rdev);radeon_device_fini (rdev); +if (dev->agp) +arch_phys_wc_del (dev->agp->agp_mtrr); +kfree (dev->agp); + dev->agp =NULL ; + done_free:kfree (rdev); dev->dev_private =NULL ;
Greg Kroah-Hartman
maintainerPATCH 81/152 macintosh: therm_windtunnel: fix regression when instantiating devices
In reply to Greg Kroah-HartmanFrom: Wolfram Sang <[email protected]>
commit 38b17afb0ebb9ecd41418d3c08bcf9198af4349d upstream.
Removing attach_adapter from this driver caused a regression for at least some machines. Those machines had the sensors described in their DT, too, so they didn't need manual creation of the sensor devices. The old code worked, though, because manual creation came first. Creation of DT devices then failed later and caused error logs, but the sensors worked nonetheless because of the manually created devices.
When removing attach_adaper, manual creation now comes later and loses the race. The sensor devices were already registered via DT, yet with another binding, so the driver could not be bound to it.
This fix refactors the code to remove the race and only manually creates devices if there are no DT nodes present. Also, the DT binding is updated to match both, the DT and manually created devices. Because we don't know which device creation will be used at runtime, the code to start the kthread is moved to do_probe() which will be called by both methods.
--- drivers/macintosh/therm_windtunnel.c | 52 ++++++++++++++++++++--------------- 1 file changed, 31 insertions(+), 21 deletions(-)
drivers/macintosh/therm_windtunnel.c +31 -21
@@ -300,9 +300,11 @@ static int control_loop(void *dummy)/* i2c probing and setup */ /************************************************************************/ -static int -do_attach (struct i2c_adapter *adapter ) +static void do_attach (struct i2c_adapter *adapter) { +struct i2c_board_info info = { }; +struct device_node *np; +/* scan 0x48-0x4f (DS1775) and 0x2c-2x2f (ADM1030) */ static const unsigned short scan_ds1775[] = {0x48 ,0x49 ,0x4a ,0x4b ,0x4c ,0x4d ,0x4e ,0x4f ,
@@ -313,25 +315,24 @@ do_attach( struct i2c_adapter *adapter )I2C_CLIENT_END }; -if (strncmp (adapter->name,"uni-n" ,5 ) ) -return 0 ; - -if ( !x.running ) { -struct i2c_board_info info; +if (x.running ||strncmp (adapter->name,"uni-n" ,5 )) +return ; -memset (&info,0 ,sizeof (struct i2c_board_info )); -strlcpy (info.type,"therm_ds1775" ,I2C_NAME_SIZE ); + np =of_find_compatible_node (adapter->dev.of_node,NULL ,"MAC,ds1775" ); +if (np) { +of_node_put (np); + }else { +strlcpy (info.type,"MAC,ds1775" ,I2C_NAME_SIZE );i2c_new_probed_device (adapter, &info, scan_ds1775,NULL ); + } -strlcpy (info.type,"therm_adm1030" ,I2C_NAME_SIZE ); + np =of_find_compatible_node (adapter->dev.of_node,NULL ,"MAC,adm1030" ); +if (np) { +of_node_put (np); + }else { +strlcpy (info.type,"MAC,adm1030" ,I2C_NAME_SIZE );i2c_new_probed_device (adapter, &info, scan_adm1030,NULL ); - -if ( x.thermostat && x.fan ) { - x.running =1 ; - x.poll_task =kthread_run (control_loop,NULL ,"g4fand" ); - } } -return 0 ; } static int
@@ -404,8 +405,8 @@ out:enum chip { ds1775, adm1030 };static const struct i2c_device_id therm_windtunnel_id[] = { - {"therm_ds1775" , ds1775 }, - {"therm_adm1030" , adm1030 }, + {"MAC,ds1775" , ds1775 }, + {"MAC,adm1030" , adm1030 }, { } };MODULE_DEVICE_TABLE (i2c, therm_windtunnel_id);
@@ -414,6 +415,7 @@ static int do_probe(struct i2c_client *cl, const struct i2c_device_id *id) { struct i2c_adapter *adapter = cl->adapter; +int ret =0 ;if ( !i2c_check_functionality (adapter,I2C_FUNC_SMBUS_WORD_DATA |I2C_FUNC_SMBUS_WRITE_BYTE ) )
@@ -421,11 +423,19 @@ do_probe(struct i2c_client *cl, const stswitch (id->driver_data) {case adm1030: -return attach_fan ( cl ); + ret =attach_fan (cl); +break ;case ds1775: -return attach_thermostat (cl); + ret =attach_thermostat (cl); +break ; } -return 0 ; + +if (!x.running && x.thermostat && x.fan) { + x.running =1 ; + x.poll_task =kthread_run (control_loop,NULL ,"g4fand" ); + } + +return ret; }static struct i2c_driver g4fan_driver = {
Greg Kroah-Hartman
maintainerPATCH 82/152 tracing: Disable trace_printk() on post poned tests
In reply to Greg Kroah-HartmanFrom: Steven Rostedt (VMware) <[email protected]>
commit 78041c0c9e935d9ce4086feeff6c569ed88ddfd4 upstream.
The tracing seftests checks various aspects of the tracing infrastructure, and one is filtering. If trace_printk() is active during a self test, it can cause the filtering to fail, which will disable that part of the trace.
To keep the selftests from failing because of trace_printk() calls, trace_printk() checks the variable tracing_selftest_running, and if set, it does not write to the tracing buffer.
As some tracers were registered earlier in boot, the selftest they triggered would fail because not all the infrastructure was set up for the full selftest. Thus, some of the tests were post poned to when their infrastructure was ready (namely file system code). The postpone code did not set the tracing_seftest_running variable, and could fail if a trace_printk() was added and executed during their run.
--- kernel/trace/trace.c | 2 ++ 1 file changed, 2 insertions(+)
kernel/trace/trace.c +2 -0
@@ -1743,6 +1743,7 @@ static __init int init_trace_selftests(vpr_info ("Running postponed tracer tests:\n" ); + tracing_selftest_running = true;list_for_each_entry_safe (p, n, &postponed_selftests, list) {/* This loop can take minutes when sanitizers are enabled, so * lets make sure we allow RCU processing.
@@ -1765,6 +1766,7 @@ static __init int init_trace_selftests(vlist_del (&p->list);kfree (p); } + tracing_selftest_running = false; out:mutex_unlock (&trace_types_lock);
Greg Kroah-Hartman
maintainerPATCH 83/152 Revert "PM / devfreq: Modify the device name as devfreq(X) for sysfs"
In reply to Greg Kroah-HartmanFrom: Orson Zhai <[email protected]>
commit 66d0e797bf095d407479c89952d42b1d96ef0a7f upstream.
This reverts commit 4585fbcb5331fc910b7e553ad3efd0dd7b320d14.
The name changing as devfreq(X) breaks some user space applications, such as Android HAL from Unisoc and Hikey [1].
The device name will be changed unexpectly after every boot depending on module init sequence. It will make trouble to setup some system configuration like selinux for Android.
So we'd like to revert it back to old naming rule before any better way being found.
[1] https://lkml.org/lkml/2018/5/8/1042
--- drivers/devfreq/devfreq.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-)
drivers/devfreq/devfreq.c +1 -3
@@ -613,7 +613,6 @@ struct devfreq *devfreq_add_device(struc {struct devfreq *devfreq;struct devfreq_governor *governor; -static atomic_t devfreq_no =ATOMIC_INIT (-1 );int err =0 ;if (!dev || !profile || !governor_name) {
@@ -677,8 +676,7 @@ struct devfreq *devfreq_add_device(struc devfreq->suspend_freq =dev_pm_opp_get_suspend_opp_freq (dev);atomic_set (&devfreq->suspend_count,0 ); -dev_set_name (&devfreq->dev,"devfreq%d" , -atomic_inc_return (&devfreq_no)); +dev_set_name (&devfreq->dev,"%s" ,dev_name (dev)); err =device_register (&devfreq->dev);if (err) {mutex_unlock (&devfreq->lock);
Greg Kroah-Hartman
maintainerPATCH 84/152 amdgpu/gmc_v9: save/restore sdpif regs during S3
In reply to Greg Kroah-HartmanFrom: Shirish S <[email protected]>
commit a3ed353cf8015ba84a0407a5dc3ffee038166ab0 upstream.
fixes S3 issue with IOMMU + S/G enabled @ 64M VRAM.
--- drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h | 1 drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 37 ++++++++++++- drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_offset.h | 2 3 files changed, 39 insertions(+), 1 deletion(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h +1 -0
@@ -157,6 +157,7 @@ struct amdgpu_gmc {uint32_t srbm_soft_reset;bool prt_warning;uint64_t stolen_size; +uint32_t sdpif_register;/* apertures */ u64 shared_aperture_start;u64 shared_aperture_end;
drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +36 -1
@@ -1383,6 +1383,19 @@ static void gmc_v9_0_init_golden_registe }/** +* gmc_v9_0_restore_registers - restores regs +* +* @adev: amdgpu_device pointer +* +* This restores register values, saved at suspend. +*/ +static void gmc_v9_0_restore_registers (struct amdgpu_device *adev) +{ +if (adev->asic_type ==CHIP_RAVEN ) +WREG32 (mmDCHUBBUB_SDPIF_MMIO_CNTRL_0, adev->gmc.sdpif_register); +} + +/** * gmc_v9_0_gart_enable - gart enable * * @adev: amdgpu_device pointer
@@ -1479,6 +1492,20 @@ static int gmc_v9_0_hw_init(void *handle }/** +* gmc_v9_0_save_registers - saves regs +* +* @adev: amdgpu_device pointer +* +* This saves potential register values that should be +* restored upon resume +*/ +static void gmc_v9_0_save_registers (struct amdgpu_device *adev) +{ +if (adev->asic_type ==CHIP_RAVEN ) + adev->gmc.sdpif_register =RREG32 (mmDCHUBBUB_SDPIF_MMIO_CNTRL_0); +} + +/** * gmc_v9_0_gart_disable - gart disable * * @adev: amdgpu_device pointer
@@ -1514,9 +1541,16 @@ static int gmc_v9_0_hw_fini(void *handlestatic int gmc_v9_0_suspend (void *handle) { +int r;struct amdgpu_device *adev = (struct amdgpu_device *)handle; -return gmc_v9_0_hw_fini (adev); + r =gmc_v9_0_hw_fini (adev); +if (r) +return r; + +gmc_v9_0_save_registers (adev); + +return 0 ; }static int gmc_v9_0_resume (void *handle)
@@ -1524,6 +1558,7 @@ static int gmc_v9_0_resume(void *handle)int r;struct amdgpu_device *adev = (struct amdgpu_device *)handle; +gmc_v9_0_restore_registers (adev); r =gmc_v9_0_hw_init (adev);if (r)return r;
drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_offset.h +2 -0
@@ -7376,6 +7376,8 @@#define mmCRTC4_CRTC_DRR_CONTROL 0x0f3e#define mmCRTC4_CRTC_DRR_CONTROL_BASE_IDX 2 +#define mmDCHUBBUB_SDPIF_MMIO_CNTRL_0 0x395d +#define mmDCHUBBUB_SDPIF_MMIO_CNTRL_0_BASE_IDX 2// addressBlock: dce_dc_fmt4_dispdec // base address: 0x2000
Greg Kroah-Hartman
maintainerPATCH 85/152 cpufreq: Fix policy initialization for internal governor drivers
In reply to Greg Kroah-HartmanFrom: Rafael J. Wysocki <[email protected]>
commit f5739cb0b56590d68d8df8a44659893b6d0084c3 upstream.
Before commit 1e4f63aecb53 ("cpufreq: Avoid creating excessively large stack frames") the initial value of the policy field in struct cpufreq_policy set by the driver's ->init() callback was implicitly passed from cpufreq_init_policy() to cpufreq_set_policy() if the default governor was neither "performance" nor "powersave". After that commit, however, cpufreq_init_policy() must take that case into consideration explicitly and handle it as appropriate, so make that happen.
--- drivers/cpufreq/cpufreq.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-)
drivers/cpufreq/cpufreq.c +10 -2
@@ -1071,9 +1071,17 @@ static int cpufreq_init_policy(struct cp pol = policy->last_policy; }else if (def_gov) { pol =cpufreq_parse_policy (def_gov->name); - }else { -return -ENODATA ; +/* +* In case the default governor is neiter "performance" +* nor "powersave", fall back to the initial policy +* value set by the driver. +*/ +if (pol ==CPUFREQ_POLICY_UNKNOWN ) + pol = policy->policy; } +if (pol !=CPUFREQ_POLICY_PERFORMANCE && + pol !=CPUFREQ_POLICY_POWERSAVE ) +return -ENODATA ; }return cpufreq_set_policy (policy, gov, pol);
Greg Kroah-Hartman
maintainerPATCH 86/152 io_uring: fix 32-bit compatability with sendmsg/recvmsg
In reply to Greg Kroah-HartmanFrom: Jens Axboe <[email protected]>
commit d876836204897b6d7d911f942084f69a1e9d5c4d upstream.
We must set MSG_CMSG_COMPAT if we're in compatability mode, otherwise the iovec import for these commands will not do the right thing and fail the command with -EINVAL.
Found by running the test suite compiled as 32-bit.
--- fs/io_uring.c | 5 +++++ 1 file changed, 5 insertions(+)
fs/io_uring.c +5 -0
@@ -1657,6 +1657,11 @@ static int io_send_recvmsg(struct io_kioelse if (force_nonblock) flags |=MSG_DONTWAIT ; +#ifdef CONFIG_COMPAT +if (req->ctx->compat) + flags |=MSG_CMSG_COMPAT ; +#endif + msg = (struct user_msghdr __user *) (unsigned long )READ_ONCE (sqe->addr);
Greg Kroah-Hartman
maintainerPATCH 87/152 netfilter: ipset: Fix "INFO: rcu detected stall in hash_xxx" reports
In reply to Greg Kroah-HartmanFrom: Jozsef Kadlecsik <[email protected]>
commit f66ee0410b1c3481ee75e5db9b34547b4d582465 upstream.
In the case of huge hash:* types of sets, due to the single spinlock of a set the processing of the whole set under spinlock protection could take too long.
There were four places where the whole hash table of the set was processed from bucket to bucket under holding the spinlock:
- During resizing a set, the original set was locked to exclude kernel side add/del element operations (userspace add/del is excluded by the nfnetlink mutex). The original set is actually just read during the resize, so the spinlocking is replaced with rcu locking of regions. However, thus there can be parallel kernel side add/del of entries. In order not to loose those operations a backlog is added and replayed after the successful resize.
- Garbage collection of timed out entries was also protected by the spinlock. In order not to lock too long, region locking is introduced and a single region is processed in one gc go. Also, the simple timer based gc running is replaced with a workqueue based solution. The internal book-keeping (number of elements, size of extensions) is moved to region level due to the region locking.
- Adding elements: when the max number of the elements is reached, the gc was called to evict the timed out entries. The new approach is that the gc is called just for the matching region, assuming that if the region (proportionally) seems to be full, then the whole set does. We could scan the other regions to check every entry under rcu locking, but for huge sets it'd mean a slowdown at adding elements.
- Listing the set header data: when the set was defined with timeout support, the garbage collector was called to clean up timed out entries to get the correct element numbers and set size values. Now the set is scanned to check non-timed out entries, without actually calling the gc for the whole set.
Thanks to Florian Westphal for helping me to solve the SOFTIRQ-safe -> SOFTIRQ-unsafe lock order issues during working on the patch.
--- include/linux/netfilter/ipset/ip_set.h | 11 net/netfilter/ipset/ip_set_core.c | 34 + net/netfilter/ipset/ip_set_hash_gen.h | 633 ++++++++++++++++++++++----------- 3 files changed, 472 insertions(+), 206 deletions(-)
include/linux/netfilter/ipset/ip_set.h +10 -1
@@ -121,6 +121,7 @@ struct ip_set_ext {u32 timeout;u8 packets_op;u8 bytes_op; +bool target; };struct ip_set ;
@@ -187,6 +188,14 @@ struct ip_set_type_variant {/* Return true if "b" set is the same as "a" * according to the create set parameters */ bool (*same_set)(const struct ip_set *a,const struct ip_set *b); +/* Region-locking is used */ +bool region_lock; +}; + +struct ip_set_region { +spinlock_t lock;/* Region lock */ +size_t ext_size;/* Size of the dynamic extensions */ +u32 elements;/* Number of elements vs timeout */ };/* The core set type structure */
@@ -681,7 +690,7 @@ ip_set_init_skbinfo(struct ip_set_skbinf }#define IP_SET_INIT_KEXT (skb, opt, set) \ - { .bytes = (skb)->len, .packets = 1, \ + { .bytes = (skb)->len, .packets = 1, .target = true,\ .timeout = ip_set_adt_opt_timeout(opt, set) }#define IP_SET_INIT_UEXT (set) \
net/netfilter/ipset/ip_set_core.c +24 -10
@@ -557,6 +557,20 @@ ip_set_rcu_get(struct net *net, ip_set_ireturn set; } +static inline void +ip_set_lock (struct ip_set *set) +{ +if (!set->variant->region_lock) +spin_lock_bh (&set->lock); +} + +static inline void +ip_set_unlock (struct ip_set *set) +{ +if (!set->variant->region_lock) +spin_unlock_bh (&set->lock); +} +int ip_set_test (ip_set_id_t index,const struct sk_buff *skb,const struct xt_action_param *par,struct ip_set_adt_opt *opt)
@@ -578,9 +592,9 @@ ip_set_test(ip_set_id_t index, const strif (ret == -EAGAIN ) {/* Type requests element to be completed */ pr_debug ("element must be completed, ADD is triggered\n" ); -spin_lock_bh (&set->lock); +ip_set_lock (set); set->variant->kadt (set, skb, par,IPSET_ADD , opt); -spin_unlock_bh (&set->lock); +ip_set_unlock (set); ret =1 ; }else {/* --return-nomatch: invert matched element */
@@ -609,9 +623,9 @@ ip_set_add(ip_set_id_t index, const stru !(opt->family == set->family || set->family == NFPROTO_UNSPEC)) return -IPSET_ERR_TYPE_MISMATCH; -spin_lock_bh (&set->lock); +ip_set_lock (set); ret = set->variant->kadt (set, skb, par,IPSET_ADD , opt); -spin_unlock_bh (&set->lock); +ip_set_unlock (set);return ret; }
@@ -631,9 +645,9 @@ ip_set_del(ip_set_id_t index, const stru !(opt->family == set->family || set->family == NFPROTO_UNSPEC)) return -IPSET_ERR_TYPE_MISMATCH; -spin_lock_bh (&set->lock); +ip_set_lock (set); ret = set->variant->kadt (set, skb, par,IPSET_DEL , opt); -spin_unlock_bh (&set->lock); +ip_set_unlock (set);return ret; }
@@ -1098,9 +1112,9 @@ ip_set_flush_set(struct ip_set *set) {pr_debug ("set: %s\n" , set->name); -spin_lock_bh (&set->lock); +ip_set_lock (set); set->variant->flush (set); -spin_unlock_bh (&set->lock); +ip_set_unlock (set); } static int ip_set_flush(struct net *net, struct sock *ctnl, struct sk_buff *skb,
@@ -1523,9 +1537,9 @@ call_ad(struct sock *ctnl, struct sk_bufbool eexist = flags &IPSET_FLAG_EXIST , retried = false; do { - spin_lock_bh(&set->lock); + ip_set_lock(set); ret = set->variant->uadt(set, tb, adt, &lineno, flags, retried); - spin_unlock_bh(&set->lock); + ip_set_unlock(set); retried = true; } while (ret == -EAGAIN && set->variant->resize &&
net/netfilter/ipset/ip_set_hash_gen.h +438 -195
1,092 lines not on this page · raw message
Greg Kroah-Hartman
maintainerPATCH 88/152 net/smc: transfer fasync_list in case of fallback
In reply to Greg Kroah-HartmanFrom: Ursula Braun <[email protected]>
commit 67f562e3e147750a02b2a91d21a163fc44a1d13e upstream.
SMC does not work together with FASTOPEN. If sendmsg() is called with flag MSG_FASTOPEN in SMC_INIT state, the SMC-socket switches to fallback mode. To handle the previous ioctl FIOASYNC call correctly in this case, it is necessary to transfer the socket wait queue fasync_list to the internal TCP socket.
--- net/smc/af_smc.c | 2 ++ 1 file changed, 2 insertions(+)
net/smc/af_smc.c +2 -0
@@ -467,6 +467,8 @@ static void smc_switch_to_fallback(strucif (smc->sk.sk_socket && smc->sk.sk_socket->file) { smc->clcsock->file = smc->sk.sk_socket->file; smc->clcsock->file->private_data = smc->clcsock; + smc->clcsock->wq.fasync_list = + smc->sk.sk_socket->wq.fasync_list; } }
Greg Kroah-Hartman
maintainerPATCH 89/152 vhost: Check docket sk_family instead of call getname
In reply to Greg Kroah-HartmanFrom: Eugenio Pérez <[email protected]>
commit 42d84c8490f9f0931786f1623191fcab397c3d64 upstream.
Doing so, we save one call to get data we already have in the struct.
Also, since there is no guarantee that getname use sockaddr_ll parameter beyond its size, we add a little bit of security here. It should do not do beyond MAX_ADDR_LEN, but syzbot found that ax25_getname writes more (72 bytes, the size of full_sockaddr_ax25, versus 20 + 32 bytes of sockaddr_ll + MAX_ADDR_LEN in syzbot repro).
--- drivers/vhost/net.c | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-)
drivers/vhost/net.c +1 -9
@@ -1414,10 +1414,6 @@ static int vhost_net_release(struct inodstatic struct socket *get_raw_socket (int fd) { -struct { -struct sockaddr_ll sa; -char buf[MAX_ADDR_LEN ]; - } uaddr;int r;struct socket *sock =sockfd_lookup (fd, &r);
@@ -1430,11 +1426,7 @@ static struct socket *get_raw_socket(int goto err; } - r = sock->ops->getname (sock, (struct sockaddr *)&uaddr.sa,0 ); -if (r <0 ) - goto err; - -if (uaddr.sa.sll_family !=AF_PACKET ) { +if (sock->sk->sk_family !=AF_PACKET ) { r = -EPFNOSUPPORT ; goto err; }
Greg Kroah-Hartman
maintainerPATCH 90/152 netfilter: ipset: Fix forceadd evaluation path
In reply to Greg Kroah-HartmanFrom: Jozsef Kadlecsik <[email protected]>
commit 8af1c6fbd9239877998c7f5a591cb2c88d41fb66 upstream.
When the forceadd option is enabled, the hash:* types should find and replace the first entry in the bucket with the new one if there are no reuseable (deleted or timed out) entries. However, the position index was just not set to zero and remained the invalid -1 if there were no reuseable entries.
