The feature of minimizing overhead of struct page associated with each
HugeTLB page aims to free its vmemmap pages (used as struct page) to save
memory, where is ~14GB/16GB per 1TB HugeTLB pages (2MB/1GB type). In
short, when a HugeTLB page is allocated or freed, the vmemmap array
representing the range associated with the page will need to be remapped.
When a page is allocated, vmemmap pages are freed after remapping. When a
page is freed, previously discarded vmemmap pages must be allocated before
remapping. More implementations and details can be found here [1].
The infrastructure of freeing vmemmap pages associated with each HugeTLB
page is already there, we can easily enable HUGETLB_PAGE_FREE_VMEMMAP for
arm64, the only thing to be fixed is flush_dcache_page() .
flush_dcache_page() need to be adapted to operate on the head page's flags
since the tail vmemmap pages are mapped with read-only after the feature
is enabled (clear operation is not permitted).
There was some discussions about this in the thread [2], but there was no
conclusion in the end. And I copied the concern proposed by Anshuman to
here and explain why those concern is superfluous. It is safe to enable
it for x86_64 as well as arm64.
1st concern:
'''
But what happens when a hot remove section's vmemmap area (which is
being teared down) is nearby another vmemmap area which is either created
or being destroyed for HugeTLB alloc/free purpose. As you mentioned
HugeTLB pages inside the hot remove section might be safe. But what about
other HugeTLB areas whose vmemmap area shares page table entries with
vmemmap entries for a section being hot removed ? Massive HugeTLB alloc
/use/free test cycle using memory just adjacent to a memory hotplug area,
which is always added and removed periodically, should be able to expose
this problem.
'''
Answer: At the time memory is removed, all HugeTLB pages either have been
migrated away or dissolved. So there is no race between memory hot remove
and free_huge_page_vmemmap(). Therefore, HugeTLB pages inside the hot
remove section is safe. Let's talk your question "what about other
HugeTLB areas whose vmemmap area shares page table entries with vmemmap
entries for a section being hot removed ?", the question is not
established. The minimal granularity size of hotplug memory 128MB (on
arm64, 4k base page), any HugeTLB smaller than 128MB is within a section,
then, there is no share PTE page tables between HugeTLB in this section
and ones in other sections and a HugeTLB page could not cross two
sections. In this case, the section cannot be freed. Any HugeTLB bigger
than 128MB (section size) whose vmemmap pages is an integer multiple of
2MB (PMD-mapped). As long as:
1) HugeTLBs are naturally aligned, power-of-two sizes
2) The HugeTLB size >= the section size
3) The HugeTLB size >= the vmemmap leaf mapping size
Then a HugeTLB will not share any leaf page table entries with *anything
else*, but will share intermediate entries. In this case, at the time
memory is removed, all HugeTLB pages either have been migrated away or
dissolved. So there is also no race between memory hot remove and
free_huge_page_vmemmap().
2nd concern:
'''
differently, not sure if ptdump would require any synchronization.
Dumping an wrong value is probably okay but crashing because a page table
entry is being freed after ptdump acquired the pointer is bad. On arm64,
ptdump() is protected against hotremove via [get|put]_online_mems().
'''
Answer: The ptdump should be fine since vmemmap_remap_free() only
exchanges PTEs or splits the PMD entry (which means allocating a PTE page
table). Both operations do not free any page tables (PTE), so ptdump
cannot run into a UAF on any page tables. The worst case is just dumping
an wrong value.
[1] https://lore.kernel.org/all/20210510030027.56044-1-songmuchun@bytedance.com/
[2] https://lore.kernel.org/all/20210518091826.36937-1-songmuchun@bytedance.com/
[songmuchun@bytedance.com: restructure the code comment inside flush_dcache_page()]
Link: https://lkml.kernel.org/r/20220414072646.21910-1-songmuchun@bytedance.com
Link: https://lkml.kernel.org/r/20220331065640.5777-2-songmuchun@bytedance.com
Signed-off-by: Muchun Song <songmuchun@bytedance.com>
Reviewed-by: Barry Song <baohua@kernel.org>
Tested-by: Barry Song <baohua@kernel.org>
Cc: Will Deacon <will@kernel.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: Bodeddula Balasubramaniam <bodeddub@amazon.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Xiongchun Duan <duanxiongchun@bytedance.com>
Cc: Fam Zheng <fam.zheng@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Pull pin control fixes from Linus Walleij:
- Fix some register offsets on Intel Alderlake
- Fix the order the UFS and SDC pins on Qualcomm SM6350
- Fix a build error in Mediatek Moore.
- Fix a pin function table in the Sunplus SP7021.
- Fix some Kconfig and static keywords on the Samsung Tesla FSD SoC.
- Fix up the EOI function for edge triggered IRQs and keep the block
clock enabled for level IRQs in the STM32 driver.
- Fix some bits and order in the Rockchip RK3308 driver.
- Handle the errorpath in the Pistachio driver probe() properly.
