Changes in kvm_util.h

16fc7cb4 - KVM: selftests: Add infrastructure for getting vCPU binary stats

Sat, 11 Jan 2025 00:50:48 +0000

KVM: selftests: Add infrastructure for getting vCPU binary statsNow that the binary stats cache infrastructure is largely scope agnostic,add support for vCPU-scoped stats. Like VM stats, open and cache thestats FD when the vCPU is created so that it's guaranteed to be valid whenvcpu_get_stats() is invoked.Account for the extra per-vCPU file descriptor in kvm_set_files_rlimit(),so that tests that create large VMs don't run afoul of resource limits.To sanity check that the infrastructure actually works, and to get a bitof bonus coverage, add an assert in x86's xapic_ipi_test to verify thatthe number of HLTs executed by the test matches the number of HLT exitsobserved by KVM.Tested-by: Manali Shukla Link: https://lore.kernel.org/r/20250111005049.1247555-9-seanjc@google.comSigned-off-by: Sean Christopherson List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

9b56532b - KVM: selftests: Adjust number of files rlimit for all "standard" VMs

Sat, 11 Jan 2025 00:50:47 +0000

KVM: selftests: Adjust number of files rlimit for all "standard" VMsMove the max vCPUs test's RLIMIT_NOFILE adjustments to common code, anduse the new helper to adjust the resource limit for non-barebones VMs bydefault. x86's recalc_apic_map_test creates 512 vCPUs, and a futurechange will open the binary stats fd for all vCPUs, which will put therecalc APIC test above some distros' default limit of 1024.Link: https://lore.kernel.org/r/20250111005049.1247555-8-seanjc@google.comSigned-off-by: Sean Christopherson List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

e65faf71 - KVM: selftests: Add struct and helpers to wrap binary stats cache

Sat, 11 Jan 2025 00:50:45 +0000

KVM: selftests: Add struct and helpers to wrap binary stats cacheAdd a struct and helpers to manage the binary stats cache, which iscurrently used only for VM-scoped stats. This will allow expanding theselftests infrastructure to provide support for vCPU-scoped binary stats,which, except for the ioctl to get the stats FD are identical to VM-scopedstats.Defer converting __vm_get_stat() to a scope-agnostic helper to a futurepatch, as getting the stats FD from KVM needs to be moved elsewherebefore it can be made completely scope-agnostic.Link: https://lore.kernel.org/r/20250111005049.1247555-6-seanjc@google.comSigned-off-by: Sean Christopherson List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

b0c3f5df - KVM: selftests: Macrofy vm_get_stat() to auto-generate stat name string

Sat, 11 Jan 2025 00:50:44 +0000

KVM: selftests: Macrofy vm_get_stat() to auto-generate stat name stringTurn vm_get_stat() into a macro that generates a string for the stat name,as opposed to taking a string. This will allow hardening stat usage inthe future to generate errors on unknown stats at compile time.No functional change intended.Link: https://lore.kernel.org/r/20250111005049.1247555-5-seanjc@google.comSigned-off-by: Sean Christopherson List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

fe85ce31 - KVM: selftests: Assert that vcpu_{g,s}et_reg() won't truncate

Thu, 28 Nov 2024 00:55:34 +0000

KVM: selftests: Assert that vcpu_{g,s}et_reg() won't truncateAssert that the register being read/written by vcpu_{g,s}et_reg() is nolarger than a uint64_t, i.e. that a selftest isn't unintentionallytruncating the value being read/written.Ideally, the assert would be done at compile-time, but that would limitthe checks to hardcoded accesses and/or require fancier compile-timeassertion infrastructure to filter out dynamic usage.Reviewed-by: Andrew Jones Link: https://lore.kernel.org/r/20241128005547.4077116-4-seanjc@google.comSigned-off-by: Sean Christopherson List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

09bb926d - KVM: selftests: Return a value from vcpu_get_reg() instead of using an out-param