--- net/netfilter/ipset/ip_set_hash_gen.h | 2 ++ 1 file changed, 2 insertions(+)
net/netfilter/ipset/ip_set_hash_gen.h +2 -0
@@ -931,6 +931,8 @@ mtype_add(struct ip_set *set, void *valu } }if (reuse || forceadd) { +if (j ==-1 ) + j =0 ; data =ahash_data (n, j, set->dsize);if (!deleted) {#ifdef IP_SET_HASH_WITH_NETS
Greg Kroah-Hartman
maintainerPATCH 91/152 netfilter: xt_hashlimit: reduce hashlimit_mutex scope for htable_put()
In reply to Greg Kroah-HartmanFrom: Cong Wang <[email protected]>
commit c4a3922d2d20c710f827d3a115ee338e8d0467df upstream.
It is unnecessary to hold hashlimit_mutex for htable_destroy() as it is already removed from the global hashtable and its refcount is already zero.
Also, switch hinfo->use to refcount_t so that we don't have to hold the mutex until it reaches zero in htable_put().
--- net/netfilter/xt_hashlimit.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-)
net/netfilter/xt_hashlimit.c +6 -6
@@ -36,6 +36,7 @@#include <linux/netfilter_ipv6/ip6_tables.h> #include <linux/mutex.h> #include <linux/kernel.h> +#include <linux/refcount.h> #include <uapi/linux/netfilter/xt_hashlimit.h> #define XT_HASHLIMIT_ALL (XT_HASHLIMIT_HASH_DIP | XT_HASHLIMIT_HASH_DPT | \
@@ -114,7 +115,7 @@ struct dsthash_ent {struct xt_hashlimit_htable {struct hlist_node node;/* global list of all htables */ -int use; +refcount_t use;u_int8_t family;bool rnd_initialized;
@@ -315,7 +316,7 @@ static int htable_create(struct net *netfor (i =0 ; i < hinfo->cfg.size; i++)INIT_HLIST_HEAD (&hinfo->hash[i]); - hinfo->use =1 ; +refcount_set (&hinfo->use,1 ); hinfo->count =0 ; hinfo->family = family; hinfo->rnd_initialized = false;
@@ -434,7 +435,7 @@ static struct xt_hashlimit_htable *htablhlist_for_each_entry (hinfo, &hashlimit_net->htables, node) {if (!strcmp (name, hinfo->name) && hinfo->family == family) { - hinfo->use++; +refcount_inc (&hinfo->use);return hinfo; } }
@@ -443,12 +444,11 @@ static struct xt_hashlimit_htable *htablstatic void htable_put (struct xt_hashlimit_htable *hinfo) { -mutex_lock (&hashlimit_mutex); -if (--hinfo->use ==0 ) { +if (refcount_dec_and_mutex_lock (&hinfo->use, &hashlimit_mutex)) {hlist_del (&hinfo->node); +mutex_unlock (&hashlimit_mutex);htable_destroy (hinfo); } -mutex_unlock (&hashlimit_mutex); }/* The algorithm used is the Simple Token Bucket Filter (TBF)
Greg Kroah-Hartman
maintainerPATCH 92/152 HID: alps: Fix an error handling path in alps_input_configured()
In reply to Greg Kroah-HartmanFrom: Christophe JAILLET <[email protected]>
commit 8d2e77b39b8fecb794e19cd006a12f90b14dd077 upstream.
They are issues:
- if 'input_allocate_device()' fails and return NULL, there is no need to free anything and 'input_free_device()' call is a no-op. It can be axed.
- 'ret' is known to be 0 at this point, so we must set it to a meaningful value before returning
--- drivers/hid/hid-alps.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/hid/hid-alps.c +1 -1
@@ -730,7 +730,7 @@ static int alps_input_configured(structif (data->has_sp) { input2 =input_allocate_device ();if (!input2) { -input_free_device (input2); + ret = -ENOMEM ; goto exit; }
Greg Kroah-Hartman
maintainerPATCH 93/152 HID: hiddev: Fix race in in hiddev_disconnect()
In reply to Greg Kroah-HartmanFrom: [email protected] <[email protected]>
commit 5c02c447eaeda29d3da121a2e17b97ccaf579b51 upstream.
Syzbot reports that "hiddev" is used after it's free in hiddev_disconnect(). The hiddev_disconnect() function sets "hiddev->exist = 0;" so hiddev_release() can free it as soon as we drop the "existancelock" lock. This patch moves the mutex_unlock(&hiddev->existancelock) until after we have finished using it.
--- drivers/hid/usbhid/hiddev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/hid/usbhid/hiddev.c +1 -1
@@ -941,9 +941,9 @@ void hiddev_disconnect(struct hid_device hiddev->exist =0 ;if (hiddev->open) { -mutex_unlock (&hiddev->existancelock);hid_hw_close (hiddev->hid);wake_up_interruptible (&hiddev->wait); +mutex_unlock (&hiddev->existancelock); }else {mutex_unlock (&hiddev->existancelock);kfree (hiddev);
Greg Kroah-Hartman
maintainerPATCH 94/152 MIPS: VPE: Fix a double free and a memory leak in release_vpe()
In reply to Greg Kroah-HartmanFrom: Christophe JAILLET <[email protected]>
commit bef8e2dfceed6daeb6ca3e8d33f9c9d43b926580 upstream.
Pointer on the memory allocated by 'alloc_progmem()' is stored in 'v->load_addr'. So this is this memory that should be freed by 'release_progmem()'.
'release_progmem()' is only a call to 'kfree()'.
With the current code, there is both a double free and a memory leak. Fix it by passing the correct pointer to 'release_progmem()'.
--- arch/mips/kernel/vpe.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/mips/kernel/vpe.c +1 -1
@@ -134,7 +134,7 @@ void release_vpe(struct vpe *v) {list_del (&v->list);if (v->load_addr) -release_progmem (v); +release_progmem (v->load_addr);kfree (v); }
Greg Kroah-Hartman
maintainerPATCH 95/152 i2c: altera: Fix potential integer overflow
In reply to Greg Kroah-HartmanFrom: Gustavo A. R. Silva <[email protected]>
commit 54498e8070e19e74498a72c7331348143e7e1f8c upstream.
Factor out 100 from the equation and do 32-bit arithmetic (3 * clk_mhz / 10) instead of 64-bit.
Notice that clk_mhz is MHz, so the multiplication will never wrap 32 bits and there is no need for div_u64().
Addresses-Coverity: 1458369 ("Unintentional integer overflow")
--- drivers/i2c/busses/i2c-altera.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/i2c/busses/i2c-altera.c +1 -1
@@ -171,7 +171,7 @@ static void altr_i2c_init(struct altr_i2/* SCL Low Time */ writel (t_low, idev->base +ALTR_I2C_SCL_LOW );/* SDA Hold Time, 300ns */ -writel (div_u64 (300 * clk_mhz,1000 ), idev->base +ALTR_I2C_SDA_HOLD ); +writel (3 * clk_mhz /10 , idev->base +ALTR_I2C_SDA_HOLD );/* Mask all master interrupt bits */ altr_i2c_int_enable (idev,ALTR_I2C_ALL_IRQ , false);
Greg Kroah-Hartman
maintainerPATCH 96/152 i2c: jz4780: silence log flood on txabrt
In reply to Greg Kroah-HartmanFrom: Wolfram Sang <[email protected]>
commit 9e661cedcc0a072d91a32cb88e0515ea26e35711 upstream.
The printout for txabrt is way too talkative and is highly annoying with scanning programs like 'i2cdetect'. Reduce it to the minimum, the rest can be gained by I2C core debugging and datasheet information. Also, make it a debug printout, it won't help the regular user.
--- drivers/i2c/busses/i2c-jz4780.c | 36 ++---------------------------------- 1 file changed, 2 insertions(+), 34 deletions(-)
drivers/i2c/busses/i2c-jz4780.c +2 -34
@@ -73,25 +73,6 @@#define JZ4780_I2C_STA_TFNF BIT(1)#define JZ4780_I2C_STA_ACT BIT(0) -static const char *const jz4780_i2c_abrt_src[] = { -"ABRT_7B_ADDR_NOACK" , -"ABRT_10ADDR1_NOACK" , -"ABRT_10ADDR2_NOACK" , -"ABRT_XDATA_NOACK" , -"ABRT_GCALL_NOACK" , -"ABRT_GCALL_READ" , -"ABRT_HS_ACKD" , -"SBYTE_ACKDET" , -"ABRT_HS_NORSTRT" , -"SBYTE_NORSTRT" , -"ABRT_10B_RD_NORSTRT" , -"ABRT_MASTER_DIS" , -"ARB_LOST" , -"SLVFLUSH_TXFIFO" , -"SLV_ARBLOST" , -"SLVRD_INTX" , -}; -#define JZ4780_I2C_INTST_IGC BIT(11)#define JZ4780_I2C_INTST_ISTT BIT(10)#define JZ4780_I2C_INTST_ISTP BIT(9)
@@ -529,21 +510,8 @@ done:static void jz4780_i2c_txabrt (struct jz4780_i2c *i2c,int src) { -int i; - -dev_err (&i2c->adap.dev,"txabrt: 0x%08x\n" , src); -dev_err (&i2c->adap.dev,"device addr=%x\n" , -jz4780_i2c_readw (i2c,JZ4780_I2C_TAR )); -dev_err (&i2c->adap.dev,"send cmd count:%d %d\n" , - i2c->cmd, i2c->cmd_buf[i2c->cmd]); -dev_err (&i2c->adap.dev,"receive data count:%d %d\n" , - i2c->cmd, i2c->data_buf[i2c->cmd]); - -for (i =0 ; i <16 ; i++) { -if (src &BIT (i)) -dev_dbg (&i2c->adap.dev,"I2C TXABRT[%d]=%s\n" , - i, jz4780_i2c_abrt_src[i]); - } +dev_dbg (&i2c->adap.dev,"txabrt: 0x%08x, cmd: %d, send: %d, recv: %d\n" , + src, i2c->cmd, i2c->cmd_buf[i2c->cmd], i2c->data_buf[i2c->cmd]); } static inline int jz4780_i2c_xfer_read(struct jz4780_i2c *i2c,
Greg Kroah-Hartman
maintainerPATCH 97/152 drm/i915/gvt: Fix orphan vgpu dmabuf_objs lifetime
In reply to Greg Kroah-HartmanFrom: Tina Zhang <[email protected]>
commit b549c252b1292aea959cd9b83537fcb9384a6112 upstream.
Deleting dmabuf item's list head after releasing its container can lead to KASAN-reported issue:
BUG: KASAN: use-after-free in __list_del_entry_valid+0x15/0xf0 Read of size 8 at addr ffff88818a4598a8 by task kworker/u8:3/13119
So fix this issue by puting deleting dmabuf_objs ahead of releasing its container.
--- drivers/gpu/drm/i915/gvt/dmabuf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/i915/gvt/dmabuf.c +1 -1
@@ -96,12 +96,12 @@ static void dmabuf_gem_object_free(struc dmabuf_obj =container_of (pos, struct intel_vgpu_dmabuf_obj, list);if (dmabuf_obj == obj) { +list_del (pos);intel_gvt_hypervisor_put_vfio_device (vgpu);idr_remove (&vgpu->object_idr, dmabuf_obj->dmabuf_id);kfree (dmabuf_obj->info);kfree (dmabuf_obj); -list_del (pos);break ; } }
Greg Kroah-Hartman
maintainerPATCH 98/152 drm/i915/gvt: Separate display reset from ALL_ENGINES reset
In reply to Greg Kroah-HartmanFrom: Tina Zhang <[email protected]>
commit 3eb55e6f753a379e293395de8d5f3be28351a7f8 upstream.
ALL_ENGINES reset doesn't clobber display with the current gvt-g supported platforms. Thus ALL_ENGINES reset shouldn't reset the display engine registers emulated by gvt-g.
This fixes guest warning like
[ 14.622026] [drm] Initialized i915 1.6.0 20200114 for 0000:00:03.0 on minor 0 [ 14.967917] fbcon: i915drmfb (fb0) is primary device [ 25.100188] [drm:drm_atomic_helper_wait_for_dependencies [drm_kms_helper]] E RROR [CRTC:51:pipe A] flip_done timed out [ 25.100860] -----------[ cut here ]----------- [ 25.100861] pll on state mismatch (expected 0, found 1) [ 25.101024] WARNING: CPU: 1 PID: 30 at drivers/gpu/drm/i915/display/intel_dis play.c:14382 verify_single_dpll_state.isra.115+0x28f/0x320 [i915] [ 25.101025] Modules linked in: intel_rapl_msr intel_rapl_common kvm_intel kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel i915 aesni_intel cr ypto_simd cryptd glue_helper cec rc_core video drm_kms_helper joydev drm input_l eds i2c_algo_bit serio_raw fb_sys_fops syscopyarea sysfillrect sysimgblt mac_hid qemu_fw_cfg sch_fq_codel parport_pc ppdev lp parport ip_tables x_tables autofs4 e1000 psmouse i2c_piix4 pata_acpi floppy [ 25.101052] CPU: 1 PID: 30 Comm: kworker/u4:1 Not tainted 5.5.0+ #1 [ 25.101053] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1 .12.1-0-ga5cab58 04/01/2014 [ 25.101055] Workqueue: events_unbound async_run_entry_fn [ 25.101092] RIP: 0010:verify_single_dpll_state.isra.115+0x28f/0x320 [i915] [ 25.101093] Code: e0 d9 ff e9 a3 fe ff ff 80 3d e9 c2 11 00 00 44 89 f6 48 c7 c7 c0 9d 88 c0 75 3b e8 eb df d9 ff e9 c7 fe ff ff e8 d1 e0 ae c4 <0f> 0b e9 7a fe ff ff 80 3d c0 c2 11 00 00 8d 71 41 89 c2 48 c7 c7 [ 25.101093] RSP: 0018:ffffb1de80107878 EFLAGS: 00010286 [ 25.101094] RAX: 0000000000000000 RBX: ffffb1de80107884 RCX: 0000000000000007 [ 25.101095] RDX: 0000000000000000 RSI: 0000000000000002 RDI: ffff94fdfdd19740 [ 25.101095] RBP: ffffb1de80107938 R08: 0000000d6bfdc7b4 R09: 000000000000002b [ 25.101096] R10: ffff94fdf82dc000 R11: 0000000000000225 R12: 00000000000001f8 [ 25.101096] R13: ffff94fdb3ca6a90 R14: ffff94fdb3ca0000 R15: 0000000000000000 [ 25.101097] FS: 0000000000000000(0000) GS:ffff94fdfdd00000(0000) knlGS:00000 00000000000 [ 25.101098] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 25.101098] CR2: 00007fbc3e2be9c8 CR3: 000000003339a003 CR4: 0000000000360ee0 [ 25.101101] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 25.101101] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 25.101102] Call Trace: [ 25.101139] intel_atomic_commit_tail+0xde4/0x1520 [i915] [ 25.101141] ? flush_workqueue_prep_pwqs+0xfa/0x130 [ 25.101142] ? flush_workqueue+0x198/0x3c0 [ 25.101174] intel_atomic_commit+0x2ad/0x320 [i915] [ 25.101209] drm_atomic_commit+0x4a/0x50 [drm] [ 25.101220] drm_client_modeset_commit_atomic+0x1c4/0x200 [drm] [ 25.101231] drm_client_modeset_commit_force+0x47/0x170 [drm] [ 25.101250] drm_fb_helper_restore_fbdev_mode_unlocked+0x4e/0xa0 [drm_kms_hel per] [ 25.101255] drm_fb_helper_set_par+0x2d/0x60 [drm_kms_helper] [ 25.101287] intel_fbdev_set_par+0x1a/0x40 [i915] [ 25.101289] ? con_is_visible+0x2e/0x60 [ 25.101290] fbcon_init+0x378/0x600 [ 25.101292] visual_init+0xd5/0x130 [ 25.101296] do_bind_con_driver+0x217/0x430 [ 25.101297] do_take_over_console+0x7d/0x1b0 [ 25.101298] do_fbcon_takeover+0x5c/0xb0 [ 25.101299] fbcon_fb_registered+0x199/0x1a0 [ 25.101301] register_framebuffer+0x22c/0x330 [ 25.101306] __drm_fb_helper_initial_config_and_unlock+0x31a/0x520 [drm_kms_h elper] [ 25.101311] drm_fb_helper_initial_config+0x35/0x40 [drm_kms_helper] [ 25.101341] intel_fbdev_initial_config+0x18/0x30 [i915] [ 25.101342] async_run_entry_fn+0x3c/0x150 [ 25.101343] process_one_work+0x1fd/0x3f0 [ 25.101344] worker_thread+0x34/0x410 [ 25.101346] kthread+0x121/0x140 [ 25.101346] ? process_one_work+0x3f0/0x3f0 [ 25.101347] ? kthread_park+0x90/0x90 [ 25.101350] ret_from_fork+0x35/0x40 [ 25.101351] --[ end trace b5b47d44cd998ba1 ]--
--- drivers/gpu/drm/i915/gvt/vgpu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/i915/gvt/vgpu.c +1 -1
@@ -560,9 +560,9 @@ void intel_gvt_reset_vgpu_locked(structintel_vgpu_reset_mmio (vgpu, dmlr);populate_pvinfo_page (vgpu); -intel_vgpu_reset_display (vgpu);if (dmlr) { +intel_vgpu_reset_display (vgpu);intel_vgpu_reset_cfg_space (vgpu);/* only reset the failsafe mode when dmlr reset */ vgpu->failsafe = false;
Greg Kroah-Hartman
maintainerPATCH 99/152 nl80211: fix potential leak in AP start
In reply to Greg Kroah-HartmanFrom: Johannes Berg <[email protected]>
commit 9951ebfcdf2b97dbb28a5d930458424341e61aa2 upstream.
If nl80211_parse_he_obss_pd() fails, we leak the previously allocated ACL memory. Free it in this case.
--- net/wireless/nl80211.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
net/wireless/nl80211.c +2 -2
@@ -4795,8 +4795,7 @@ static int nl80211_start_ap(struct sk_bu err =nl80211_parse_he_obss_pd ( info->attrs[NL80211_ATTR_HE_OBSS_PD ], ¶ms.he_obss_pd); -if (err) -return err; + goto out; }nl80211_calculate_ap_params (¶ms);
@@ -4818,6 +4817,7 @@ static int nl80211_start_ap(struct sk_bu }wdev_unlock (wdev); +out:kfree (params.acl);return err;
Greg Kroah-Hartman
maintainerPATCH 100/152 mac80211: Remove a redundant mutex unlock
In reply to Greg Kroah-HartmanFrom: Andrei Otcheretianski <[email protected]>
commit 0daa63ed4c6c4302790ce67b7a90c0997ceb7514 upstream.
The below-mentioned commit changed the code to unlock *inside* the function, but previously the unlock was *outside*. It failed to remove the outer unlock, however, leading to double unlock.
Fix this.
--- net/mac80211/mlme.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-)
net/mac80211/mlme.c +1 -5
@@ -2959,7 +2959,7 @@ static void ieee80211_rx_mgmt_auth(struc (auth_transaction == 2 && ifmgd->auth_data->expected_transaction == 2)) {if (!ieee80211_mark_sta_auth (sdata, bssid)) - goto out_err; +return ;/* ignore frame -- wait for timeout */ } else if (ifmgd->auth_data->algorithm == WLAN_AUTH_SAE && auth_transaction == 2) {sdata_info (sdata,"SAE peer confirmed\n" );
@@ -2967,10 +2967,6 @@ static void ieee80211_rx_mgmt_auth(struc }cfg80211_rx_mlme_mgmt (sdata->dev, (u8 *)mgmt, len); -return ; - out_err: -mutex_unlock (&sdata->local->sta_mtx); -/* ignore frame -- wait for timeout */ }#define case_WLAN (type) \
Greg Kroah-Hartman
maintainerPATCH 101/152 kbuild: fix DT binding schema rule to detect command line changes
In reply to Greg Kroah-HartmanFrom: Masahiro Yamada <[email protected]>
commit 7a04960560640ac5b0b89461f7757322b57d0c7a upstream.
This if_change_rule is not working properly; it cannot detect any command line change.
The reason is because cmd-check in scripts/Kbuild.include compares $(cmd_$@) and $(cmd_$1), but cmd_dtc_dt_yaml does not exist here.
For if_change_rule to work properly, the stem part of cmd_* and rule_* must match. Because this cmd_and_fixdep invokes cmd_dtc, this rule must be named rule_dtc.
--- scripts/Makefile.lib | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
scripts/Makefile.lib +2 -2
@@ -305,13 +305,13 @@ DT_TMP_SCHEMA := $(objtree)/$(DT_BINDING quiet_cmd_dtb_check = CHECK $@ cmd_dtb_check = $(DT_CHECKER) -u $(srctree)/$(DT_BINDING_DIR) -p $(DT_TMP_SCHEMA) $@ ; -define rule_dtc_dt_yaml +define rule_dtc $(call cmd_and_fixdep,dtc,yaml) $(call cmd,dtb_check) endef $(obj)/%.dt.yaml: $(src)/%.dts $(DTC) $(DT_TMP_SCHEMA) FORCE - $(call if_changed_rule,dtc_dt_yaml) + $(call if_changed_rule,dtc) dtc-tmp = $(subst $(comma),_,$(dot-target).dts.tmp)
Greg Kroah-Hartman
maintainerPATCH 102/152 hv_netvsc: Fix unwanted wakeup in netvsc_attach()
In reply to Greg Kroah-HartmanFrom: Haiyang Zhang <[email protected]>
commit f6f13c125e05603f68f5bf31f045b95e6d493598 upstream.
When netvsc_attach() is called by operations like changing MTU, etc., an extra wakeup may happen while netvsc_attach() calling rndis_filter_device_add() which sends rndis messages when queue is stopped in netvsc_detach(). The completion message will wake up queue 0.
We can reproduce the issue by changing MTU etc., then the wake_queue counter from "ethtool -S" will increase beyond stop_queue counter:
stop_queue: 0
wake_queue: 1
The issue causes queue wake up, and counter increment, no other ill effects in current code. So we didn't see any network problem for now.
To fix this, initialize tx_disable to true, and set it to false when the NIC is ready to be attached or registered.
