kernel.git/arch/arm64/include, branch linux-6.4.y

arm64: sdei: abort running SDEI handlers during crash

2023-09-13T07:48:42Z

commit 5cd474e57368f0957c343bb21e309cf82826b1ef upstream. Interrupts are blocked in SDEI context, per the SDEI spec: "The client interrupts cannot preempt the event handler." If we crashed in the SDEI handler-running context (as with ACPI's AGDI) then we need to clean up the SDEI state before proceeding to the crash kernel so that the crash kernel can have working interrupts. Track the active SDEI handler per-cpu so that we can COMPLETE_AND_RESUME the handler, discarding the interrupted context. Fixes: f5df26961853 ("arm64: kernel: Add arch-specific SDEI entry code and CPU masking") Signed-off-by: D Scott Phillips Cc: stable@vger.kernel.org Reviewed-by: James Morse Tested-by: Mihai Carabas Link: https://lore.kernel.org/r/20230627002939.2758-1-scott@os.amperecomputing.com Signed-off-by: Will Deacon Signed-off-by: Greg Kroah-Hartman

arm64/ptrace: Ensure that SME is set up for target when writing SSVE state

2023-08-23T15:32:53Z

commit 5d0a8d2fba50e9c07cde4aad7fba28c008b07a5b upstream. When we use NT_ARM_SSVE to either enable streaming mode or change the vector length for a process we do not currently do anything to ensure that there is storage allocated for the SME specific register state. If the task had not previously used SME or we changed the vector length then the task will not have had TIF_SME set or backing storage for ZA/ZT allocated, resulting in inconsistent register sizes when saving state and spurious traps which flush the newly set register state. We should set TIF_SME to disable traps and ensure that storage is allocated for ZA and ZT if it is not already allocated. This requires modifying sme_alloc() to make the flush of any existing register state optional so we don't disturb existing state for ZA and ZT. Fixes: e12310a0d30f ("arm64/sme: Implement ptrace support for streaming mode SVE registers") Reported-by: David Spickett Signed-off-by: Mark Brown Cc: # 5.19.x Link: https://lore.kernel.org/r/20230810-arm64-fix-ptrace-race-v1-1-a5361fad2bd6@kernel.org Signed-off-by: Catalin Marinas Signed-off-by: Greg Kroah-Hartman

