diff options
Diffstat (limited to 'virt')
-rw-r--r-- | virt/kvm/arm/arch_timer.c | 608 | ||||
-rw-r--r-- | virt/kvm/arm/arm.c | 64 | ||||
-rw-r--r-- | virt/kvm/arm/hyp/vgic-v3-sr.c | 6 | ||||
-rw-r--r-- | virt/kvm/arm/mmu.c | 145 | ||||
-rw-r--r-- | virt/kvm/arm/trace.h | 107 | ||||
-rw-r--r-- | virt/kvm/arm/vgic/vgic-its.c | 31 | ||||
-rw-r--r-- | virt/kvm/arm/vgic/vgic-v3.c | 8 | ||||
-rw-r--r-- | virt/kvm/arm/vgic/vgic.c | 14 | ||||
-rw-r--r-- | virt/kvm/coalesced_mmio.c | 3 | ||||
-rw-r--r-- | virt/kvm/eventfd.c | 13 | ||||
-rw-r--r-- | virt/kvm/irqchip.c | 4 | ||||
-rw-r--r-- | virt/kvm/kvm_main.c | 106 | ||||
-rw-r--r-- | virt/kvm/vfio.c | 4 |
13 files changed, 743 insertions, 370 deletions
diff --git a/virt/kvm/arm/arch_timer.c b/virt/kvm/arm/arch_timer.c index b07ac4614e1c..3417f2dbc366 100644 --- a/virt/kvm/arm/arch_timer.c +++ b/virt/kvm/arm/arch_timer.c @@ -25,6 +25,7 @@ #include <clocksource/arm_arch_timer.h> #include <asm/arch_timer.h> +#include <asm/kvm_emulate.h> #include <asm/kvm_hyp.h> #include <kvm/arm_vgic.h> @@ -34,7 +35,9 @@ static struct timecounter *timecounter; static unsigned int host_vtimer_irq; +static unsigned int host_ptimer_irq; static u32 host_vtimer_irq_flags; +static u32 host_ptimer_irq_flags; static DEFINE_STATIC_KEY_FALSE(has_gic_active_state); @@ -52,12 +55,34 @@ static bool kvm_timer_irq_can_fire(struct arch_timer_context *timer_ctx); static void kvm_timer_update_irq(struct kvm_vcpu *vcpu, bool new_level, struct arch_timer_context *timer_ctx); static bool kvm_timer_should_fire(struct arch_timer_context *timer_ctx); +static void kvm_arm_timer_write(struct kvm_vcpu *vcpu, + struct arch_timer_context *timer, + enum kvm_arch_timer_regs treg, + u64 val); +static u64 kvm_arm_timer_read(struct kvm_vcpu *vcpu, + struct arch_timer_context *timer, + enum kvm_arch_timer_regs treg); u64 kvm_phys_timer_read(void) { return timecounter->cc->read(timecounter->cc); } +static void get_timer_map(struct kvm_vcpu *vcpu, struct timer_map *map) +{ + if (has_vhe()) { + map->direct_vtimer = vcpu_vtimer(vcpu); + map->direct_ptimer = vcpu_ptimer(vcpu); + map->emul_ptimer = NULL; + } else { + map->direct_vtimer = vcpu_vtimer(vcpu); + map->direct_ptimer = NULL; + map->emul_ptimer = vcpu_ptimer(vcpu); + } + + trace_kvm_get_timer_map(vcpu->vcpu_id, map); +} + static inline bool userspace_irqchip(struct kvm *kvm) { return static_branch_unlikely(&userspace_irqchip_in_use) && @@ -78,20 +103,27 @@ static void soft_timer_cancel(struct hrtimer *hrt) static irqreturn_t kvm_arch_timer_handler(int irq, void *dev_id) { struct kvm_vcpu *vcpu = *(struct kvm_vcpu **)dev_id; - struct arch_timer_context *vtimer; + struct arch_timer_context *ctx; + struct timer_map map; /* * We may see a timer interrupt after vcpu_put() has been called which * sets the CPU's vcpu pointer to NULL, because even though the timer - * has been disabled in vtimer_save_state(), the hardware interrupt + * has been disabled in timer_save_state(), the hardware interrupt * signal may not have been retired from the interrupt controller yet. */ if (!vcpu) return IRQ_HANDLED; - vtimer = vcpu_vtimer(vcpu); - if (kvm_timer_should_fire(vtimer)) - kvm_timer_update_irq(vcpu, true, vtimer); + get_timer_map(vcpu, &map); + + if (irq == host_vtimer_irq) + ctx = map.direct_vtimer; + else + ctx = map.direct_ptimer; + + if (kvm_timer_should_fire(ctx)) + kvm_timer_update_irq(vcpu, true, ctx); if (userspace_irqchip(vcpu->kvm) && !static_branch_unlikely(&has_gic_active_state)) @@ -122,7 +154,9 @@ static u64 kvm_timer_compute_delta(struct arch_timer_context *timer_ctx) static bool kvm_timer_irq_can_fire(struct arch_timer_context *timer_ctx) { - return !(timer_ctx->cnt_ctl & ARCH_TIMER_CTRL_IT_MASK) && + WARN_ON(timer_ctx && timer_ctx->loaded); + return timer_ctx && + !(timer_ctx->cnt_ctl & ARCH_TIMER_CTRL_IT_MASK) && (timer_ctx->cnt_ctl & ARCH_TIMER_CTRL_ENABLE); } @@ -132,21 +166,22 @@ static bool kvm_timer_irq_can_fire(struct arch_timer_context *timer_ctx) */ static u64 kvm_timer_earliest_exp(struct kvm_vcpu *vcpu) { - u64 min_virt = ULLONG_MAX, min_phys = ULLONG_MAX; - struct arch_timer_context *vtimer = vcpu_vtimer(vcpu); - struct arch_timer_context *ptimer = vcpu_ptimer(vcpu); + u64 min_delta = ULLONG_MAX; + int i; - if (kvm_timer_irq_can_fire(vtimer)) - min_virt = kvm_timer_compute_delta(vtimer); + for (i = 0; i < NR_KVM_TIMERS; i++) { + struct arch_timer_context *ctx = &vcpu->arch.timer_cpu.timers[i]; - if (kvm_timer_irq_can_fire(ptimer)) - min_phys = kvm_timer_compute_delta(ptimer); + WARN(ctx->loaded, "timer %d loaded\n", i); + if (kvm_timer_irq_can_fire(ctx)) + min_delta = min(min_delta, kvm_timer_compute_delta(ctx)); + } /* If none of timers can fire, then return 0 */ - if ((min_virt == ULLONG_MAX) && (min_phys == ULLONG_MAX)) + if (min_delta == ULLONG_MAX) return 0; - return min(min_virt, min_phys); + return min_delta; } static enum hrtimer_restart kvm_bg_timer_expire(struct hrtimer *hrt) @@ -173,41 +208,58 @@ static enum hrtimer_restart kvm_bg_timer_expire(struct hrtimer *hrt) return HRTIMER_NORESTART; } -static enum hrtimer_restart kvm_phys_timer_expire(struct hrtimer *hrt) +static enum hrtimer_restart kvm_hrtimer_expire(struct hrtimer *hrt) { - struct arch_timer_context *ptimer; - struct arch_timer_cpu *timer; + struct arch_timer_context *ctx; struct kvm_vcpu *vcpu; u64 ns; - timer = container_of(hrt, struct arch_timer_cpu, phys_timer); - vcpu = container_of(timer, struct kvm_vcpu, arch.timer_cpu); - ptimer = vcpu_ptimer(vcpu); + ctx = container_of(hrt, struct arch_timer_context, hrtimer); + vcpu = ctx->vcpu; + + trace_kvm_timer_hrtimer_expire(ctx); /* * Check that the timer has really expired from the guest's * PoV (NTP on the host may have forced it to expire * early). If not ready, schedule for a later time. */ - ns = kvm_timer_compute_delta(ptimer); + ns = kvm_timer_compute_delta(ctx); if (unlikely(ns)) { hrtimer_forward_now(hrt, ns_to_ktime(ns)); return HRTIMER_RESTART; } - kvm_timer_update_irq(vcpu, true, ptimer); + kvm_timer_update_irq(vcpu, true, ctx); return HRTIMER_NORESTART; } static bool kvm_timer_should_fire(struct arch_timer_context *timer_ctx) { + enum kvm_arch_timers index; u64 cval, now; + if (!timer_ctx) + return false; + + index = arch_timer_ctx_index(timer_ctx); + if (timer_ctx->loaded) { - u32 cnt_ctl; + u32 cnt_ctl = 0; + + switch (index) { + case TIMER_VTIMER: + cnt_ctl = read_sysreg_el0(cntv_ctl); + break; + case TIMER_PTIMER: + cnt_ctl = read_sysreg_el0(cntp_ctl); + break; + case NR_KVM_TIMERS: + /* GCC is braindead */ + cnt_ctl = 0; + break; + } - /* Only the virtual timer can be loaded so far */ - cnt_ctl = read_sysreg_el0(cntv_ctl); return (cnt_ctl & ARCH_TIMER_CTRL_ENABLE) && (cnt_ctl & ARCH_TIMER_CTRL_IT_STAT) && !(cnt_ctl & ARCH_TIMER_CTRL_IT_MASK); @@ -224,13 +276,13 @@ static bool kvm_timer_should_fire(struct arch_timer_context *timer_ctx) bool kvm_timer_is_pending(struct kvm_vcpu *vcpu) { - struct arch_timer_context *vtimer = vcpu_vtimer(vcpu); - struct arch_timer_context *ptimer = vcpu_ptimer(vcpu); + struct timer_map map; - if (kvm_timer_should_fire(vtimer)) - return true; + get_timer_map(vcpu, &map); - return kvm_timer_should_fire(ptimer); + return kvm_timer_should_fire(map.direct_vtimer) || + kvm_timer_should_fire(map.direct_ptimer) || + kvm_timer_should_fire(map.emul_ptimer); } /* @@ -269,77 +321,70 @@ static void kvm_timer_update_irq(struct kvm_vcpu *vcpu, bool new_level, } } -/* Schedule the background timer for the emulated timer. */ -static void phys_timer_emulate(struct kvm_vcpu *vcpu) +static void timer_emulate(struct arch_timer_context *ctx) { - struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu; - struct arch_timer_context *ptimer = vcpu_ptimer(vcpu); + bool should_fire = kvm_timer_should_fire(ctx); + + trace_kvm_timer_emulate(ctx, should_fire); + + if (should_fire) { + kvm_timer_update_irq(ctx->vcpu, true, ctx); + return; + } /* * If the timer can fire now, we don't need to have a soft timer * scheduled for the future. If the timer cannot fire at all, * then we also don't need a soft timer. */ - if (kvm_timer_should_fire(ptimer) || !kvm_timer_irq_can_fire(ptimer)) { - soft_timer_cancel(&timer->phys_timer); + if (!kvm_timer_irq_can_fire(ctx)) { + soft_timer_cancel(&ctx->hrtimer); return; } - soft_timer_start(&timer->phys_timer, kvm_timer_compute_delta(ptimer)); + soft_timer_start(&ctx->hrtimer, kvm_timer_compute_delta(ctx)); } -/* - * Check if there was a change in the timer state, so that we should either - * raise or lower the line level to the GIC or schedule a background timer to - * emulate the physical timer. - */ -static void kvm_timer_update_state(struct kvm_vcpu *vcpu) +static void timer_save_state(struct arch_timer_context *ctx) { - struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu; - struct arch_timer_context *vtimer = vcpu_vtimer(vcpu); - struct arch_timer_context *ptimer = vcpu_ptimer(vcpu); - bool level; + struct arch_timer_cpu *timer = vcpu_timer(ctx->vcpu); + enum kvm_arch_timers index = arch_timer_ctx_index(ctx); + unsigned long flags; - if (unlikely(!timer->enabled)) + if (!timer->enabled) return; - /* - * The vtimer virtual interrupt is a 'mapped' interrupt, meaning part - * of its lifecycle is offloaded to the hardware, and we therefore may - * not have lowered the irq.level value before having to signal a new - * interrupt, but have to signal an interrupt every time the level is - * asserted. - */ - level = kvm_timer_should_fire(vtimer); - kvm_timer_update_irq(vcpu, level, vtimer); + local_irq_save(flags); - phys_timer_emulate(vcpu); + if (!ctx->loaded) + goto out; - if (kvm_timer_should_fire(ptimer) != ptimer->irq.level) - kvm_timer_update_irq(vcpu, !ptimer->irq.level, ptimer); -} + switch (index) { + case TIMER_VTIMER: + ctx->cnt_ctl = read_sysreg_el0(cntv_ctl); + ctx->cnt_cval = read_sysreg_el0(cntv_cval); -static void vtimer_save_state(struct kvm_vcpu *vcpu) -{ - struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu; - struct arch_timer_context *vtimer = vcpu_vtimer(vcpu); - unsigned long flags; + /* Disable the timer */ + write_sysreg_el0(0, cntv_ctl); + isb(); - local_irq_save(flags); + break; + case TIMER_PTIMER: + ctx->cnt_ctl = read_sysreg_el0(cntp_ctl); + ctx->cnt_cval = read_sysreg_el0(cntp_cval); - if (!vtimer->loaded) - goto out; + /* Disable the timer */ + write_sysreg_el0(0, cntp_ctl); + isb(); - if (timer->enabled) { - vtimer->cnt_ctl = read_sysreg_el0(cntv_ctl); - vtimer->cnt_cval = read_sysreg_el0(cntv_cval); + break; + case NR_KVM_TIMERS: + BUG(); } - /* Disable the virtual timer */ - write_sysreg_el0(0, cntv_ctl); - isb(); + trace_kvm_timer_save_state(ctx); - vtimer->loaded = false; + ctx->loaded = false; out: local_irq_restore(flags); } @@ -349,67 +394,72 @@ out: * thread is removed from its waitqueue and made runnable when there's a timer * interrupt to handle. */ -void kvm_timer_schedule(struct kvm_vcpu *vcpu) +static void kvm_timer_blocking(struct kvm_vcpu *vcpu) { - struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu; - struct arch_timer_context *vtimer = vcpu_vtimer(vcpu); - struct arch_timer_context *ptimer = vcpu_ptimer(vcpu); - - vtimer_save_state(vcpu); + struct arch_timer_cpu *timer = vcpu_timer(vcpu); + struct timer_map map; - /* - * No need to schedule a background timer if any guest timer has - * already expired, because kvm_vcpu_block will return before putting - * the thread to sleep. - */ - if (kvm_timer_should_fire(vtimer) || kvm_timer_should_fire(ptimer)) - return; + get_timer_map(vcpu, &map); /* - * If both timers are not capable of raising interrupts (disabled or + * If no timers are capable of raising interrupts (disabled or * masked), then there's no more work for us to do. */ - if (!kvm_timer_irq_can_fire(vtimer) && !kvm_timer_irq_can_fire(ptimer)) + if (!kvm_timer_irq_can_fire(map.direct_vtimer) && + !kvm_timer_irq_can_fire(map.direct_ptimer) && + !kvm_timer_irq_can_fire(map.emul_ptimer)) return; /* - * The guest timers have not yet expired, schedule a background timer. + * At least one guest time will expire. Schedule a background timer. * Set the earliest expiration time among the guest timers. */ soft_timer_start(&timer->bg_timer, kvm_timer_earliest_exp(vcpu)); } -static void vtimer_restore_state(struct kvm_vcpu *vcpu) +static void kvm_timer_unblocking(struct kvm_vcpu *vcpu) { - struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu; - struct arch_timer_context *vtimer = vcpu_vtimer(vcpu); + struct arch_timer_cpu *timer = vcpu_timer(vcpu); + + soft_timer_cancel(&timer->bg_timer); +} + +static void timer_restore_state(struct arch_timer_context *ctx) +{ + struct arch_timer_cpu *timer = vcpu_timer(ctx->vcpu); + enum kvm_arch_timers index = arch_timer_ctx_index(ctx); unsigned long flags; + if (!timer->enabled) + return; + local_irq_save(flags); - if (vtimer->loaded) + if (ctx->loaded) goto out; - if (timer->enabled) { - write_sysreg_el0(vtimer->cnt_cval, cntv_cval); + switch (index) { + case TIMER_VTIMER: + write_sysreg_el0(ctx->cnt_cval, cntv_cval); isb(); - write_sysreg_el0(vtimer->cnt_ctl, cntv_ctl); + write_sysreg_el0(ctx->cnt_ctl, cntv_ctl); + break; + case TIMER_PTIMER: + write_sysreg_el0(ctx->cnt_cval, cntp_cval); + isb(); + write_sysreg_el0(ctx->cnt_ctl, cntp_ctl); + break; + case NR_KVM_TIMERS: + BUG(); } - vtimer->loaded = true; + trace_kvm_timer_restore_state(ctx); + + ctx->loaded = true; out: local_irq_restore(flags); } -void kvm_timer_unschedule(struct kvm_vcpu *vcpu) -{ - struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu; - - vtimer_restore_state(vcpu); - - soft_timer_cancel(&timer->bg_timer); -} - static void set_cntvoff(u64 cntvoff) { u32 low = lower_32_bits(cntvoff); @@ -425,23 +475,32 @@ static void set_cntvoff(u64 cntvoff) kvm_call_hyp(__kvm_timer_set_cntvoff, low, high); } -static inline void set_vtimer_irq_phys_active(struct kvm_vcpu *vcpu, bool active) +static inline void set_timer_irq_phys_active(struct arch_timer_context *ctx, bool active) { int r; - r = irq_set_irqchip_state(host_vtimer_irq, IRQCHIP_STATE_ACTIVE, active); + r = irq_set_irqchip_state(ctx->host_timer_irq, IRQCHIP_STATE_ACTIVE, active); WARN_ON(r); } -static void kvm_timer_vcpu_load_gic(struct kvm_vcpu *vcpu) +static void kvm_timer_vcpu_load_gic(struct arch_timer_context *ctx) { - struct arch_timer_context *vtimer = vcpu_vtimer(vcpu); - bool phys_active; + struct kvm_vcpu *vcpu = ctx->vcpu; + bool phys_active = false; + + /* + * Update the timer output so that it is likely to match the + * state we're about to restore. If the timer expires between + * this point and the register restoration, we'll take the + * interrupt anyway. + */ + kvm_timer_update_irq(ctx->vcpu, kvm_timer_should_fire(ctx), ctx); if (irqchip_in_kernel(vcpu->kvm)) - phys_active = kvm_vgic_map_is_active(vcpu, vtimer->irq.irq); - else - phys_active = vtimer->irq.level; - set_vtimer_irq_phys_active(vcpu, phys_active); + phys_active = kvm_vgic_map_is_active(vcpu, ctx->irq.irq); + + phys_active |= ctx->irq.level; + + set_timer_irq_phys_active(ctx, phys_active); } static void kvm_timer_vcpu_load_nogic(struct kvm_vcpu *vcpu) @@ -466,28 +525,32 @@ static void kvm_timer_vcpu_load_nogic(struct kvm_vcpu *vcpu) void kvm_timer_vcpu_load(struct kvm_vcpu *vcpu) { - struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu; - struct arch_timer_context *vtimer = vcpu_vtimer(vcpu); - struct arch_timer_context *ptimer = vcpu_ptimer(vcpu); + struct arch_timer_cpu *timer = vcpu_timer(vcpu); + struct timer_map map; if (unlikely(!timer->enabled)) return; - if (static_branch_likely(&has_gic_active_state)) - kvm_timer_vcpu_load_gic(vcpu); - else + get_timer_map(vcpu, &map); + + if (static_branch_likely(&has_gic_active_state)) { + kvm_timer_vcpu_load_gic(map.direct_vtimer); + if (map.direct_ptimer) + kvm_timer_vcpu_load_gic(map.direct_ptimer); + } else { kvm_timer_vcpu_load_nogic(vcpu); + } - set_cntvoff(vtimer->cntvoff); + set_cntvoff(map.direct_vtimer->cntvoff); - vtimer_restore_state(vcpu); + kvm_timer_unblocking(vcpu); - /* Set the background timer for the physical timer emulation. */ - phys_timer_emulate(vcpu); + timer_restore_state(map.direct_vtimer); + if (map.direct_ptimer) + timer_restore_state(map.direct_ptimer); - /* If the timer fired while we weren't running, inject it now */ - if (kvm_timer_should_fire(ptimer) != ptimer->irq.level) - kvm_timer_update_irq(vcpu, !ptimer->irq.level, ptimer); + if (map.emul_ptimer) + timer_emulate(map.emul_ptimer); } bool kvm_timer_should_notify_user(struct kvm_vcpu *vcpu) @@ -509,15 +572,20 @@ bool kvm_timer_should_notify_user(struct kvm_vcpu *vcpu) void kvm_timer_vcpu_put(struct kvm_vcpu *vcpu) { - struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu; + struct arch_timer_cpu *timer = vcpu_timer(vcpu); + struct timer_map map; if (unlikely(!timer->enabled)) return; - vtimer_save_state(vcpu); + get_timer_map(vcpu, &map); + + timer_save_state(map.direct_vtimer); + if (map.direct_ptimer) + timer_save_state(map.direct_ptimer); /* - * Cancel the physical timer emulation, because the only case where we + * Cancel soft timer emulation, because the only case where we * need it after a vcpu_put is in the context of a sleeping VCPU, and * in that case we already factor in the deadline for the physical * timer when scheduling the bg_timer. @@ -525,7 +593,11 @@ void kvm_timer_vcpu_put(struct kvm_vcpu *vcpu) * In any case, we re-schedule the hrtimer for the physical timer when * coming back to the VCPU thread in kvm_timer_vcpu_load(). */ - soft_timer_cancel(&timer->phys_timer); + if (map.emul_ptimer) + soft_timer_cancel(&map.emul_ptimer->hrtimer); + + if (swait_active(kvm_arch_vcpu_wq(vcpu))) + kvm_timer_blocking(vcpu); /* * The kernel may decide to run userspace after calling vcpu_put, so @@ -534,8 +606,7 @@ void kvm_timer_vcpu_put(struct kvm_vcpu *vcpu) * counter of non-VHE case. For VHE, the virtual counter uses a fixed * virtual offset of zero, so no need to zero CNTVOFF_EL2 register. */ - if (!has_vhe()) - set_cntvoff(0); + set_cntvoff(0); } /* @@ -550,7 +621,7 @@ static void unmask_vtimer_irq_user(struct kvm_vcpu *vcpu) if (!kvm_timer_should_fire(vtimer)) { kvm_timer_update_irq(vcpu, false, vtimer); if (static_branch_likely(&has_gic_active_state)) - set_vtimer_irq_phys_active(vcpu, false); + set_timer_irq_phys_active(vtimer, false); else enable_percpu_irq(host_vtimer_irq, host_vtimer_irq_flags); } @@ -558,7 +629,7 @@ static void unmask_vtimer_irq_user(struct kvm_vcpu *vcpu) void kvm_timer_sync_hwstate(struct kvm_vcpu *vcpu) { - struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu; + struct arch_timer_cpu *timer = vcpu_timer(vcpu); if (unlikely(!timer->enabled)) return; @@ -569,9 +640,10 @@ void kvm_timer_sync_hwstate(struct kvm_vcpu *vcpu) int kvm_timer_vcpu_reset(struct kvm_vcpu *vcpu) { - struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu; - struct arch_timer_context *vtimer = vcpu_vtimer(vcpu); - struct arch_timer_context *ptimer = vcpu_ptimer(vcpu); + struct arch_timer_cpu *timer = vcpu_timer(vcpu); + struct timer_map map; + + get_timer_map(vcpu, &map); /* * The bits in CNTV_CTL are architecturally reset to UNKNOWN for ARMv8 @@ -579,12 +651,22 @@ int kvm_timer_vcpu_reset(struct kvm_vcpu *vcpu) * resets the timer to be disabled and unmasked and is compliant with * the ARMv7 architecture. */ - vtimer->cnt_ctl = 0; - ptimer->cnt_ctl = 0; - kvm_timer_update_state(vcpu); + vcpu_vtimer(vcpu)->cnt_ctl = 0; + vcpu_ptimer(vcpu)->cnt_ctl = 0; - if (timer->enabled && irqchip_in_kernel(vcpu->kvm)) - kvm_vgic_reset_mapped_irq(vcpu, vtimer->irq.irq); + if (timer->enabled) { + kvm_timer_update_irq(vcpu, false, vcpu_vtimer(vcpu)); + kvm_timer_update_irq(vcpu, false, vcpu_ptimer(vcpu)); + + if (irqchip_in_kernel(vcpu->kvm)) { + kvm_vgic_reset_mapped_irq(vcpu, map.direct_vtimer->irq.irq); + if (map.direct_ptimer) + kvm_vgic_reset_mapped_irq(vcpu, map.direct_ptimer->irq.irq); + } + } + + if (map.emul_ptimer) + soft_timer_cancel(&map.emul_ptimer->hrtimer); return 0; } @@ -610,56 +692,76 @@ static void update_vtimer_cntvoff(struct kvm_vcpu *vcpu, u64 cntvoff) void kvm_timer_vcpu_init(struct kvm_vcpu *vcpu) { - struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu; + struct arch_timer_cpu *timer = vcpu_timer(vcpu); struct arch_timer_context *vtimer = vcpu_vtimer(vcpu); struct arch_timer_context *ptimer = vcpu_ptimer(vcpu); /* Synchronize cntvoff across all vtimers of a VM. */ update_vtimer_cntvoff(vcpu, kvm_phys_timer_read()); - vcpu_ptimer(vcpu)->cntvoff = 0; + ptimer->cntvoff = 0; hrtimer_init(&timer->bg_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); timer->bg_timer.function = kvm_bg_timer_expire; - hrtimer_init(&timer->phys_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); - timer->phys_timer.function = kvm_phys_timer_expire; + hrtimer_init(&vtimer->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); + hrtimer_init(&ptimer->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); + vtimer->hrtimer.function = kvm_hrtimer_expire; + ptimer->hrtimer.function = kvm_hrtimer_expire; vtimer->irq.irq = default_vtimer_irq.irq; ptimer->irq.irq = default_ptimer_irq.irq; + + vtimer->host_timer_irq = host_vtimer_irq; + ptimer->host_timer_irq = host_ptimer_irq; + + vtimer->host_timer_irq_flags = host_vtimer_irq_flags; + ptimer->host_timer_irq_flags = host_ptimer_irq_flags; + + vtimer->vcpu = vcpu; + ptimer->vcpu = vcpu; } static void kvm_timer_init_interrupt(void *info) { enable_percpu_irq(host_vtimer_irq, host_vtimer_irq_flags); + enable_percpu_irq(host_ptimer_irq, host_ptimer_irq_flags); } int kvm_arm_timer_set_reg(struct kvm_vcpu *vcpu, u64 regid, u64 value) { - struct arch_timer_context *vtimer = vcpu_vtimer(vcpu); - struct arch_timer_context *ptimer = vcpu_ptimer(vcpu); + struct arch_timer_context *timer; + bool level; switch (regid) { case KVM_REG_ARM_TIMER_CTL: - vtimer->cnt_ctl = value & ~ARCH_TIMER_CTRL_IT_STAT; + timer = vcpu_vtimer(vcpu); + kvm_arm_timer_write(vcpu, timer, TIMER_REG_CTL, value); break; case KVM_REG_ARM_TIMER_CNT: + timer = vcpu_vtimer(vcpu); update_vtimer_cntvoff(vcpu, kvm_phys_timer_read() - value); break; case KVM_REG_ARM_TIMER_CVAL: - vtimer->cnt_cval = value; + timer = vcpu_vtimer(vcpu); + kvm_arm_timer_write(vcpu, timer, TIMER_REG_CVAL, value); break; case KVM_REG_ARM_PTIMER_CTL: - ptimer->cnt_ctl = value & ~ARCH_TIMER_CTRL_IT_STAT; + timer = vcpu_ptimer(vcpu); + kvm_arm_timer_write(vcpu, timer, TIMER_REG_CTL, value); break; case KVM_REG_ARM_PTIMER_CVAL: - ptimer->cnt_cval = value; + timer = vcpu_ptimer(vcpu); + kvm_arm_timer_write(vcpu, timer, TIMER_REG_CVAL, value); break; default: return -1; } - kvm_timer_update_state(vcpu); + level = kvm_timer_should_fire(timer); + kvm_timer_update_irq(vcpu, level, timer); + timer_emulate(timer); + return 0; } @@ -679,26 +781,113 @@ static u64 read_timer_ctl(struct arch_timer_context *timer) u64 kvm_arm_timer_get_reg(struct kvm_vcpu *vcpu, u64 regid) { - struct arch_timer_context *ptimer = vcpu_ptimer(vcpu); - struct arch_timer_context *vtimer = vcpu_vtimer(vcpu); - switch (regid) { case KVM_REG_ARM_TIMER_CTL: - return read_timer_ctl(vtimer); + return kvm_arm_timer_read(vcpu, + vcpu_vtimer(vcpu), TIMER_REG_CTL); case KVM_REG_ARM_TIMER_CNT: - return kvm_phys_timer_read() - vtimer->cntvoff; + return kvm_arm_timer_read(vcpu, + vcpu_vtimer(vcpu), TIMER_REG_CNT); case