--- drivers/net/hyperv/netvsc.c | 2 +- drivers/net/hyperv/netvsc_drv.c | 3 +++ 2 files changed, 4 insertions(+), 1 deletion(-)
drivers/net/hyperv/netvsc.c +1 -1
@@ -99,7 +99,7 @@ static struct netvsc_device *alloc_net_dinit_waitqueue_head (&net_device->wait_drain); net_device->destroy = false; - net_device->tx_disable = false; + net_device->tx_disable = true; net_device->max_pkt =RNDIS_MAX_PKT_DEFAULT ; net_device->pkt_align =RNDIS_PKT_ALIGN_DEFAULT ;
drivers/net/hyperv/netvsc_drv.c +3 -0
@@ -973,6 +973,7 @@ static int netvsc_attach(struct net_devi }/* In any case device is now ready */ + nvdev->tx_disable = false;netif_device_attach (ndev);/* Note: enable and attach happen when sub-channels setup */
@@ -2350,6 +2351,8 @@ static int netvsc_probe(struct hv_deviceelse net->max_mtu =ETH_DATA_LEN ; + nvdev->tx_disable = false; + ret =register_netdevice (net);if (ret !=0 ) {pr_err ("Unable to register netdev.\n" );
Greg Kroah-Hartman
maintainerPATCH 103/152 usb: charger: assign specific number for enum value
In reply to Greg Kroah-HartmanFrom: Peter Chen <[email protected]>
commit ca4b43c14cd88d28cfc6467d2fa075aad6818f1d upstream.
To work properly on every architectures and compilers, the enum value needs to be specific numbers.
--- include/uapi/linux/usb/charger.h | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-)
include/uapi/linux/usb/charger.h +8 -8
@@ -14,18 +14,18 @@* ACA (Accessory Charger Adapters) */ enum usb_charger_type { -UNKNOWN_TYPE , -SDP_TYPE , -DCP_TYPE , -CDP_TYPE , -ACA_TYPE , +UNKNOWN_TYPE =0 , +SDP_TYPE =1 , +DCP_TYPE =2 , +CDP_TYPE =3 , +ACA_TYPE =4 , };/* USB charger state */ enum usb_charger_state { -USB_CHARGER_DEFAULT , -USB_CHARGER_PRESENT , -USB_CHARGER_ABSENT , +USB_CHARGER_DEFAULT =0 , +USB_CHARGER_PRESENT =1 , +USB_CHARGER_ABSENT =2 , };#endif /* _UAPI__LINUX_USB_CHARGER_H */
Greg Kroah-Hartman
maintainerPATCH 104/152 nvme-pci: Hold cq_poll_lock while completing CQEs
In reply to Greg Kroah-HartmanFrom: Bijan Mottahedeh <[email protected]>
commit 9515743bfb39c61aaf3d4f3219a645c8d1fe9a0e upstream.
Completions need to consumed in the same order the controller submitted them, otherwise future completion entries may overwrite ones we haven't handled yet. Hold the nvme queue's poll lock while completing new CQEs to prevent another thread from freeing command tags for reuse out-of-order.
--- drivers/nvme/host/pci.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/nvme/host/pci.c +1 -1
@@ -1084,9 +1084,9 @@ static int nvme_poll(struct blk_mq_hw_ctspin_lock (&nvmeq->cq_poll_lock); found =nvme_process_cq (nvmeq, &start, &end,-1 ); +nvme_complete_cqes (nvmeq, start, end);spin_unlock (&nvmeq->cq_poll_lock); -nvme_complete_cqes (nvmeq, start, end);return found; }
Greg Kroah-Hartman
maintainerPATCH 105/152 s390/qeth: vnicc Fix EOPNOTSUPP precedence
In reply to Greg Kroah-HartmanFrom: Alexandra Winter <[email protected]>
commit 6f3846f0955308b6d1b219419da42b8de2c08845 upstream.
When getting or setting VNICC parameters, the error code EOPNOTSUPP should have precedence over EBUSY.
EBUSY is used because vnicc feature and bridgeport feature are mutually exclusive, which is a temporary condition.
Whereas EOPNOTSUPP indicates that the HW does not support all or parts of the vnicc feature.
This issue causes the vnicc sysfs params to show 'blocked by bridgeport' for HW that does not support VNICC at all.
--- drivers/s390/net/qeth_l2_main.c | 29 +++++++++++++---------------- 1 file changed, 13 insertions(+), 16 deletions(-)
drivers/s390/net/qeth_l2_main.c +13 -16
@@ -1846,15 +1846,14 @@ int qeth_l2_vnicc_set_state(struct qeth_QETH_CARD_TEXT (card,2 ,"vniccsch" ); -/* do not change anything if BridgePort is enabled */ -if (qeth_bridgeport_is_in_use (card)) -return -EBUSY ; -/* check if characteristic and enable/disable are supported */ if (!(card->options.vnicc.sup_chars & vnicc) || !(card->options.vnicc.set_char_sup & vnicc))return -EOPNOTSUPP ; +if (qeth_bridgeport_is_in_use (card)) +return -EBUSY ; +/* set enable/disable command and store wanted characteristic */ if (state) { cmd =IPA_VNICC_ENABLE ;
@@ -1900,14 +1899,13 @@ int qeth_l2_vnicc_get_state(struct qeth_QETH_CARD_TEXT (card,2 ,"vniccgch" ); -/* do not get anything if BridgePort is enabled */ -if (qeth_bridgeport_is_in_use (card)) -return -EBUSY ; -/* check if characteristic is supported */ if (!(card->options.vnicc.sup_chars & vnicc))return -EOPNOTSUPP ; +if (qeth_bridgeport_is_in_use (card)) +return -EBUSY ; +/* if card is ready, query current VNICC state */ if (qeth_card_hw_is_reachable (card)) rc =qeth_l2_vnicc_query_chars (card);
@@ -1925,15 +1923,14 @@ int qeth_l2_vnicc_set_timeout(struct qetQETH_CARD_TEXT (card,2 ,"vniccsto" ); -/* do not change anything if BridgePort is enabled */ -if (qeth_bridgeport_is_in_use (card)) -return -EBUSY ; -/* check if characteristic and set_timeout are supported */ if (!(card->options.vnicc.sup_chars &QETH_VNICC_LEARNING ) || !(card->options.vnicc.getset_timeout_sup &QETH_VNICC_LEARNING ))return -EOPNOTSUPP ; +if (qeth_bridgeport_is_in_use (card)) +return -EBUSY ; +/* do we need to do anything? */ if (card->options.vnicc.learning_timeout == timeout)return rc;
@@ -1962,14 +1959,14 @@ int qeth_l2_vnicc_get_timeout(struct qetQETH_CARD_TEXT (card,2 ,"vniccgto" ); -/* do not get anything if BridgePort is enabled */ -if (qeth_bridgeport_is_in_use (card)) -return -EBUSY ; -/* check if characteristic and get_timeout are supported */ if (!(card->options.vnicc.sup_chars &QETH_VNICC_LEARNING ) || !(card->options.vnicc.getset_timeout_sup &QETH_VNICC_LEARNING ))return -EOPNOTSUPP ; + +if (qeth_bridgeport_is_in_use (card)) +return -EBUSY ; +/* if card is ready, get timeout. Otherwise, just return stored value */ *timeout = card->options.vnicc.learning_timeout;if (qeth_card_hw_is_reachable (card))
Greg Kroah-Hartman
maintainerPATCH 106/152 net: netlink: cap max groups which will be considered in netlink_bind()
In reply to Greg Kroah-HartmanFrom: Nikolay Aleksandrov <[email protected]>
commit 3a20773beeeeadec41477a5ba872175b778ff752 upstream.
Since nl_groups is a u32 we can't bind more groups via ->bind (netlink_bind) call, but netlink has supported more groups via setsockopt() for a long time and thus nlk->ngroups could be over 32. Recently I added support for per-vlan notifications and increased the groups to 33 for NETLINK_ROUTE which exposed an old bug in the netlink_bind() code causing out-of-bounds access on archs where unsigned long is 32 bits via test_bit() on a local variable. Fix this by capping the maximum groups in netlink_bind() to BITS_PER_TYPE(u32), effectively capping them at 32 which is the minimum of allocated groups and the maximum groups which can be bound via netlink_bind().
--- net/netlink/af_netlink.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
net/netlink/af_netlink.c +3 -2
@@ -1014,7 +1014,8 @@ static int netlink_bind(struct socket *sif (nlk->netlink_bind && groups) {int group; -for (group =0 ; group < nlk->ngroups; group++) { +/* nl_groups is a u32, so cap the maximum groups we can bind */ +for (group =0 ; group <BITS_PER_TYPE (u32); group++) {if (!test_bit (group, &groups))continue ; err = nlk->netlink_bind (net, group +1 );
@@ -1033,7 +1034,7 @@ static int netlink_bind(struct socket *s netlink_insert(sk, nladdr->nl_pid) :netlink_autobind (sock);if (err) { -netlink_undo_bind (nlk->ngroups, groups, sk); +netlink_undo_bind (BITS_PER_TYPE (u32), groups, sk); goto unlock; } }
Greg Kroah-Hartman
maintainerPATCH 107/152 net: atlantic: fix use after free kasan warn
In reply to Greg Kroah-HartmanFrom: Pavel Belous <[email protected]>
commit a4980919ad6a7be548d499bc5338015e1a9191c6 upstream.
skb->len is used to calculate statistics after xmit invocation.
Under a stress load it may happen that skb will be xmited, rx interrupt will come and skb will be freed, all before xmit function is even returned.
Eventually, skb->len will access unallocated area.
Moving stats calculation into tx_clean routine.
--- drivers/net/ethernet/aquantia/atlantic/aq_nic.c | 4 ---- drivers/net/ethernet/aquantia/atlantic/aq_ring.c | 7 +++++-- 2 files changed, 5 insertions(+), 6 deletions(-)
drivers/net/ethernet/aquantia/atlantic/aq_nic.c +0 -4
@@ -598,10 +598,6 @@ int aq_nic_xmit(struct aq_nic_s *self, sif (likely (frags)) { err = self->aq_hw_ops->hw_ring_tx_xmit (self->aq_hw, ring, frags); -if (err >=0 ) { - ++ring->stats.tx.packets; - ring->stats.tx.bytes += skb->len; - } }else { err =NETDEV_TX_BUSY ; }
drivers/net/ethernet/aquantia/atlantic/aq_ring.c +5 -2
@@ -243,9 +243,12 @@ bool aq_ring_tx_clean(struct aq_ring_s * } } -if (unlikely (buff->is_eop)) -dev_kfree_skb_any (buff->skb); +if (unlikely (buff->is_eop)) { + ++self->stats.rx.packets; + self->stats.tx.bytes += buff->skb->len; +dev_kfree_skb_any (buff->skb); + } buff->pa =0U ; buff->eop_index =0xffffU ; self->sw_head =aq_ring_next_dx (self, self->sw_head);
Greg Kroah-Hartman
maintainerPATCH 108/152 net: atlantic: fix potential error handling
In reply to Greg Kroah-HartmanFrom: Pavel Belous <[email protected]>
commit 380ec5b9af7f0d57dbf6ac067fd9f33cff2fef71 upstream.
Code inspection found that in case of mapping error we do return current 'ret' value. But beside error, it is used to count number of descriptors allocated for the packet. In that case map_skb function could return '1'.
Changing it to return zero (number of mapped descriptors for skb)
--- drivers/net/ethernet/aquantia/atlantic/aq_nic.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/net/ethernet/aquantia/atlantic/aq_nic.c +3 -1
@@ -467,8 +467,10 @@ static unsigned int aq_nic_map_skb(struc dx_buff->len, DMA_TO_DEVICE); -if (unlikely (dma_mapping_error (aq_nic_get_dev (self), dx_buff->pa))) +if (unlikely (dma_mapping_error (aq_nic_get_dev (self), dx_buff->pa))) { + ret =0 ; goto exit; + } first = dx_buff; dx_buff->len_pkt = skb->len;
Greg Kroah-Hartman
maintainerPATCH 109/152 net: atlantic: fix out of range usage of active_vlans array
In reply to Greg Kroah-HartmanFrom: Dmitry Bogdanov <[email protected]>
commit 5a292c89a84d49b598f8978f154bdda48b1072c0 upstream.
fix static checker warning:
drivers/net/ethernet/aquantia/atlantic/aq_filters.c:166 aq_check_approve_fvlan()
error: passing untrusted data to 'test_bit()'
--- drivers/net/ethernet/aquantia/atlantic/aq_filters.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/aquantia/atlantic/aq_filters.c +1 -1
@@ -158,7 +158,7 @@ aq_check_approve_fvlan(struct aq_nic_s * }if ((aq_nic->ndev->features &NETIF_F_HW_VLAN_CTAG_FILTER ) && - (!test_bit (be16_to_cpu (fsp->h_ext.vlan_tci), + (!test_bit (be16_to_cpu (fsp->h_ext.vlan_tci) &VLAN_VID_MASK , aq_nic->active_vlans))) {netdev_err (aq_nic->ndev,"ethtool: unknown vlan-id specified" );
Greg Kroah-Hartman
maintainerPATCH 110/152 net/smc: no peer ID in CLC decline for SMCD
In reply to Greg Kroah-HartmanFrom: Ursula Braun <[email protected]>
commit 369537c97024dca99303a8d4d6ab38b4f54d3909 upstream.
Just SMCR requires a CLC Peer ID, but not SMCD. The field should be zero for SMCD.
--- net/smc/smc_clc.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
net/smc/smc_clc.c +3 -1
@@ -372,7 +372,9 @@ int smc_clc_send_decline(struct smc_sock dclc.hdr.length =htons (sizeof (struct smc_clc_msg_decline )); dclc.hdr.version =SMC_CLC_V1 ; dclc.hdr.flag = (peer_diag_info ==SMC_CLC_DECL_SYNCERR ) ?1 :0 ; -memcpy (dclc.id_for_peer, local_systemid,sizeof (local_systemid)); +if (smc->conn.lgr && !smc->conn.lgr->is_smcd) +memcpy (dclc.id_for_peer, local_systemid, +sizeof (local_systemid)); dclc.peer_diagnosis =htonl (peer_diag_info);memcpy (dclc.trl.eyecatcher,SMC_EYECATCHER ,sizeof (SMC_EYECATCHER ));
Greg Kroah-Hartman
maintainerPATCH 111/152 net: ena: make ena rxfh support ETH_RSS_HASH_NO_CHANGE
In reply to Greg Kroah-HartmanFrom: Arthur Kiyanovski <[email protected]>
commit 470793a78ce344bd53d31e0c2d537f71ba957547 upstream.
As the name suggests ETH_RSS_HASH_NO_CHANGE is received upon changing the key or indirection table using ethtool while keeping the same hash function.
Also add a function for retrieving the current hash function from the ena-com layer.
--- drivers/net/ethernet/amazon/ena/ena_com.c | 5 +++++ drivers/net/ethernet/amazon/ena/ena_com.h | 8 ++++++++ drivers/net/ethernet/amazon/ena/ena_ethtool.c | 3 +++ 3 files changed, 16 insertions(+)
drivers/net/ethernet/amazon/ena/ena_com.c +5 -0
@@ -1059,6 +1059,11 @@ static void ena_com_hash_key_fill_defaul hash_key->keys_num =sizeof (hash_key->key) /sizeof (u32); } +int ena_com_get_current_hash_function (struct ena_com_dev *ena_dev) +{ +return ena_dev->rss.hash_func; +} +static int ena_com_hash_key_allocate (struct ena_com_dev *ena_dev) {struct ena_rss *rss = &ena_dev->rss;
drivers/net/ethernet/amazon/ena/ena_com.h +8 -0
@@ -656,6 +656,14 @@ int ena_com_rss_init(struct ena_com_dev*/ void ena_com_rss_destroy (struct ena_com_dev *ena_dev); +/* ena_com_get_current_hash_function - Get RSS hash function +* @ena_dev: ENA communication layer struct +* +* Return the current hash function. +* @return: 0 or one of the ena_admin_hash_functions values. +*/ +int ena_com_get_current_hash_function (struct ena_com_dev *ena_dev); +/* ena_com_fill_hash_function - Fill RSS hash function * @ena_dev: ENA communication layer struct * @func: The hash function (Toeplitz or crc)
drivers/net/ethernet/amazon/ena/ena_ethtool.c +3 -0
@@ -736,6 +736,9 @@ static int ena_set_rxfh(struct net_devic }switch (hfunc) { +case ETH_RSS_HASH_NO_CHANGE : + func =ena_com_get_current_hash_function (ena_dev); +break ;case ETH_RSS_HASH_TOP : func =ENA_ADMIN_TOEPLITZ ;break ;
Greg Kroah-Hartman
maintainerPATCH 112/152 selftests: Install settings files to fix TIMEOUT failures
In reply to Greg Kroah-HartmanFrom: Michael Ellerman <[email protected]>
commit b9167c8078c3527de6da241c8a1a75a9224ed90a upstream.
Commit 852c8cbf34d3 ("selftests/kselftest/runner.sh: Add 45 second timeout per test") added a 45 second timeout for tests, and also added a way for tests to customise the timeout via a settings file.
For example the ftrace tests take multiple minutes to run, so they were given longer in commit b43e78f65b1d ("tracing/selftests: Turn off timeout setting").
This works when the tests are run from the source tree. However if the tests are installed with "make -C tools/testing/selftests install", the settings files are not copied into the install directory. When the tests are then run from the install directory the longer timeouts are not applied and the tests timeout incorrectly.
So add the settings files to TEST_FILES of the appropriate Makefiles to cause the settings files to be installed using the existing install logic.
--- tools/testing/selftests/ftrace/Makefile | 2 +- tools/testing/selftests/livepatch/Makefile | 2 ++ tools/testing/selftests/rseq/Makefile | 2 ++ tools/testing/selftests/rtc/Makefile | 2 ++ 4 files changed, 7 insertions(+), 1 deletion(-)
tools/testing/selftests/ftrace/Makefile +1 -1
@@ -2,7 +2,7 @@ all: TEST_PROGS := ftracetest -TEST_FILES := test.d +TEST_FILES := test.d settings EXTRA_CLEAN := $(OUTPUT)/logs/* include ../lib.mk
tools/testing/selftests/livepatch/Makefile +2 -0
@@ -6,4 +6,6 @@ TEST_PROGS := \ test-callbacks.sh \ test-shadow-vars.sh +TEST_FILES := settings + include ../lib.mk
tools/testing/selftests/rseq/Makefile +2 -0
@@ -19,6 +19,8 @@ TEST_GEN_PROGS_EXTENDED = librseq.so TEST_PROGS = run_param_test.sh +TEST_FILES := settings + include ../lib.mk $(OUTPUT)/librseq.so: rseq.c rseq.h rseq-*.h
tools/testing/selftests/rtc/Makefile +2 -0
@@ -6,4 +6,6 @@ TEST_GEN_PROGS = rtctest TEST_GEN_PROGS_EXTENDED = setdate +TEST_FILES := settings + include ../lib.mk
Greg Kroah-Hartman
maintainerPATCH 113/152 kbuild: remove header compile test
In reply to Greg Kroah-HartmanFrom: Masahiro Yamada <[email protected]>
commit fcbb8461fd2376ba3782b5b8bd440c929b8e4980 upstream.
There are both positive and negative options about this feature. At first, I thought it was a good idea, but actually Linus stated a negative opinion (https://lkml.org/lkml/2019/9/29/227). I admit it is ugly and annoying.
The baseline I'd like to keep is the compile-test of uapi headers. (Otherwise, kernel developers have no way to ensure the correctness of the exported headers.)
I will maintain a small build rule in usr/include/Makefile. Remove the other header test functionality.
[ added to 5.4.y due to start of build warnings from backported patches
because of this feature - gregkh]
--- Documentation/kbuild/makefiles.rst | 17 Makefile | 1 drivers/gpu/drm/i915/Kconfig.debug | 1 include/Kbuild | 1185 ------------------------------------- init/Kconfig | 22 scripts/Makefile.build | 9 scripts/Makefile.lib | 14 usr/include/Makefile | 12 8 files changed, 9 insertions(+), 1252 deletions(-)
Documentation/kbuild/makefiles.rst +0 -17
@@ -1115,23 +1115,6 @@ When kbuild executes, the following step In this example, extra-y is used to list object files that shall be built, but shall not be linked as part of built-in.a. - header-test-y - - header-test-y specifies headers (`*.h`) in the current directory that - should be compile tested to ensure they are self-contained, - i.e. compilable as standalone units. If CONFIG_HEADER_TEST is enabled, - this builds them as part of extra-y. - - header-test-pattern-y - - This works as a weaker version of header-test-y, and accepts wildcard - patterns. The typical usage is:: - - header-test-pattern-y += *.h - - This specifies all the files that matches to `*.h` in the current - directory, but the files in 'header-test-' are excluded. - 6.7 Commands useful for building a boot image ---------------------------------------------
Makefile +0 -1
@@ -618,7 +618,6 @@ ifeq ($(KBUILD_EXTMOD),) init-y := init/ drivers-y := drivers/ sound/ drivers-$(CONFIG_SAMPLES) += samples/ -drivers-$(CONFIG_KERNEL_HEADER_TEST) += include/ net-y := net/ libs-y := lib/ core-y := usr/
drivers/gpu/drm/i915/Kconfig.debug +0 -1
@@ -7,7 +7,6 @@ config DRM_I915_WERROR # We use the dependency on !COMPILE_TEST to not be enabled in # allmodconfig or allyesconfig configurations depends on !COMPILE_TEST - select HEADER_TEST default n help Add -Werror to the build flags for (and only for) i915.ko.
include/Kbuild deleted +0 -1185
1,185 lines not on this page · raw message
init/Kconfig +1 -21
@@ -105,29 +105,9 @@ config COMPILE_TEST here. If you are a user/distributor, say N here to exclude useless drivers to be distributed. -config HEADER_TEST - bool "Compile test headers that should be standalone compilable" - help - Compile test headers listed in header-test-y target to ensure they are - self-contained, i.e. compilable as standalone units. - - If you are a developer or tester and want to ensure the requested - headers are self-contained, say Y here. Otherwise, choose N. - -config KERNEL_HEADER_TEST - bool "Compile test kernel headers" - depends on HEADER_TEST - help - Headers in include/ are used to build external moduls. - Compile test them to ensure they are self-contained, i.e. - compilable as standalone units. - - If you are a developer or tester and want to ensure the headers - in include/ are self-contained, say Y here. Otherwise, choose N. - config UAPI_HEADER_TEST bool "Compile test UAPI headers" - depends on HEADER_TEST && HEADERS_INSTALL && CC_CAN_LINK + depends on HEADERS_INSTALL && CC_CAN_LINK help Compile test headers exported to user-space to ensure they are self-contained, i.e. compilable as standalone units.
scripts/Makefile.build +0 -9
@@ -283,15 +283,6 @@ quiet_cmd_cc_lst_c = MKLST $@ $(obj)/%.lst: $(src)/%.c FORCE $(call if_changed_dep,cc_lst_c) -# header test (header-test-y, header-test-m target) -# --------------------------------------------------------------------------- - -quiet_cmd_cc_s_h = CC $@ - cmd_cc_s_h = $(CC) $(c_flags) -S -o $@ -x c /dev/null -include $< - -$(obj)/%.h.s: $(src)/%.h FORCE - $(call if_changed_dep,cc_s_h) - # Compile assembler sources (.S) # ---------------------------------------------------------------------------
scripts/Makefile.lib +0 -14
@@ -65,20 +65,6 @@ extra-y += $(patsubst %.dtb,%.dt.yaml, $ extra-$(CONFIG_OF_ALL_DTBS) += $(patsubst %.dtb,%.dt.yaml, $(dtb-)) endif -# Test self-contained headers - -# Wildcard searches in $(srctree)/$(src)/, but not in $(objtree)/$(obj)/. -# Stale generated headers are often left over, so pattern matching should -# be avoided. Please notice $(srctree)/$(src)/ and $(objtree)/$(obj) point -# to the same location for in-tree building. So, header-test-pattern-y should -# be used with care. -header-test-y += $(filter-out $(header-test-), \ - $(patsubst $(srctree)/$(src)/%, %, \ - $(wildcard $(addprefix $(srctree)/$(src)/, \ - $(header-test-pattern-y))))) - -extra-$(CONFIG_HEADER_TEST) += $(addsuffix .s, $(header-test-y) $(header-test-m)) - # Add subdir path extra-y := $(addprefix $(obj)/,$(extra-y))
usr/include/Makefile +8 -4
@@ -99,9 +99,13 @@ endif # asm-generic/*.h is used by asm/*.h, and should not be included directly header-test- += asm-generic/% -# The rest are compile-tested -header-test-y += $(filter-out $(header-test-), \ - $(patsubst $(obj)/%,%, $(wildcard \ - $(addprefix $(obj)/, *.h */*.h */*/*.h */*/*/*.h)))) +extra-y := $(patsubst %.h,%.hdrtest, $(filter-out $(header-test-), \ + $(patsubst $(obj)/%,%, $(shell find $(obj) -name '*.h')))) + +quiet_cmd_hdrtest = HDRTEST $< + cmd_hdrtest = $(CC) $(c_flags) -S -o /dev/null -x c /dev/null -include $<; touch $@ + +$(obj)/%.hdrtest: $(obj)/%.h FORCE + $(call if_changed_dep,hdrtest) clean-files += $(filter-out Makefile, $(notdir $(wildcard $(obj)/*)))
Greg Kroah-Hartman
maintainerPATCH 114/152 kbuild: move headers_check rule to usr/include/Makefile
In reply to Greg Kroah-HartmanFrom: Masahiro Yamada <[email protected]>
commit 7ecaf069da52e472d393f03e79d721aabd724166 upstream.