* tag 'pinctrl-v5.18-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl:
pinctrl: pistachio: fix use of irq_of_parse_and_map()
pinctrl: stm32: Keep pinctrl block clock enabled when LEVEL IRQ requested
pinctrl: rockchip: sort the rk3308_mux_recalced_data entries
pinctrl: rockchip: fix RK3308 pinmux bits
pinctrl: stm32: Do not call stm32_gpio_get() for edge triggered IRQs in EOI
pinctrl: Fix an error in pin-function table of SP7021
pinctrl: samsung: fix missing GPIOLIB on ARM64 Exynos config
pinctrl: mediatek: moore: Fix build error
pinctrl: qcom: sm6350: fix order of UFS & SDC pins
pinctrl: alderlake: Fix register offsets for ADL-N variant
pinctrl: samsung: staticize fsd_pin_ctrl
Pull powerpc fixes from Michael Ellerman:
- Partly revert a change to our timer_interrupt() that caused lockups
with high res timers disabled.
- Fix a bug in KVM TCE handling that could corrupt kernel memory.
- Two commits fixing Power9/Power10 perf alternative event selection.
Thanks to Alexey Kardashevskiy, Athira Rajeev, David Gibson, Frederic
Barrat, Madhavan Srinivasan, Miguel Ojeda, and Nicholas Piggin.
* tag 'powerpc-5.18-3' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
powerpc/perf: Fix 32bit compile
powerpc/perf: Fix power10 event alternatives
powerpc/perf: Fix power9 event alternatives
KVM: PPC: Fix TCE handling for VFIO
powerpc/time: Always set decrementer in timer_interrupt()
Pull perf fixes from Borislav Petkov:
- Add Sapphire Rapids CPU support
- Fix a perf vmalloc-ed buffer mapping error (PERF_USE_VMALLOC in use)
* tag 'perf_urgent_for_v5.18_rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/x86/cstate: Add SAPPHIRERAPIDS_X CPU support
perf/core: Fix perf_mmap fail when CONFIG_PERF_USE_VMALLOC enabled
Pull ARC fixes from Vineet Gupta:
- Assorted fixes
* tag 'arc-5.18-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc:
ARC: remove redundant READ_ONCE() in cmpxchg loop
ARC: atomic: cleanup atomic-llsc definitions
arc: drop definitions of pgd_index() and pgd_offset{, _k}() entirely
ARC: dts: align SPI NOR node name with dtschema
ARC: Remove a redundant memset()
ARC: fix typos in comments
ARC: entry: fix syscall_trace_exit argument
Pull xen fixes from Juergen Gross:
"A simple cleanup patch and a refcount fix for Xen on Arm"
* tag 'for-linus-5.18-rc4-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
arm/xen: Fix some refcount leaks
xen: Convert kmap() to kmap_local_page()
Add a struct page forward declaration to cacheflush_32.h.
Fixes this build warning:
CC drivers/crypto/xilinx/zynqmp-sha.o
In file included from arch/sparc/include/asm/cacheflush.h:11,
from include/linux/cacheflush.h:5,
from drivers/crypto/xilinx/zynqmp-sha.c:6:
arch/sparc/include/asm/cacheflush_32.h:38:37: warning: 'struct page' declared inside parameter list will not be visible outside of this definition or declaration
38 | void sparc_flush_page_to_ram(struct page *page);
Exposed by commit 0e03b8fd29 ("crypto: xilinx - Turn SHA into a
tristate and allow COMPILE_TEST") but not Fixes: that commit because the
underlying problem is older.
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Reported-by: kernel test robot <lkp@intel.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: David S. Miller <davem@davemloft.net>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: sparclinux@vger.kernel.org
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull kvm fixes from Paolo Bonzini:
"The main and larger change here is a workaround for AMD's lack of
cache coherency for encrypted-memory guests.
I have another patch pending, but it's waiting for review from the
architecture maintainers.
RISC-V:
- Remove 's' & 'u' as valid ISA extension
- Do not allow disabling the base extensions 'i'/'m'/'a'/'c'
x86:
- Fix NMI watchdog in guests on AMD
- Fix for SEV cache incoherency issues
- Don't re-acquire SRCU lock in complete_emulated_io()
- Avoid NULL pointer deref if VM creation fails
- Fix race conditions between APICv disabling and vCPU creation
- Bugfixes for disabling of APICv
- Preserve BSP MSR_KVM_POLL_CONTROL across suspend/resume
selftests:
- Do not use bitfields larger than 32-bits, they differ between GCC
and clang"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
kvm: selftests: introduce and use more page size-related constants
kvm: selftests: do not use bitfields larger than 32-bits for PTEs
KVM: SEV: add cache flush to solve SEV cache incoherency issues
KVM: SVM: Flush when freeing encrypted pages even on SME_COHERENT CPUs
KVM: SVM: Simplify and harden helper to flush SEV guest page(s)
KVM: selftests: Silence compiler warning in the kvm_page_table_test
KVM: x86/pmu: Update AMD PMC sample period to fix guest NMI-watchdog
x86/kvm: Preserve BSP MSR_KVM_POLL_CONTROL across suspend/resume
KVM: SPDX style and spelling fixes
KVM: x86: Skip KVM_GUESTDBG_BLOCKIRQ APICv update if APICv is disabled
KVM: x86: Pend KVM_REQ_APICV_UPDATE during vCPU creation to fix a race
KVM: nVMX: Defer APICv updates while L2 is active until L1 is active
KVM: x86: Tag APICv DISABLE inhibit, not ABSENT, if APICv is disabled
KVM: Initialize debugfs_dentry when a VM is created to avoid NULL deref
KVM: Add helpers to wrap vcpu->srcu_idx and yell if it's abused
KVM: RISC-V: Use kvm_vcpu.srcu_idx, drop RISC-V's unnecessary copy
KVM: x86: Don't re-acquire SRCU lock in complete_emulated_io()
RISC-V: KVM: Restrict the extensions that can be disabled
RISC-V: KVM: Remove 's' & 'u' as valid ISA extension
Pull RISC-V fixes Palmer Dabbelt:
- A pair of build fixes for the recent cpuidle driver
- A fix for systems without sv57 that manifests as a crash
early in boot
* tag 'riscv-for-linus-5.18-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:
RISC-V: cpuidle: fix Kconfig select for RISCV_SBI_CPUIDLE
RISC-V: mm: Fix set_satp_mode() for platform not having Sv57
cpuidle: riscv: support non-SMP config
Pull arm64 fixes from Will Deacon:
"There's no real pattern to the fixes, but the main one fixes our
pmd_leaf() definition to resolve a NULL dereference on the migration
path.
