e3d8fe099304cc3608f72d7c9db15842d5c5dca6
9 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
Greg Kroah-Hartman
|
e3d8fe0993 |
Merge 5.15.91 into android14-5.15
Changes in 5.15.91 memory: tegra: Remove clients SID override programming memory: atmel-sdramc: Fix missing clk_disable_unprepare in atmel_ramc_probe() memory: mvebu-devbus: Fix missing clk_disable_unprepare in mvebu_devbus_probe() dmaengine: ti: k3-udma: Do conditional decrement of UDMA_CHAN_RT_PEER_BCNT_REG arm64: dts: imx8mp-phycore-som: Remove invalid PMIC property ARM: dts: imx6ul-pico-dwarf: Use 'clock-frequency' ARM: dts: imx7d-pico: Use 'clock-frequency' ARM: dts: imx6qdl-gw560x: Remove incorrect 'uart-has-rtscts' arm64: dts: imx8mm-beacon: Fix ecspi2 pinmux ARM: imx: add missing of_node_put() HID: intel_ish-hid: Add check for ishtp_dma_tx_map arm64: dts: imx8mm-venice-gw7901: fix USB2 controller OC polarity soc: imx8m: Fix incorrect check for of_clk_get_by_name() reset: uniphier-glue: Use reset_control_bulk API reset: uniphier-glue: Fix possible null-ptr-deref EDAC/highbank: Fix memory leak in highbank_mc_probe() firmware: arm_scmi: Harden shared memory access in fetch_response firmware: arm_scmi: Harden shared memory access in fetch_notification tomoyo: fix broken dependency on *.conf.default RDMA/core: Fix ib block iterator counter overflow IB/hfi1: Reject a zero-length user expected buffer IB/hfi1: Reserve user expected TIDs IB/hfi1: Fix expected receive setup error exit issues IB/hfi1: Immediately remove invalid memory from hardware IB/hfi1: Remove user expected buffer invalidate race affs: initialize fsdata in affs_truncate() PM: AVS: qcom-cpr: Fix an error handling path in cpr_probe() arm64: dts: qcom: msm8992: Don't use sfpb mutex arm64: dts: qcom: msm8992-libra: Add CPU regulators arm64: dts: qcom: msm8992-libra: Fix the memory map phy: ti: fix Kconfig warning and operator precedence NFSD: fix use-after-free in nfsd4_ssc_setup_dul() ARM: dts: at91: sam9x60: fix the ddr clock for sam9x60 amd-xgbe: TX Flow Ctrl Registers are h/w ver dependent amd-xgbe: Delay AN timeout during KR training bpf: Fix pointer-leak due to insufficient speculative store bypass mitigation phy: rockchip-inno-usb2: Fix missing clk_disable_unprepare() in rockchip_usb2phy_power_on() net: nfc: Fix use-after-free in local_cleanup() net: wan: Add checks for NULL for utdm in undo_uhdlc_init and unmap_si_regs net: enetc: avoid deadlock in enetc_tx_onestep_tstamp() sch_htb: Avoid grafting on htb_destroy_class_offload when destroying htb gpio: use raw spinlock for gpio chip shadowed data gpio: mxc: Protect GPIO irqchip RMW with bgpio spinlock gpio: mxc: Always set GPIOs used as interrupt source to INPUT mode wifi: rndis_wlan: Prevent buffer overflow in rndis_query_oid pinctrl/rockchip: Use temporary variable for struct device pinctrl/rockchip: add error handling for pull/drive register getters pinctrl: rockchip: fix reading pull type on rk3568 net: stmmac: Fix queue statistics reading net/sched: sch_taprio: fix possible use-after-free l2tp: Serialize access to sk_user_data with sk_callback_lock l2tp: Don't sleep and disable BH under writer-side sk_callback_lock l2tp: convert l2tp_tunnel_list to idr l2tp: close all race conditions in l2tp_tunnel_register() octeontx2-pf: Avoid use of GFP_KERNEL in atomic context net: usb: sr9700: Handle negative len net: mdio: validate parameter addr in mdiobus_get_phy() HID: check empty report_list in hid_validate_values() HID: check empty report_list in bigben_probe() net: stmmac: fix invalid call to mdiobus_get_phy() pinctrl: rockchip: fix mux route data for rk3568 HID: revert CHERRY_MOUSE_000C quirk usb: gadget: f_fs: Prevent race during ffs_ep0_queue_wait usb: gadget: f_fs: Ensure ep0req is dequeued before free_request Bluetooth: Fix possible deadlock in rfcomm_sk_state_change net: ipa: disable ipa interrupt during suspend net/mlx5: E-switch, Fix setting of reserved fields on MODIFY_SCHEDULING_ELEMENT net: mlx5: eliminate anonymous module_init & module_exit drm/panfrost: fix GENERIC_ATOMIC64 dependency dmaengine: Fix double increment of client_count in dma_chan_get() net: macb: fix PTP TX timestamp failure due to packet