Currently, some sanity checks for uapi headers are done by scripts/headers_check.pl, which is wired up to the 'headers_check' target in the top Makefile.
It is true compiling headers has better test coverage, but there are still several headers excluded from the compile test. I like to keep headers_check.pl for a while, but we can delete a lot of code by moving the build rule to usr/include/Makefile.
--- Makefile | 11 +++-------- lib/Kconfig.debug | 11 ----------- scripts/Makefile.headersinst | 18 ------------------ usr/include/Makefile | 9 ++++++--- 4 files changed, 9 insertions(+), 40 deletions(-)
Makefile +3 -8
@@ -1195,19 +1195,15 @@ headers: $(version_h) scripts_unifdef ua $(Q)$(MAKE) $(hdr-inst)=include/uapi $(Q)$(MAKE) $(hdr-inst)=arch/$(SRCARCH)/include/uapi +# Deprecated. It is no-op now. PHONY += headers_check -headers_check: headers - $(Q)$(MAKE) $(hdr-inst)=include/uapi HDRCHECK=1 - $(Q)$(MAKE) $(hdr-inst)=arch/$(SRCARCH)/include/uapi HDRCHECK=1 +headers_check: + @: ifdef CONFIG_HEADERS_INSTALL prepare: headers endif -ifdef CONFIG_HEADERS_CHECK -all: headers_check -endif - PHONY += scripts_unifdef scripts_unifdef: scripts_basic $(Q)$(MAKE) $(build)=scripts scripts/unifdef
@@ -1475,7 +1471,6 @@ help: @echo ' versioncheck - Sanity check on version.h usage' @echo ' includecheck - Check for duplicate included header files' @echo ' export_report - List the usages of all exported symbols' - @echo ' headers_check - Sanity check on exported headers' @echo ' headerdep - Detect inclusion cycles in headers' @echo ' coccicheck - Check with Coccinelle' @echo ''
lib/Kconfig.debug +0 -11
@@ -299,17 +299,6 @@ config HEADERS_INSTALL user-space program samples. It is also needed by some features such as uapi header sanity checks. -config HEADERS_CHECK - bool "Run sanity checks on uapi headers when building 'all'" - depends on HEADERS_INSTALL - help - This option will run basic sanity checks on uapi headers when - building the 'all' target, for example, ensure that they do not - attempt to include files which were not exported, etc. - - If you're making modifications to header files which are - relevant for userspace, say 'Y'. - config OPTIMIZE_INLINING def_bool y help
scripts/Makefile.headersinst +0 -18
@@ -56,9 +56,6 @@ new-dirs := $(filter-out $(existing $(if $(new-dirs), $(shell mkdir -p $(new-dirs))) # Rules - -ifndef HDRCHECK - quiet_cmd_install = HDRINST $@ cmd_install = $(CONFIG_SHELL) $(srctree)/scripts/headers_install.sh $< $@
@@ -81,21 +78,6 @@ existing-headers := $(filter $(old-heade -include $(foreach f,$(existing-headers),$(dir $(f)).$(notdir $(f)).cmd) -else - -quiet_cmd_check = HDRCHK $< - cmd_check = $(PERL) $(srctree)/scripts/headers_check.pl $(dst) $(SRCARCH) $<; touch $@ - -check-files := $(addsuffix .chk, $(all-headers)) - -$(check-files): $(dst)/%.chk : $(dst)/% $(srctree)/scripts/headers_check.pl - $(call cmd,check) - -__headers: $(check-files) - @: - -endif - PHONY += FORCE FORCE:
usr/include/Makefile +6 -3
@@ -99,11 +99,14 @@ endif # asm-generic/*.h is used by asm/*.h, and should not be included directly header-test- += asm-generic/% -extra-y := $(patsubst %.h,%.hdrtest, $(filter-out $(header-test-), \ - $(patsubst $(obj)/%,%, $(shell find $(obj) -name '*.h')))) +extra-y := $(patsubst $(obj)/%.h,%.hdrtest, $(shell find $(obj) -name '*.h')) quiet_cmd_hdrtest = HDRTEST $< - cmd_hdrtest = $(CC) $(c_flags) -S -o /dev/null -x c /dev/null -include $<; touch $@ + cmd_hdrtest = \ + $(CC) $(c_flags) -S -o /dev/null -x c /dev/null \ + $(if $(filter-out $(header-test-), $*.h), -include $<); \ + $(PERL) $(srctree)/scripts/headers_check.pl $(obj) $(SRCARCH) $<; \ + touch $@ $(obj)/%.hdrtest: $(obj)/%.h FORCE $(call if_changed_dep,hdrtest)
Greg Kroah-Hartman
maintainerPATCH 115/152 kbuild: remove unneeded variable, single-all
In reply to Greg Kroah-HartmanFrom: Masahiro Yamada <[email protected]>
commit 35e046a203ee3bc8ba9ae3561b50de02646dfb81 upstream.
When single-build is set, everything in $(MAKECMDGOALS) is a single target. You can use $(MAKECMDGOALS) to list out the single targets.
--- Makefile | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-)
Makefile +2 -4
@@ -1761,11 +1761,9 @@ tools/%: FORCE ifdef single-build -single-all := $(filter $(single-targets), $(MAKECMDGOALS)) - # .ko is special because modpost is needed -single-ko := $(sort $(filter %.ko, $(single-all))) -single-no-ko := $(sort $(patsubst %.ko,%.mod, $(single-all))) +single-ko := $(sort $(filter %.ko, $(MAKECMDGOALS))) +single-no-ko := $(sort $(patsubst %.ko,%.mod, $(MAKECMDGOALS))) $(single-ko): single_modpost @:
Greg Kroah-Hartman
maintainerPATCH 116/152 kbuild: make single target builds even faster
In reply to Greg Kroah-HartmanFrom: Masahiro Yamada <[email protected]>
commit b1fbfcb4a20949df08dd995927cdc5ad220c128d upstream.
Commit 2dffd23f81a3 ("kbuild: make single target builds much faster") made the situation much better.
To improve it even more, apply the similar idea to the top Makefile. Trim unrelated directories from build-dirs.
The single build code must be moved above the 'descend' target.
--- Makefile | 90 ++++++++++++++++++++++++++++++++------------------------------- 1 file changed, 47 insertions(+), 43 deletions(-)
Makefile +47 -43
@@ -1635,6 +1635,50 @@ help: PHONY += prepare endif # KBUILD_EXTMOD +# Single targets +# --------------------------------------------------------------------------- +# To build individual files in subdirectories, you can do like this: +# +# make foo/bar/baz.s +# +# The supported suffixes for single-target are listed in 'single-targets' +# +# To build only under specific subdirectories, you can do like this: +# +# make foo/bar/baz/ + +ifdef single-build + +# .ko is special because modpost is needed +single-ko := $(sort $(filter %.ko, $(MAKECMDGOALS))) +single-no-ko := $(sort $(patsubst %.ko,%.mod, $(MAKECMDGOALS))) + +$(single-ko): single_modpost + @: +$(single-no-ko): descend + @: + +ifeq ($(KBUILD_EXTMOD),) +# For the single build of in-tree modules, use a temporary file to avoid +# the situation of modules_install installing an invalid modules.order. +MODORDER := .modules.tmp +endif + +PHONY += single_modpost +single_modpost: $(single-no-ko) + $(Q){ $(foreach m, $(single-ko), echo $(extmod-prefix)$m;) } > $(MODORDER) + $(Q)$(MAKE) -f $(srctree)/scripts/Makefile.modpost + +KBUILD_MODULES := 1 + +export KBUILD_SINGLE_TARGETS := $(addprefix $(extmod-prefix), $(single-no-ko)) + +# trim unrelated directories +build-dirs := $(foreach d, $(build-dirs), \ + $(if $(filter $(d)/%, $(KBUILD_SINGLE_TARGETS)), $(d))) + +endif + # Handle descending into subdirectories listed in $(build-dirs) # Preset locale variables to speed up the build process. Limit locale # tweaks to this spot to avoid wrong language settings when running
@@ -1643,7 +1687,9 @@ endif # KBUILD_EXTMOD PHONY += descend $(build-dirs) descend: $(build-dirs) $(build-dirs): prepare - $(Q)$(MAKE) $(build)=$@ single-build=$(single-build) need-builtin=1 need-modorder=1 + $(Q)$(MAKE) $(build)=$@ \ + single-build=$(if $(filter-out $@/, $(single-no-ko)),1) \ + need-builtin=1 need-modorder=1 clean-dirs := $(addprefix _clean_, $(clean-dirs)) PHONY += $(clean-dirs) clean
@@ -1747,48 +1793,6 @@ tools/%: FORCE $(Q)mkdir -p $(objtree)/tools $(Q)$(MAKE) LDFLAGS= MAKEFLAGS="$(tools_silent) $(filter --j% -j,$(MAKEFLAGS))" O=$(abspath $(objtree)) subdir=tools -C $(srctree)/tools/ $* -# Single targets -# --------------------------------------------------------------------------- -# To build individual files in subdirectories, you can do like this: -# -# make foo/bar/baz.s -# -# The supported suffixes for single-target are listed in 'single-targets' -# -# To build only under specific subdirectories, you can do like this: -# -# make foo/bar/baz/ - -ifdef single-build - -# .ko is special because modpost is needed -single-ko := $(sort $(filter %.ko, $(MAKECMDGOALS))) -single-no-ko := $(sort $(patsubst %.ko,%.mod, $(MAKECMDGOALS))) - -$(single-ko): single_modpost - @: -$(single-no-ko): descend - @: - -ifeq ($(KBUILD_EXTMOD),) -# For the single build of in-tree modules, use a temporary file to avoid -# the situation of modules_install installing an invalid modules.order. -MODORDER := .modules.tmp -endif - -PHONY += single_modpost -single_modpost: $(single-no-ko) - $(Q){ $(foreach m, $(single-ko), echo $(extmod-prefix)$m;) } > $(MODORDER) - $(Q)$(MAKE) -f $(srctree)/scripts/Makefile.modpost - -KBUILD_MODULES := 1 - -export KBUILD_SINGLE_TARGETS := $(addprefix $(extmod-prefix), $(single-no-ko)) - -single-build = $(if $(filter-out $@/, $(single-no-ko)),1) - -endif - # FIXME Should go into a make.lib or something # ===========================================================================
Greg Kroah-Hartman
maintainerPATCH 117/152 namei: only return -ECHILD from follow_dotdot_rcu()
In reply to Greg Kroah-HartmanFrom: Aleksa Sarai <[email protected]>
commit 2b98149c2377bff12be5dd3ce02ae0506e2dd613 upstream.
It's over-zealous to return hard errors under RCU-walk here, given that a REF-walk will be triggered for all other cases handling ".." under RCU.
The original purpose of this check was to ensure that if a rename occurs such that a directory is moved outside of the bind-mount which the resolution started in, it would be detected and blocked to avoid being able to mess with paths outside of the bind-mount. However, triggering a new REF-walk is just as effective a solution.
--- fs/namei.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
fs/namei.c +1 -1
@@ -1360,7 +1360,7 @@ static int follow_dotdot_rcu(struct name nd->path.dentry = parent; nd->seq = seq;if (unlikely (!path_connected (&nd->path))) -return -ENOENT ; +return -ECHILD ;break ; }else {struct mount *mnt =real_mount (nd->path.mnt);
Greg Kroah-Hartman
maintainerPATCH 118/152 mwifiex: drop most magic numbers from mwifiex_process_tdls_action_frame()
In reply to Greg Kroah-HartmanFrom: Brian Norris <[email protected]>
commit 70e5b8f445fd27fde0c5583460e82539a7242424 upstream.
Before commit 1e58252e334d ("mwifiex: Fix heap overflow in mmwifiex_process_tdls_action_frame()"), mwifiex_process_tdls_action_frame() already had too many magic numbers. But this commit just added a ton more, in the name of checking for buffer overflows. That seems like a really bad idea.
Let's make these magic numbers a little less magic, by
(a) factoring out 'pos[1]' as 'ie_len'
(b) using 'sizeof' on the appropriate source or destination fields where possible, instead of bare numbers
(c) dropping redundant checks, per below.
Regarding redundant checks: the beginning of the loop has this:
if (pos + 2 + pos[1] > end)
break;but then individual 'case's include stuff like this:
if (pos > end - 3)
return;
if (pos[1] != 1)
return;
Note that the second 'return' (validating the length, pos[1]) combined with the above condition (ensuring 'pos + 2 + length' doesn't exceed 'end'), makes the first 'return' (whose 'if' can be reworded as 'pos > end - pos[1] - 2') redundant. Rather than unwind the magic numbers there, just drop those conditions.
--- drivers/net/wireless/marvell/mwifiex/tdls.c | 75 ++++++++++------------------ 1 file changed, 28 insertions(+), 47 deletions(-)
drivers/net/wireless/marvell/mwifiex/tdls.c +28 -47
@@ -894,7 +894,7 @@ void mwifiex_process_tdls_action_frame(su8 *peer, *pos, *end;u8 i, action, basic;u16 cap =0 ; -int ie_len =0 ; +int ies_len =0 ;if (len < (sizeof (struct ethhdr ) +3 ))return ;
@@ -916,7 +916,7 @@ void mwifiex_process_tdls_action_frame(s pos = buf +sizeof (struct ethhdr ) +4 ;/* payload 1+ category 1 + action 1 + dialog 1 */ cap =get_unaligned_le16 (pos); - ie_len = len -sizeof (struct ethhdr ) -TDLS_REQ_FIX_LEN ; + ies_len = len -sizeof (struct ethhdr ) -TDLS_REQ_FIX_LEN ; pos +=2 ;break ;
@@ -926,7 +926,7 @@ void mwifiex_process_tdls_action_frame(s/* payload 1+ category 1 + action 1 + dialog 1 + status code 2*/ pos = buf +sizeof (struct ethhdr ) +6 ; cap =get_unaligned_le16 (pos); - ie_len = len -sizeof (struct ethhdr ) -TDLS_RESP_FIX_LEN ; + ies_len = len -sizeof (struct ethhdr ) -TDLS_RESP_FIX_LEN ; pos +=2 ;break ;
@@ -934,7 +934,7 @@ void mwifiex_process_tdls_action_frame(sif (len < (sizeof (struct ethhdr ) +TDLS_CONFIRM_FIX_LEN ))return ; pos = buf +sizeof (struct ethhdr ) +TDLS_CONFIRM_FIX_LEN ; - ie_len = len -sizeof (struct ethhdr ) -TDLS_CONFIRM_FIX_LEN ; + ies_len = len -sizeof (struct ethhdr ) -TDLS_CONFIRM_FIX_LEN ;break ;default :mwifiex_dbg (priv->adapter,ERROR ,"Unknown TDLS frame type.\n" );
@@ -947,33 +947,33 @@ void mwifiex_process_tdls_action_frame(s sta_ptr->tdls_cap.capab =cpu_to_le16 (cap); -for (end = pos + ie_len; pos +1 < end; pos +=2 + pos[1 ]) { -if (pos +2 + pos[1 ] > end) +for (end = pos + ies_len; pos +1 < end; pos +=2 + pos[1 ]) { +u8 ie_len = pos[1 ]; + +if (pos +2 + ie_len > end)break ;switch (*pos) {case WLAN_EID_SUPP_RATES : -if (pos[1 ] >32 ) +if (ie_len >sizeof (sta_ptr->tdls_cap.rates))return ; - sta_ptr->tdls_cap.rates_len = pos[1 ]; -for (i =0 ; i < pos[1 ]; i++) + sta_ptr->tdls_cap.rates_len = ie_len; +for (i =0 ; i < ie_len; i++) sta_ptr->tdls_cap.rates[i] = pos[i +2 ];break ;case WLAN_EID_EXT_SUPP_RATES : -if (pos[1 ] >32 ) +if (ie_len >sizeof (sta_ptr->tdls_cap.rates))return ; basic = sta_ptr->tdls_cap.rates_len; -if (pos[1 ] >32 - basic) +if (ie_len >sizeof (sta_ptr->tdls_cap.rates) - basic)return ; -for (i =0 ; i < pos[1 ]; i++) +for (i =0 ; i < ie_len; i++) sta_ptr->tdls_cap.rates[basic + i] = pos[i +2 ]; - sta_ptr->tdls_cap.rates_len += pos[1 ]; + sta_ptr->tdls_cap.rates_len += ie_len;break ;case WLAN_EID_HT_CAPABILITY : -if (pos > end -sizeof (struct ieee80211_ht_cap ) -2 ) -return ; -if (pos[1 ] !=sizeof (struct ieee80211_ht_cap )) +if (ie_len !=sizeof (struct ieee80211_ht_cap ))return ;/* copy the ie's value into ht_capb*/ memcpy((u8 *)&sta_ptr->tdls_cap.ht_capb, pos + 2,
@@ -981,59 +981,45 @@ void mwifiex_process_tdls_action_frame(s sta_ptr->is_11n_enabled =1 ;break ;case WLAN_EID_HT_OPERATION : -if (pos > end - -sizeof (struct ieee80211_ht_operation ) -2 ) -return ; -if (pos[1 ] !=sizeof (struct ieee80211_ht_operation )) +if (ie_len !=sizeof (struct ieee80211_ht_operation ))return ;/* copy the ie's value into ht_oper*/ memcpy (&sta_ptr->tdls_cap.ht_oper, pos +2 ,sizeof (struct ieee80211_ht_operation ));break ;case WLAN_EID_BSS_COEX_2040 : -if (pos > end -3 ) -return ; -if (pos[1 ] !=1 ) +if (ie_len !=sizeof (pos[2 ]))return ; sta_ptr->tdls_cap.coex_2040 = pos[2 ];break ;case WLAN_EID_EXT_CAPABILITY : -if (pos > end -sizeof (struct ieee_types_header )) -return ; -if (pos[1 ] <sizeof (struct ieee_types_header )) +if (ie_len <sizeof (struct ieee_types_header ))return ; -if (pos[1 ] >8 ) +if (ie_len >8 )return ;memcpy ((u8 *)&sta_ptr->tdls_cap.extcap, pos,sizeof (struct ieee_types_header ) + -min_t (u8, pos[1 ],8 )); +min_t (u8, ie_len,8 ));break ;case WLAN_EID_RSN : -if (pos > end -sizeof (struct ieee_types_header )) +if (ie_len <sizeof (struct ieee_types_header ))return ; -if (pos[1 ] <sizeof (struct ieee_types_header )) -return ; -if (pos[1 ] >IEEE_MAX_IE_SIZE - +if (ie_len >IEEE_MAX_IE_SIZE -sizeof (struct ieee_types_header ))return ;memcpy ((u8 *)&sta_ptr->tdls_cap.rsn_ie, pos,sizeof (struct ieee_types_header ) + -min_t (u8, pos[1 ],IEEE_MAX_IE_SIZE - +min_t (u8, ie_len,IEEE_MAX_IE_SIZE -sizeof (struct ieee_types_header )));break ;case WLAN_EID_QOS_CAPA : -if (pos > end -3 ) -return ; -if (pos[1 ] !=1 ) +if (ie_len !=sizeof (pos[2 ]))return ; sta_ptr->tdls_cap.qos_info = pos[2 ];break ;case WLAN_EID_VHT_OPERATION :if (priv->adapter->is_hw_11ac_capable) { -if (pos > end - -sizeof (struct ieee80211_vht_operation ) -2 ) -return ; -if (pos[1 ] != +if (ie_len !=sizeof (struct ieee80211_vht_operation ))return ;/* copy the ie's value into vhtoper*/
@@ -1043,10 +1029,7 @@ void mwifiex_process_tdls_action_frame(sbreak ;case WLAN_EID_VHT_CAPABILITY :if (priv->adapter->is_hw_11ac_capable) { -if (pos > end - -sizeof (struct ieee80211_vht_cap ) -2 ) -return ; -if (pos[1 ] !=sizeof (struct ieee80211_vht_cap )) +if (ie_len !=sizeof (struct ieee80211_vht_cap ))return ;/* copy the ie's value into vhtcap*/ memcpy((u8 *)&sta_ptr->tdls_cap.vhtcap, pos + 2,
@@ -1056,9 +1039,7 @@ void mwifiex_process_tdls_action_frame(sbreak ;case WLAN_EID_AID :if (priv->adapter->is_hw_11ac_capable) { -if (pos > end -4 ) -return ; -if (pos[1 ] !=2 ) +if (ie_len !=sizeof (u16))return ; sta_ptr->tdls_cap.aid =get_unaligned_le16 ((pos +2 ));
Greg Kroah-Hartman
maintainerPATCH 119/152 mwifiex: delete unused mwifiex_get_intf_num()
In reply to Greg Kroah-HartmanFrom: Brian Norris <[email protected]>
commit 1c9f329b084b7b8ea6d60d91a202e884cdcf6aae upstream.