- Fix PMU event validation in the absence of any event counters
- Fix allmodconfig build using clang in conjunction with binutils
- Fix definitions of pXd_leaf() to handle PROT_NONE entries
- More typo fixes"
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
arm64: mm: fix p?d_leaf()
arm64: fix typos in comments
arm64: Improve HAVE_DYNAMIC_FTRACE_WITH_REGS selection for clang
arm_pmu: Validate single/group leader events
The of_find_compatible_node() function returns a node pointer with
refcount incremented, We should use of_node_put() on it when done
Add the missing of_node_put() to release the refcount.
Fixes: 9b08aaa319 ("ARM: XEN: Move xen_early_init() before efi_init()")
Fixes: b2371587fe ("arm/xen: Read extended regions from DT and init Xen resource")
Signed-off-by: Miaoqian Lin <linmq006@gmail.com>
Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
Signed-off-by: Stefano Stabellini <stefano.stabellini@xilinx.com>
There can be lots of build errors when building cpuidle-riscv-sbi.o.
They are all caused by a kconfig problem with this warning:
WARNING: unmet direct dependencies detected for RISCV_SBI_CPUIDLE
Depends on [n]: CPU_IDLE [=y] && RISCV [=y] && RISCV_SBI [=n]
Selected by [y]:
- SOC_VIRT [=y] && CPU_IDLE [=y]
so make the 'select' of RISCV_SBI_CPUIDLE also depend on RISCV_SBI.
Fixes: c5179ef1ca ("RISC-V: Enable RISC-V SBI CPU Idle driver for QEMU virt machine")
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Reported-by: kernel test robot <lkp@intel.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
Cc: stable@vger.kernel.org
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
When Sv57 is not available the satp.MODE test in set_satp_mode() will
fail and lead to pgdir re-programming for Sv48. The pgdir re-programming
will fail as well due to pre-existing pgdir entry used for Sv57 and as
a result kernel fails to boot on RISC-V platform not having Sv57.
To fix above issue, we should clear the pgdir memory in set_satp_mode()
before re-programming.
Fixes: 011f09d120 ("riscv: mm: Set sv57 on defaultly")
Reported-by: Mayuresh Chitale <mchitale@ventanamicro.com>
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
Reviewed-by: Atish Patra <atishp@rivosinc.com>
Cc: stable@vger.kernel.org
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
Flush the CPU caches when memory is reclaimed from an SEV guest (where
reclaim also includes it being unmapped from KVM's memslots). Due to lack
of coherency for SEV encrypted memory, failure to flush results in silent
data corruption if userspace is malicious/broken and doesn't ensure SEV
guest memory is properly pinned and unpinned.
Cache coherency is not enforced across the VM boundary in SEV (AMD APM
vol.2 Section 15.34.7). Confidential cachelines, generated by confidential
VM guests have to be explicitly flushed on the host side. If a memory page
containing dirty confidential cachelines was released by VM and reallocated
to another user, the cachelines may corrupt the new user at a later time.
KVM takes a shortcut by assuming all confidential memory remain pinned
until the end of VM lifetime. Therefore, KVM does not flush cache at
mmu_notifier invalidation events. Because of this incorrect assumption and
the lack of cache flushing, malicous userspace can crash the host kernel:
creating a malicious VM and continuously allocates/releases unpinned
confidential memory pages when the VM is running.
Add cache flush operations to mmu_notifier operations to ensure that any
physical memory leaving the guest VM get flushed. In particular, hook
mmu_notifier_invalidate_range_start and mmu_notifier_release events and
flush cache accordingly. The hook after releasing the mmu lock to avoid
contention with other vCPUs.
Cc: stable@vger.kernel.org
Suggested-by: Sean Christpherson <seanjc@google.com>
Reported-by: Mingwei Zhang <mizhang@google.com>
Signed-off-by: Mingwei Zhang <mizhang@google.com>
Message-Id: <20220421031407.2516575-4-mizhang@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Use clflush_cache_range() to flush the confidential memory when
SME_COHERENT is supported in AMD CPU. Cache flush is still needed since
SME_COHERENT only support cache invalidation at CPU side. All confidential
cache lines are still incoherent with DMA devices.