padding virtio-net: correctly enable callback during start_xmit l2tp: prevent lockdep issue in l2tp_tunnel_register() HID: betop: check shape of output reports cifs: fix potential deadlock in cache_refresh_path() dmaengine: xilinx_dma: call of_node_put() when breaking out of for_each_child_of_node() phy: phy-can-transceiver: Skip warning if no "max-bitrate" drm/amd/display: fix issues with driver unload nvme-pci: fix timeout request state check tcp: avoid the lookup process failing to get sk in ehash table octeontx2-pf: Fix the use of GFP_KERNEL in atomic context on rt ptdma: pt_core_execute_cmd() should use spinlock device property: fix of node refcount leak in fwnode_graph_get_next_endpoint() w1: fix deadloop in __w1_remove_master_device() w1: fix WARNING after calling w1_process() driver core: Fix test_async_probe_init saves device in wrong array selftests/net: toeplitz: fix race on tpacket_v3 block close net: dsa: microchip: ksz9477: port map correction in ALU table entry register thermal/core: Remove duplicate information when an error occurs thermal/core: Rename 'trips' to 'num_trips' thermal: Validate new state in cur_state_store() thermal/core: fix error code in __thermal_cooling_device_register() thermal: core: call put_device() only after device_register() fails net: stmmac: enable all safety features by default tcp: fix rate_app_limited to default to 1 scsi: iscsi: Fix multiple iSCSI session unbind events sent to userspace cpufreq: Add Tegra234 to cpufreq-dt-platdev blocklist kcsan: test: don't put the expect array on the stack cpufreq: Add SM6375 to cpufreq-dt-platdev blocklist ASoC: fsl_micfil: Correct the number of steps on SX controls net: usb: cdc_ether: add support for Thales Cinterion PLS62-W modem drm: Add orientation quirk for Lenovo ideapad D330-10IGL s390/debug: add _ASM_S390_ prefix to header guard s390: expicitly align _edata and _end symbols on page boundary perf/x86/msr: Add Emerald Rapids perf/x86/intel/uncore: Add Emerald Rapids cpufreq: armada-37xx: stop using 0 as NULL pointer ASoC: fsl_ssi: Rename AC'97 streams to avoid collisions with AC'97 CODEC ASoC: fsl-asoc-card: Fix naming of AC'97 CODEC widgets spi: spidev: remove debug messages that access spidev->spi without locking KVM: s390: interrupt: use READ_ONCE() before cmpxchg() scsi: hisi_sas: Set a port invalid only if there are no devices attached when refreshing port id r8152: add vendor/device ID pair for Microsoft Devkit platform/x86: touchscreen_dmi: Add info for the CSL Panther Tab HD platform/x86: asus-nb-wmi: Add alternate mapping for KEY_SCREENLOCK lockref: stop doing cpu_relax in the cmpxchg loop firmware: coreboot: Check size of table entry and use flex-array drm/i915: Allow switching away via vga-switcheroo if uninitialized Revert "selftests/bpf: check null propagation only neither reg is PTR_TO_BTF_ID" drm/i915: Remove unused variable x86: ACPI: cstate: Optimize C3 entry on AMD CPUs fs: reiserfs: remove useless new_opts in reiserfs_remount sysctl: add a new register_sysctl_init() interface kernel/panic: move panic sysctls to its own file panic: unset panic_on_warn inside panic() ubsan: no need to unset panic_on_warn in ubsan_epilogue() kasan: no need to unset panic_on_warn in end_report() exit: Add and use make_task_dead. objtool: Add a missing comma to avoid string concatenation hexagon: Fix function name in die() h8300: Fix build errors from do_exit() to make_task_dead() transition csky: Fix function name in csky_alignment() and die() ia64: make IA64_MCA_RECOVERY bool instead of tristate panic: Separate sysctl logic from CONFIG_SMP exit: Put an upper limit on how often we can oops exit: Expose "oops_count" to sysfs exit: Allow oops_limit to be disabled panic: Consolidate open-coded panic_on_warn checks panic: Introduce warn_limit panic: Expose "warn_count" to sysfs docs: Fix path paste-o for /sys/kernel/warn_count exit: Use READ_ONCE() for all oops/warn limit reads Bluetooth: hci_sync: cancel cmd_timer if hci_open failed drm/amdgpu: complete gfxoff allow signal during suspend without delay scsi: hpsa: Fix allocation size for scsi_host_alloc() KVM: SVM: fix tsc scaling cache logic module: Don't wait for