Thu, 28 Nov 2024 00:55:33 +0000

KVM: selftests: Return a value from vcpu_get_reg() instead of using an out-paramReturn a uint64_t from vcpu_get_reg() instead of having the caller providea pointer to storage, as none of the vcpu_get_reg() usage in KVM selftestsaccesses a register larger than 64 bits, and vcpu_set_reg() only accepts a64-bit value. If a use case comes along that needs to get a register thatis larger than 64 bits, then a utility can be added to assert success andtake a void pointer, but until then, forcing an out param yields ugly codeand prevents feeding the output of vcpu_get_reg() into vcpu_set_reg().Reviewed-by: Andrew Jones Acked-by: Claudio Imbrenda Link: https://lore.kernel.org/r/20241128005547.4077116-3-seanjc@google.comSigned-off-by: Sean Christopherson List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

215b3cb7 - KVM: selftests: Add a test for coalesced MMIO (and PIO on x86)

Wed, 28 Aug 2024 18:14:45 +0000

KVM: selftests: Add a test for coalesced MMIO (and PIO on x86)Add a test to verify that KVM correctly exits (or not) when a vCPU'scoalesced I/O ring is full (or isn't). Iterate over all legal startingpoints in the ring (with an empty ring), and verify that KVM doesn't exituntil the ring is full.Opportunistically verify that KVM exits immediately on non-coalesced I/O,either because the MMIO/PIO region was never registered, or because aprevious region was unregistered.This is a regression test for a KVM bug where KVM would prematurely exitdue to bad math resulting in a false positive if the first entry in thering was before the halfway mark. See commit 92f6d4130497 ("KVM: Fixcoalesced_mmio_has_room() to avoid premature userspace exit").Enable the test for x86, arm64, and risc-v, i.e. all architectures excepts390, which doesn't have MMIO.On x86, which has both MMIO and PIO, interleave MMIO and PIO into the samering, as KVM shouldn't exit until a non-coalesced I/O is encountered,regardless of whether the ring is filled with MMIO, PIO, or both.Lastly, wrap the coalesced I/O ring in a structure to prepare for apotential future where KVM supports multiple ring buffers beyond KVM's"default" built-in buffer.Link: https://lore.kernel.org/all/20240820133333.1724191-1-ilstam@amazon.comCc: Ilias Stamatis Cc: Marc Zyngier Cc: Oliver Upton Cc: Anup Patel Link: https://lore.kernel.org/r/20240828181446.652474-2-seanjc@google.comSigned-off-by: Sean Christopherson List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

ce3b90bd - KVM: selftests: Remove unused kvm_memcmp_hva_gva()

Fri, 02 Aug 2024 20:08:53 +0000

KVM: selftests: Remove unused kvm_memcmp_hva_gva()Remove sefltests' kvm_memcmp_hva_gva(), which has literally never had asingle user since it was introduced by commit 783e9e51266eb ("kvm:selftests: add API testing infrastructure").Link: https://lore.kernel.org/r/20240802200853.336512-1-seanjc@google.comSigned-off-by: Sean Christopherson List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

3a085fbf - KVM: selftests: Move GDT, IDT, and TSS fields to x86's kvm_vm_arch

Thu, 14 Mar 2024 23:26:22 +0000

KVM: selftests: Move GDT, IDT, and TSS fields to x86's kvm_vm_archNow that kvm_vm_arch exists, move the GDT, IDT, and TSS fields to x86'simplementation, as the structures are firmly x86-only.Reviewed-by: Ackerley Tng Link: https://lore.kernel.org/r/20240314232637.2538648-4-seanjc@google.comSigned-off-by: Sean Christopherson List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

f54884f9 - KVM: sefltests: Add kvm_util_types.h to hold common types, e.g. vm_vaddr_t

Thu, 14 Mar 2024 23:26:21 +0000

KVM: sefltests: Add kvm_util_types.h to hold common types, e.g. vm_vaddr_tMove the base types unique to KVM selftests out of kvm_util.h and into anew header, kvm_util_types.h. This will allow kvm_util_arch.h, i.e. corearch headers, to reference common types, e.g. vm_vaddr_t and vm_paddr_t.No functional change intended.Reviewed-by: Ackerley Tng Link: https://lore.kernel.org/r/20240314232637.2538648-3-seanjc@google.comSigned-off-by: Sean Christopherson List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

2b7deea3 - Revert "kvm: selftests: move base kvm_util.h declarations to kvm_util_base.h"