Cc: stable@vger.kerel.org
Fixes: add5e2f045 ("KVM: SVM: Add support for the SEV-ES VMSA")
Reviewed-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Mingwei Zhang <mizhang@google.com>
Message-Id: <20220421031407.2516575-3-mizhang@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Rework sev_flush_guest_memory() to explicitly handle only a single page,
and harden it to fall back to WBINVD if VM_PAGE_FLUSH fails. Per-page
flushing is currently used only to flush the VMSA, and in its current
form, the helper is completely broken with respect to flushing actual
guest memory, i.e. won't work correctly for an arbitrary memory range.
VM_PAGE_FLUSH takes a host virtual address, and is subject to normal page
walks, i.e. will fault if the address is not present in the host page
tables or does not have the correct permissions. Current AMD CPUs also
do not honor SMAP overrides (undocumented in kernel versions of the APM),
so passing in a userspace address is completely out of the question. In
other words, KVM would need to manually walk the host page tables to get
the pfn, ensure the pfn is stable, and then use the direct map to invoke
VM_PAGE_FLUSH. And the latter might not even work, e.g. if userspace is
particularly evil/clever and backs the guest with Secret Memory (which
unmaps memory from the direct map).
Signed-off-by: Sean Christopherson <seanjc@google.com>
Fixes: add5e2f045 ("KVM: SVM: Add support for the SEV-ES VMSA")
Reported-by: Mingwei Zhang <mizhang@google.com>
Cc: stable@vger.kernel.org
Signed-off-by: Mingwei Zhang <mizhang@google.com>
Message-Id: <20220421031407.2516575-2-mizhang@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
NMI-watchdog is one of the favorite features of kernel developers,
but it does not work in AMD guest even with vPMU enabled and worse,
the system misrepresents this capability via /proc.
This is a PMC emulation error. KVM does not pass the latest valid
value to perf_event in time when guest NMI-watchdog is running, thus
the perf_event corresponding to the watchdog counter will enter the
old state at some point after the first guest NMI injection, forcing
the hardware register PMC0 to be constantly written to 0x800000000001.
Meanwhile, the running counter should accurately reflect its new value
based on the latest coordinated pmc->counter (from vPMC's point of view)
rather than the value written directly by the guest.
Fixes: 168d918f26 ("KVM: x86: Adjust counter sample period after a wrmsr")
Reported-by: Dongli Cao <caodongli@kingsoft.com>
Signed-off-by: Like Xu <likexu@tencent.com>
Reviewed-by: Yanan Wang <wangyanan55@huawei.com>
Tested-by: Yanan Wang <wangyanan55@huawei.com>
Reviewed-by: Jim Mattson <jmattson@google.com>
Message-Id: <20220409015226.38619-1-likexu@tencent.com>
Cc: stable@vger.kernel.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
MSR_KVM_POLL_CONTROL is cleared on reset, thus reverting guests to
host-side polling after suspend/resume. Non-bootstrap CPUs are
restored correctly by the haltpoll driver because they are hot-unplugged
during suspend and hot-plugged during resume; however, the BSP
is not hotpluggable and remains in host-sde polling mode after
the guest resume. The makes the guest pay for the cost of vmexits
every time the guest enters idle.
Fix it by recording BSP's haltpoll state and resuming it during guest
resume.
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
Message-Id: <1650267752-46796-1-git-send-email-wanpengli@tencent.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Skip the APICv inhibit update for KVM_GUESTDBG_BLOCKIRQ if APICv is
disabled at the module level to avoid having to acquire the mutex and
potentially process all vCPUs. The DISABLE inhibit will (barring bugs)
never be lifted, so piling on more inhibits is unnecessary.
Fixes: cae72dcc3b ("KVM: x86: inhibit APICv when KVM_GUESTDBG_BLOCKIRQ active")
Cc: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20220420013732.3308816-5-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Make a KVM_REQ_APICV_UPDATE request when creating a vCPU with an
in-kernel local APIC and APICv enabled at the module level. Consuming
kvm_apicv_activated() and stuffing vcpu->arch.apicv_active directly can
race with __kvm_set_or_clear_apicv_inhibit(), as vCPU creation happens
before the vCPU is fully onlined, i.e. it won't get the request made to
"all" vCPUs. If APICv is globally inhibited between setting apicv_active
and onlining the vCPU, the vCPU will end up running with APICv enabled
and trigger KVM's sanity check.
Mark APICv as active during vCPU creation if APICv is enabled at the
module level, both to be optimistic about it's final state, e.g. to avoid
additional VMWRITEs on VMX, and because there are likely bugs lurking
since KVM checks apicv_active in multiple vCPU creation paths. While
keeping the current behavior of consuming kvm_apicv_activated() is
arguably safer from a regression perspective, force apicv_active so that
vCPU creation runs with deterministic state and so that if there are bugs,
they are found sooner than later, i.e. not when some crazy race condition
is hit.