KVM: arm64: Handle kvm_arm_init failure correctly in finalize_pkvm

2023-08-03T08:25:41Z

[ Upstream commit fa729bc7c9c8c17a2481358c841ef8ca920485d3 ] Currently there is no synchronisation between finalize_pkvm() and kvm_arm_init() initcalls. The finalize_pkvm() proceeds happily even if kvm_arm_init() fails resulting in the following warning on all the CPUs and eventually a HYP panic: | kvm [1]: IPA Size Limit: 48 bits | kvm [1]: Failed to init hyp memory protection | kvm [1]: error initializing Hyp mode: -22 | | | | WARNING: CPU: 0 PID: 0 at arch/arm64/kvm/pkvm.c:226 _kvm_host_prot_finalize+0x30/0x50 | Modules linked in: | CPU: 0 PID: 0 Comm: swapper/0 Not tainted 6.4.0 #237 | Hardware name: FVP Base RevC (DT) | pstate: 634020c5 (nZCv daIF +PAN -UAO +TCO +DIT -SSBS BTYPE=--) | pc : _kvm_host_prot_finalize+0x30/0x50 | lr : __flush_smp_call_function_queue+0xd8/0x230 | | Call trace: | _kvm_host_prot_finalize+0x3c/0x50 | on_each_cpu_cond_mask+0x3c/0x6c | pkvm_drop_host_privileges+0x4c/0x78 | finalize_pkvm+0x3c/0x5c | do_one_initcall+0xcc/0x240 | do_initcall_level+0x8c/0xac | do_initcalls+0x54/0x94 | do_basic_setup+0x1c/0x28 | kernel_init_freeable+0x100/0x16c | kernel_init+0x20/0x1a0 | ret_from_fork+0x10/0x20 | Failed to finalize Hyp protection: -22 | dtb=fvp-base-revc.dtb | kvm [95]: nVHE hyp BUG at: arch/arm64/kvm/hyp/nvhe/mem_protect.c:540! | kvm [95]: nVHE call trace: | kvm [95]: [] __kvm_nvhe_hyp_panic+0xac/0xf8 | kvm [95]: [] __kvm_nvhe_handle_host_mem_abort+0x1a0/0x2ac | kvm [95]: [] __kvm_nvhe_handle_trap+0x4c/0x160 | kvm [95]: [] __kvm_nvhe___skip_pauth_save+0x4/0x4 | kvm [95]: ---[ end nVHE call trace ]--- | kvm [95]: Hyp Offset: 0xfffe8db00ffa0000 | Kernel panic - not syncing: HYP panic: | PS:a34023c9 PC:0000f250710b973c ESR:00000000f2000800 | FAR:ffff000800cb00d0 HPFAR:000000000880cb00 PAR:0000000000000000 | VCPU:0000000000000000 | CPU: 3 PID: 95 Comm: kworker/u16:2 Tainted: G W 6.4.0 #237 | Hardware name: FVP Base RevC (DT) | Workqueue: rpciod rpc_async_schedule | Call trace: | dump_backtrace+0xec/0x108 | show_stack+0x18/0x2c | dump_stack_lvl+0x50/0x68 | dump_stack+0x18/0x24 | panic+0x138/0x33c | nvhe_hyp_panic_handler+0x100/0x184 | new_slab+0x23c/0x54c | ___slab_alloc+0x3e4/0x770 | kmem_cache_alloc_node+0x1f0/0x278 | __alloc_skb+0xdc/0x294 | tcp_stream_alloc_skb+0x2c/0xf0 | tcp_sendmsg_locked+0x3d0/0xda4 | tcp_sendmsg+0x38/0x5c | inet_sendmsg+0x44/0x60 | sock_sendmsg+0x1c/0x34 | xprt_sock_sendmsg+0xdc/0x274 | xs_tcp_send_request+0x1ac/0x28c | xprt_transmit+0xcc/0x300 | call_transmit+0x78/0x90 | __rpc_execute+0x114/0x3d8 | rpc_async_schedule+0x28/0x48 | process_one_work+0x1d8/0x314 | worker_thread+0x248/0x474 | kthread+0xfc/0x184 | ret_from_fork+0x10/0x20 | SMP: stopping secondary CPUs | Kernel Offset: 0x57c5cb460000 from 0xffff800080000000 | PHYS_OFFSET: 0x80000000 | CPU features: 0x00000000,1035b7a3,ccfe773f | Memory Limit: none | ---[ end Kernel panic - not syncing: HYP panic: | PS:a34023c9 PC:0000f250710b973c ESR:00000000f2000800 | FAR:ffff000800cb00d0 HPFAR:000000000880cb00 PAR:0000000000000000 | VCPU:0000000000000000 ]--- Fix it by checking for the successfull initialisation of kvm_arm_init() in finalize_pkvm() before proceeding any futher. Fixes: 87727ba2bb05 ("KVM: arm64: Ensure CPU PMU probes before pKVM host de-privilege") Cc: Will Deacon Cc: Marc Zyngier Cc: Oliver Upton Cc: James Morse Cc: Suzuki K Poulose Cc: Zenghui Yu Signed-off-by: Sudeep Holla Acked-by: Marc Zyngier Link: https://lore.kernel.org/r/20230704193243.3300506-1-sudeep.holla@arm.com Signed-off-by: Oliver Upton Signed-off-by: Sasha Levin