GOING modules tracing: Make sure trace_printk() can output as soon as it can be used trace_events_hist: add check for return value of 'create_hist_field' ftrace/scripts: Update the instructions for ftrace-bisect.sh cifs: Fix oops due to uncleared server->smbd_conn in reconnect i2c: mv64xxx: Remove shutdown method from driver i2c: mv64xxx: Add atomic_xfer method to driver ksmbd: add smbd max io size parameter ksmbd: add max connections parameter ksmbd: do not sign response to session request for guest login ksmbd: downgrade ndr version error message to debug ksmbd: limit pdu length size according to connection status ovl: fail on invalid uid/gid mapping at copy up KVM: x86/vmx: Do not skip segment attributes if unusable bit is set KVM: arm64: GICv4.1: Fix race with doorbell on VPE activation/deactivation thermal: intel: int340x: Protect trip temperature from concurrent updates ipv6: fix reachability confirmation with proxy_ndp ARM: 9280/1: mm: fix warning on phys_addr_t to void pointer assignment EDAC/device: Respect any driver-supplied workqueue polling value EDAC/qcom: Do not pass llcc_driv_data as edac_device_ctl_info's pvt_info net: mana: Fix IRQ name - add PCI and queue number scsi: ufs: core: Fix devfreq deadlocks i2c: designware: use casting of u64 in clock multiplication to avoid overflow netlink: prevent potential spectre v1 gadgets net: fix UaF in netns ops registration error path drm/i915/selftest: fix intel_selftest_modify_policy argument types netfilter: nft_set_rbtree: Switch to node list walk for overlap detection netfilter: nft_set_rbtree: skip elements in transaction from garbage collection netlink: annotate data races around nlk->portid netlink: annotate data races around dst_portid and dst_group netlink: annotate data races around sk_state ipv4: prevent potential spectre v1 gadget in ip_metrics_convert() ipv4: prevent potential spectre v1 gadget in fib_metrics_match() netfilter: conntrack: fix vtag checks for ABORT/SHUTDOWN_COMPLETE netrom: Fix use-after-free of a listening socket. net/sched: sch_taprio: do not schedule in taprio_reset() sctp: fail if no bound addresses can be used for a given scope riscv/kprobe: Fix instruction simulation of JALR nvme: fix passthrough csi check gpio: mxc: Unlock on error path in mxc_flip_edge() ravb: Rename "no_ptp_cfg_active" and "ptp_cfg_active" variables net: ravb: Fix lack of register setting after system resumed for Gen3 net: ravb: Fix possible hang if RIS2_QFF1 happen net: mctp: mark socks as dead on unhash, prevent re-add thermal: intel: int340x: Add locking to int340x_thermal_get_trip_type() net/tg3: resolve deadlock in tg3_reset_task() during EEH net: mdio-mux-meson-g12a: force internal PHY off on mux switch treewide: fix up files incorrectly marked executable tools: gpio: fix -c option of gpio-event-mon Revert "Input: synaptics - switch touchpad on HP Laptop 15-da3001TU to RMI mode" cpufreq: Move to_gov_attr_set() to cpufreq.h cpufreq: governor: Use kobject release() method to free dbs_data kbuild: Allow kernel installation packaging to override pkg-config block: fix and cleanup bio_check_ro x86/i8259: Mark legacy PIC interrupts with IRQ_LEVEL netfilter: conntrack: unify established states for SCTP paths perf/x86/amd: fix potential integer overflow on shift of a int Linux 5.15.91 Change-Id: I3349d802533097ac86e5c680fbd40c00c9719ec7 Signed-off-by: Greg Kroah-Hartman <gregkh@google.com> |
||
Marc Zyngier
|
036093c08d |
KVM: arm64: GICv4.1: Fix race with doorbell on VPE activation/deactivation
commit ef3691683d7bfd0a2acf48812e4ffe894f10bfa8 upstream.
To save the vgic LPI pending state with GICv4.1, the VPEs must all be
unmapped from the ITSs so that the sGIC caches can be flushed.
The opposite is done once the state is saved.
This is all done by using the activate/deactivate irqdomain callbacks
directly from the vgic code. Crutially, this is done without holding
the irqdesc lock for the interrupts that represent the VPE. And these
callbacks are changing the state of the irqdesc. What could possibly
go wrong?
If a doorbell fires while we are messing with the irqdesc state,
it will acquire the lock and change the interrupt state concurrently.