Commit 7afb94da3cd8 ("mwifiex: update set_mac_address logic") fixed the only user of this function, partly because the author seems to have noticed that, as written, it's on the borderline between highly misleading and buggy.
Anyway, no sense in keeping dead code around: let's drop it.
--- drivers/net/wireless/marvell/mwifiex/main.h | 13 ------------- 1 file changed, 13 deletions(-)
drivers/net/wireless/marvell/mwifiex/main.h +0 -13
@@ -1295,19 +1295,6 @@ mwifiex_copy_rates(u8 *dest, u32 pos, u8return pos; } -/* This function return interface number with the same bss_type. -*/ -static inline u8 -mwifiex_get_intf_num (struct mwifiex_adapter *adapter,u8 bss_type) -{ -u8 i, num =0 ; - -for (i =0 ; i < adapter->priv_num; i++) -if (adapter->priv[i] && adapter->priv[i]->bss_type == bss_type) - num++; -return num; -} -/* * This function returns the correct private structure pointer based * upon the BSS type and BSS number.
Greg Kroah-Hartman
maintainerPATCH 120/152 KVM: SVM: Override default MMIO mask if memory encryption is enabled
In reply to Greg Kroah-HartmanFrom: Tom Lendacky <[email protected]>
commit 52918ed5fcf05d97d257f4131e19479da18f5d16 upstream.
The KVM MMIO support uses bit 51 as the reserved bit to cause nested page faults when a guest performs MMIO. The AMD memory encryption support uses a CPUID function to define the encryption bit position. Given this, it is possible that these bits can conflict.
Use svm_hardware_setup() to override the MMIO mask if memory encryption support is enabled. Various checks are performed to ensure that the mask is properly defined and rsvd_bits() is used to generate the new mask (as was done prior to the change that necessitated this patch).
--- arch/x86/kvm/svm.c | 43 +++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 43 insertions(+)
arch/x86/kvm/svm.c +43 -0
@@ -1298,6 +1298,47 @@ static void shrink_ple_window(struct kvm } } +/* +* The default MMIO mask is a single bit (excluding the present bit), +* which could conflict with the memory encryption bit. Check for +* memory encryption support and override the default MMIO mask if +* memory encryption is enabled. +*/ +static __init void svm_adjust_mmio_mask (void ) +{ +unsigned int enc_bit, mask_bit; +u64 msr, mask; + +/* If there is no memory encryption support, use existing mask */ +if (cpuid_eax (0x80000000 ) <0x8000001f ) +return ; + +/* If memory encryption is not enabled, use existing mask */ +rdmsrl (MSR_K8_SYSCFG , msr); +if (!(msr &MSR_K8_SYSCFG_MEM_ENCRYPT )) +return ; + + enc_bit =cpuid_ebx (0x8000001f ) &0x3f ; + mask_bit = boot_cpu_data.x86_phys_bits; + +/* Increment the mask bit if it is the same as the encryption bit */ +if (enc_bit == mask_bit) + mask_bit++; + +/* +* If the mask bit location is below 52, then some bits above the +* physical addressing limit will always be reserved, so use the +* rsvd_bits() function to generate the mask. This mask, along with +* the present bit, will be used to generate a page fault with +* PFER.RSV = 1. +* +* If the mask bit location is 52 (or above), then clear the mask. +*/ + mask = (mask_bit <52 ) ?rsvd_bits (mask_bit,51 ) |PT_PRESENT_MASK :0 ; + +kvm_mmu_set_mmio_spte_mask (mask, mask,PT_WRITABLE_MASK |PT_USER_MASK ); +} +static __init int svm_hardware_setup (void ) {int cpu;
@@ -1352,6 +1393,8 @@ static __init int svm_hardware_setup(voi } } +svm_adjust_mmio_mask (); +for_each_possible_cpu (cpu) { r =svm_cpu_init (cpu);if (r)
Greg Kroah-Hartman
maintainerPATCH 121/152 KVM: Check for a bad hva before dropping into the ghc slow path
In reply to Greg Kroah-HartmanFrom: Sean Christopherson <[email protected]>
commit fcfbc617547fc6d9552cb6c1c563b6a90ee98085 upstream.
When reading/writing using the guest/host cache, check for a bad hva before checking for a NULL memslot, which triggers the slow path for handing cross-page accesses. Because the memslot is nullified on error by __kvm_gfn_to_hva_cache_init(), if the bad hva is encountered after crossing into a new page, then the kvm_{read,write}_guest() slow path could potentially write/access the first chunk prior to detecting the bad hva.
Arguably, performing a partial access is semantically correct from an architectural perspective, but that behavior is certainly not intended. In the original implementation, memslot was not explicitly nullified and therefore the partial access behavior varied based on whether the memslot itself was null, or if the hva was simply bad. The current behavior was introduced as a seemingly unintentional side effect in commit f1b9dd5eb86c ("kvm: Disallow wraparound in kvm_gfn_to_hva_cache_init"), which justified the change with "since some callers don't check the return code from this function, it sit seems prudent to clear ghc->memslot in the event of an error".
Regardless of intent, the partial access is dependent on _not_ checking the result of the cache initialization, which is arguably a bug in its own right, at best simply weird.
--- virt/kvm/kvm_main.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-)
virt/kvm/kvm_main.c +6 -6
@@ -2275,12 +2275,12 @@ int kvm_write_guest_offset_cached(structif (slots->generation != ghc->generation)__kvm_gfn_to_hva_cache_init (slots, ghc, ghc->gpa, ghc->len); -if (unlikely (!ghc->memslot)) -return kvm_write_guest (kvm, gpa, data, len); -if (kvm_is_error_hva (ghc->hva))return -EFAULT ; +if (unlikely (!ghc->memslot)) +return kvm_write_guest (kvm, gpa, data, len); + r =__copy_to_user ((void __user *)ghc->hva + offset, data, len);if (r)return -EFAULT ;
@@ -2308,12 +2308,12 @@ int kvm_read_guest_cached(struct kvm *kvif (slots->generation != ghc->generation)__kvm_gfn_to_hva_cache_init (slots, ghc, ghc->gpa, ghc->len); -if (unlikely (!ghc->memslot)) -return kvm_read_guest (kvm, ghc->gpa, data, len); -if (kvm_is_error_hva (ghc->hva))return -EFAULT ; +if (unlikely (!ghc->memslot)) +return kvm_read_guest (kvm, ghc->gpa, data, len); + r =__copy_from_user (data, (void __user *)ghc->hva, len);if (r)return -EFAULT ;
Greg Kroah-Hartman
maintainerPATCH 122/152 sched/fair: Optimize select_idle_cpu
In reply to Greg Kroah-HartmanFrom: Cheng Jian <[email protected]>
commit 60588bfa223ff675b95f866249f90616613fbe31 upstream.
select_idle_cpu() will scan the LLC domain for idle CPUs, it's always expensive. so the next commit :
1ad3aaf3fcd2 ("sched/core: Implement new approach to scale select_idle_cpu()")
introduces a way to limit how many CPUs we scan.
But it consume some CPUs out of 'nr' that are not allowed for the task and thus waste our attempts. The function always return nr_cpumask_bits, and we can't find a CPU which our task is allowed to run.
Cpumask may be too big, similar to select_idle_core(), use per_cpu_ptr 'select_idle_mask' to prevent stack overflow.
--- kernel/sched/fair.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-)
kernel/sched/fair.c +4 -3
@@ -5933,6 +5933,7 @@ static inline int select_idle_smt(struct*/ static int select_idle_cpu (struct task_struct *p,struct sched_domain *sd,int target) { +struct cpumask *cpus =this_cpu_cpumask_var_ptr (select_idle_mask);struct sched_domain *this_sd;u64 avg_cost, avg_idle;u64 time, cost;
@@ -5964,11 +5965,11 @@ static int select_idle_cpu(struct task_s time =cpu_clock (this); -for_each_cpu_wrap (cpu,sched_domain_span (sd), target) { +cpumask_and (cpus,sched_domain_span (sd), p->cpus_ptr); + +for_each_cpu_wrap (cpu, cpus, target) {if (!--nr)return si_cpu; -if (!cpumask_test_cpu (cpu, p->cpus_ptr)) -continue ;if (available_idle_cpu (cpu))break ;if (si_cpu ==-1 &&sched_idle_cpu (cpu))
Greg Kroah-Hartman
maintainerPATCH 123/152 f2fs: fix to add swap extent correctly
In reply to Greg Kroah-HartmanFrom: Chao Yu <[email protected]>
commit 3e5e479a39ce9ed60cd63f7565cc1d9da77c2a4e upstream.
As Youling reported in mailing list:
https://www.linuxquestions.org/questions/linux-newbie-8/the-file-system-f2fs-is-broken-4175666043/
https://www.linux.org/threads/the-file-system-f2fs-is-broken.26490/
There is a test case can corrupt f2fs image:
- dd if=/dev/zero of=/swapfile bs=1M count=4096
- chmod 600 /swapfile
- mkswap /swapfile
- swapon --discard /swapfile
The root cause is f2fs_swap_activate() intends to return zero value to setup_swap_extents() to enable SWP_FS mode (swap file goes through fs), in this flow, setup_swap_extents() setups swap extent with wrong block address range, result in discard_swap() erasing incorrect address.
Because f2fs_swap_activate() has pinned swapfile, its data block address will not change, it's safe to let swap to handle IO through raw device, so we can get rid of SWAP_FS mode and initial swap extents inside f2fs_swap_activate(), by this way, later discard_swap() can trim in right address range.
--- fs/f2fs/data.c | 32 +++++++++++++++++++++++++------- 1 file changed, 25 insertions(+), 7 deletions(-)
fs/f2fs/data.c +25 -7
@@ -3030,7 +3030,8 @@ int f2fs_migrate_page(struct address_spa#ifdef CONFIG_SWAP /* Copied from generic_swapfile_activate() to check any holes */ -static int check_swap_activate (struct file *swap_file,unsigned int max) +static int check_swap_activate (struct swap_info_struct *sis, +struct file *swap_file,sector_t *span) {struct address_space *mapping = swap_file->f_mapping;struct inode *inode = mapping->host;
@@ -3041,6 +3042,8 @@ static int check_swap_activate(struct fisector_t last_block;sector_t lowest_block =-1 ;sector_t highest_block =0 ; +int nr_extents =0 ; +int ret; blkbits = inode->i_blkbits; blocks_per_page =PAGE_SIZE >> blkbits;
@@ -3052,7 +3055,8 @@ static int check_swap_activate(struct fi probe_block =0 ; page_no =0 ; last_block =i_size_read (inode) >> blkbits; -while ((probe_block + blocks_per_page) <= last_block && page_no < max) { +while ((probe_block + blocks_per_page) <= last_block && + page_no < sis->max) {unsigned block_in_page;sector_t first_block;
@@ -3092,13 +3096,27 @@ static int check_swap_activate(struct fi highest_block = first_block; } +/* +* We found a PAGE_SIZE-length, PAGE_SIZE-aligned run of blocks +*/ + ret =add_swap_extent (sis, page_no,1 , first_block); +if (ret <0 ) + goto out; + nr_extents += ret; page_no++; probe_block += blocks_per_page; reprobe:continue ; } -return 0 ; - + ret = nr_extents; + *span =1 + highest_block - lowest_block; +if (page_no ==0 ) + page_no =1 ;/* force Empty message */ + sis->max = page_no; + sis->pages = page_no -1 ; + sis->highest_bit = page_no -1 ; +out: +return ret; bad_bmap:pr_err ("swapon: swapfile has holes\n" );return -EINVAL ;
@@ -3120,14 +3138,14 @@ static int f2fs_swap_activate(struct swaif (ret)return ret; - ret =check_swap_activate (file, sis->max); -if (ret) + ret =check_swap_activate (sis, file, span); +if (ret <0 )return ret;set_inode_flag (inode,FI_PIN_FILE );f2fs_precache_extents (inode);f2fs_update_time (F2FS_I_SB (inode),REQ_TIME ); -return 0 ; +return ret; }static void f2fs_swap_deactivate (struct file *file)
Greg Kroah-Hartman
maintainerPATCH 124/152 RDMA/hns: Simplify the calculation and usage of wqe idx for post verbs
In reply to Greg Kroah-HartmanFrom: Yixian Liu <[email protected]>
commit 4768820243d71d49f1044b3f911ac3d52bdb79af upstream.
Currently, the wqe idx is calculated repeatly everywhere it is used. This patch defines wqe_idx and calculated it only once, then just use it as needed.
--- drivers/infiniband/hw/hns/hns_roce_device.h | 3 - drivers/infiniband/hw/hns/hns_roce_hw_v1.c | 37 ++++++++++-------------- drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 43 +++++++++++----------------- 3 files changed, 35 insertions(+), 48 deletions(-)
drivers/infiniband/hw/hns/hns_roce_device.h +1 -2
@@ -425,7 +425,7 @@ struct hns_roce_mr_table {struct hns_roce_wq {u64 *wrid;/* Work request ID */ spinlock_t lock; -int wqe_cnt;/* WQE num */ +u32 wqe_cnt;/* WQE num */ u32 max_post;int max_gs;int offset;
@@ -658,7 +658,6 @@ struct hns_roce_qp {u8 sdb_en;u32 doorbell_qpn;u32 sq_signal_bits; -u32 sq_next_wqe;struct hns_roce_wq sq;struct ib_umem *umem;
drivers/infiniband/hw/hns/hns_roce_hw_v1.c +16 -21
@@ -74,8 +74,8 @@ static int hns_roce_v1_post_send(structunsigned long flags =0 ;void *wqe =NULL ;__le32 doorbell[2 ]; +u32 wqe_idx =0 ;int nreq =0 ; -u32 ind =0 ;int ret =0 ;u8 *smac;int loopback;
@@ -88,7 +88,7 @@ static int hns_roce_v1_post_send(struct }spin_lock_irqsave (&qp->sq.lock, flags); - ind = qp->sq_next_wqe; +for (nreq =0 ; wr; ++nreq, wr = wr->next) {if (hns_roce_wq_overflow (&qp->sq, nreq, qp->ibqp.send_cq)) { ret = -ENOMEM ;
@@ -96,6 +96,8 @@ static int hns_roce_v1_post_send(struct goto out; } + wqe_idx = (qp->sq.head + nreq) & (qp->sq.wqe_cnt -1 ); +if (unlikely (wr->num_sge > qp->sq.max_gs)) {dev_err (dev,"num_sge=%d > qp->sq.max_gs=%d\n" , wr->num_sge, qp->sq.max_gs);
@@ -104,9 +106,8 @@ static int hns_roce_v1_post_send(struct goto out; } - wqe =get_send_wqe (qp, ind & (qp->sq.wqe_cnt -1 )); - qp->sq.wrid[(qp->sq.head + nreq) & (qp->sq.wqe_cnt -1 )] = - wr->wr_id; + wqe =get_send_wqe (qp, wqe_idx); + qp->sq.wrid[wqe_idx] = wr->wr_id;/* Corresponding to the RC and RD type wqe process separately */ if (ibqp->qp_type ==IB_QPT_GSI ) {
@@ -210,7 +211,6 @@ static int hns_roce_v1_post_send(structcpu_to_le32 ((wr->sg_list[1 ].addr) >>32 ); ud_sq_wqe->l_key1 =cpu_to_le32 (wr->sg_list[1 ].lkey); - ind++; }else if (ibqp->qp_type ==IB_QPT_RC ) {u32 tmp_len =0 ;
@@ -308,7 +308,6 @@ static int hns_roce_v1_post_send(struct ctrl->flag |=cpu_to_le32 (wr->num_sge <<HNS_ROCE_WQE_SGE_NUM_BIT ); } - ind++; } }
@@ -336,7 +335,6 @@ out: doorbell[1 ] = sq_db.u32_8;hns_roce_write64_k (doorbell, qp->sq.db_reg_l); - qp->sq_next_wqe = ind; }spin_unlock_irqrestore (&qp->sq.lock, flags);
@@ -348,12 +346,6 @@ static int hns_roce_v1_post_recv(structconst struct ib_recv_wr *wr, const struct ib_recv_wr **bad_wr) { - int ret = 0; -int nreq =0 ; -int ind =0 ; -int i =0 ; -u32 reg_val; -unsigned long flags =0 ; struct hns_roce_rq_wqe_ctrl *ctrl = NULL;struct hns_roce_wqe_data_seg *scat =NULL ;struct hns_roce_qp *hr_qp =to_hr_qp (ibqp);
@@ -361,9 +353,14 @@ static int hns_roce_v1_post_recv(structstruct device *dev = &hr_dev->pdev->dev;struct hns_roce_rq_db rq_db;__le32 doorbell[2 ] = {0 }; +unsigned long flags =0 ; +unsigned int wqe_idx; +int ret =0 ; +int nreq =0 ; +int i =0 ; +u32 reg_val;spin_lock_irqsave (&hr_qp->rq.lock, flags); - ind = hr_qp->rq.head & (hr_qp->rq.wqe_cnt -1 );for (nreq =0 ; wr; ++nreq, wr = wr->next) { if (hns_roce_wq_overflow(&hr_qp->rq, nreq,
@@ -373,6 +370,8 @@ static int hns_roce_v1_post_recv(struct goto out; } + wqe_idx = (hr_qp->rq.head + nreq) & (hr_qp->rq.wqe_cnt -1 ); +if (unlikely (wr->num_sge > hr_qp->rq.max_gs)) {dev_err (dev,"rq:num_sge=%d > qp->sq.max_gs=%d\n" , wr->num_sge, hr_qp->rq.max_gs);
@@ -381,7 +380,7 @@ static int hns_roce_v1_post_recv(struct goto out; } - ctrl =get_recv_wqe (hr_qp, ind); + ctrl =get_recv_wqe (hr_qp, wqe_idx); roce_set_field(ctrl->rwqe_byte_12, RQ_WQE_CTRL_RWQE_BYTE_12_RWQE_SGE_NUM_M,
@@ -393,9 +392,7 @@ static int hns_roce_v1_post_recv(structfor (i =0 ; i < wr->num_sge; i++)set_data_seg (scat + i, wr->sg_list + i); - hr_qp->rq.wrid[ind] = wr->wr_id; - - ind = (ind +1 ) & (hr_qp->rq.wqe_cnt -1 ); + hr_qp->rq.wrid[wqe_idx] = wr->wr_id; } out:
@@ -2702,7 +2699,6 @@ static int hns_roce_v1_m_sqp(struct ib_q hr_qp->rq.tail =0 ; hr_qp->sq.head =0 ; hr_qp->sq.tail =0 ; - hr_qp->sq_next_wqe =0 ; }kfree (context);
@@ -3316,7 +3312,6 @@ static int hns_roce_v1_m_qp(struct ib_qp hr_qp->rq.tail =0 ; hr_qp->sq.head =0 ; hr_qp->sq.tail =0 ; - hr_qp->sq_next_wqe =0 ; } out:kfree (context);
drivers/infiniband/hw/hns/hns_roce_hw_v2.c +18 -25
137 lines not on this page · raw message
Greg Kroah-Hartman
maintainerPATCH 125/152 RDMA/hns: Bugfix for posting a wqe with sge
In reply to Greg Kroah-HartmanFrom: Lijun Ou <[email protected]>
commit 468d020e2f02867b8ec561461a1689cd4365e493 upstream.
Driver should first check whether the sge is valid, then fill the valid sge and the caculated total into hardware, otherwise invalid sges will cause an error.
--- drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 41 +++++++++++++++++------------ 1 file changed, 25 insertions(+), 16 deletions(-)
drivers/infiniband/hw/hns/hns_roce_hw_v2.c +25 -16
@@ -110,7 +110,7 @@ static void set_atomic_seg(struct hns_ro }static void set_extend_sge (struct hns_roce_qp *qp,const struct ib_send_wr *wr, -unsigned int *sge_ind) +unsigned int *sge_ind,int valid_num_sge) {struct hns_roce_v2_wqe_data_seg *dseg;struct ib_sge *sg;
@@ -123,7 +123,7 @@ static void set_extend_sge(struct hns_roif (qp->ibqp.qp_type ==IB_QPT_RC || qp->ibqp.qp_type ==IB_QPT_UC ) num_in_wqe =HNS_ROCE_V2_UC_RC_SGE_NUM_IN_WQE ; - extend_sge_num = wr->num_sge - num_in_wqe; + extend_sge_num = valid_num_sge - num_in_wqe; sg = wr->sg_list + num_in_wqe; shift = qp->hr_buf.page_shift;
@@ -159,14 +159,16 @@ static void set_extend_sge(struct hns_rostatic int set_rwqe_data_seg (struct ib_qp *ibqp,const struct ib_send_wr *wr,struct hns_roce_v2_rc_send_wqe *rc_sq_wqe,void *wqe,unsigned int *sge_ind, +int valid_num_sge,const struct ib_send_wr **bad_wr) {struct hns_roce_dev *hr_dev =to_hr_dev (ibqp->device);struct hns_roce_v2_wqe_data_seg *dseg = wqe;struct hns_roce_qp *qp =to_hr_qp (ibqp); +int j =0 ;int i; -if (wr->send_flags &IB_SEND_INLINE && wr->num_sge) { +if (wr->send_flags &IB_SEND_INLINE && valid_num_sge) {if (le32_to_cpu (rc_sq_wqe->msg_len) > hr_dev->caps.max_sq_inline) { *bad_wr = wr;
@@ -190,7 +192,7 @@ static int set_rwqe_data_seg(struct ib_qroce_set_bit (rc_sq_wqe->byte_4,V2_RC_SEND_WQE_BYTE_4_INLINE_S ,1 ); }else { -if (wr->num_sge <=HNS_ROCE_V2_UC_RC_SGE_NUM_IN_WQE ) { +if (valid_num_sge <=HNS_ROCE_V2_UC_RC_SGE_NUM_IN_WQE ) {for (i =0 ; i < wr->num_sge; i++) {if (likely (wr->sg_list[i].length)) {set_data_seg_v2 (dseg, wr->sg_list + i);
@@ -203,19 +205,21 @@ static int set_rwqe_data_seg(struct ib_qV2_RC_SEND_WQE_BYTE_20_MSG_START_SGE_IDX_S , (*sge_ind) & (qp->sge.sge_cnt - 1)); -for (i =0 ; i <HNS_ROCE_V2_UC_RC_SGE_NUM_IN_WQE ; i++) { +for (i =0 ; i < wr->num_sge && + j <HNS_ROCE_V2_UC_RC_SGE_NUM_IN_WQE ; i++) {if (likely (wr->sg_list[i].length)) {set_data_seg_v2 (dseg, wr->sg_list + i); dseg++; + j++; } } -set_extend_sge (qp, wr, sge_ind); +set_extend_sge (qp, wr, sge_ind, valid_num_sge); }roce_set_field (rc_sq_wqe->byte_16,V2_RC_SEND_WQE_BYTE_16_SGE_NUM_M , -V2_RC_SEND_WQE_BYTE_16_SGE_NUM_S , wr->num_sge); +V2_RC_SEND_WQE_BYTE_16_SGE_NUM_S , valid_num_sge); }return 0 ;
@@ -243,6 +247,7 @@ static int hns_roce_v2_post_send(structunsigned int sge_idx;unsigned int wqe_idx;unsigned long flags; +int valid_num_sge;void *wqe =NULL ;bool loopback;int attr_mask;
@@ -292,8 +297,16 @@ static int hns_roce_v2_post_send(struct qp->sq.wrid[wqe_idx] = wr->wr_id; owner_bit = ~(((qp->sq.head + nreq) >>ilog2 (qp->sq.wqe_cnt)) &0x1 ); + valid_num_sge =0 ; tmp_len =0 ; +for (i =0 ; i < wr->num_sge; i++) { +if (likely (wr->sg_list[i].length)) { + tmp_len += wr->sg_list[i].length; + valid_num_sge++; + } + } +/* Corresponding to the QP type, wqe process separately */ if (ibqp->qp_type ==IB_QPT_GSI ) { ud_sq_wqe = wqe;
@@ -329,9 +342,6 @@ static int hns_roce_v2_post_send(structV2_UD_SEND_WQE_BYTE_4_OPCODE_S , HNS_ROCE_V2_WQE_OP_SEND); -for (i =0 ; i < wr->num_sge; i++) - tmp_len += wr->sg_list[i].length; - ud_sq_wqe->msg_len =cpu_to_le32 (le32_to_cpu (ud_sq_wqe->msg_len) + tmp_len);
@@ -367,7 +377,7 @@ static int hns_roce_v2_post_send(structroce_set_field (ud_sq_wqe->byte_16,V2_UD_SEND_WQE_BYTE_16_SGE_NUM_M ,V2_UD_SEND_WQE_BYTE_16_SGE_NUM_S , - wr->num_sge); + valid_num_sge); roce_set_field(ud_sq_wqe->byte_20, V2_UD_SEND_WQE_BYTE_20_MSG_START_SGE_IDX_M,
@@ -422,12 +432,10 @@ static int hns_roce_v2_post_send(structmemcpy (&ud_sq_wqe->dgid[0 ], &ah->av.dgid[0 ],GID_LEN_V2 ); -set_extend_sge (qp, wr, &sge_idx); +set_extend_sge (qp, wr, &sge_idx, valid_num_sge); }else if (ibqp->qp_type ==IB_QPT_RC ) { rc_sq_wqe = wqe;memset (rc_sq_wqe,0 ,sizeof (*rc_sq_wqe)); -for (i =0 ; i < wr->num_sge; i++) - tmp_len += wr->sg_list[i].length; rc_sq_wqe->msg_len =cpu_to_le32 (le32_to_cpu (rc_sq_wqe->msg_len) + tmp_len);
@@ -548,10 +556,11 @@ static int hns_roce_v2_post_send(structroce_set_field (rc_sq_wqe->byte_16,V2_RC_SEND_WQE_BYTE_16_SGE_NUM_M ,V2_RC_SEND_WQE_BYTE_16_SGE_NUM_S , - wr->num_sge); + valid_num_sge); }else if (wr->opcode !=IB_WR_REG_MR ) { ret =set_rwqe_data_seg (ibqp, wr, rc_sq_wqe, - wqe, &sge_idx, bad_wr); + wqe, &sge_idx, + valid_num_sge, bad_wr);if (ret) goto out; }
Greg Kroah-Hartman
maintainerPATCH 126/152 drivers: net: xgene: Fix the order of the arguments of alloc_etherdev_mqs()
In reply to Greg Kroah-HartmanFrom: Christophe JAILLET <[email protected]>
commit 5a44c71ccda60a50073c5d7fe3f694cdfa3ab0c2 upstream.
'alloc_etherdev_mqs()' expects first 'tx', then 'rx'. The semantic here looks reversed.
Reorder the arguments passed to 'alloc_etherdev_mqs()' in order to keep the correct semantic.
In fact, this is a no-op because both XGENE_NUM_[RT]X_RING are 8.
--- drivers/net/ethernet/apm/xgene/xgene_enet_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/apm/xgene/xgene_enet_main.c +1 -1
@@ -2020,7 +2020,7 @@ static int xgene_enet_probe(struct platfint ret; ndev =alloc_etherdev_mqs (sizeof (struct xgene_enet_pdata ), -XGENE_NUM_RX_RING ,XGENE_NUM_TX_RING ); +XGENE_NUM_TX_RING ,XGENE_NUM_RX_RING );if (!ndev)return -ENOMEM ;
Greg Kroah-Hartman
maintainerPATCH 127/152 ima: ima/lsm policy rule loading logic bug fixes
In reply to Greg Kroah-HartmanFrom: Janne Karhunen <[email protected]>
commit 483ec26eed42bf050931d9a5c5f9f0b5f2ad5f3b upstream.
Keep the ima policy rules around from the beginning even if they appear invalid at the time of loading, as they may become active after an lsm policy load. However, loading a custom IMA policy with unknown LSM labels is only safe after we have transitioned from the "built-in" policy rules to a custom IMA policy.
Patch also fixes the rule re-use during the lsm policy reload and makes some prints a bit more human readable.
Changelog:
v4:
- Do not allow the initial policy load refer to non-existing lsm rules. v3:
- Fix too wide policy rule matching for non-initialized LSMs v2:
- Fix log prints
--- security/integrity/ima/ima_policy.c | 44 +++++++++++++++++++++--------------- 1 file changed, 26 insertions(+), 18 deletions(-)
security/integrity/ima/ima_policy.c +26 -18
@@ -263,7 +263,7 @@ static void ima_lsm_free_rule(struct imastatic struct ima_rule_entry *ima_lsm_copy_rule (struct ima_rule_entry *entry) {struct ima_rule_entry *nentry; -int i, result; +int i; nentry =kmalloc (sizeof (*nentry),GFP_KERNEL );if (!nentry)
@@ -277,7 +277,7 @@ static struct ima_rule_entry *ima_lsm_co memset(nentry->lsm, 0, FIELD_SIZEOF(struct ima_rule_entry, lsm));for (i =0 ; i <MAX_LSM_RULES ; i++) { -if (!entry->lsm[i].rule) +if (!entry->lsm[i].args_p)continue ; nentry->lsm[i].type = entry->lsm[i].type;
@@ -286,13 +286,13 @@ static struct ima_rule_entry *ima_lsm_coif (!nentry->lsm[i].args_p) goto out_err; - result =security_filter_rule_init (nentry->lsm[i].type, - Audit_equal, - nentry->lsm[i].args_p, - &nentry->lsm[i].rule); -if (result == -EINVAL ) -pr_warn ("ima: rule for LSM \'%d\' is undefined\n" , - entry->lsm[i].type); +security_filter_rule_init (nentry->lsm[i].type, + Audit_equal, + nentry->lsm[i].args_p, + &nentry->lsm[i].rule); +if (!nentry->lsm[i].rule) +pr_warn ("rule for LSM \'%s\' is undefined\n" , + (char *)entry->lsm[i].args_p); }return nentry;
@@ -329,7 +329,7 @@ static void ima_lsm_update_rules(void)list_for_each_entry_safe (entry, e, &ima_policy_rules, list) { needs_update =0 ;for (i =0 ; i <MAX_LSM_RULES ; i++) { -if (entry->lsm[i].rule) { +if (entry->lsm[i].args_p) { needs_update =1 ;break ; }
@@ -339,8 +339,7 @@ static void ima_lsm_update_rules(void) result =ima_lsm_update_rule (entry);if (result) { -pr_err ("ima: lsm rule update error %d\n" , - result); +pr_err ("lsm rule update error %d\n" , result);return ; } }
@@ -357,7 +356,7 @@ int ima_lsm_policy_change(struct notifie }/** -* ima_match_rules - determine whether an inode matches the measure rule. +* ima_match_rules - determine whether an inode matches the policy rule. * @rule: a pointer to a rule * @inode: a pointer to an inode * @cred: a pointer to a credentials structure for user validation
@@ -415,9 +414,12 @@ static bool ima_match_rules(struct ima_rint rc =0 ;u32 osid; -if (!rule->lsm[i].rule) -continue ; - +if (!rule->lsm[i].rule) { +if (!rule->lsm[i].args_p) +continue ; +else +return false; + }switch (i) {case LSM_OBJ_USER :case LSM_OBJ_ROLE :
@@ -822,8 +824,14 @@ static int ima_lsm_rule_init(struct ima_ entry->lsm[lsm_rule].args_p, &entry->lsm[lsm_rule].rule);if (!entry->lsm[lsm_rule].rule) { -kfree (entry->lsm[lsm_rule].args_p); -return -EINVAL ; +pr_warn ("rule for LSM \'%s\' is undefined\n" , + (char *)entry->lsm[lsm_rule].args_p); + +if (ima_rules == &ima_default_rules) { +kfree (entry->lsm[lsm_rule].args_p); + result = -EINVAL ; + }else + result =0 ; }return result;
Greg Kroah-Hartman
maintainerPATCH 128/152 kprobes: Set unoptimized flag after unoptimizing code
In reply to Greg Kroah-HartmanFrom: Masami Hiramatsu <[email protected]>
commit f66c0447cca1281116224d474cdb37d6a18e4b5b upstream.
Set the unoptimized flag after confirming the code is completely unoptimized. Without this fix, when a kprobe hits the intermediate modified instruction (the first byte is replaced by an INT3, but later bytes can still be a jump address operand) while unoptimizing, it can return to the middle byte of the modified code, which causes an invalid instruction exception in the kernel.
Usually, this is a rare case, but if we put a probe on the function call while text patching, it always causes a kernel panic as below:
# echo p text_poke+5 > kprobe_events # echo 1 > events/kprobes/enable # echo 0 > events/kprobes/enable
invalid opcode: 0000 [#1] PREEMPT SMP PTI
RIP: 0010:text_poke+0x9/0x50 Call Trace: arch_unoptimize_kprobe+0x22/0x28 arch_unoptimize_kprobes+0x39/0x87 kprobe_optimizer+0x6e/0x290 process_one_work+0x2a0/0x610 worker_thread+0x28/0x3d0 ? process_one_work+0x610/0x610 kthread+0x10d/0x130 ? kthread_park+0x80/0x80 ret_from_fork+0x3a/0x50
text_poke() is used for patching the code in optprobes.
This can happen even if we blacklist text_poke() and other functions, because there is a small time window during which we show the intermediate code to other CPUs.
[ mingo: Edited the changelog. ]
--- kernel/kprobes.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
kernel/kprobes.c +3 -1
@@ -510,6 +510,8 @@ static void do_unoptimize_kprobes(void)arch_unoptimize_kprobes (&unoptimizing_list, &freeing_list);/* Loop free_list for disarming */ list_for_each_entry_safe (op, tmp, &freeing_list, list) { +/* Switching from detour code to origin */ + op->kp.flags &= ~KPROBE_FLAG_OPTIMIZED ;/* Disarm probes if marked disabled */ if (kprobe_disabled (&op->kp))arch_disarm_kprobe (&op->kp);
@@ -649,6 +651,7 @@ static void force_unoptimize_kprobe(stru {lockdep_assert_cpus_held ();arch_unoptimize_kprobe (op); + op->kp.flags &= ~KPROBE_FLAG_OPTIMIZED ;if (kprobe_disabled (&op->kp))arch_disarm_kprobe (&op->kp); }
@@ -676,7 +679,6 @@ static void unoptimize_kprobe(struct kprreturn ; } - op->kp.flags &= ~KPROBE_FLAG_OPTIMIZED ;if (!list_empty (&op->list)) {/* Dequeue from the optimization queue */ list_del_init (&op->list);
Greg Kroah-Hartman
maintainerPATCH 129/152 lib/vdso: Make __arch_update_vdso_data() logic understandable
In reply to Greg Kroah-HartmanFrom: Thomas Gleixner <[email protected]>
commit 9a6b55ac4a44060bcb782baf002859b2a2c63267 upstream.
The function name suggests that this is a boolean checking whether the architecture asks for an update of the VDSO data, but it works the other way round. To spare further confusion invert the logic.
--- include/asm-generic/vdso/vsyscall.h | 4 ++-- kernel/time/vsyscall.c | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-)
include/asm-generic/vdso/vsyscall.h +2 -2
@@ -12,9 +12,9 @@ static __always_inline struct vdso_data#endif /* __arch_get_k_vdso_data */ #ifndef __arch_update_vdso_data -static __always_inline int __arch_update_vdso_data(void) +static __always_inline bool __arch_update_vdso_data(void) { -return 0 ; +return true; }#endif /* __arch_update_vdso_data */
kernel/time/vsyscall.c +1 -1
@@ -84,7 +84,7 @@ void update_vsyscall(struct timekeeper *struct vdso_timestamp *vdso_ts;u64 nsec; -if (__arch_update_vdso_data ()) { +if (!__arch_update_vdso_data ()) {/* * Some architectures might want to skip the update of the * data page.
Greg Kroah-Hartman
maintainerPATCH 130/152 lib/vdso: Update coarse timekeeper unconditionally
In reply to Greg Kroah-HartmanFrom: Thomas Gleixner <[email protected]>
commit 9f24c540f7f8eb3a981528da9a9a636a5bdf5987 upstream.
The low resolution parts of the VDSO, i.e.:
clock_gettime(CLOCK_*_COARSE), clock_getres(), time()
can be used even if there is no VDSO capable clocksource.
But if an architecture opts out of the VDSO data update then this information becomes stale. This affects ARM when there is no architected timer available. The lack of update causes userspace to use stale data forever.
Make the update of the low resolution parts unconditional and only skip the update of the high resolution parts if the architecture requests it.
--- kernel/time/vsyscall.c | 37 +++++++++++++++++-------------------- 1 file changed, 17 insertions(+), 20 deletions(-)
kernel/time/vsyscall.c +17 -20
@@ -28,11 +28,6 @@ static inline void update_vdso_data(stru vdata[CS_RAW ].mult = tk->tkr_raw.mult; vdata[CS_RAW ].shift = tk->tkr_raw.shift; -/* CLOCK_REALTIME */ - vdso_ts = &vdata[CS_HRES_COARSE ].basetime[CLOCK_REALTIME ]; - vdso_ts->sec = tk->xtime_sec; - vdso_ts->nsec = tk->tkr_mono.xtime_nsec; -/* CLOCK_MONOTONIC */ vdso_ts = &vdata[CS_HRES_COARSE ].basetime[CLOCK_MONOTONIC ]; vdso_ts->sec = tk->xtime_sec + tk->wall_to_monotonic.tv_sec;
@@ -70,12 +65,6 @@ static inline void update_vdso_data(stru vdso_ts = &vdata[CS_HRES_COARSE ].basetime[CLOCK_TAI ]; vdso_ts->sec = tk->xtime_sec + (s64 )tk->tai_offset; vdso_ts->nsec = tk->tkr_mono.xtime_nsec; - -/* -* Read without the seqlock held by clock_getres(). -* Note: No need to have a second copy. -*/ -WRITE_ONCE (vdata[CS_HRES_COARSE ].hrtimer_res, hrtimer_resolution); }void update_vsyscall (struct timekeeper *tk)
@@ -84,20 +73,17 @@ void update_vsyscall(struct timekeeper *struct vdso_timestamp *vdso_ts;u64 nsec; -if (!__arch_update_vdso_data ()) { -/* -* Some architectures might want to skip the update of the -* data page. -*/ -return ; - } -/* copy vsyscall data */ vdso_write_begin (vdata); vdata[CS_HRES_COARSE ].clock_mode =__arch_get_clock_mode (tk); vdata[CS_RAW ].clock_mode =__arch_get_clock_mode (tk); +/* CLOCK_REALTIME also required for time() */ + vdso_ts = &vdata[CS_HRES_COARSE ].basetime[CLOCK_REALTIME ]; + vdso_ts->sec = tk->xtime_sec; + vdso_ts->nsec = tk->tkr_mono.xtime_nsec; +/* CLOCK_REALTIME_COARSE */ vdso_ts = &vdata[CS_HRES_COARSE ].basetime[CLOCK_REALTIME_COARSE ]; vdso_ts->sec = tk->xtime_sec;
@@ -110,7 +96,18 @@ void update_vsyscall(struct timekeeper * nsec = nsec + tk->wall_to_monotonic.tv_nsec; vdso_ts->sec +=__iter_div_u64_rem (nsec,NSEC_PER_SEC , &vdso_ts->nsec); -update_vdso_data (vdata, tk); +/* +* Read without the seqlock held by clock_getres(). +* Note: No need to have a second copy. +*/ +WRITE_ONCE (vdata[CS_HRES_COARSE ].hrtimer_res, hrtimer_resolution); + +/* +* Architectures can opt out of updating the high resolution part +* of the VDSO. +*/ +if (__arch_update_vdso_data ()) +update_vdso_data (vdata, tk);__arch_update_vsyscall (vdata, tk);
Greg Kroah-Hartman
maintainerPATCH 131/152 pwm: omap-dmtimer: put_device() after of_find_device_by_node()
In reply to Greg Kroah-HartmanFrom: Uwe Kleine-König <[email protected]>
commit c7cb3a1dd53f63c64fb2b567d0be130b92a44d91 upstream.
This was found by coccicheck:
drivers/pwm/pwm-omap-dmtimer.c:304:2-8: ERROR: missing put_device; call of_find_device_by_node on line 255, but without a corresponding object release within this function.
--- drivers/pwm/pwm-omap-dmtimer.c | 21 +++++++++++++++------ 1 file changed, 15 insertions(+), 6 deletions(-)
drivers/pwm/pwm-omap-dmtimer.c +15 -6
@@ -256,7 +256,7 @@ static int pwm_omap_dmtimer_probe(structif (!timer_pdev) {dev_err (&pdev->dev,"Unable to find Timer pdev\n" ); ret = -ENODEV ; - goto put; + goto err_find_timer_pdev; } timer_pdata =dev_get_platdata (&timer_pdev->dev);
@@ -264,7 +264,7 @@ static int pwm_omap_dmtimer_probe(structdev_dbg (&pdev->dev,"dmtimer pdata structure NULL, deferring probe\n" ); ret = -EPROBE_DEFER ; - goto put; + goto err_platdata; } pdata = timer_pdata->timer_ops;
@@ -283,19 +283,19 @@ static int pwm_omap_dmtimer_probe(struct !pdata->write_counter) {dev_err (&pdev->dev,"Incomplete dmtimer pdata structure\n" ); ret = -EINVAL ; - goto put; + goto err_platdata; }if (!of_get_property (timer,"ti,timer-pwm" ,NULL )) {dev_err (&pdev->dev,"Missing ti,timer-pwm capability\n" ); ret = -ENODEV ; - goto put; + goto err_timer_property; } dm_timer = pdata->request_by_node (timer);if (!dm_timer) { ret = -EPROBE_DEFER ; - goto put; + goto err_request_timer; } omap =devm_kzalloc (&pdev->dev,sizeof (*omap),GFP_KERNEL );
@@ -352,7 +352,14 @@ err_pwmchip_add: err_alloc_omap: pdata->free (dm_timer); -put: +err_request_timer: + +err_timer_property: +err_platdata: + +put_device (&timer_pdev->dev); +err_find_timer_pdev: +of_node_put (timer);return ret;
@@ -372,6 +379,8 @@ static int pwm_omap_dmtimer_remove(struc omap->pdata->free (omap->dm_timer); +put_device (&omap->dm_timer_pdev->dev); +mutex_destroy (&omap->mutex);return 0 ;
Greg Kroah-Hartman
maintainerPATCH 132/152 perf hists browser: Restore ESC as "Zoom out" of DSO/thread/etc
In reply to Greg Kroah-HartmanFrom: Arnaldo Carvalho de Melo <[email protected]>
commit 3f7774033e6820d25beee5cf7aefa11d4968b951 upstream.
We need to set actions->ms.map since 599a2f38a989 ("perf hists browser:
Check sort keys before hot key actions"), as in that patch we bail out if map is NULL.
--- tools/perf/ui/browsers/hists.c | 1 + 1 file changed, 1 insertion(+)
tools/perf/ui/browsers/hists.c +1 -0
@@ -3054,6 +3054,7 @@ static int perf_evsel__hists_browse(strucontinue ; } + actions->ms.map = map; top =pstack__peek (browser->pstack);if (top == &browser->hists->dso_filter) {/*
Greg Kroah-Hartman
maintainerPATCH 133/152 perf ui gtk: Add missing zalloc object
In reply to Greg Kroah-HartmanFrom: Jiri Olsa <[email protected]>
commit 604e2139a1026793b8c2172bd92c7e9d039a5cf0 upstream.
When we moved zalloc.o to the library we missed gtk library which needs it compiled in, otherwise the missing __zfree symbol will cause the library to fail to load.
Adding the zalloc object to the gtk library build.
--- tools/perf/ui/gtk/Build | 5 +++++ 1 file changed, 5 insertions(+)
tools/perf/ui/gtk/Build +5 -0
@@ -7,3 +7,8 @@ gtk-y += util.o gtk-y += helpline.o gtk-y += progress.o gtk-y += annotate.o +gtk-y += zalloc.o + +$(OUTPUT)ui/gtk/zalloc.o: ../lib/zalloc.c FORCE + $(call rule_mkdir) + $(call if_changed_dep,cc_o_c)
Greg Kroah-Hartman
maintainerPATCH 134/152 x86/resctrl: Check monitoring static key in the MBM overflow handler
In reply to Greg Kroah-HartmanFrom: Xiaochen Shen <[email protected]>
commit 536a0d8e79fb928f2735db37dda95682b6754f9a upstream.
Currently, there are three static keys in the resctrl file system:rdt_mon_enable_key and rdt_alloc_enable_key indicate if the monitoring feature and the allocation feature are enabled, respectively. The rdt_enable_key is enabled when either the monitoring feature or the allocation feature is enabled.
If no monitoring feature is present (either hardware doesn't support a monitoring feature or the feature is disabled by the kernel command line option "rdt="), rdt_enable_key is still enabled but rdt_mon_enable_key is disabled.
MBM is a monitoring feature. The MBM overflow handler intends to check if the monitoring feature is not enabled for fast return.
So check the rdt_mon_enable_key in it instead of the rdt_enable_key as former is the more accurate check.