arm64: set __exception_irq_entry with __irq_entry as a default

2023-07-27T06:56:47Z

[ Upstream commit f6794950f0e5ba37e3bbedda4d6ab0aad7395dd3 ] filter_irq_stacks() is supposed to cut entries which are related irq entries from its call stack. And in_irqentry_text() which is called by filter_irq_stacks() uses __irqentry_text_start/end symbol to find irq entries in callstack. But it doesn't work correctly as without "CONFIG_FUNCTION_GRAPH_TRACER", arm64 kernel doesn't include gic_handle_irq which is entry point of arm64 irq between __irqentry_text_start and __irqentry_text_end as we discussed in below link. https://lore.kernel.org/all/CACT4Y+aReMGLYua2rCLHgFpS9io5cZC04Q8GLs-uNmrn1ezxYQ@mail.gmail.com/#t This problem can makes unintentional deep call stack entries especially in KASAN enabled situation as below. [ 2479.383395]I[0:launcher-loader: 1719] Stack depot reached limit capacity [ 2479.383538]I[0:launcher-loader: 1719] WARNING: CPU: 0 PID: 1719 at lib/stackdepot.c:129 __stack_depot_save+0x464/0x46c [ 2479.385693]I[0:launcher-loader: 1719] pstate: 624000c5 (nZCv daIF +PAN -UAO +TCO -DIT -SSBS BTYPE=--) [ 2479.385724]I[0:launcher-loader: 1719] pc : __stack_depot_save+0x464/0x46c [ 2479.385751]I[0:launcher-loader: 1719] lr : __stack_depot_save+0x460/0x46c [ 2479.385774]I[0:launcher-loader: 1719] sp : ffffffc0080073c0 [ 2479.385793]I[0:launcher-loader: 1719] x29: ffffffc0080073e0 x28: ffffffd00b78a000 x27: 0000000000000000 [ 2479.385839]I[0:launcher-loader: 1719] x26: 000000000004d1dd x25: ffffff891474f000 x24: 00000000ca64d1dd [ 2479.385882]I[0:launcher-loader: 1719] x23: 0000000000000200 x22: 0000000000000220 x21: 0000000000000040 [ 2479.385925]I[0:launcher-loader: 1719] x20: ffffffc008007440 x19: 0000000000000000 x18: 0000000000000000 [ 2479.385969]I[0:launcher-loader: 1719] x17: 2065726568207475 x16: 000000000000005e x15: 2d2d2d2d2d2d2d20 [ 2479.386013]I[0:launcher-loader: 1719] x14: 5d39313731203a72 x13: 00000000002f6b30 x12: 00000000002f6af8 [ 2479.386057]I[0:launcher-loader: 1719] x11: 00000000ffffffff x10: ffffffb90aacf000 x9 : e8a74a6c16008800 [ 2479.386101]I[0:launcher-loader: 1719] x8 : e8a74a6c16008800 x7 : 00000000002f6b30 x6 : 00000000002f6af8 [ 2479.386145]I[0:launcher-loader: 1719] x5 : ffffffc0080070c8 x4 : ffffffd00b192380 x3 : ffffffd0092b313c [ 2479.386189]I[0:launcher-loader: 1719] x2 : 0000000000000001 x1 : 0000000000000004 x0 : 0000000000000022 [ 2479.386231]I[0:launcher-loader: 1719] Call trace: [ 2479.386248]I[0:launcher-loader: 1719] __stack_depot_save+0x464/0x46c [ 2479.386273]I[0:launcher-loader: 1719] kasan_save_stack+0x58/0x70 [ 2479.386303]I[0:launcher-loader: 1719] save_stack_info+0x34/0x138 [ 2479.386331]I[0:launcher-loader: 1719] kasan_save_free_info+0x18/0x24 [ 2479.386358]I[0:launcher-loader: 1719] ____kasan_slab_free+0x16c/0x170 [ 2479.386385]I[0:launcher-loader: 1719] __kasan_slab_free+0x10/0x20 [ 2479.386410]I[0:launcher-loader: 1719] kmem_cache_free+0x238/0x53c [ 2479.386435]I[0:launcher-loader: 1719] mempool_free_slab+0x1c/0x28 [ 2479.386460]I[0:launcher-loader: 1719] mempool_free+0x7c/0x1a0 [ 2479.386484]I[0:launcher-loader: 1719] bvec_free+0x34/0x80 [ 2479.386514]I[0:launcher-loader: 1719] bio_free+0x60/0x98 [ 2479.386540]I[0:launcher-loader: 1719] bio_put+0x50/0x21c [ 2479.386567]I[0:launcher-loader: 1719] f2fs_write_end_io+0x4ac/0x4d0 [ 2479.386594]I[0:launcher-loader: 1719] bio_endio+0x2dc/0x300 [ 2479.386622]I[0:launcher-loader: 1719] __dm_io_complete+0x324/0x37c [ 2479.386650]I[0:launcher-loader: 1719] dm_io_dec_pending+0x60/0xa4 [ 2479.386676]I[0:launcher-loader: 1719] clone_endio+0xf8/0x2f0 [ 2479.386700]I[0:launcher-loader: 1719] bio_endio+0x2dc/0x300 [ 2479.386727]I[0:launcher-loader: 1719] blk_update_request+0x258/0x63c [ 2479.386754]I[0:launcher-loader: 1719] scsi_end_request+0x50/0x304 [ 2479.386782]I[0:launcher-loader: 1719] scsi_io_completion+0x88/0x160 [ 2479.386808]I[0:launcher-loader: 1719] scsi_finish_command+0x17c/0x194 [ 2479.386833]I[0:launcher-loader: 1719] scsi_complete+0xcc/0x158 [ 2479.386859]I[0:launcher-loader: 1719] blk_mq_complete_request+0x4c/0x5c [ 2479.386885]I[0:launcher-loader: 