Thu, 14 Mar 2024 23:26:20 +0000

Revert "kvm: selftests: move base kvm_util.h declarations to kvm_util_base.h"Effectively revert the movement of code from kvm_util.h => kvm_util_base.h,as the TL;DR of the justification for the move was to avoid #idefs and/orcircular dependencies between what ended up being ucall_common.h and whatwas (and now again, is), kvm_util.h.But avoiding #ifdef and circular includes is trivial: don't do that. Thecost of removing kvm_util_base.h is a few extra includes of ucall_common.h,but that cost is practically nothing. On the other hand, having a "base"version of a header that is really just the header itself is confusing,and makes it weird/hard to choose names for headers that actually are"base" headers, e.g. to hold core KVM selftests typedefs.For all intents and purposes, this reverts commit7d9a662ed9f0403e7b94940dceb81552b8edb931.Reviewed-by: Ackerley Tng Link: https://lore.kernel.org/r/20240314232637.2538648-2-seanjc@google.comSigned-off-by: Sean Christopherson List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

96c1a628 - kvm: selftests: move ucall declarations into ucall_common.h

Fri, 10 Dec 2021 16:46:12 +0000

kvm: selftests: move ucall declarations into ucall_common.hNow that core kvm_util declarations have special home inkvm_util_base.h, move ucall-related declarations out into a separateheader.Signed-off-by: Michael Roth Message-Id: <20211210164620.11636-3-michael.roth@amd.com>Signed-off-by: Paolo Bonzini List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

7d9a662e - kvm: selftests: move base kvm_util.h declarations to kvm_util_base.h

Fri, 10 Dec 2021 16:46:11 +0000

kvm: selftests: move base kvm_util.h declarations to kvm_util_base.hBetween helper macros and interfaces that will be introduced insubsequent patches, much of kvm_util.h would end up being declarationsspecific to ucall. Ideally these could be separated out into a separateheader since they are not strictly required for writing guest tests andare mostly self-contained interfaces other than a reliance on a fewcore declarations like struct kvm_vm. This doesn't make a bigdifference as far as how tests will be compiled/written since all theseinterfaces will still be packaged up into a single/common libkvm.a usedby all tests, but it is still nice to be able to compartmentalize toimprove readabilty and reduce merge conflicts in the future for commontasks like adding new interfaces to kvm_util.h.Furthermore, some of the ucall declarations will be arch-specific,requiring various #ifdef'ery in kvm_util.h. Ideally these declarationscould live in separate arch-specific headers, e.g.include//ucall.h, which would handle arch-specific declarationsas well as pulling in common ucall-related declarations shared by allarchs.One simple way to do this would be to #include ucall.h at the bottom ofkvm_util.h, after declarations it relies upon like struct kvm_vm.This is brittle however, and doesn't scale easily to other sets ofinterfaces that may be added in the future.Instead, move all declarations currently in kvm_util.h intokvm_util_base.h, then have kvm_util.h #include it. With this change,non-base declarations can be selectively moved/introduced into separateheaders, which can then be included in kvm_util.h so that individualtests don't need to be touched. Subsequent patches will then moveucall-related declarations into a separate header to meet the abovegoals.Signed-off-by: Michael Roth Message-Id: <20211210164620.11636-2-michael.roth@amd.com>Signed-off-by: Paolo Bonzini List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

3e06cdf1 - KVM: selftests: Add initial support for RISC-V 64-bit

Tue, 05 Oct 2021 12:39:56 +0000

KVM: selftests: Add initial support for RISC-V 64-bitWe add initial support for RISC-V 64-bit in KVM selftests usingwhich we can cross-compile and run arch independent tests such as:demand_paging_testdirty_log_testkvm_create_max_vcpus,kvm_page_table_testset_memory_region_testkvm_binary_stats_testAll VM guest modes defined in kvm_util.h require at least 48-bitguest virtual address so to use KVM RISC-V selftests hardwareneed to support at least Sv48 MMU for guest (i.e. VS-mode).Signed-off-by: Anup Patel Reviewed-and-tested-by: Atish Patra List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