WARNING: CPU: 0 PID: 484 at arch/x86/kvm/x86.c:9877 vcpu_enter_guest+0x2ae3/0x3ee0 arch/x86/kvm/x86.c:9877
Modules linked in:
CPU: 0 PID: 484 Comm: syz-executor361 Not tainted 5.16.13 #2
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1ubuntu1~cloud0 04/01/2014
RIP: 0010:vcpu_enter_guest+0x2ae3/0x3ee0 arch/x86/kvm/x86.c:9877
Call Trace:
<TASK>
vcpu_run arch/x86/kvm/x86.c:10039 [inline]
kvm_arch_vcpu_ioctl_run+0x337/0x15e0 arch/x86/kvm/x86.c:10234
kvm_vcpu_ioctl+0x4d2/0xc80 arch/x86/kvm/../../../virt/kvm/kvm_main.c:3727
vfs_ioctl fs/ioctl.c:51 [inline]
__do_sys_ioctl fs/ioctl.c:874 [inline]
__se_sys_ioctl fs/ioctl.c:860 [inline]
__x64_sys_ioctl+0x16d/0x1d0 fs/ioctl.c:860
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x38/0x90 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x44/0xae
The bug was hit by a syzkaller spamming VM creation with 2 vCPUs and a
call to KVM_SET_GUEST_DEBUG.
r0 = openat$kvm(0xffffffffffffff9c, &(0x7f0000000000), 0x0, 0x0)
r1 = ioctl$KVM_CREATE_VM(r0, 0xae01, 0x0)
ioctl$KVM_CAP_SPLIT_IRQCHIP(r1, 0x4068aea3, &(0x7f0000000000)) (async)
r2 = ioctl$KVM_CREATE_VCPU(r1, 0xae41, 0x0) (async)
r3 = ioctl$KVM_CREATE_VCPU(r1, 0xae41, 0x400000000000002)
ioctl$KVM_SET_GUEST_DEBUG(r3, 0x4048ae9b, &(0x7f00000000c0)={0x5dda9c14aa95f5c5})
ioctl$KVM_RUN(r2, 0xae80, 0x0)
Reported-by: Gaoning Pan <pgn@zju.edu.cn>
Reported-by: Yongkang Jia <kangel@zju.edu.cn>
Fixes: 8df14af42f ("kvm: x86: Add support for dynamic APICv activation")
Cc: stable@vger.kernel.org
Cc: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20220420013732.3308816-4-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Defer APICv updates that occur while L2 is active until nested VM-Exit,
i.e. until L1 regains control. vmx_refresh_apicv_exec_ctrl() assumes L1
is active and (a) stomps all over vmcs02 and (b) neglects to ever updated
vmcs01. E.g. if vmcs12 doesn't enable the TPR shadow for L2 (and thus no
APICv controls), L1 performs nested VM-Enter APICv inhibited, and APICv
becomes unhibited while L2 is active, KVM will set various APICv controls
in vmcs02 and trigger a failed VM-Entry. The kicker is that, unless
running with nested_early_check=1, KVM blames L1 and chaos ensues.
In all cases, ignoring vmcs02 and always deferring the inhibition change
to vmcs01 is correct (or at least acceptable). The ABSENT and DISABLE
inhibitions cannot truly change while L2 is active (see below).
IRQ_BLOCKING can change, but it is firmly a best effort debug feature.
Furthermore, only L2's APIC is accelerated/virtualized to the full extent
possible, e.g. even if L1 passes through its APIC to L2, normal MMIO/MSR
interception will apply to the virtual APIC managed by KVM.
The exception is the SELF_IPI register when x2APIC is enabled, but that's
an acceptable hole.
Lastly, Hyper-V's Auto EOI can technically be toggled if L1 exposes the
MSRs to L2, but for that to work in any sane capacity, L1 would need to
pass through IRQs to L2 as well, and IRQs must be intercepted to enable
virtual interrupt delivery. I.e. exposing Auto EOI to L2 and enabling
VID for L2 are, for all intents and purposes, mutually exclusive.
Lack of dynamic toggling is also why this scenario is all but impossible
to encounter in KVM's current form. But a future patch will pend an
APICv update request _during_ vCPU creation to plug a race where a vCPU
that's being created doesn't get included in the "all vCPUs request"
because it's not yet visible to other vCPUs. If userspaces restores L2
after VM creation (hello, KVM selftests), the first KVM_RUN will occur
while L2 is active and thus service the APICv update request made during
VM creation.
Cc: stable@vger.kernel.org
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220420013732.3308816-3-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Set the DISABLE inhibit, not the ABSENT inhibit, if APICv is disabled via
module param. A recent refactoring to add a wrapper for setting/clearing
inhibits unintentionally changed the flag, probably due to a copy+paste
goof.
Fixes: 4f4c4a3ee5 ("KVM: x86: Trace all APICv inhibit changes and capture overall status")
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20220420013732.3308816-2-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add wrappers to acquire/release KVM's SRCU lock when stashing the index
in vcpu->src_idx, along with rudimentary detection of illegal usage,
e.g. re-acquiring SRCU and thus overwriting vcpu->src_idx. Because the
SRCU index is (currently) either 0 or 1, illegal nesting bugs can go
unnoticed for quite some time and only cause problems when the nested
lock happens to get a different index.
Wrap the WARNs in PROVE_RCU=y, and make them ONCE, otherwise KVM will
likely yell so loudly that it will bring the kernel to its knees.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Tested-by: Fabiano Rosas <farosas@linux.ibm.com>
Message-Id: <20220415004343.2203171-4-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Use the generic kvm_vcpu's srcu_idx instead of using an indentical field
in RISC-V's version of kvm_vcpu_arch. Generic KVM very intentionally
does not touch vcpu->srcu_idx, i.e. there's zero chance of running afoul
of common code.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220415004343.2203171-3-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Don't re-acquire SRCU in complete_emulated_io() now that KVM acquires the
lock in kvm_arch_vcpu_ioctl_run(). More importantly, don't overwrite
vcpu->srcu_idx. If the index acquired by complete_emulated_io() differs
from the one acquired by kvm_arch_vcpu_ioctl_run(), KVM will effectively
leak a lock and hang if/when synchronize_srcu() is invoked for the
relevant grace period.