1719] scsi_done_internal+0xf4/0x1e0 [ 2479.386910]I[0:launcher-loader: 1719] scsi_done+0x14/0x20 [ 2479.386935]I[0:launcher-loader: 1719] ufshcd_compl_one_cqe+0x578/0x71c [ 2479.386963]I[0:launcher-loader: 1719] ufshcd_mcq_poll_cqe_nolock+0xc8/0x150 [ 2479.386991]I[0:launcher-loader: 1719] ufshcd_intr+0x868/0xc0c [ 2479.387017]I[0:launcher-loader: 1719] __handle_irq_event_percpu+0xd0/0x348 [ 2479.387044]I[0:launcher-loader: 1719] handle_irq_event_percpu+0x24/0x74 [ 2479.387068]I[0:launcher-loader: 1719] handle_irq_event+0x74/0xe0 [ 2479.387091]I[0:launcher-loader: 1719] handle_fasteoi_irq+0x174/0x240 [ 2479.387118]I[0:launcher-loader: 1719] handle_irq_desc+0x7c/0x2c0 [ 2479.387147]I[0:launcher-loader: 1719] generic_handle_domain_irq+0x1c/0x28 [ 2479.387174]I[0:launcher-loader: 1719] gic_handle_irq+0x64/0x158 [ 2479.387204]I[0:launcher-loader: 1719] call_on_irq_stack+0x2c/0x54 [ 2479.387231]I[0:launcher-loader: 1719] do_interrupt_handler+0x70/0xa0 [ 2479.387258]I[0:launcher-loader: 1719] el1_interrupt+0x34/0x68 [ 2479.387283]I[0:launcher-loader: 1719] el1h_64_irq_handler+0x18/0x24 [ 2479.387308]I[0:launcher-loader: 1719] el1h_64_irq+0x68/0x6c [ 2479.387332]I[0:launcher-loader: 1719] blk_attempt_bio_merge+0x8/0x170 [ 2479.387356]I[0:launcher-loader: 1719] blk_mq_attempt_bio_merge+0x78/0x98 [ 2479.387383]I[0:launcher-loader: 1719] blk_mq_submit_bio+0x324/0xa40 [ 2479.387409]I[0:launcher-loader: 1719] __submit_bio+0x104/0x138 [ 2479.387436]I[0:launcher-loader: 1719] submit_bio_noacct_nocheck+0x1d0/0x4a0 [ 2479.387462]I[0:launcher-loader: 1719] submit_bio_noacct+0x618/0x804 [ 2479.387487]I[0:launcher-loader: 1719] submit_bio+0x164/0x180 [ 2479.387511]I[0:launcher-loader: 1719] f2fs_submit_read_bio+0xe4/0x1c4 [ 2479.387537]I[0:launcher-loader: 1719] f2fs_mpage_readpages+0x888/0xa4c [ 2479.387563]I[0:launcher-loader: 1719] f2fs_readahead+0xd4/0x19c [ 2479.387587]I[0:launcher-loader: 1719] read_pages+0xb0/0x4ac [ 2479.387614]I[0:launcher-loader: 1719] page_cache_ra_unbounded+0x238/0x288 [ 2479.387642]I[0:launcher-loader: 1719] do_page_cache_ra+0x60/0x6c [ 2479.387669]I[0:launcher-loader: 1719] page_cache_ra_order+0x318/0x364 [ 2479.387695]I[0:launcher-loader: 1719] ondemand_readahead+0x30c/0x3d8 [ 2479.387722]I[0:launcher-loader: 1719] page_cache_sync_ra+0xb4/0xc8 [ 2479.387749]I[0:launcher-loader: 1719] filemap_read+0x268/0xd24 [ 2479.387777]I[0:launcher-loader: 1719] f2fs_file_read_iter+0x1a0/0x62c [ 2479.387806]I[0:launcher-loader: 1719] vfs_read+0x258/0x34c [ 2479.387831]I[0:launcher-loader: 1719] ksys_pread64+0x8c/0xd0 [ 2479.387857]I[0:launcher-loader: 1719] __arm64_sys_pread64+0x48/0x54 [ 2479.387881]I[0:launcher-loader: 1719] invoke_syscall+0x58/0x158 [ 2479.387909]I[0:launcher-loader: 1719] el0_svc_common+0xf0/0x134 [ 2479.387935]I[0:launcher-loader: 1719] do_el0_svc+0x44/0x114 [ 2479.387961]I[0:launcher-loader: 1719] el0_svc+0x2c/0x80 [ 2479.387985]I[0:launcher-loader: 1719] el0t_64_sync_handler+0x48/0x114 [ 2479.388010]I[0:launcher-loader: 1719] el0t_64_sync+0x190/0x194 [ 2479.388038]I[0:launcher-loader: 1719] Kernel panic - not syncing: kernel: panic_on_warn set ... So let's set __exception_irq_entry with __irq_entry as a default. Applying this patch, we can see gic_hande_irq is included in Systemp.map as below. * Before ffffffc008010000 T __do_softirq ffffffc008010000 T __irqentry_text_end ffffffc008010000 T __irqentry_text_start ffffffc008010000 T __softirqentry_text_start ffffffc008010000 T _stext ffffffc00801066c T __softirqentry_text_end ffffffc008010670 T __entry_text_start * After ffffffc008010000 T __irqentry_text_start ffffffc008010000 T _stext ffffffc008010000 t gic_handle_irq ffffffc00801013c t gic_handle_irq ffffffc008010294 T __irqentry_text_end ffffffc008010298 T __do_softirq ffffffc008010298 T __softirqentry_text_start ffffffc008010904 T __softirqentry_text_end ffffffc008010908 T __entry_text_start Signed-off-by: Youngmin Nam Signed-off-by: SEO HOYOUNG Reviewed-by: Mark Rutland Link: https://lore.kernel.org/r/20230424010436.779733-1-youngmin.nam@samsung.com Signed-off-by: Catalin Marinas Signed-off-by: Sasha Levin