88209c10 - KVM: selftests: Add IRQ GSI routing library functions

Tue, 09 Nov 2021 02:39:03 +0000

KVM: selftests: Add IRQ GSI routing library functionsAdd an architecture independent wrapper function for creating andwriting IRQ GSI routing tables. Also add a function to add irqchipentries.Signed-off-by: Ricardo Koller Acked-by: Andrew Jones Signed-off-by: Marc Zyngier Link: https://lore.kernel.org/r/20211109023906.1091208-15-ricarkol@google.com List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

227895ed - KVM: selftests: Add kvm_irq_line library function

Tue, 09 Nov 2021 02:38:53 +0000

KVM: selftests: Add kvm_irq_line library functionAdd an architecture independent wrapper function for the KVM_IRQ_LINEioctl.Signed-off-by: Ricardo Koller Acked-by: Andrew Jones Signed-off-by: Marc Zyngier Link: https://lore.kernel.org/r/20211109023906.1091208-5-ricarkol@google.com List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

aa674de1 - KVM: selftests: arm64: Add support for various modes with 16kB page size

Mon, 27 Dec 2021 12:48:09 +0000

KVM: selftests: arm64: Add support for various modes with 16kB page sizeThe 16kB page size is not a popular choice, due to only a few CPUsactually implementing support for it. However, it can lead to someinteresting performance improvements given the right uarch choices.Add support for this page size for various PA/VA combinations.Signed-off-by: Marc Zyngier Reviewed-by: Andrew Jones Link: https://lore.kernel.org/r/20211227124809.1335409-7-maz@kernel.org List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

e7f58a6b - KVM: selftests: arm64: Add support for VM_MODE_P36V48_{4K,64K}

Mon, 27 Dec 2021 12:48:08 +0000

KVM: selftests: arm64: Add support for VM_MODE_P36V48_{4K,64K}Some of the arm64 systems out there have an IPA space that ispositively tiny. Nonetheless, they make great KVM hosts.Add support for 36bit IPA support with 4kB pages, which makessome of the fruity machines happy. Whilst we're at it, add supportfor 64kB pages as well, though these boxes have no support for it.Reviewed-by: Andrew Jones Signed-off-by: Marc Zyngier Link: https://lore.kernel.org/r/20211227124809.1335409-6-maz@kernel.org List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

357c628e - KVM: selftests: arm64: Introduce a variable default IPA size

Mon, 27 Dec 2021 12:48:05 +0000

KVM: selftests: arm64: Introduce a variable default IPA sizeContrary to popular belief, there is no such thing as a defaultIPA size on arm64. Anything goes, and implementations are theusual Wild West.The selftest infrastructure default to 40bit IPA, which obviouslydoesn't work for some systems out there.Turn VM_MODE_DEFAULT from a constant into a variable, and letguest_modes_append_default() populate it, depending on whatthe HW can do. In order to preserve the current behaviour, westill pick 40bits IPA as the default if it is available, andthe largest supported IPA space otherwise.Signed-off-by: Marc Zyngier Reviewed-by: Andrew Jones Link: https://lore.kernel.org/r/20211227124809.1335409-3-maz@kernel.org List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h

577e022b - selftests: KVM: Fix non-x86 compiling

Tue, 14 Dec 2021 15:18:42 +0000

selftests: KVM: Fix non-x86 compilingAttempting to compile on a non-x86 architecture fails withinclude/kvm_util.h: In function â€˜vm_compute_max_gfnâ€™:include/kvm_util.h:79:21: error: dereferencing pointer to incomplete type â€˜struct kvm_vmâ€™ return ((1ULL << vm->pa_bits) >> vm->page_shift) - 1; ^~This is because the declaration of struct kvm_vm is inlib/kvm_util_internal.h as an effort to make it private tothe test lib code. We can still provide arch specific functions,though, by making the generic function symbols weak. Do that tofix the compile error.Fixes: c8cc43c1eae2 ("selftests: KVM: avoid failures due to reserved HyperTransport region")Cc: stable@vger.kernel.orgSigned-off-by: Andrew Jones Message-Id: <20211214151842.848314-1-drjones@redhat.com>Signed-off-by: Paolo Bonzini List of files: /linux-6.15/tools/testing/selftests/kvm/include/kvm_util.h