[ bp: Massage commit message. ]
--- arch/x86/kernel/cpu/resctrl/internal.h | 1 + arch/x86/kernel/cpu/resctrl/monitor.c | 4 ++-- 2 files changed, 3 insertions(+), 2 deletions(-)
arch/x86/kernel/cpu/resctrl/internal.h +1 -0
@@ -57,6 +57,7 @@ static inline struct rdt_fs_context *rdt }DECLARE_STATIC_KEY_FALSE (rdt_enable_key); +DECLARE_STATIC_KEY_FALSE (rdt_mon_enable_key);/** * struct mon_evt - Entry in the event list of a resource
arch/x86/kernel/cpu/resctrl/monitor.c +2 -2
@@ -514,7 +514,7 @@ void mbm_handle_overflow(struct work_strmutex_lock (&rdtgroup_mutex); -if (!static_branch_likely (&rdt_enable_key)) +if (!static_branch_likely (&rdt_mon_enable_key)) goto out_unlock; d =get_domain_from_cpu (cpu, &rdt_resources_all[RDT_RESOURCE_L3 ]);
@@ -543,7 +543,7 @@ void mbm_setup_overflow_handler(struct runsigned long delay =msecs_to_jiffies (delay_ms);int cpu; -if (!static_branch_likely (&rdt_enable_key)) +if (!static_branch_likely (&rdt_mon_enable_key))return ; cpu =cpumask_any (&dom->cpu_mask); dom->mbm_work_cpu = cpu;
Greg Kroah-Hartman
maintainerPATCH 135/152 KVM: x86: Remove spurious kvm_mmu_unload() from vcpu destruction path
In reply to Greg Kroah-HartmanFrom: Sean Christopherson <[email protected]>
commit 9d979c7e6ff43ca3200ffcb74f57415fd633a2da upstream.
x86 does not load its MMU until KVM_RUN, which cannot be invoked until after vCPU creation succeeds. Given that kvm_arch_vcpu_destroy() is called if and only if vCPU creation fails, it is impossible for the MMU to be loaded.
Note, the bogus kvm_mmu_unload() call was added during an unrelated refactoring of vCPU allocation, i.e. was presumably added as an opportunstic "fix" for a perceived leak.
--- arch/x86/kvm/x86.c | 4 ---- 1 file changed, 4 deletions(-)
arch/x86/kvm/x86.c +0 -4
@@ -9194,10 +9194,6 @@ void kvm_arch_vcpu_destroy(struct kvm_vc { vcpu->arch.apf.msr_val =0 ; -vcpu_load (vcpu); -kvm_mmu_unload (vcpu); -vcpu_put (vcpu); -kvm_arch_vcpu_free (vcpu); }
Greg Kroah-Hartman
maintainerPATCH 136/152 KVM: x86: Remove spurious clearing of async #PF MSR
In reply to Greg Kroah-HartmanFrom: Sean Christopherson <[email protected]>
commit 208050dac5ef4de5cb83ffcafa78499c94d0b5ad upstream.
Remove a bogus clearing of apf.msr_val from kvm_arch_vcpu_destroy().
apf.msr_val is only set to a non-zero value by kvm_pv_enable_async_pf(), which is only reachable by kvm_set_msr_common(), i.e. by writing MSR_KVM_ASYNC_PF_EN. KVM does not autonomously write said MSR, i.e. can only be written via KVM_SET_MSRS or KVM_RUN. Since KVM_SET_MSRS and KVM_RUN are vcpu ioctls, they require a valid vcpu file descriptor. kvm_arch_vcpu_destroy() is only called if KVM_CREATE_VCPU fails, and KVM declares KVM_CREATE_VCPU successful once the vcpu fd is installed and thus visible to userspace. Ergo, apf.msr_val cannot be non-zero when kvm_arch_vcpu_destroy() is called.
--- arch/x86/kvm/x86.c | 2 -- 1 file changed, 2 deletions(-)
arch/x86/kvm/x86.c +0 -2
@@ -9192,8 +9192,6 @@ void kvm_arch_vcpu_postcreate(struct kvmvoid kvm_arch_vcpu_destroy (struct kvm_vcpu *vcpu) { - vcpu->arch.apf.msr_val =0 ; -kvm_arch_vcpu_free (vcpu); }
Greg Kroah-Hartman
maintainerPATCH 137/152 rcu: Allow only one expedited GP to run concurrently with wakeups
In reply to Greg Kroah-HartmanFrom: Neeraj Upadhyay <[email protected]>
commit 4bc6b745e5cbefed92c48071e28a5f41246d0470 upstream.
The current expedited RCU grace-period code expects that a task requesting an expedited grace period cannot awaken until that grace period has reached the wakeup phase. However, it is possible for a long preemption to result in the waiting task never sleeping. For example, consider the following sequence of events:
1. Task A starts an expedited grace period by invoking
synchronize_rcu_expedited(). It proceeds normally up to the wait_event() near the end of that function, and is then preempted (or interrupted or whatever).
2. The expedited grace period completes, and a kworker task starts
the awaken phase, having incremented the counter and acquired the rcu_state structure's .exp_wake_mutex. This kworker task is then preempted or interrupted or whatever.
3. Task A resumes and enters wait_event(), which notes that the
expedited grace period has completed, and thus doesn't sleep.
4. Task B starts an expedited grace period exactly as did Task A,
complete with the preemption (or whatever delay) just before the call to wait_event().
5. The expedited grace period completes, and another kworker
task starts the awaken phase, having incremented the counter. However, it blocks when attempting to acquire the rcu_state structure's .exp_wake_mutex because step 2's kworker task has not yet released it.
6. Steps 4 and 5 repeat, resulting in overflow of the rcu_node
structure's ->exp_wq[] array.
In theory, this is harmless. Tasks waiting on the various ->exp_wq[] array will just be spuriously awakened, but they will just sleep again on noting that the rcu_state structure's ->expedited_sequence value has not advanced far enough.
In practice, this wastes CPU time and is an accident waiting to happen. This commit therefore moves the rcu_exp_gp_seq_end() call that officially ends the expedited grace period (along with associate tracing) until after the ->exp_wake_mutex has been acquired. This prevents Task A from awakening prematurely, thus preventing more than one expedited grace period from being in flight during a previous expedited grace period's wakeup phase.
--- kernel/rcu/tree_exp.h | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-)
kernel/rcu/tree_exp.h +5 -6
@@ -540,14 +540,13 @@ static void rcu_exp_wait_wake(unsigned lstruct rcu_node *rnp;synchronize_sched_expedited_wait (); -rcu_exp_gp_seq_end (); -trace_rcu_exp_grace_period (rcu_state.name, s,TPS ("end" )); -/* -* Switch over to wakeup mode, allowing the next GP, but -only- the -* next GP, to proceed. -*/ +// Switch over to wakeup mode, allowing the next GP to proceed. +// End the previous grace period only after acquiring the mutex +// to ensure that only one GP runs concurrently with wakeups. mutex_lock (&rcu_state.exp_wake_mutex); +rcu_exp_gp_seq_end (); +trace_rcu_exp_grace_period (rcu_state.name, s,TPS ("end" ));rcu_for_each_node_breadth_first (rnp) {if (ULONG_CMP_LT (READ_ONCE (rnp->exp_seq_rq), s)) {
Greg Kroah-Hartman
maintainerPATCH 138/152 ubifs: Fix ino_t format warnings in orphan_delete()
In reply to Greg Kroah-HartmanFrom: Geert Uytterhoeven <[email protected]>
commit 155fc6ba488a8bdfd1d3be3d7ba98c9cec2b2429 upstream.
On alpha and s390x:
fs/ubifs/debug.h:158:11: warning: format ‘%lu’ expects argument of type ‘long unsigned int’, but argument 4 has type ‘ino_t {aka unsigned int}’ [-Wformat=]
...
fs/ubifs/orphan.c:132:3: note: in expansion of macro ‘dbg_gen’
dbg_gen("deleted twice ino %lu", orph->inum);
...
fs/ubifs/orphan.c:140:3: note: in expansion of macro ‘dbg_gen’
dbg_gen("delete later ino %lu", orph->inum);__kernel_ino_t is "unsigned long" on most architectures, but not on alpha and s390x, where it is "unsigned int". Hence when printing an ino_t, it should always be cast to "unsigned long" first.
Fix this by re-adding the recently removed casts.
--- fs/ubifs/orphan.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
fs/ubifs/orphan.c +2 -2
@@ -129,7 +129,7 @@ static void __orphan_drop(struct ubifs_istatic void orphan_delete (struct ubifs_info *c,struct ubifs_orphan *orph) {if (orph->del) { -dbg_gen ("deleted twice ino %lu" , orph->inum); +dbg_gen ("deleted twice ino %lu" , (unsigned long )orph->inum);return ; }
@@ -137,7 +137,7 @@ static void orphan_delete(struct ubifs_i orph->del =1 ; orph->dnext = c->orph_dnext; c->orph_dnext = orph; -dbg_gen ("delete later ino %lu" , orph->inum); +dbg_gen ("delete later ino %lu" , (unsigned long )orph->inum);return ; }
Greg Kroah-Hartman
maintainerPATCH 139/152 thermal: db8500: Depromote debug print
In reply to Greg Kroah-HartmanFrom: Linus Walleij <[email protected]>
commit c56dcfa3d4d0f49f0c37cd24886aa86db7aa7f30 upstream.
We are not interested in getting this debug print on our console all the time.
--- drivers/thermal/db8500_thermal.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/thermal/db8500_thermal.c +2 -2
@@ -152,8 +152,8 @@ static irqreturn_t prcmu_high_irq_handledb8500_thermal_update_config (th, idx,THERMAL_TREND_RAISING , next_low, next_high); -dev_info (&th->tz->device, -"PRCMU set max %ld, min %ld\n" , next_high, next_low); +dev_dbg (&th->tz->device, +"PRCMU set max %ld, min %ld\n" , next_high, next_low); }else if (idx == num_points -1 )/* So we roof out 1 degree over the max point */ th->interpolated_temp = db8500_thermal_points[idx] +1 ;
Greg Kroah-Hartman
maintainerPATCH 140/152 thermal: brcmstb_thermal: Do not use DT coefficients
In reply to Greg Kroah-HartmanFrom: Florian Fainelli <[email protected]>
commit e1ff6fc22f19e2af8adbad618526b80067911d40 upstream.
At the time the brcmstb_thermal driver and its binding were merged, the DT binding did not make the coefficients properties a mandatory one, therefore all users of the brcmstb_thermal driver out there have a non functional implementation with zero coefficients. Even if these properties were provided, the formula used for computation is incorrect.
The coefficients are entirely process specific (right now, only 28nm is supported) and not board or SoC specific, it is therefore appropriate to hard code them in the driver given the compatibility string we are probed with which has to be updated whenever a new process is introduced.
We remove the existing coefficients definition since subsequent patches are going to add support for a new process and will introduce new coefficients as well.
--- drivers/thermal/broadcom/brcmstb_thermal.c | 31 ++++++++--------------------- 1 file changed, 9 insertions(+), 22 deletions(-)
drivers/thermal/broadcom/brcmstb_thermal.c +9 -22
@@ -49,7 +49,7 @@#define AVS_TMON_TP_TEST_ENABLE 0x20/* Default coefficients */ -#define AVS_TMON_TEMP_SLOPE -487 +#define AVS_TMON_TEMP_SLOPE 487#define AVS_TMON_TEMP_OFFSET 410040/* HW related temperature constants */
@@ -108,23 +108,12 @@ struct brcmstb_thermal_priv {struct thermal_zone_device *thermal; }; -static void avs_tmon_get_coeffs (struct thermal_zone_device *tz,int *slope, -int *offset) -{ - *slope =thermal_zone_get_slope (tz); - *offset =thermal_zone_get_offset (tz); -} -/* Convert a HW code to a temperature reading (millidegree celsius) */ static inline int avs_tmon_code_to_temp (struct thermal_zone_device *tz,u32 code) { -const int val = code &AVS_TMON_TEMP_MASK ; -int slope, offset; - -avs_tmon_get_coeffs (tz, &slope, &offset); - -return slope * val + offset; +return (AVS_TMON_TEMP_OFFSET - + (int )((code &AVS_TMON_TEMP_MAX ) *AVS_TMON_TEMP_SLOPE )); }/*
@@ -136,20 +125,18 @@ static inline int avs_tmon_code_to_temp(static inline u32 avs_tmon_temp_to_code (struct thermal_zone_device *tz,int temp,bool low) { -int slope, offset; -if (temp <AVS_TMON_TEMP_MIN ) -return AVS_TMON_TEMP_MAX ;/* Maximum code value */ - -avs_tmon_get_coeffs (tz, &slope, &offset); +return AVS_TMON_TEMP_MAX ;/* Maximum code value */ -if (temp >= offset) +if (temp >=AVS_TMON_TEMP_OFFSET )return 0 ;/* Minimum code value */ if (low) -return (u32)(DIV_ROUND_UP (offset - temp,abs (slope))); +return (u32)(DIV_ROUND_UP (AVS_TMON_TEMP_OFFSET - temp, +AVS_TMON_TEMP_SLOPE ));else -return (u32)((offset - temp) /abs (slope)); +return (u32)((AVS_TMON_TEMP_OFFSET - temp) / +AVS_TMON_TEMP_SLOPE ); }static int brcmstb_get_temp (void *data,int *temp)
Greg Kroah-Hartman
maintainerPATCH 141/152 netfilter: nft_tunnel: no need to call htons() when dumping ports
In reply to Greg Kroah-HartmanFrom: Xin Long <[email protected]>
commit cf3e204a1ca5442190018a317d9ec181b4639bd6 upstream.
info->key.tp_src and tp_dst are __be16, when using nla_put_be16() to dump them, htons() is not needed, so remove it in this patch.
--- net/netfilter/nft_tunnel.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
net/netfilter/nft_tunnel.c +2 -2
@@ -505,8 +505,8 @@ static int nft_tunnel_opts_dump(struct sstatic int nft_tunnel_ports_dump (struct sk_buff *skb,struct ip_tunnel_info *info) { -if (nla_put_be16 (skb,NFTA_TUNNEL_KEY_SPORT ,htons (info->key.tp_src)) <0 || -nla_put_be16 (skb,NFTA_TUNNEL_KEY_DPORT ,htons (info->key.tp_dst)) <0 ) +if (nla_put_be16 (skb,NFTA_TUNNEL_KEY_SPORT , info->key.tp_src) <0 || +nla_put_be16 (skb,NFTA_TUNNEL_KEY_DPORT , info->key.tp_dst) <0 )return -1 ;return 0 ;
Greg Kroah-Hartman
maintainerPATCH 142/152 netfilter: nf_flowtable: fix documentation
In reply to Greg Kroah-HartmanFrom: Matteo Croce <[email protected]>
commit 78e06cf430934fc3768c342cbebdd1013dcd6fa7 upstream.
In the flowtable documentation there is a missing semicolon, the command as is would give this error:
nftables.conf:5:27-33: Error: syntax error, unexpected devices, expecting newline or semicolon
hook ingress priority 0 devices = { br0, pppoe-data };
^^^^^^^
nftables.conf:4:12-13: Error: invalid hook (null)
flowtable ft {
^^--- Documentation/networking/nf_flowtable.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Documentation/networking/nf_flowtable.txt +1 -1
@@ -76,7 +76,7 @@ flowtable and add one rule to your forwa table inet x { flowtable f { - hook ingress priority 0 devices = { eth0, eth1 }; + hook ingress priority 0; devices = { eth0, eth1 }; } chain y { type filter hook forward priority 0; policy accept;
Greg Kroah-Hartman
maintainerPATCH 143/152 bus: tegra-aconnect: Remove PM_CLK dependency
In reply to Greg Kroah-HartmanFrom: Sameer Pujar <[email protected]>
commit 2f56acf818a08a9187ac8ec6e3d994fc13dc368d upstream.
The ACONNECT bus driver does not use pm-clk interface anymore and hence the dependency can be removed from its Kconfig option.
--- drivers/bus/Kconfig | 1 - 1 file changed, 1 deletion(-)
drivers/bus/Kconfig +0 -1
@@ -138,7 +138,6 @@ config TEGRA_ACONNECT tristate "Tegra ACONNECT Bus Driver" depends on ARCH_TEGRA_210_SOC depends on OF && PM - select PM_CLK help Driver for the Tegra ACONNECT bus which is used to interface with the devices inside the Audio Processing Engine (APE) for Tegra210.
Greg Kroah-Hartman
maintainerPATCH 144/152 xfs: clear kernel only flags in XFS_IOC_ATTRMULTI_BY_HANDLE
In reply to Greg Kroah-HartmanFrom: Christoph Hellwig <[email protected]>
commit 953aa9d136f53e226448dbd801a905c28f8071bf upstream.
Don't allow passing arbitrary flags as they change behavior including memory allocation that the call stack is not prepared for.
--- fs/xfs/libxfs/xfs_attr.h | 7 +++++-- fs/xfs/xfs_ioctl.c | 2 ++ fs/xfs/xfs_ioctl32.c | 2 ++ 3 files changed, 9 insertions(+), 2 deletions(-)
fs/xfs/libxfs/xfs_attr.h +5 -2
@@ -26,7 +26,7 @@ struct xfs_attr_list_context;*========================================================================*/ -#define ATTR_DONTFOLLOW 0x0001/* -- unused, from IRIX -- */ +#define ATTR_DONTFOLLOW 0x0001/* -- ignored, from IRIX -- */ #define ATTR_ROOT 0x0002/* use attrs in root (trusted) namespace */ #define ATTR_TRUST 0x0004/* -- unused, from IRIX -- */ #define ATTR_SECURE 0x0008/* use attrs in security namespace */
@@ -37,7 +37,10 @@ struct xfs_attr_list_context;#define ATTR_KERNOVAL 0x2000/* [kernel] get attr size only, not value */ #define ATTR_INCOMPLETE 0x4000/* [kernel] return INCOMPLETE attr keys */ -#define ATTR_ALLOC 0x8000/* allocate xattr buffer on demand */ +#define ATTR_ALLOC 0x8000/* [kernel] allocate xattr buffer on demand */ + +#define ATTR_KERNEL_FLAGS \ + (ATTR_KERNOTIME | ATTR_KERNOVAL | ATTR_INCOMPLETE | ATTR_ALLOC)#define XFS_ATTR_FLAGS \ { ATTR_DONTFOLLOW, "DONTFOLLOW" }, \
fs/xfs/xfs_ioctl.c +2 -0
@@ -536,6 +536,8 @@ xfs_attrmulti_by_handle( error =0 ;for (i =0 ; i < am_hreq.opcount; i++) { + ops[i].am_flags &= ~ATTR_KERNEL_FLAGS ; + ops[i].am_error =strncpy_from_user ((char *)attr_name, ops[i].am_attrname,MAXNAMELEN );if (ops[i].am_error ==0 || ops[i].am_error ==MAXNAMELEN )
fs/xfs/xfs_ioctl32.c +2 -0
@@ -450,6 +450,8 @@ xfs_compat_attrmulti_by_handle( error =0 ;for (i =0 ; i < am_hreq.opcount; i++) { + ops[i].am_flags &= ~ATTR_KERNEL_FLAGS ; + ops[i].am_error =strncpy_from_user ((char *)attr_name,compat_ptr (ops[i].am_attrname),MAXNAMELEN );
Greg Kroah-Hartman
maintainerPATCH 145/152 padata: always acquire cpu_hotplug_lock before pinst->lock
In reply to Greg Kroah-HartmanFrom: Daniel Jordan <[email protected]>
commit 38228e8848cd7dd86ccb90406af32de0cad24be3 upstream.
lockdep complains when padata's paths to update cpumasks via CPU hotplug and sysfs are both taken:
# echo 0 > /sys/devices/system/cpu/cpu1/online # echo ff > /sys/kernel/pcrypt/pencrypt/parallel_cpumask
======================================================
WARNING: possible circular locking dependency detected 5.4.0-rc8-padata-cpuhp-v3+ #1 Not tainted
------------------------------------------------------
bash/205 is trying to acquire lock:
ffffffff8286bcd0 (cpu_hotplug_lock.rw_sem){++++}, at: padata_set_cpumask+0x2b/0x120
but task is already holding lock:
ffff8880001abfa0 (&pinst->lock){+.+.}, at: padata_set_cpumask+0x26/0x120
which lock already depends on the new lock.
padata doesn't take cpu_hotplug_lock and pinst->lock in a consistent order. Which should be first? CPU hotplug calls into padata with cpu_hotplug_lock already held, so it should have priority.
--- kernel/padata.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
kernel/padata.c +2 -2
@@ -643,8 +643,8 @@ int padata_set_cpumask(struct padata_insstruct cpumask *serial_mask, *parallel_mask;int err = -EINVAL ; -mutex_lock (&pinst->lock);get_online_cpus (); +mutex_lock (&pinst->lock);switch (cpumask_type) {case PADATA_CPU_PARALLEL :
@@ -662,8 +662,8 @@ int padata_set_cpumask(struct padata_ins err =__padata_set_cpumasks (pinst, parallel_mask, serial_mask); out: -put_online_cpus ();mutex_unlock (&pinst->lock); +put_online_cpus ();return err; }
Greg Kroah-Hartman
maintainerPATCH 146/152 locking/lockdep: Fix lockdep_stats indentation problem
In reply to Greg Kroah-HartmanFrom: Waiman Long <[email protected]>
commit a030f9767da1a6bbcec840fc54770eb11c2414b6 upstream.
It was found that two lines in the output of /proc/lockdep_stats have indentation problem:
# cat /proc/lockdep_stats
:
in-process chains: 25057
stack-trace entries: 137827 [max: 524288]
number of stack traces: 7973
number of stack hash chains: 6355
combined max dependencies: 1356414598
hardirq-safe locks: 57
hardirq-unsafe locks: 1286
:All the numbers displayed in /proc/lockdep_stats except the two stack trace numbers are formatted with a field with of 11. To properly align all the numbers, a field width of 11 is now added to the two stack trace numbers.
--- kernel/locking/lockdep_proc.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
kernel/locking/lockdep_proc.c +2 -2
@@ -286,9 +286,9 @@ static int lockdep_stats_show(struct seqseq_printf (m," stack-trace entries: %11lu [max: %lu]\n" , nr_stack_trace_entries,MAX_STACK_TRACE_ENTRIES );#if defined(CONFIG_TRACE_IRQFLAGS ) && defined(CONFIG_PROVE_LOCKING ) -seq_printf (m," number of stack traces: %llu\n" , +seq_printf (m," number of stack traces: %11llu\n" ,lockdep_stack_trace_count ()); -seq_printf (m," number of stack hash chains: %llu\n" , +seq_printf (m," number of stack hash chains: %11llu\n" ,lockdep_stack_hash_count ());#endif seq_printf(m, " combined max dependencies: %11u\n",
Greg Kroah-Hartman
maintainerPATCH 147/152 mm/debug.c: always print flags in dump_page()
In reply to Greg Kroah-HartmanFrom: Vlastimil Babka <[email protected]>
commit 5b57b8f22709f07c0ab5921c94fd66e8c59c3e11 upstream.