KVM: arm64: vgic-v4: Make the doorbell request robust w.r.t preemption

2023-07-27T06:56:43Z

commit b321c31c9b7b309dcde5e8854b741c8e6a9a05f0 upstream. Xiang reports that VMs occasionally fail to boot on GICv4.1 systems when running a preemptible kernel, as it is possible that a vCPU is blocked without requesting a doorbell interrupt. The issue is that any preemption that occurs between vgic_v4_put() and schedule() on the block path will mark the vPE as nonresident and *not* request a doorbell irq. This occurs because when the vcpu thread is resumed on its way to block, vcpu_load() will make the vPE resident again. Once the vcpu actually blocks, we don't request a doorbell anymore, and the vcpu won't be woken up on interrupt delivery. Fix it by tracking that we're entering WFI, and key the doorbell request on that flag. This allows us not to make the vPE resident when going through a preempt/schedule cycle, meaning we don't lose any state. Cc: stable@vger.kernel.org Fixes: 8e01d9a396e6 ("KVM: arm64: vgic-v4: Move the GICv4 residency flow to be driven by vcpu_load/put") Reported-by: Xiang Chen Suggested-by: Zenghui Yu Tested-by: Xiang Chen Co-developed-by: Oliver Upton Signed-off-by: Marc Zyngier Acked-by: Zenghui Yu Link: https://lore.kernel.org/r/20230713070657.3873244-1-maz@kernel.org Signed-off-by: Oliver Upton Signed-off-by: Greg Kroah-Hartman