Fixes: 8d25b7beca ("KVM: x86: pull kvm->srcu read-side to kvm_arch_vcpu_ioctl_run")
Cc: stable@vger.kernel.org
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20220415004343.2203171-2-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The "read_bhrb" global symbol is only called under CONFIG_PPC64 of
arch/powerpc/perf/core-book3s.c but it is compiled for both 32 and 64 bit
anyway (and LLVM fails to link this on 32bit).
This fixes it by moving bhrb.o to obj64 targets.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220421025756.571995-1-aik@ozlabs.ru
When scheduling a group of events, there are constraint checks done to
make sure all events can go in a group. Example, one of the criteria is
that events in a group cannot use the same PMC. But platform specific
PMU supports alternative event for some of the event codes. During
perf_event_open(), if any event group doesn't match constraint check
criteria, further lookup is done to find alternative event.
By current design, the array of alternatives events in PMU code is
expected to be sorted by column 0. This is because in
find_alternative() the return criteria is based on event code
comparison. ie. "event < ev_alt[i][0])". This optimisation is there
since find_alternative() can be called multiple times. In power10 PMU
code, the alternative event array is not sorted properly and hence there
is breakage in finding alternative event.
To work with existing logic, fix the alternative event array to be
sorted by column 0 for power10-pmu.c
Results:
In case where an alternative event is not chosen when we could, events
will be multiplexed. ie, time sliced where it could actually run
concurrently.
Example, in power10 PM_INST_CMPL_ALT(0x00002) has alternative event,
PM_INST_CMPL(0x500fa). Without the fix, if a group of events with PMC1
to PMC4 is used along with PM_INST_CMPL_ALT, it will be time sliced
since all programmable PMC's are consumed already. But with the fix,
when it picks alternative event on PMC5, all events will run
concurrently.
Before:
# perf stat -e r00002,r100fc,r200fa,r300fc,r400fc
Performance counter stats for 'system wide':
328668935 r00002 (79.94%)
56501024 r100fc (79.95%)
49564238 r200fa (79.95%)
376 r300fc (80.19%)
660 r400fc (79.97%)
4.039150522 seconds time elapsed
With the fix, since alternative event is chosen to run on PMC6, events
will be run concurrently.
After:
# perf stat -e r00002,r100fc,r200fa,r300fc,r400fc
Performance counter stats for 'system wide':
23596607 r00002
4907738 r100fc
2283608 r200fa
135 r300fc
248 r400fc
1.664671390 seconds time elapsed
Fixes: a64e697cef ("powerpc/perf: power10 Performance Monitoring support")
Signed-off-by: Athira Rajeev <atrajeev@linux.vnet.ibm.com>
Reviewed-by: Madhavan Srinivasan <maddy@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220419114828.89843-2-atrajeev@linux.vnet.ibm.com
When scheduling a group of events, there are constraint checks done to
make sure all events can go in a group. Example, one of the criteria is
that events in a group cannot use the same PMC. But platform specific
PMU supports alternative event for some of the event codes. During
perf_event_open(), if any event group doesn't match constraint check
criteria, further lookup is done to find alternative event.
By current design, the array of alternatives events in PMU code is
expected to be sorted by column 0. This is because in
find_alternative() the return criteria is based on event code
comparison. ie. "event < ev_alt[i][0])". This optimisation is there
since find_alternative() can be called multiple times. In power9 PMU
code, the alternative event array is not sorted properly and hence there
is breakage in finding alternative events.
To work with existing logic, fix the alternative event array to be
sorted by column 0 for power9-pmu.c
Results:
With alternative events, multiplexing can be avoided. That is, for
example, in power9 PM_LD_MISS_L1 (0x3e054) has alternative event,
PM_LD_MISS_L1_ALT (0x400f0). This is an identical event which can be
programmed in a different PMC.
Before:
# perf stat -e r3e054,r300fc
Performance counter stats for 'system wide':
1057860 r3e054 (50.21%)
379 r300fc (49.79%)
0.944329741 seconds time elapsed
Since both the events are using PMC3 in this case, they are
multiplexed here.
After:
# perf stat -e r3e054,r300fc
Performance counter stats for 'system wide':
1006948 r3e054
182 r300fc
Fixes: 91e0bd1e62 ("powerpc/perf: Add PM_LD_MISS_L1 and PM_BR_2PATH to power9 event list")
Signed-off-by: Athira Rajeev <atrajeev@linux.vnet.ibm.com>
Reviewed-by: Madhavan Srinivasan <maddy@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220419114828.89843-1-atrajeev@linux.vnet.ibm.com
The LoPAPR spec defines a guest visible IOMMU with a variable page size.
Currently QEMU advertises 4K, 64K, 2M, 16MB pages, a Linux VM picks
the biggest (16MB). In the case of a passed though PCI device, there is
a hardware IOMMU which does not support all pages sizes from the above -
P8 cannot do 2MB and P9 cannot do 16MB. So for each emulated
16M IOMMU page we may create several smaller mappings ("TCEs") in
the hardware IOMMU.