Commit 76a1850e4572 ("mm/debug.c: __dump_page() prints an extra line") inadvertently removed printing of page flags for pages that are neither anon nor ksm nor have a mapping. Fix that.
Using pr_cont() again would be a solution, but the commit explicitly removed its use. Avoiding the danger of mixing up split lines from multiple CPUs might be beneficial for near-panic dumps like this, so fix this without reintroducing pr_cont().
--- mm/debug.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)
mm/debug.c +5 -3
@@ -47,6 +47,7 @@ void __dump_page(struct page *page, consstruct address_space *mapping;bool page_poisoned =PagePoisoned (page);int mapcount; +char *type ="" ;/* * If struct page is poisoned don't access Page*() functions as that
@@ -78,9 +79,9 @@ void __dump_page(struct page *page, cons page,page_ref_count (page), mapcount, page->mapping, page_to_pgoff(page));if (PageKsm (page)) -pr_warn ("ksm flags: %#lx(%pGp)\n" , page->flags, &page->flags); + type ="ksm " ;else if (PageAnon (page)) -pr_warn ("anon flags: %#lx(%pGp)\n" , page->flags, &page->flags); + type ="anon " ;else if (mapping) {if (mapping->host && mapping->host->i_dentry.first) {struct dentry *dentry;
@@ -88,10 +89,11 @@ void __dump_page(struct page *page, conspr_warn ("%ps name:\"%pd\"\n" , mapping->a_ops, dentry); }else pr_warn ("%ps\n" , mapping->a_ops); -pr_warn ("flags: %#lx(%pGp)\n" , page->flags, &page->flags); }BUILD_BUG_ON (ARRAY_SIZE (pageflag_names) != __NR_PAGEFLAGS +1 ); +pr_warn ("%sflags: %#lx(%pGp)\n" , type, page->flags, &page->flags); + hex_only: print_hex_dump(KERN_WARNING, "raw: ", DUMP_PREFIX_NONE, 32, sizeof(unsigned long), page,
Greg Kroah-Hartman
maintainerPATCH 148/152 mm/gup: allow FOLL_FORCE for get_user_pages_fast()
In reply to Greg Kroah-HartmanFrom: John Hubbard <[email protected]>
commit f4000fdf435b8301a11cf85237c561047f8c4c72 upstream.
Commit 817be129e6f2 ("mm: validate get_user_pages_fast flags") allowed only FOLL_WRITE and FOLL_LONGTERM to be passed to get_user_pages_fast(). This, combined with the fact that get_user_pages_fast() falls back to "slow gup", which *does* accept FOLL_FORCE, leads to an odd situation:
if you need FOLL_FORCE, you cannot call get_user_pages_fast().
There does not appear to be any reason for filtering out FOLL_FORCE. There is nothing in the _fast() implementation that requires that we avoid writing to the pages. So it appears to have been an oversight.
Fix by allowing FOLL_FORCE to be set for get_user_pages_fast().
--- mm/gup.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
mm/gup.c +2 -1
@@ -2401,7 +2401,8 @@ int get_user_pages_fast(unsigned long stunsigned long addr, len, end;int nr =0 , ret =0 ; -if (WARN_ON_ONCE (gup_flags & ~(FOLL_WRITE |FOLL_LONGTERM ))) +if (WARN_ON_ONCE (gup_flags & ~(FOLL_WRITE |FOLL_LONGTERM | +FOLL_FORCE )))return -EINVAL ; start =untagged_addr (start) &PAGE_MASK ;
Greg Kroah-Hartman
maintainerPATCH 149/152 mm/huge_memory.c: use head to check huge zero page
In reply to Greg Kroah-HartmanFrom: Wei Yang <[email protected]>
commit cb829624867b5ab10bc6a7036d183b1b82bfe9f8 upstream.
The page could be a tail page, if this is the case, this BUG_ON will never be triggered.
--- mm/huge_memory.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
mm/huge_memory.c +1 -1
@@ -2712,7 +2712,7 @@ int split_huge_page_to_list(struct pageunsigned long flags;pgoff_t end; -VM_BUG_ON_PAGE (is_huge_zero_page (page), page); +VM_BUG_ON_PAGE (is_huge_zero_page (head), head);VM_BUG_ON_PAGE (!PageLocked (page), page);VM_BUG_ON_PAGE (!PageCompound (page), page);
Greg Kroah-Hartman
maintainerPATCH 150/152 mm, thp: fix defrag setting if newline is not used
In reply to Greg Kroah-HartmanFrom: David Rientjes <[email protected]>
commit f42f25526502d851d0e3ca1e46297da8aafce8a7 upstream.
If thp defrag setting "defer" is used and a newline is *not* used when writing to the sysfs file, this is interpreted as the "defer+madvise" option.
This is because we do prefix matching and if five characters are written without a newline, the current code ends up comparing to the first five bytes of the "defer+madvise" option and using that instead.
Use the more appropriate sysfs_streq() that handles the trailing newline for us. Since this doubles as a nice cleanup, do it in enabled_store() as well.
The current implementation relies on prefix matching: the number of bytes compared is either the number of bytes written or the length of the option being compared. With a newline, "defer\n" does not match "defer+"madvise"; without a newline, however, "defer" is considered to match "defer+madvise" (prefix matching is only comparing the first five bytes). End result is that writing "defer" is broken unless it has an additional trailing character.
This means that writing "madv" in the past would match and set "madvise". With strict checking, that no longer is the case but it is unlikely anybody is currently doing this.
--- mm/huge_memory.c | 24 ++++++++---------------- 1 file changed, 8 insertions(+), 16 deletions(-)
mm/huge_memory.c +8 -16
@@ -177,16 +177,13 @@ static ssize_t enabled_store(struct kobj {ssize_t ret = count; -if (!memcmp ("always" , buf, -min (sizeof ("always" )-1 , count))) { +if (sysfs_streq (buf,"always" )) {clear_bit (TRANSPARENT_HUGEPAGE_REQ_MADV_FLAG , &transparent_hugepage_flags);set_bit (TRANSPARENT_HUGEPAGE_FLAG , &transparent_hugepage_flags); - }else if (!memcmp ("madvise" , buf, -min (sizeof ("madvise" )-1 , count))) { + }else if (sysfs_streq (buf,"madvise" )) {clear_bit (TRANSPARENT_HUGEPAGE_FLAG , &transparent_hugepage_flags);set_bit (TRANSPARENT_HUGEPAGE_REQ_MADV_FLAG , &transparent_hugepage_flags); - }else if (!memcmp ("never" , buf, -min (sizeof ("never" )-1 , count))) { + }else if (sysfs_streq (buf,"never" )) {clear_bit (TRANSPARENT_HUGEPAGE_FLAG , &transparent_hugepage_flags);clear_bit (TRANSPARENT_HUGEPAGE_REQ_MADV_FLAG , &transparent_hugepage_flags); }else
@@ -250,32 +247,27 @@ static ssize_t defrag_store(struct kobje struct kobj_attribute *attr, const char *buf, size_t count) { - if (!memcmp("always", buf, - min(sizeof("always")-1, count))) { + if (sysfs_streq(buf, "always")) { clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_KSWAPD_FLAG, &transparent_hugepage_flags); clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_KSWAPD_OR_MADV_FLAG, &transparent_hugepage_flags); clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_REQ_MADV_FLAG, &transparent_hugepage_flags); set_bit(TRANSPARENT_HUGEPAGE_DEFRAG_DIRECT_FLAG, &transparent_hugepage_flags); - } else if (!memcmp("defer+madvise", buf, -min (sizeof ("defer+madvise" )-1 , count))) { + } else if (sysfs_streq(buf, "defer+madvise")) { clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_DIRECT_FLAG, &transparent_hugepage_flags); clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_KSWAPD_FLAG, &transparent_hugepage_flags); clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_REQ_MADV_FLAG, &transparent_hugepage_flags); set_bit(TRANSPARENT_HUGEPAGE_DEFRAG_KSWAPD_OR_MADV_FLAG, &transparent_hugepage_flags); - }else if (!memcmp ("defer" , buf, -min (sizeof ("defer" )-1 , count))) { + } else if (sysfs_streq(buf, "defer")) { clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_DIRECT_FLAG, &transparent_hugepage_flags); clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_KSWAPD_OR_MADV_FLAG, &transparent_hugepage_flags); clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_REQ_MADV_FLAG, &transparent_hugepage_flags); set_bit(TRANSPARENT_HUGEPAGE_DEFRAG_KSWAPD_FLAG, &transparent_hugepage_flags); - }else if (!memcmp ("madvise" , buf, -min (sizeof ("madvise" )-1 , count))) { + } else if (sysfs_streq(buf, "madvise")) { clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_DIRECT_FLAG, &transparent_hugepage_flags); clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_KSWAPD_FLAG, &transparent_hugepage_flags); clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_KSWAPD_OR_MADV_FLAG, &transparent_hugepage_flags); set_bit(TRANSPARENT_HUGEPAGE_DEFRAG_REQ_MADV_FLAG, &transparent_hugepage_flags); - }else if (!memcmp ("never" , buf, -min (sizeof ("never" )-1 , count))) { + } else if (sysfs_streq(buf, "never")) { clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_DIRECT_FLAG, &transparent_hugepage_flags); clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_KSWAPD_FLAG, &transparent_hugepage_flags); clear_bit(TRANSPARENT_HUGEPAGE_DEFRAG_KSWAPD_OR_MADV_FLAG, &transparent_hugepage_flags);
Greg Kroah-Hartman
maintainerPATCH 151/152 kvm: nVMX: VMWRITE checks VMCS-link pointer before VMCS field
In reply to Greg Kroah-HartmanFrom: Jim Mattson <[email protected]>
commit dd2d6042b7f4a5440705b4ffc6c4c2dba81a43b7 upstream.
According to the SDM, a VMWRITE in VMX non-root operation with an invalid VMCS-link pointer results in VMfailInvalid before the validity of the VMCS field in the secondary source operand is checked.
For consistency, modify both handle_vmwrite and handle_vmread, even though there was no problem with the latter.
--- arch/x86/kvm/vmx/nested.c | 59 +++++++++++++++++++--------------------------- 1 file changed, 25 insertions(+), 34 deletions(-)
arch/x86/kvm/vmx/nested.c +25 -34
@@ -4609,32 +4609,28 @@ static int handle_vmread(struct kvm_vcpu {unsigned long field;u64 field_value; +struct vcpu_vmx *vmx =to_vmx (vcpu);unsigned long exit_qualification =vmcs_readl (EXIT_QUALIFICATION );u32 vmx_instruction_info =vmcs_read32 (VMX_INSTRUCTION_INFO );int len;gva_t gva =0 ; -struct vmcs12 *vmcs12; +struct vmcs12 *vmcs12 =is_guest_mode (vcpu) ?get_shadow_vmcs12 (vcpu) + :get_vmcs12 (vcpu);struct x86_exception e;short offset;if (!nested_vmx_check_permission (vcpu))return 1 ; -if (to_vmx (vcpu)->nested.current_vmptr ==-1ull ) +/* +* In VMX non-root operation, when the VMCS-link pointer is -1ull, +* any VMREAD sets the ALU flags for VMfailInvalid. +*/ +if (vmx->nested.current_vmptr ==-1ull || + (is_guest_mode (vcpu) && +get_vmcs12 (vcpu)->vmcs_link_pointer ==-1ull ))return nested_vmx_failInvalid (vcpu); -if (!is_guest_mode (vcpu)) - vmcs12 =get_vmcs12 (vcpu); -else { -/* -* When vmcs->vmcs_link_pointer is -1ull, any VMREAD -* to shadowed-field sets the ALU flags for VMfailInvalid. -*/ -if (get_vmcs12 (vcpu)->vmcs_link_pointer ==-1ull ) -return nested_vmx_failInvalid (vcpu); - vmcs12 =get_shadow_vmcs12 (vcpu); - } -/* Decode instruction info and find the field to read */ field =kvm_register_readl (vcpu, (((vmx_instruction_info) >>28 ) &0xf ));
@@ -4713,13 +4709,20 @@ static int handle_vmwrite(struct kvm_vcp*/ u64 field_value =0 ;struct x86_exception e; -struct vmcs12 *vmcs12; +struct vmcs12 *vmcs12 =is_guest_mode (vcpu) ?get_shadow_vmcs12 (vcpu) + :get_vmcs12 (vcpu);short offset;if (!nested_vmx_check_permission (vcpu))return 1 ; -if (vmx->nested.current_vmptr ==-1ull ) +/* +* In VMX non-root operation, when the VMCS-link pointer is -1ull, +* any VMWRITE sets the ALU flags for VMfailInvalid. +*/ +if (vmx->nested.current_vmptr ==-1ull || + (is_guest_mode (vcpu) && +get_vmcs12 (vcpu)->vmcs_link_pointer ==-1ull ))return nested_vmx_failInvalid (vcpu);if (vmx_instruction_info & (1u <<10 ))
@@ -4747,24 +4750,12 @@ static int handle_vmwrite(struct kvm_vcpreturn nested_vmx_failValid (vcpu,VMXERR_VMWRITE_READ_ONLY_VMCS_COMPONENT ); -if (!is_guest_mode (vcpu)) { - vmcs12 =get_vmcs12 (vcpu); - -/* -* Ensure vmcs12 is up-to-date before any VMWRITE that dirties -* vmcs12, else we may crush a field or consume a stale value. -*/ -if (!is_shadow_field_rw (field)) -copy_vmcs02_to_vmcs12_rare (vcpu, vmcs12); - }else { -/* -* When vmcs->vmcs_link_pointer is -1ull, any VMWRITE -* to shadowed-field sets the ALU flags for VMfailInvalid. -*/ -if (get_vmcs12 (vcpu)->vmcs_link_pointer ==-1ull ) -return nested_vmx_failInvalid (vcpu); - vmcs12 =get_shadow_vmcs12 (vcpu); - } +/* +* Ensure vmcs12 is up-to-date before any VMWRITE that dirties +* vmcs12, else we may crush a field or consume a stale value. +*/ +if (!is_guest_mode (vcpu) && !is_shadow_field_rw (field)) +copy_vmcs02_to_vmcs12_rare (vcpu, vmcs12); offset =vmcs_field_to_offset (field);if (offset <0 )
Greg Kroah-Hartman
maintainerPATCH 152/152 kvm: nVMX: VMWRITE checks unsupported field before read-only field
In reply to Greg Kroah-HartmanFrom: Jim Mattson <[email protected]>
commit 693e02cc24090c379217138719d9d84e50036b24 upstream.
According to the SDM, VMWRITE checks to see if the secondary source operand corresponds to an unsupported VMCS field before it checks to see if the secondary source operand corresponds to a VM-exit information field and the processor does not support writing to VM-exit information fields.
--- arch/x86/kvm/vmx/nested.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-)
arch/x86/kvm/vmx/nested.c +6 -5
@@ -4741,6 +4741,12 @@ static int handle_vmwrite(struct kvm_vcp field =kvm_register_readl (vcpu, (((vmx_instruction_info) >>28 ) &0xf )); + + offset =vmcs_field_to_offset (field); +if (offset <0 ) +return nested_vmx_failValid (vcpu, +VMXERR_UNSUPPORTED_VMCS_COMPONENT ); +/* * If the vCPU supports "VMWRITE to any supported field in the * VMCS," then the "read-only" fields are actually read/write.
@@ -4757,11 +4763,6 @@ static int handle_vmwrite(struct kvm_vcpif (!is_guest_mode (vcpu) && !is_shadow_field_rw (field))copy_vmcs02_to_vmcs12_rare (vcpu, vmcs12); - offset =vmcs_field_to_offset (field); -if (offset <0 ) -return nested_vmx_failValid (vcpu, -VMXERR_UNSUPPORTED_VMCS_COMPONENT ); -/* * Some Intel CPUs intentionally drop the reserved bits of the AR byte * fields on VMWRITE. Emulate this behavior to ensure consistent KVM
Jon Hunter
maintainerQuoted 9 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 5.4.24 release. There are 152 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 Thu, 05 Mar 2020 17:42:10 +0000. Anything received after that time might be too late.
The whole patch series can be found in one patch at:
Quoted 8 lines
https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.4.24-rc1.gzor in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.gitlinux-5.4.y
and the diffstat can be found below.thanks,
greg k-h
All tests passing for Tegra ...
Test results for stable-v5.4:
13 builds: 13 pass, 0 fail
22 boots: 22 pass, 0 fail
40 tests: 40 pass, 0 failLinux version: 5.4.24-rc1-g1254e88b4fc1
Boards tested: tegra124-jetson-tk1, tegra186-p2771-0000,
tegra194-p2972-0000, tegra20-ventana,
tegra210-p2371-2180, tegra210-p3450-0000,
tegra30-cardhu-a04Cheers
Jon
Signature · 1 line
nvpublic
Quoted 9 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 5.4.24 release. There are 152 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 Thu, 05 Mar 2020 17:42:10 +0000. Anything received after that time might be too late.
The whole patch series can be found in one patch at:
Quoted 9 lines
https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.4.24-rc1.gzor in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.gitlinux-5.4.y
and the diffstat can be found below.thanks,
greg k-h
Compiled and booted on my test system. No dmesg regressions.
thanks,
-- Shuah
Quoted 10 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 5.4.24 release. There are 152 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 Thu, 05 Mar 2020 17:42:10 +0000. Anything received after that time might be too late.
The whole patch series can be found in one patch at:
Quoted 8 lines
https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.4.24-rc1.gzor in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.gitlinux-5.4.y
and the diffstat can be found below.thanks,
greg k-h
Results from Linaro’s test farm.
Regressions detected on x86_64 and i386.
Test failure output:
CVE-2017-5715: VULN (IBRS+IBPB or retpoline+IBPB+RSB filling, is needed to mitigate the vulnerability)
Test description:
CVE-2017-5715 branch target injection (Spectre Variant 2)
Impact: Kernel
Mitigation 1: new opcode via microcode update that should be used by up to date compilers to protect the BTB (by flushing indirect branch predictors)
Mitigation 2: introducing "retpoline" into compilers, and recompile software/OS with it
Performance impact of the mitigation: high for mitigation 1, medium for mitigation 2, depending on your CPU
ref:
https://github.com/speed47/spectre-meltdown-checker
https://qa-reports.linaro.org/lkft/linux-stable-rc-5.4-oe/build/v5.4.23-153-g1254e88b4fc1/testrun/1264789/log
https://qa-reports.linaro.org/lkft/linux-stable-rc-5.4-oe/tests/spectre-meltdown-checker-test/CVE-2017-5715
Summary
------------------------------------------------------------------------
kernel: 5.4.24-rc1
git repo: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git
git branch: linux-5.4.y
git commit: 1254e88b4fc1470d152f494c3590bb6a33ab33eb
git describe: v5.4.23-153-g1254e88b4fc1
Test details: https://qa-reports.linaro.org/lkft/linux-stable-rc-5.4-oe/build/v5.4.23-153-g1254e88b4fc1
Regressions (compared to build v5.4.23)
------------------------------------------------------------------------
i386:
spectre-meltdown-checker-test:
* CVE-2017-5715
x86:
spectre-meltdown-checker-test:
* CVE-2017-5715
No fixes (compared to build v5.4.23)
Ran 27633 total tests in the following environments and test suites.
Environments
--------------
- dragonboard-410c
- hi6220-hikey
- i386
- juno-r2
- nxp-ls2088
- qemu_arm
- qemu_arm64
- qemu_i386
- qemu_x86_64
- x15
- x86
Test Suites
-----------
* build
* install-android-platform-tools-r2600
* kselftest
* libgpiod
* linux-log-parser
* perf
* ltp-cap_bounds-tests
* ltp-commands-tests
* ltp-containers-tests
* ltp-cpuhotplug-tests
* ltp-cve-tests
* ltp-dio-tests
* ltp-fcntl-locktests-tests
* ltp-filecaps-tests
* ltp-fs-tests
* ltp-fs_bind-tests
* ltp-fs_perms_simple-tests
* ltp-fsx-tests
* ltp-hugetlb-tests
* ltp-io-tests
* ltp-ipc-tests
* ltp-math-tests
* ltp-mm-tests
* ltp-nptl-tests
* ltp-pty-tests
* ltp-sched-tests
* ltp-securebits-tests
* ltp-syscalls-tests
* network-basic-tests
* spectre-meltdown-checker-test
* v4l2-compliance
* kvm-unit-tests
* libhugetlbfs
* ltp-crypto-tests
* ltp-cap_bounds-64k-page_size-tests
* ltp-cap_bounds-kasan-tests
* ltp-commands-64k-page_size-tests
* ltp-commands-kasan-tests
* ltp-containers-64k-page_size-tests
* ltp-containers-kasan-tests
* ltp-cpuhotplug-64k-page_size-tests
* ltp-cpuhotplug-kasan-tests
* ltp-crypto-64k-page_size-tests
* ltp-crypto-kasan-tests
* ltp-cve-64k-page_size-tests
* ltp-cve-kasan-tests
* ltp-dio-64k-page_size-tests
* ltp-dio-kasan-tests
* ltp-fcntl-locktests-64k-page_size-tests
* ltp-fcntl-locktests-kasan-tests
* ltp-filecaps-64k-page_size-tests
* ltp-filecaps-kasan-tests
* ltp-fs-64k-page_size-tests
* ltp-fs-kasan-tests
* ltp-fs_bind-64k-page_size-tests
* ltp-fs_bind-kasan-tests
* ltp-fs_perms_simple-64k-page_size-tests
* ltp-fs_perms_simple-kasan-tests
* ltp-fsx-64k-page_size-tests
* ltp-fsx-kasan-tests
* ltp-hugetlb-64k-page_size-tests
* ltp-hugetlb-kasan-tests
* ltp-io-64k-page_size-tests
* ltp-io-kasan-tests
* ltp-ipc-64k-page_size-tests
* ltp-ipc-kasan-tests
* ltp-math-64k-page_size-tests
* ltp-math-kasan-tests
* ltp-mm-kasan-tests
* ltp-nptl-64k-page_size-tests
* ltp-pty-64k-page_size-tests
* ltp-sched-64k-page_size-tests
* ltp-sched-kasan-tests
* ltp-securebits-64k-page_size-tests
* ltp-syscalls-64k-page_size-tests
* ltp-syscalls-compat-tests
* ltp-syscalls-kasan-tests
* ltp-open-posix-tests
* ssuite
* kselftest-vsyscall-mode-native
* kselftest-vsyscall-mode-none
* ltp-nptl-kasan-tests
* ltp-pty-kasan-tests
* ltp-securebits-kasan-tests
Signature · 2 lines
Linaro LKFT https://lkft.linaro.org
Quoted 8 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 5.4.24 release. There are 152 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 Thu, 05 Mar 2020 17:42:10 +0000. Anything received after that time might be too late.
Build results:
total: 158 pass: 158 fail: 0 Qemu test results: total: 422 pass: 422 fail: 0
Guenter