| Message ID | 20191022015925.31916-1-sean.j.christopherson@intel.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | KVM: Refactor vCPU creation | expand |
Hi Sean, On Mon, Oct 21, 2019 at 06:58:40PM -0700, Sean Christopherson wrote: > *************************** DISCLAIMER ********************************** > The non-x86 arch specific patches are completely untested. Although the > changes are conceptually straightforward, I'm not remotely confident that > the patches are bug free, e.g. checkpatch caught several blatant typos > that would break compilation. > ************************************************************************* > > The end goal of this series is to strip down the interface between common > KVM code and arch specific code so that there is precisely one arch hook > for creating a vCPU and one hook for destroying a vCPU. In addition to > cleaning up the code base, simplifying the interface gives architectures > more freedom to organize their vCPU creation code. > > KVM's vCPU creation code is comically messy. kvm_vm_ioctl_create_vcpu() > calls three separate arch hooks: init(), create() and setup(). The init() > call is especially nasty as it's hidden away in a common KVM function, > kvm_init_vcpu(), that for all intents and purposes must be immediately > invoked after the vcpu object is allocated. > > Not to be outdone, vCPU destruction also has three arch hooks: uninit(), > destroy() and free(), the latter of which isn't actually invoked by common > KVM code, but the hook declaration still exists because architectures are > relying on its forward declaration. > > Eliminating the extra arch hooks is relatively straightforward, just > tedious. For the most part, there is no fundamental constraint that > necessitated the proliferation of arch hooks, rather they crept in over > time, usually when x86-centric code was moved out of generic KVM and into > x86 code. > > E.g. kvm_arch_vcpu_setup() was added to allow x86 to do vcpu_load(), which > can only be done after preempt_notifier initialization, but adding setup() > overlooked the fact that the preempt_notifier was only initialized after > kvm_arch_vcpu_create() because preemption support was added when x86's MMU > setup (the vcpu_load() user) was called from common KVM code. > > For all intents and purposes, there is no true functional change in this > series. The order of some allocations will change, and a few memory leaks > are fixed, but the actual functionality of a guest should be unaffected. > > Patches 01-03 are bug fixes in error handling paths that were found by > inspection when refactoring the associated code. > > Patches 04-43 refactor each arch implementation so that the unwanted arch > hooks can be dropped without a functional change, e.g. move code out of > kvm_arch_vcpu_setup() so that all implementations are empty, then drop the > functions and caller. > > Patches 44-45 are minor clean up to eliminate kvm_vcpu_uninit(). > > > The net result is to go from this: > > vcpu = kvm_arch_vcpu_create(kvm, id); > | > |-> kvm_vcpu_init() > | > |-> kvm_arch_vcpu_init() > > if (IS_ERR(vcpu)) { > r = PTR_ERR(vcpu); > goto vcpu_decrement; > } > > preempt_notifier_init(&vcpu->preempt_notifier, &kvm_preempt_ops); > > r = kvm_arch_vcpu_setup(vcpu); > if (r) > goto vcpu_destroy; > > to this: > > r = kvm_arch_vcpu_precreate(kvm, id); > if (r) > goto vcpu_decrement; > > vcpu = kmem_cache_zalloc(kvm_vcpu_cache, GFP_KERNEL); > if (!vcpu) { > r = -ENOMEM; > goto vcpu_decrement; > } > > page = alloc_page(GFP_KERNEL | __GFP_ZERO); > if (!page) { > r = -ENOMEM; > goto vcpu_free; > } > vcpu->run = page_address(page); > > kvm_vcpu_init(vcpu, kvm, id); > > r = kvm_arch_vcpu_create(vcpu); > if (r) > goto vcpu_free_run_page; > What a fantastically welcome piece of work! Thanks for doing this, many's the time I waded through all those calls to ensure a patch was doing the right thing. Modulo the nit in patch 42, the arm64 changes survive a guest boot + hackbench and build fine. The lack of changing the arm-specific destroy function to a void also causes a series of warnings for a 32-bit arm build, but otherwise builds fine. You can add my: Acked-by: Christoffer Dall <christoffer.dall@arm.com> To the arm/arm64 and generic parts. Thanks, Christoffer
*************************** DISCLAIMER ********************************** The non-x86 arch specific patches are completely untested. Although the changes are conceptually straightforward, I'm not remotely confident that the patches are bug free, e.g. checkpatch caught several blatant typos that would break compilation. ************************************************************************* The end goal of this series is to strip down the interface between common KVM code and arch specific code so that there is precisely one arch hook for creating a vCPU and one hook for destroying a vCPU. In addition to cleaning up the code base, simplifying the interface gives architectures more freedom to organize their vCPU creation code. KVM's vCPU creation code is comically messy. kvm_vm_ioctl_create_vcpu() calls three separate arch hooks: init(), create() and setup(). The init() call is especially nasty as it's hidden away in a common KVM function, kvm_init_vcpu(), that for all intents and purposes must be immediately invoked after the vcpu object is allocated. Not to be outdone, vCPU destruction also has three arch hooks: uninit(), destroy() and free(), the latter of which isn't actually invoked by common KVM code, but the hook declaration still exists because architectures are relying on its forward declaration. Eliminating the extra arch hooks is relatively straightforward, just tedious. For the most part, there is no fundamental constraint that necessitated the proliferation of arch hooks, rather they crept in over time, usually when x86-centric code was moved out of generic KVM and into x86 code. E.g. kvm_arch_vcpu_setup() was added to allow x86 to do vcpu_load(), which can only be done after preempt_notifier initialization, but adding setup() overlooked the fact that the preempt_notifier was only initialized after kvm_arch_vcpu_create() because preemption support was added when x86's MMU setup (the vcpu_load() user) was called from common KVM code. For all intents and purposes, there is no true functional change in this series. The order of some allocations will change, and a few memory leaks are fixed, but the actual functionality of a guest should be unaffected. Patches 01-03 are bug fixes in error handling paths that were found by inspection when refactoring the associated code. Patches 04-43 refactor each arch implementation so that the unwanted arch hooks can be dropped without a functional change, e.g. move code out of kvm_arch_vcpu_setup() so that all implementations are empty, then drop the functions and caller. Patches 44-45 are minor clean up to eliminate kvm_vcpu_uninit(). The net result is to go from this: vcpu = kvm_arch_vcpu_create(kvm, id); | |-> kvm_vcpu_init() | |-> kvm_arch_vcpu_init() if (IS_ERR(vcpu)) { r = PTR_ERR(vcpu); goto vcpu_decrement; } preempt_notifier_init(&vcpu->preempt_notifier, &kvm_preempt_ops); r = kvm_arch_vcpu_setup(vcpu); if (r) goto vcpu_destroy; to this: r = kvm_arch_vcpu_precreate(kvm, id); if (r) goto vcpu_decrement; vcpu = kmem_cache_zalloc(kvm_vcpu_cache, GFP_KERNEL); if (!vcpu) { r = -ENOMEM; goto vcpu_decrement; } page = alloc_page(GFP_KERNEL | __GFP_ZERO); if (!page) { r = -ENOMEM; goto vcpu_free; } vcpu->run = page_address(page); kvm_vcpu_init(vcpu, kvm, id); r = kvm_arch_vcpu_create(vcpu); if (r) goto vcpu_free_run_page; Sean Christopherson (45): KVM: PPC: Book3S HV: Uninit vCPU if vcore creation fails KVM: PPC: Book3S PR: Free shared page if mmu initialization fails KVM: x86: Free wbinvd_dirty_mask if vCPU creation fails KVM: VMX: Allocate VPID after initializing VCPU KVM: VMX: Use direct vcpu pointer during vCPU create/free KVM: SVM: Use direct vcpu pointer during vCPU create/free KVM: x86: Allocate vcpu struct in common x86 code KVM: x86: Move FPU allocation to common x86 code KVM: x86: Move allocation of pio_data page down a few lines KVM: x86: Move kvm_vcpu_init() invocation to common code KVM: PPC: e500mc: Add build-time assert that vcpu is at offset 0 KVM: PPC: Allocate vcpu struct in common PPC code KVM: PPC: Book3S PR: Allocate book3s and shadow vcpu after common init KVM: PPC: e500mc: Move reset of oldpir below call to kvm_vcpu_init() KVM: PPC: Move kvm_vcpu_init() invocation to common code KVM: MIPS: Use kvm_vcpu_cache to allocate vCPUs KVM: MIPS: Drop kvm_arch_vcpu_free() KVM: PPC: Drop kvm_arch_vcpu_free() KVM: arm: Drop kvm_arch_vcpu_free() KVM: x86: Remove spurious kvm_mmu_unload() from vcpu destruction path KVM: x86: Remove spurious clearing of async #PF MSR KVM: x86: Drop kvm_arch_vcpu_free() KVM: Remove kvm_arch_vcpu_free() declaration KVM: Add kvm_arch_vcpu_precreate() to handle pre-allocation issues KVM: s390: Move guts of kvm_arch_vcpu_init() into kvm_arch_vcpu_create() KVM: s390: Invoke kvm_vcpu_init() before allocating sie_page KVM: MIPS: Invoke kvm_vcpu_uninit() immediately prior to freeing vcpu KVM: x86: Invoke kvm_vcpu_uninit() immediately prior to freeing vcpu KVM: Introduce kvm_vcpu_destroy() KVM: Move vcpu alloc and init invocation to common code KVM: Unexport kvm_vcpu_cache and kvm_{un}init_vcpu() KVM: Move initialization of preempt notifier to kvm_vcpu_init() KVM: x86: Move guts of kvm_arch_vcpu_setup() into kvm_arch_vcpu_create() KVM: MIPS: Move .vcpu_setup() call to kvm_arch_vcpu_create() KVM: s390: Manually invoke vcpu setup during kvm_arch_vcpu_create() KVM: PPC: BookE: Setup vcpu during kvmppc_core_vcpu_create() KVM: Drop kvm_arch_vcpu_setup() KVM: x86: Move all vcpu init code into kvm_arch_vcpu_create() KVM: MIPS: Move all vcpu init code into kvm_arch_vcpu_create() KVM: ARM: Move all vcpu init code into kvm_arch_vcpu_create() KVM: PPC: Move all vcpu init code into kvm_arch_vcpu_create() KVM: arm64: Free sve_state via arm specific hook KVM: Drop kvm_arch_vcpu_init() and kvm_arch_vcpu_uninit() KVM: Move putting of vcpu->pid to kvm_vcpu_destroy() KVM: Move vcpu->run page allocation out of kvm_vcpu_init() arch/arm/include/asm/kvm_host.h | 2 +- arch/arm/kvm/guest.c | 5 - arch/arm64/include/asm/kvm_host.h | 2 +- arch/arm64/kvm/guest.c | 5 - arch/arm64/kvm/reset.c | 2 +- arch/mips/kvm/mips.c | 84 ++++------- arch/powerpc/include/asm/kvm_ppc.h | 6 +- arch/powerpc/kvm/book3s.c | 9 +- arch/powerpc/kvm/book3s_hv.c | 27 +--- arch/powerpc/kvm/book3s_pr.c | 33 ++--- arch/powerpc/kvm/booke.c | 65 ++++---- arch/powerpc/kvm/e500.c | 34 +---- arch/powerpc/kvm/e500mc.c | 32 ++-- arch/powerpc/kvm/powerpc.c | 70 ++++----- arch/s390/include/asm/kvm_host.h | 1 - arch/s390/kvm/kvm-s390.c | 110 +++++++------- arch/x86/include/asm/kvm_host.h | 2 +- arch/x86/kvm/svm.c | 52 +------ arch/x86/kvm/vmx/vmx.c | 72 +++------ arch/x86/kvm/x86.c | 230 ++++++++++++++--------------- include/linux/kvm_host.h | 13 +- virt/kvm/arm/arm.c | 76 ++++------ virt/kvm/kvm_main.c | 71 +++++---- 23 files changed, 387 insertions(+), 616 deletions(-)