The code wrongly uses the emulated TCE index instead of hardware TCE
index in error handling. The problem is easier to see on POWER8 with
multi-level TCE tables (when only the first level is preallocated)
as hash mode uses real mode TCE hypercalls handlers.
The kernel starts using indirect tables when VMs get bigger than 128GB
(depends on the max page order).
The very first real mode hcall is going to fail with H_TOO_HARD as
in the real mode we cannot allocate memory for TCEs (we can in the virtual
mode) but on the way out the code attempts to clear hardware TCEs using
emulated TCE indexes which corrupts random kernel memory because
it_offset==1<<59 is subtracted from those indexes and the resulting index
is out of the TCE table bounds.
This fixes kvmppc_clear_tce() to use the correct TCE indexes.
While at it, this fixes TCE cache invalidation which uses emulated TCE
indexes instead of the hardware ones. This went unnoticed as 64bit DMA
is used these days and VMs map all RAM in one go and only then do DMA
and this is when the TCE cache gets populated.
Potentially this could slow down mapping, however normally 16MB
emulated pages are backed by 64K hardware pages so it is one write to
the "TCE Kill" per 256 updates which is not that bad considering the size
of the cache (1024 TCEs or so).
Fixes: ca1fc489cf ("KVM: PPC: Book3S: Allow backing bigger guest IOMMU pages with smaller physical pages")
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Tested-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Frederic Barrat <fbarrat@linux.ibm.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220420050840.328223-1-aik@ozlabs.ru
The Samsung pinctrl drivers depend on OF_GPIO, which is part of GPIOLIB.
ARMv7 Exynos platform selects GPIOLIB and Samsung pinctrl drivers. ARMv8
Exynos selects only the latter leading to possible wrong configuration
on ARMv8 build:
WARNING: unmet direct dependencies detected for PINCTRL_EXYNOS
Depends on [n]: PINCTRL [=y] && OF_GPIO [=n] && (ARCH_EXYNOS [=y] || ARCH_S5PV210 || COMPILE_TEST [=y])
Selected by [y]:
- ARCH_EXYNOS [=y]
Always select the GPIOLIB from the Samsung pinctrl drivers to fix the
issue. This requires removing of OF_GPIO dependency (to avoid recursive
dependency), so add dependency on OF for COMPILE_TEST cases.
Reported-by: Necip Fazil Yildiran <fazilyildiran@gmail.com>
Fixes: eed6b3eb20 ("arm64: Split out platform options to separate Kconfig")
Cc: <stable@vger.kernel.org>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20220420141407.470955-1-krzysztof.kozlowski@linaro.org
This is a partial revert of commit 0faf20a1ad ("powerpc/64s/interrupt:
Don't enable MSR[EE] in irq handlers unless perf is in use").
Prior to that commit, we always set the decrementer in
timer_interrupt(), to clear the timer interrupt. Otherwise we could end
up continuously taking timer interrupts.
When high res timers are enabled there is no problem seen with leaving
the decrementer untouched in timer_interrupt(), because it will be
programmed via hrtimer_interrupt() -> tick_program_event() ->
clockevents_program_event() -> decrementer_set_next_event().
However with CONFIG_HIGH_RES_TIMERS=n or booting with highres=off, we
see a stall/lockup, because tick_nohz_handler() does not cause a
reprogram of the decrementer, leading to endless timer interrupts.
Example trace:
[ 1.898617][ T7] Freeing initrd memory: 2624K^M
[ 22.680919][ C1] rcu: INFO: rcu_sched detected stalls on CPUs/tasks:^M
[ 22.682281][ C1] rcu: 0-....: (25 ticks this GP) idle=073/0/0x1 softirq=10/16 fqs=1050 ^M
[ 22.682851][ C1] (detected by 1, t=2102 jiffies, g=-1179, q=476)^M
[ 22.683649][ C1] Sending NMI from CPU 1 to CPUs 0:^M
[ 22.685252][ C0] NMI backtrace for cpu 0^M
[ 22.685649][ C0] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.16.0-rc2-00185-g0faf20a1ad16 #145^M
[ 22.686393][ C0] NIP: c000000000016d64 LR: c000000000f6cca4 CTR: c00000000019c6e0^M
[ 22.686774][ C0] REGS: c000000002833590 TRAP: 0500 Not tainted (5.16.0-rc2-00185-g0faf20a1ad16)^M
[ 22.687222][ C0] MSR: 8000000000009033 <SF,EE,ME,IR,DR,RI,LE> CR: 24000222 XER: 00000000^M
[ 22.688297][ C0] CFAR: c00000000000c854 IRQMASK: 0 ^M
...
[ 22.692637][ C0] NIP [c000000000016d64] arch_local_irq_restore+0x174/0x250^M
[ 22.694443][ C0] LR [c000000000f6cca4] __do_softirq+0xe4/0x3dc^M
[ 22.695762][ C0] Call Trace:^M
[ 22.696050][ C0] [c000000002833830] [c000000000f6cc80] __do_softirq+0xc0/0x3dc (unreliable)^M
[ 22.697377][ C0] [c000000002833920] [c000000000151508] __irq_exit_rcu+0xd8/0x130^M
[ 22.698739][ C0] [c000000002833950] [c000000000151730] irq_exit+0x20/0x40^M
[ 22.699938][ C0] [c000000002833970] [c000000000027f40] timer_interrupt+0x270/0x460^M
[ 22.701119][ C0] [c0000000028339d0] [c0000000000099a8] decrementer_common_virt+0x208/0x210^M
Possibly this should be fixed in the lowres timing code, but that would
be a generic change and could take some time and may not backport
easily, so for now make the programming of the decrementer unconditional
again in timer_interrupt() to avoid the stall/lockup.