KVM: arm64: Correctly handle page aging notifiers for unaligned memslot

2023-07-27T06:56:42Z

commit df6556adf27b7372cfcd97e1c0afb0d516c8279f upstream. Userspace is allowed to select any PAGE_SIZE aligned hva to back guest memory. This is even the case with hugepages, although it is a rather suboptimal configuration as PTE level mappings are used at stage-2. The arm64 page aging handlers have an assumption that the specified range is exactly one page/block of memory, which in the aforementioned case is not necessarily true. All together this leads to the WARN() in kvm_age_gfn() firing. However, the WARN is only part of the issue as the table walkers visit at most a single leaf PTE. For hugepage-backed memory in a memslot that isn't hugepage-aligned, page aging entirely misses accesses to the hugepage beyond the first page in the memslot. Add a new walker dedicated to handling page aging MMU notifiers capable of walking a range of PTEs. Convert kvm(_test)_age_gfn() over to the new walker and drop the WARN that caught the issue in the first place. The implementation of this walker was inspired by the test_clear_young() implementation by Yu Zhao [*], but repurposed to address a bug in the existing aging implementation. Cc: stable@vger.kernel.org # v5.15 Fixes: 056aad67f836 ("kvm: arm/arm64: Rework gpa callback handlers") Link: https://lore.kernel.org/kvmarm/20230526234435.662652-6-yuzhao@google.com/ Co-developed-by: Yu Zhao Signed-off-by: Yu Zhao Reported-by: Reiji Watanabe Reviewed-by: Marc Zyngier Reviewed-by: Shaoqin Huang Link: https://lore.kernel.org/r/20230627235405.4069823-1-oliver.upton@linux.dev Signed-off-by: Oliver Upton Signed-off-by: Greg Kroah-Hartman

arm64: sme: Use STR P to clear FFR context field in streaming SVE mode

2023-07-19T14:36:25Z

[ Upstream commit 893b24181b4c4bf1fa2841b1ed192e5413a97cb1 ] The FFR is a predicate register which can vary between 16 and 256 bits in size depending upon the configured vector length. When saving the SVE state in streaming SVE mode, the FFR register is inaccessible and so commit 9f5848665788 ("arm64/sve: Make access to FFR optional") simply clears the FFR field of the in-memory context structure. Unfortunately, it achieves this using an unconditional 8-byte store and so if the SME vector length is anything other than 64 bytes in size we will either fail to clear the entire field or, worse, we will corrupt memory immediately following the structure. This has led to intermittent kfence splats in CI [1] and can trigger kmalloc Redzone corruption messages when running the 'fp-stress' kselftest: | ============================================================================= | BUG kmalloc-1k (Not tainted): kmalloc Redzone overwritten | ----------------------------------------------------------------------------- | | 0xffff000809bf1e22-0xffff000809bf1e27 @offset=7714. First byte 0x0 instead of 0xcc | Allocated in do_sme_acc+0x9c/0x220 age=2613 cpu=1 pid=531 | __kmalloc+0x8c/0xcc | do_sme_acc+0x9c/0x220 | ... Replace the 8-byte store with a store of a predicate register which has been zero-initialised with PFALSE, ensuring that the entire field is cleared in memory. [1] https://lore.kernel.org/r/CA+G9fYtU7HsV0R0dp4XEH5xXHSJFw8KyDf5VQrLLfMxWfxQkag@mail.gmail.com Cc: Mark Brown Cc: Mark Rutland Cc: Naresh Kamboju Fixes: 9f5848665788 ("arm64/sve: Make access to FFR optional") Reported-by: Linux Kernel Functional Testing Signed-off-by: Will Deacon Reviewed-by: Mark Brown Tested-by: Anders Roxell Link: https://lore.kernel.org/r/20230628155605.22296-1-will@kernel.org Signed-off-by: Catalin Marinas Signed-off-by: Sasha Levin

Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm

2023-06-23T00:54:10Z

Pull kvm fixes from Paolo Bonzini: "ARM: - Correctly save/restore PMUSERNR_EL0 when host userspace is using PMU counters directly - Fix GICv2 emulation on GICv3 after the locking rework - Don't use smp_processor_id() in kvm_pmu_probe_armpmu(), and document why Generic: - Avoid setting page table entries pointing to a deleted memslot if a host page table entry is changed concurrently with the deletion" * tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: KVM: Avoid illegal stage2 mapping on invalid memory slot KVM: arm64: Use raw_smp_processor_id() in kvm_pmu_probe_armpmu() KVM: arm64: Restore GICv2-on-GICv3 functionality KVM: arm64: PMU: Don't overwrite PMUSERENR with vcpu loaded KVM: arm64: PMU: Restore the host's PMUSERENR_EL0

KVM: arm64: PMU: Don't overwrite PMUSERENR with vcpu loaded

2023-06-04T16:19:36Z

Currently, with VHE, KVM sets ER, CR, SW and EN bits of PMUSERENR_EL0 to 1 on vcpu_load(), and saves and restores the register value for the host on vcpu_load() and vcpu_put(). If the value of those bits are cleared on a pCPU with a vCPU loaded (armv8pmu_start() would do that when PMU counters are programmed for the guest), PMU access from the guest EL0 might be trapped to the guest EL1 directly regardless of the current PMUSERENR_EL0 value of the vCPU. Fix this by not letting armv8pmu_start() overwrite PMUSERENR_EL0 on the pCPU where PMUSERENR_EL0 for the guest is loaded, and instead updating the saved shadow register value for the host so that the value can be restored on vcpu_put() later. While vcpu_{put,load}() are manipulating PMUSERENR_EL0, disable IRQs to prevent a race condition between these processes and IPIs that attempt to update PMUSERENR_EL0 for the host EL0. Suggested-by: Mark Rutland Suggested-by: Marc Zyngier Fixes: 83a7a4d643d3 ("arm64: perf: Enable PMU counter userspace access for perf event") Signed-off-by: Reiji Watanabe Signed-off-by: Marc Zyngier Link: https://lore.kernel.org/r/20230603025035.3781797-3-reijiw@google.com

Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm

2023-06-04T11:16:53Z

Pull kvm fixes from Paolo Bonzini: "ARM: - Address some fallout of the locking rework, this time affecting the way the vgic is configured - Fix an issue where the page table walker frees a subtree and then proceeds with walking what it has just freed... - Check that a given PA donated to the guest is actually memory (only affecting pKVM) - Correctly handle MTE CMOs by Set/Way - Fix the reported address of a watchpoint forwarded to userspace - Fix the freeing of the root of stage-2 page tables - Stop creating spurious PMU events to perform detection of the default PMU and use the existing PMU list instead x86: - Fix a memslot lookup bug in the NX recovery thread that could theoretically let userspace bypass the NX hugepage mitigation - Fix a s/BLOCKING/PENDING bug in SVM's vNMI support - Account exit stats for fastpath VM-Exits that never leave the super tight run-loop - Fix an out-of-bounds bug in the optimized APIC map code, and add a regression test for the race" * tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: KVM: selftests: Add test for race in kvm_recalculate_apic_map() KVM: x86: Bail from kvm_recalculate_phys_map() if x2APIC ID is out-of-bounds KVM: x86: Account fastpath-only VM-Exits in vCPU stats KVM: SVM: vNMI pending bit is V_NMI_PENDING_MASK not V_NMI_BLOCKING_MASK KVM: x86/mmu: Grab memslot for correct address space in NX recovery worker KVM: arm64: Document default vPMU behavior on heterogeneous systems KVM: arm64: Iterate arm_pmus list to probe for default PMU KVM: arm64: Drop last page ref in kvm_pgtable_stage2_free_removed() KVM: arm64: Populate fault info for watchpoint KVM: arm64: Reload PTE after invoking walker callback on preorder traversal KVM: arm64: Handle trap of tagged Set/Way CMOs arm64: Add missing Set/Way CMO encodings KVM: arm64: Prevent unconditional donation of unmapped regions from the host KVM: arm64: vgic: Fix a comment KVM: arm64: vgic: Fix locking comment KVM: arm64: vgic: Wrap vgic_its_create() with config_lock KVM: arm64: vgic: Fix a circular locking issue