Since we don't hole the lock, curruption occurs in on the interrupt
state. Oh well.
While acquiring the lock would fix this (and this was Shanker's
initial approach), this is still a layering violation we could do
without. A better approach is actually to free the VPE interrupt,
do what we have to do, and re-request it.
It is more work, but this usually happens only once in the lifetime
of the VM and we don't really care about this sort of overhead.
Fixes:
|
||
|
Marc Zyngier
|
e3031d06f7 |
BACKPORT: KVM: Use 'unsigned long' as kvm_for_each_vcpu()'s index
Everywhere we use kvm_for_each_vpcu(), we use an int as the vcpu index. Unfortunately, we're about to move rework the iterator, which requires this to be upgrade to an unsigned long. Let's bite the bullet and repaint all of it in one go. Signed-off-by: Marc Zyngier <maz@kernel.org> Message-Id: <20211116160403.4074052-7-maz@kernel.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> (cherry picked from commit 46808a4cb89708c2e5b264eb9d1035762581921b) [willdeacon@: Drop riscv hunks; drop x86 hunks for code that isn't present; convert kvm_hyperv_tsc_notifier() as well] Signed-off-by: Will Deacon <willdeacon@google.com> Bug: 233587962 Bug: 233588291 Change-Id: I78a3ae244de22575684dd47a9823b9fc61b6fa7d |
||
Jia He
|
da5b0ba023 |
UPSTREAM: KVM: arm64: vgic: Add memcg accounting to vgic allocations
Inspired by commit
|
||
Zenghui Yu
|
12df742921 |
KVM: arm64: GICv4.1: Restore VLPI pending state to physical side
When setting the forwarding path of a VLPI (switch to the HW mode), we can also transfer the pending state from irq->pending_latch to VPT (especially in migration, the pending states of VLPIs are restored into kvm’s vgic first). And we currently send "INT+VSYNC" to trigger a VLPI to pending. Signed-off-by: Zenghui Yu <yuzenghui@huawei.com> Signed-off-by: Shenming Lu <lushenming@huawei.com> Signed-off-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/20210322060158.1584-6-lushenming@huawei.com |
||
Shenming Lu
|
80317fe4a6 |
KVM: arm64: GICv4.1: Add function to get VLPI state
With GICv4.1 and the vPE unmapped, which indicates the invalidation of any VPT caches associated with the vPE, we can get the VLPI state by peeking at the VPT. So we add a function for this. Signed-off-by: Shenming Lu <lushenming@huawei.com> Signed-off-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/20210322060158.1584-4-lushenming@huawei.com |
||
|
Shenming Lu
|
57e3cebd02 |
KVM: arm64: Delay the polling of the GICR_VPENDBASER.Dirty bit
In order to reduce the impact of the VPT parsing happening on the GIC, we can split the vcpu reseidency in two phases: - programming GICR_VPENDBASER: this still happens in vcpu_load() - checking for the VPT parsing to be complete: this can happen on vcpu entry (in kvm_vgic_flush_hwstate()) This allows the GIC and the CPU to work in parallel, rewmoving some of the entry overhead. Suggested-by: Marc Zyngier <maz@kernel.org> Signed-off-by: Shenming Lu <lushenming@huawei.com> Signed-off-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/20201128141857.983-3-lushenming@huawei.com |
||
|
Marc Zyngier
|
a3f574cd65 |
KVM: arm64: vgic-v4: Plug race between non-residency and v4.1 doorbell
When making a vPE non-resident because it has hit a blocking WFI,
the doorbell can fire at any time after the write to the RD.
Crucially, it can fire right between the write to GICR_VPENDBASER
and the write to the pending_last field in the its_vpe structure.
This means that we would overwrite pending_last with stale data,
and potentially not wakeup until some unrelated event (such as
a timer interrupt) puts the vPE back on the CPU.
GICv4 isn't affected by this as we actively mask the doorbell on
entering the guest, while GICv4.1 automatically manages doorbell
delivery without any hypervisor-driven masking.
Use the vpe_lock to synchronize such update, which solves the
problem altogether.
Fixes:
|
||
|
Marc Zyngier
|
9ed24f4b71 |
KVM: arm64: Move virt/kvm/arm to arch/arm64
Now that the 32bit KVM/arm host is a distant memory, let's move the whole of the KVM/arm64 code into the arm64 tree. As they said in the song: Welcome Home (Sanitarium). Signed-off-by: Marc Zyngier <maz@kernel.org> Acked-by: Will Deacon <will@kernel.org> Link: https://lore.kernel.org/r/20200513104034.74741-1-maz@kernel.org |