Fixes: 0faf20a1ad ("powerpc/64s/interrupt: Don't enable MSR[EE] in irq handlers unless perf is in use")
Reported-by: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Link: https://lore.kernel.org/r/20220420141657.771442-1-mpe@ellerman.id.au
Pull xtensa fixes from Max Filippov:
- fix patching CPU selection in patch_text
- fix potential deadlock in ISS platform serial driver
- fix potential register clobbering in coprocessor exception handler
* tag 'xtensa-20220416' of https://github.com/jcmvbkbc/linux-xtensa:
xtensa: fix a7 clobbering in coprocessor context load/store
arch: xtensa: platforms: Fix deadlock in rs_close()
xtensa: patch_text: Fixup last cpu should be master
The first "if" condition in __memcpy_flushcache is supposed to align the
"dest" variable to 8 bytes and copy data up to this alignment. However,
this condition may misbehave if "size" is greater than 4GiB.
The statement min_t(unsigned, size, ALIGN(dest, 8) - dest); casts both
arguments to unsigned int and selects the smaller one. However, the
cast truncates high bits in "size" and it results in misbehavior.
For example:
suppose that size == 0x100000001, dest == 0x200000002
min_t(unsigned, size, ALIGN(dest, 8) - dest) == min_t(0x1, 0xe) == 0x1;
...
dest += 0x1;
so we copy just one byte "and" dest remains unaligned.
This patch fixes the bug by replacing unsigned with size_t.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Currently, the config isa register allows us to disable all allowed
single letter ISA extensions. It shouldn't be the case as vmm shouldn't
be able to disable base extensions (imac).
These extensions should always be enabled as long as they are enabled
in the host ISA.
Signed-off-by: Atish Patra <atishp@rivosinc.com>
Signed-off-by: Anup Patel <anup@brainfault.org>
Fixes: 92ad82002c ("RISC-V: KVM: Implement
KVM_GET_ONE_REG/KVM_SET_ONE_REG ioctls")
There are no ISA extension defined as 's' & 'u' in RISC-V specifications.
The misa register defines 's' & 'u' bit as Supervisor/User privilege mode
enabled. But it should not appear in the ISA extension in the device tree.
Remove those from the allowed ISA extension for kvm.
Fixes: a33c72faf2 ("RISC-V: KVM: Implement VCPU create, init and
destroy functions")
Signed-off-by: Atish Patra <atishp@rivosinc.com>
Signed-off-by: Anup Patel <anup@brainfault.org>
Huge page backed vmalloc memory could benefit performance in many cases.
However, some users of vmalloc may not be ready to handle huge pages for
various reasons: hardware constraints, potential pages split, etc.
VM_NO_HUGE_VMAP was introduced to allow vmalloc users to opt-out huge
pages. However, it is not easy to track down all the users that require
the opt-out, as the allocation are passed different stacks and may cause
issues in different layers.
To address this issue, replace VM_NO_HUGE_VMAP with an opt-in flag,
VM_ALLOW_HUGE_VMAP, so that users that benefit from huge pages could ask
specificially.
Also, remove vmalloc_no_huge() and add opt-in helper vmalloc_huge().
Fixes: fac54e2bfb ("x86/Kconfig: Select HAVE_ARCH_HUGE_VMALLOC with HAVE_ARCH_HUGE_VMAP")
Link: https://lore.kernel.org/netdev/14444103-d51b-0fb3-ee63-c3f182f0b546@molgen.mpg.de/"
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Song Liu <song@kernel.org>
Reviewed-by: Rik van Riel <riel@surriel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch reverts commit 7082a29c22 ("ARC: use ACCESS_ONCE in cmpxchg
loop").
It is not necessary to use READ_ONCE() because cmpxchg contains barrier. We
can get it from commit d57f727264 ("ARC: add compiler barrier to LLSC
based cmpxchg").
Signed-off-by: Bang Li <libang.linuxer@gmail.com>
Signed-off-by: Vineet Gupta <vgupta@kernel.org>
They were in <asm/pgtables.h> and have been removed from there in
974b9b2c68 ("mm: consolidate pte_index() and pte_offset_*() definitions")
in favor of the generic version. But that missed that the same definitons
also existed in <asm/pgtable-levels.h>, where they were (inadvertently?)
introduced in fe6cb7b043 ("ARC: mm: disintegrate pgtable.h into levels
and flags").
Fixes: 974b9b2c68 ("mm: consolidate pte_index() and pte_offset_*() definitions")
Fixes: fe6cb7b043 ("ARC: mm: disintegrate pgtable.h into levels and flags")
Signed-off-by: Rolf Eike Beer <eb@emlix.com>
Signed-off-by: Vineet Gupta <vgupta@kernel.org>
disasm_instr() already call memset(0) on its 2nd argument, so there is no
need to clear it explicitly before calling this function.
Remove the redundant memset().
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Vineet Gupta <vgupta@kernel.org>