KVM_REG_ARM_TIMER_CVAL: - return vtimer->cnt_cval; + return kvm_arm_timer_read(vcpu, + vcpu_vtimer(vcpu), TIMER_REG_CVAL); case KVM_REG_ARM_PTIMER_CTL: - return read_timer_ctl(ptimer); - case KVM_REG_ARM_PTIMER_CVAL: - return ptimer->cnt_cval; + return kvm_arm_timer_read(vcpu, + vcpu_ptimer(vcpu), TIMER_REG_CTL); case KVM_REG_ARM_PTIMER_CNT: - return kvm_phys_timer_read(); + return kvm_arm_timer_read(vcpu, + vcpu_vtimer(vcpu), TIMER_REG_CNT); + case KVM_REG_ARM_PTIMER_CVAL: + return kvm_arm_timer_read(vcpu, + vcpu_ptimer(vcpu), TIMER_REG_CVAL); } return (u64)-1; } +static u64 kvm_arm_timer_read(struct kvm_vcpu *vcpu, + struct arch_timer_context *timer, + enum kvm_arch_timer_regs treg) +{ + u64 val; + + switch (treg) { + case TIMER_REG_TVAL: + val = kvm_phys_timer_read() - timer->cntvoff - timer->cnt_cval; + break; + + case TIMER_REG_CTL: + val = read_timer_ctl(timer); + break; + + case TIMER_REG_CVAL: + val = timer->cnt_cval; + break; + + case TIMER_REG_CNT: + val = kvm_phys_timer_read() - timer->cntvoff; + break; + + default: + BUG(); + } + + return val; +} + +u64 kvm_arm_timer_read_sysreg(struct kvm_vcpu *vcpu, + enum kvm_arch_timers tmr, + enum kvm_arch_timer_regs treg) +{ + u64 val; + + preempt_disable(); + kvm_timer_vcpu_put(vcpu); + + val = kvm_arm_timer_read(vcpu, vcpu_get_timer(vcpu, tmr), treg); + + kvm_timer_vcpu_load(vcpu); + preempt_enable(); + + return val; +} + +static void kvm_arm_timer_write(struct kvm_vcpu *vcpu, + struct arch_timer_context *timer, + enum kvm_arch_timer_regs treg, + u64 val) +{ + switch (treg) { + case TIMER_REG_TVAL: + timer->cnt_cval = val - kvm_phys_timer_read() - timer->cntvoff; + break; + + case TIMER_REG_CTL: + timer->cnt_ctl = val & ~ARCH_TIMER_CTRL_IT_STAT; + break; + + case TIMER_REG_CVAL: + timer->cnt_cval = val; + break; + + default: + BUG(); + } +} + +void kvm_arm_timer_write_sysreg(struct kvm_vcpu *vcpu, + enum kvm_arch_timers tmr, + enum kvm_arch_timer_regs treg, + u64 val) +{ + preempt_disable(); + kvm_timer_vcpu_put(vcpu); + + kvm_arm_timer_write(vcpu, vcpu_get_timer(vcpu, tmr), treg, val); + + kvm_timer_vcpu_load(vcpu); + preempt_enable(); +} + static int kvm_timer_starting_cpu(unsigned int cpu) { kvm_timer_init_interrupt(NULL); @@ -724,6 +913,8 @@ int kvm_timer_hyp_init(bool has_gic) return -ENODEV; } + /* First, do the virtual EL1 timer irq */ + if (info->virtual_irq <= 0) { kvm_err("kvm_arch_timer: invalid virtual timer IRQ: %d\n", info->virtual_irq); @@ -734,15 +925,15 @@ int kvm_timer_hyp_init(bool has_gic) host_vtimer_irq_flags = irq_get_trigger_type(host_vtimer_irq); if (host_vtimer_irq_flags != IRQF_TRIGGER_HIGH && host_vtimer_irq_flags != IRQF_TRIGGER_LOW) { - kvm_err("Invalid trigger for IRQ%d, assuming level low\n", + kvm_err("Invalid trigger for vtimer IRQ%d, assuming level low\n", host_vtimer_irq); host_vtimer_irq_flags = IRQF_TRIGGER_LOW; } err = request_percpu_irq(host_vtimer_irq, kvm_arch_timer_handler, - "kvm guest timer", kvm_get_running_vcpus()); + "kvm guest vtimer", kvm_get_running_vcpus()); if (err) { - kvm_err("kvm_arch_timer: can't request interrupt %d (%d)\n", + kvm_err("kvm_arch_timer: can't request vtimer interrupt %d (%d)\n", host_vtimer_irq, err); return err; } @@ -760,6 +951,43 @@ int kvm_timer_hyp_init(bool has_gic) kvm_debug("virtual timer IRQ%d\n", host_vtimer_irq); + /* Now let's do the physical EL1 timer irq */ + + if (info->physical_irq > 0) { + host_ptimer_irq = info->physical_irq; + host_ptimer_irq_flags = irq_get_trigger_type(host_ptimer_irq); + if (host_ptimer_irq_flags != IRQF_TRIGGER_HIGH && + host_ptimer_irq_flags != IRQF_TRIGGER_LOW) { + kvm_err("Invalid trigger for ptimer IRQ%d, assuming level low\n", + host_ptimer_irq); + host_ptimer_irq_flags = IRQF_TRIGGER_LOW; + } + + err = request_percpu_irq(host_ptimer_irq, kvm_arch_timer_handler, + "kvm guest ptimer", kvm_get_running_vcpus()); + if (err) { + kvm_err("kvm_arch_timer: can't request ptimer interrupt %d (%d)\n", + host_ptimer_irq, err); + return err; + } + + if (has_gic) { + err = irq_set_vcpu_affinity(host_ptimer_irq, + kvm_get_running_vcpus()); + if (err) { + kvm_err("kvm_arch_timer: error setting vcpu affinity\n"); + goto out_free_irq; + } + } + + kvm_debug("physical timer IRQ%d\n", host_ptimer_irq); + } else if (has_vhe()) { + kvm_err("kvm_arch_timer: invalid physical timer IRQ: %d\n", + info->physical_irq); + err = -ENODEV; + goto out_free_irq; + } + cpuhp_setup_state(CPUHP_AP_KVM_ARM_TIMER_STARTING, "kvm/arm/timer:starting", kvm_timer_starting_cpu, kvm_timer_dying_cpu); @@ -771,7 +999,7 @@ out_free_irq: void kvm_timer_vcpu_terminate(struct kvm_vcpu *vcpu) { - struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu; + struct arch_timer_cpu *timer = vcpu_timer(vcpu); soft_timer_cancel(&timer->bg_timer); } @@ -807,16 +1035,18 @@ bool kvm_arch_timer_get_input_level(int vintid) if (vintid == vcpu_vtimer(vcpu)->irq.irq) timer = vcpu_vtimer(vcpu); + else if (vintid == vcpu_ptimer(vcpu)->irq.irq) + timer = vcpu_ptimer(vcpu); else - BUG(); /* We only map the vtimer so far */ + BUG(); return kvm_timer_should_fire(timer); } int kvm_timer_enable(struct kvm_vcpu *vcpu) { - struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu; - struct arch_timer_context *vtimer = vcpu_vtimer(vcpu); + struct arch_timer_cpu *timer = vcpu_timer(vcpu); + struct timer_map map; int ret; if (timer->enabled) @@ -834,19 +1064,33 @@ int kvm_timer_enable(struct kvm_vcpu *vcpu) return -EINVAL; } - ret = kvm_vgic_map_phys_irq(vcpu, host_vtimer_irq, vtimer->irq.irq, + get_timer_map(vcpu, &map); + + ret = kvm_vgic_map_phys_irq(vcpu, + map.direct_vtimer->host_timer_irq, + map.direct_vtimer->irq.irq, kvm_arch_timer_get_input_level); if (ret) return ret; + if (map.direct_ptimer) { + ret = kvm_vgic_map_phys_irq(vcpu, + map.direct_ptimer->host_timer_irq, + map.direct_ptimer->irq.irq, + kvm_arch_timer_get_input_level); + } + + if (ret) + return ret; + no_vgic: timer->enabled = 1; return 0; } /* - * On VHE system, we only need to configure trap on physical timer and counter - * accesses in EL0 and EL1 once, not for every world switch. + * On VHE system, we only need to configure the EL2 timer trap register once, + * not for every world switch. * The host kernel runs at EL2 with HCR_EL2.TGE == 1, * and this makes those bits have no effect for the host kernel execution. */ @@ -857,11 +1101,11 @@ void kvm_timer_init_vhe(void) u64 val; /* - * Disallow physical timer access for the guest. - * Physical counter access is allowed. + * VHE systems allow the guest direct access to the EL1 physical + * timer/counter. */ val = read_sysreg(cnthctl_el2); - val &= ~(CNTHCTL_EL1PCEN << cnthctl_shift); + val |= (CNTHCTL_EL1PCEN << cnthctl_shift); val |= (CNTHCTL_EL1PCTEN << cnthctl_shift); write_sysreg(val, cnthctl_el2); } diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c index 9c486fad3f9f..99c37384ba7b 100644 --- a/virt/kvm/arm/arm.c +++ b/virt/kvm/arm/arm.c @@ -65,7 +65,6 @@ static DEFINE_PER_CPU(struct kvm_vcpu *, kvm_arm_running_vcpu); /* The VMID used in the VTTBR */ static atomic64_t kvm_vmid_gen = ATOMIC64_INIT(1); static u32 kvm_next_vmid; -static unsigned int kvm_vmid_bits __read_mostly; static DEFINE_SPINLOCK(kvm_vmid_lock); static bool vgic_present; @@ -142,7 +141,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) kvm_vgic_early_init(kvm); /* Mark the initial VMID generation invalid */ - kvm->arch.vmid_gen = 0; + kvm->arch.vmid.vmid_gen = 0; /* The maximum number of VCPUs is limited by the host's GIC model */ kvm->arch.max_vcpus = vgic_present ? @@ -336,13 +335,11 @@ int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu) void kvm_arch_vcpu_blocking(struct kvm_vcpu *vcpu) { - kvm_timer_schedule(vcpu); kvm_vgic_v4_enable_doorbell(vcpu); } void kvm_arch_vcpu_unblocking(struct kvm_vcpu *vcpu) { - kvm_timer_unschedule(vcpu); kvm_vgic_v4_disable_doorbell(vcpu); } @@ -472,37 +469,31 @@ void force_vm_exit(const cpumask_t *mask) /** * need_new_vmid_gen - check that the VMID is still valid - * @kvm: The VM's VMID to check + * @vmid: The VMID to check * * return true if there is a new generation of VMIDs being used * - * The hardware supports only 256 values with the value zero reserved for the - * host, so we check if an assigned value belongs to a previous generation, - * which which requires us to assign a new value. If we're the first to use a - * VMID for the new generation, we must flush necessary caches and TLBs on all - * CPUs. + * The hardware supports a limited set of values with the value zero reserved + * for the host, so we check if an assigned value belongs to a previous + * generation, which which requires us to assign a new value. If we're the + * first to use a VMID for the new generation, we must flush necessary caches + * and TLBs on all CPUs. */ -static bool need_new_vmid_gen(struct kvm *kvm) +static bool need_new_vmid_gen(struct kvm_vmid *vmid) { u64 current_vmid_gen = atomic64_read(&kvm_vmid_gen); smp_rmb(); /* Orders read of kvm_vmid_gen and kvm->arch.vmid */ - return unlikely(READ_ONCE(kvm->arch.vmid_gen) != current_vmid_gen); + return unlikely(READ_ONCE(vmid->vmid_gen) != current_vmid_gen); } /** - * update_vttbr - Update the VTTBR with a valid VMID before the guest runs - * @kvm The guest that we are about to run - * - * Called from kvm_arch_vcpu_ioctl_run before entering the guest to ensure the - * VM has a valid VMID, otherwise assigns a new one and flushes corresponding - * caches and TLBs. + * update_vmid - Update the vmid with a valid VMID for the current generation + * @kvm: The guest that struct vmid belongs to + * @vmid: The stage-2 VMID information struct */ -static void update_vttbr(struct kvm *kvm) +static void update_vmid(struct kvm_vmid *vmid) { - phys_addr_t pgd_phys; - u64 vmid, cnp = kvm_cpu_has_cnp() ? VTTBR_CNP_BIT : 0; - - if (!need_new_vmid_gen(kvm)) + if (!need_new_vmid_gen(vmid)) return; spin_lock(&kvm_vmid_lock); @@ -512,7 +503,7 @@ static void update_vttbr(struct kvm *kvm) * already allocated a valid vmid for this vm, then this vcpu should * use the same vmid. */ - if (!need_new_vmid_gen(kvm)) { + if (!need_new_vmid_gen(vmid)) { spin_unlock(&kvm_vmid_lock); return; } @@ -536,18 +527,12 @@ static void update_vttbr(struct kvm *kvm) kvm_call_hyp(__kvm_flush_vm_context); } - kvm->arch.vmid = kvm_next_vmid; + vmid->vmid = kvm_next_vmid; kvm_next_vmid++; - kvm_next_vmid &= (1 << kvm_vmid_bits) - 1; - - /* update vttbr to be used with the new vmid */ - pgd_phys = virt_to_phys(kvm->arch.pgd); - BUG_ON(pgd_phys & ~kvm_vttbr_baddr_mask(kvm)); - vmid = ((u64)(kvm->arch.vmid) << VTTBR_VMID_SHIFT) & VTTBR_VMID_MASK(kvm_vmid_bits); - kvm->arch.vttbr = kvm_phys_to_vttbr(pgd_phys) | vmid | cnp; + kvm_next_vmid &= (1 << kvm_get_vmid_bits()) - 1; smp_wmb(); - WRITE_ONCE(kvm->arch.vmid_gen, atomic64_read(&kvm_vmid_gen)); + WRITE_ONCE(vmid->vmid_gen, atomic64_read(&kvm_vmid_gen)); spin_unlock(&kvm_vmid_lock); } @@ -700,7 +685,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) */ cond_resched(); - update_vttbr(vcpu->kvm); + update_vmid(&vcpu->kvm->arch.vmid); check_vcpu_requests(vcpu); @@ -749,7 +734,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) */ smp_store_mb(vcpu->mode, IN_GUEST_MODE); - if (ret <= 0 || need_new_vmid_gen(vcpu->kvm) || + if (ret <= 0 || need_new_vmid_gen(&vcpu->kvm->arch.vmid) || kvm_request_pending(vcpu)) { vcpu->mode = OUTSIDE_GUEST_MODE; isb(); /* Ensure work in x_flush_hwstate is committed */ @@ -775,7 +760,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) ret = kvm_vcpu_run_vhe(vcpu); kvm_arm_vhe_guest_exit(); } else { - ret = kvm_call_hyp(__kvm_vcpu_run_nvhe, vcpu); + ret = kvm_call_hyp_ret(__kvm_vcpu_run_nvhe, vcpu); } vcpu->mode = OUTSIDE_GUEST_MODE; @@ -1427,10 +1412,6 @@ static inline void hyp_cpu_pm_exit(void) static int init_common_resources(void) { - /* set size of VMID supported by CPU */ - kvm_vmid_bits = kvm_get_vmid_bits(); - kvm_info("%d-bit VMID\n", kvm_vmid_bits); - kvm_set_ipa_limit(); return 0; @@ -1571,6 +1552,7 @@ static int init_hyp_mode(void) kvm_cpu_context_t *cpu_ctxt; cpu_ctxt = per_cpu_ptr(&kvm_host_cpu_state, cpu); + kvm_init_host_cpu_context(cpu_ctxt, cpu); err = create_hyp_mappings(cpu_ctxt, cpu_ctxt + 1, PAGE_HYP); if (err) { @@ -1581,7 +1563,7 @@ static int init_hyp_mode(void) err = hyp_map_aux_data(); if (err) - kvm_err("Cannot map host auxilary data: %d\n", err); + kvm_err("Cannot map host auxiliary data: %d\n", err); return 0; diff --git a/virt/kvm/arm/hyp/vgic-v3-sr.c b/virt/kvm/arm/hyp/vgic-v3-sr.c index 9652c453480f..370bd6c5e6cb 100644 --- a/virt/kvm/arm/hyp/vgic-v3-sr.c +++ b/virt/kvm/arm/hyp/vgic-v3-sr.c @@ -222,11 +222,11 @@ void __hyp_text __vgic_v3_save_state(struct kvm_vcpu *vcpu) } } - if (used_lrs) { + if (used_lrs || cpu_if->its_vpe.its_vm) { int i; u32 elrsr; - elrsr = read_gicreg(ICH_ELSR_EL2); + elrsr = read_gicreg(ICH_ELRSR_EL2); write_gicreg(cpu_if->vgic_hcr & ~ICH_HCR_EN, ICH_HCR_EL2); @@ -247,7 +247,7 @@ void __hyp_text __vgic_v3_restore_state(struct kvm_vcpu *vcpu) u64 used_lrs = vcpu->arch.vgic_cpu.used_lrs; int i; - if (used_lrs) { + if (used_lrs || cpu_if->its_vpe.its_vm) { write_gicreg(cpu_if->vgic_hcr, ICH_HCR_EL2); for (i = 0; i < used_lrs; i++) diff --git a/virt/kvm/arm/mmu.c b/virt/kvm/arm/mmu.c index e9d28a7ca673..27c958306449 100644 --- a/virt/kvm/arm/mmu.c +++ b/virt/kvm/arm/mmu.c @@ -102,8 +102,7 @@ static bool kvm_is_device_pfn(unsigned long pfn) * @addr: IPA * @pmd: pmd pointer for IPA * - * Function clears a PMD entry, flushes addr 1st and 2nd stage TLBs. Marks all - * pages in the range dirty. + * Function clears a PMD entry, flushes addr 1st and 2nd stage TLBs. */ static void stage2_dissolve_pmd(struct kvm *kvm, phys_addr_t addr, pmd_t *pmd) { @@ -121,8 +120,7 @@ static void stage2_dissolve_pmd(struct kvm *kvm, phys_addr_t addr, pmd_t *pmd) * @addr: IPA * @pud: pud pointer for IPA * - * Function clears a PUD entry, flushes addr 1st and 2nd stage TLBs. Marks all - * pages in the range dirty. + * Function clears a PUD entry, flushes addr 1st and 2nd stage TLBs. */ static void stage2_dissolve_pud(struct kvm *kvm, phys_addr_t addr, pud_t *pudp) { @@ -899,15 +897,15 @@ int create_hyp_exec_mappings(phys_addr_t phys_addr, size_t size, * kvm_alloc_stage2_pgd - allocate level-1 table for stage-2 translation. * @kvm: The KVM struct pointer for the VM. * - * Allocates only the stage-2 HW PGD level table(s) (can support either full - * 40-bit input addresses or limited to 32-bit input addresses). Clears the - * allocated pages. + * Allocates only the stage-2 HW PGD level table(s) of size defined by + * stage2_pgd_size(kvm). * * Note we don't need locking here as this is only called when the VM is * created, which can only be done once. */ int kvm_alloc_stage2_pgd(struct kvm *kvm) { + phys_addr_t pgd_phys; pgd_t *pgd; if (kvm->arch.pgd != NULL) { @@ -920,7 +918,12 @@ int kvm_alloc_stage2_pgd(struct kvm *kvm) if (!pgd) return -ENOMEM; + pgd_phys = virt_to_phys(pgd); + if (WARN_ON(pgd_phys & ~kvm_vttbr_baddr_mask(kvm))) + return -EINVAL; + kvm->arch.pgd = pgd; + kvm->arch.pgd_phys = pgd_phys; return 0; } @@ -1008,6 +1011,7 @@ void kvm_free_stage2_pgd(struct kvm *kvm) unmap_stage2_range(kvm, 0, kvm_phys_size(kvm)); pgd = READ_ONCE(kvm->arch.pgd); kvm->arch.pgd = NULL; + kvm->arch.pgd_phys = 0; } spin_unlock(&kvm->mmu_lock); @@ -1060,25 +1064,43 @@ static int stage2_set_pmd_huge(struct kvm *kvm, struct kvm_mmu_memory_cache { pmd_t *pmd, old_pmd; +retry: pmd = stage2_get_pmd(kvm, cache, addr); VM_BUG_ON(!pmd); old_pmd = *pmd; + /* + * Multiple vcpus faulting on the same PMD entry, can + * lead to them sequentially updating the PMD with the + * same value. Following the break-before-make + * (pmd_clear() followed by tlb_flush()) process can + * hinder forward progress due to refaults generated + * on missing translations. + * + * Skip updating the page table if the entry is + * unchanged. + */ + if (pmd_val(old_pmd) == pmd_val(*new_pmd)) + return 0; + if (pmd_present(old_pmd)) { /* - * Multiple vcpus faulting on the same PMD entry, can - * lead to them sequentially updating the PMD with the - * same value. Following the break-before-make - * (pmd_clear() followed by tlb_flush()) process can - * hinder forward progress due to refaults generated - * on missing translations. + * If we already have PTE level mapping for this block, + * we must unmap it to avoid inconsistent TLB state and + * leaking the table page. We could end up in this situation + * if the memory slot was marked for dirty logging and was + * reverted, leaving PTE level mappings for the pages accessed + * during the period. So, unmap the PTE level mapping for this + * block and retry, as we could have released the upper level + * table in the process. * - * Skip updating the page table if the entry is - * unchanged. + * Normal THP split/merge follows mmu_notifier callbacks and do + * get handled accordingly. */ - if (pmd_val(old_pmd) == pmd_val(*new_pmd)) - return 0; - + if (!pmd_thp_or_huge(old_pmd)) { + unmap_stage2_range(kvm, addr & S2_PMD_MASK, S2_PMD_SIZE); + goto retry; + } /* * Mapping in huge pages should only happen through a * fault. If a page is merged into a transparent huge @@ -1090,8 +1112,7 @@ static int stage2_set_pmd_huge(struct kvm *kvm, struct kvm_mmu_memory_cache * should become splitting first, unmapped, merged, * and mapped back in on-demand. */ - VM_BUG_ON(pmd_pfn(old_pmd) != pmd_pfn(*new_pmd)); - + WARN_ON_ONCE(pmd_pfn(old_pmd) != pmd_pfn(*new_pmd)); pmd_clear(pmd); kvm_tlb_flush_vmid_ipa(kvm, addr); } else { @@ -1107,6 +1128,7 @@ static int stage2_set_pud_huge(struct kvm *kvm, struct kvm_mmu_memory_cache *cac { pud_t *pudp, old_pud; +retry: pudp = stage2_get_pud(kvm, cache, addr); VM_BUG_ON(!pudp); @@ -1114,14 +1136,23 @@ static int stage2_set_pud_huge(struct kvm *kvm, struct kvm_mmu_memory_cache *cac /* * A large number of vcpus faulting on the same stage 2 entry, - * can lead to a refault due to the - * stage2_pud_clear()/tlb_flush(). Skip updating the page - * tables if there is no change. + * can lead to a refault due to the stage2_pud_clear()/tlb_flush(). + * Skip updating the page tables if there is no change. */ if (pud_val(old_pud) == pud_val(*new_pudp)) return 0; if (stage2_pud_present(kvm, old_pud)) { + /* + * If we already have table level mapping for this block, unmap + * the range for this block and retry. + */ + if (!stage2_pud_huge(kvm, old_pud)) { + unmap_stage2_range(kvm, addr & S2_PUD_MASK, S2_PUD_SIZE); + goto retry; + } + + WARN_ON_ONCE(kvm_pud_pfn(old_pud) != kvm_pud_pfn(*new_pudp)); stage2_pud_clear(kvm, pudp); kvm_tlb_flush_vmid_ipa(kvm, addr); } else { @@ -1396,14 +1427,6 @@ static bool transparent_hugepage_adjust(kvm_pfn_t *pfnp, phys_addr_t *ipap) return false; } -static bool kvm_is_write_fault(struct kvm_vcpu *vcpu) -{ - if (kvm_vcpu_trap_is_iabt(vcpu)) - return false; - - return kvm_vcpu_dabt_iswrite(vcpu); -} - /** * stage2_wp_ptes - write protect PMD range * @pmd: pointer to pmd entry @@ -1452,13 +1475,11 @@ static void stage2_wp_pmds(struct kvm *kvm, pud_t *pud, } /** - * stage2_wp_puds - write protect PGD range - * @pgd: pointer to pgd entry - * @addr: range start address - * @end: range end address - * - * Process PUD entries, for a huge PUD we cause a panic. - */ + * stage2_wp_puds - write protect PGD range + * @pgd: pointer to pgd entry + * @addr: range start address + * @end: range end address + */ static void stage2_wp_puds(struct kvm *kvm, pgd_t *pgd, phys_addr_t addr, phys_addr_t end) { @@ -1595,51 +1616,51 @@ static void kvm_send_hwpoison_signal(unsigned long address, send_sig_mceerr(BUS_MCEERR_AR, (void __user *)address, lsb, current); } -static bool fault_supports_stage2_pmd_mappings(struct kvm_memory_slot *memslot, - unsigned long hva) +static bool fault_supports_stage2_huge_mapping(struct kvm_memory_slot *memslot, + unsigned long hva, + unsigned long map_size) { - gpa_t gpa_start, gpa_end; + gpa_t gpa_start; hva_t uaddr_start, uaddr_end; size_t size; size = memslot->npages * PAGE_SIZE; gpa_start = memslot->base_gfn << PAGE_SHIFT; - gpa_end = gpa_start + size; uaddr_start = memslot->userspace_addr; uaddr_end = uaddr_start + size; /* * Pages belonging to memslots that don't have the same alignment - * within a PMD for userspace and IPA cannot be mapped with stage-2 - * PMD entries, because we'll end up mapping the wrong pages. + * within a PMD/PUD for userspace and IPA cannot be mapped with stage-2 + * PMD/PUD entries, because we'll end up mapping the wrong pages. * * Consider a layout like the following: * * memslot->userspace_addr: * +-----+--------------------+--------------------+---+ - * |abcde|fgh Stage-1 PMD | Stage-1 PMD tv|xyz| + * |abcde|fgh Stage-1 block | Stage-1 block tv|xyz| * +-----+--------------------+--------------------+---+ * * memslot->base_gfn << PAGE_SIZE: * +---+--------------------+--------------------+-----+ - * |abc|def Stage-2 PMD | Stage-2 PMD |tvxyz| + * |abc|def Stage-2 block | Stage-2 block |tvxyz| * +---+--------------------+--------------------+-----+ * - * If we create those stage-2 PMDs, we'll end up with this incorrect + * If we create those stage-2 blocks, we'll end up with this incorrect * mapping: * d -> f * e -> g * f -> h */ - if ((gpa_start & ~S2_PMD_MASK) != (uaddr_start & ~S2_PMD_MASK)) + if ((gpa_start & (map_size - 1)) != (uaddr_start & (map_size - 1))) return false; /* * Next, let's make sure we're not trying to map anything not covered - * by the memslot. This means we have to prohibit PMD size mappings - * for the beginning and end of a non-PMD aligned and non-PMD sized + * by the memslot. This means we have to prohibit block size mappings + * for the beginning and end of a non-block aligned and non-block sized * memory slot (illustrated by the head and tail parts of the * userspace view above containing pages 'abcde' and 'xyz', * respectively). @@ -1648,8 +1669,8 @@ static bool fault_supports_stage2_pmd_mappings(struct kvm_memory_slot *memslot, * userspace_addr or the base_gfn, as both are equally aligned (per * the check above) and equally sized. */ - return (hva & S2_PMD_MASK) >= uaddr_start && - (hva & S2_PMD_MASK) + S2_PMD_SIZE <= uaddr_end; + return (hva & ~(map_size - 1)) >= uaddr_start && + (hva & ~(map_size - 1)) + map_size <= uaddr_end; } static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa, @@ -1678,12 +1699,6 @@ static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa, return -EFAULT; } - if (!fault_supports_stage2_pmd_mappings(memslot, hva)) - force_pte = true; - - if (logging_active) - force_pte = true; - /* Let's check if we will get back a huge page backed by hugetlbfs */ down_read(¤t->mm->mmap_sem); vma = find_vma_intersection(current->mm, hva, hva + 1); @@ -1694,6 +1709,12 @@ static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa, } vma_pagesize = vma_kernel_pagesize(vma); + if (logging_active || + !fault_supports_stage2_huge_mapping(memslot, hva, vma_pagesize)) { + force_pte = true; + vma_pagesize = PAGE_SIZE; + } + /* * The stage2 has a minimum of 2 level table (For arm64 see * kvm_arm_setup_stage2()). Hence, we are guaranteed that we can @@ -1701,11 +1722,9 @@ static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa, * As for PUD huge maps, we must make sure that we have at least * 3 levels, i.e, PMD is not folded. */ - if ((vma_pagesize == PMD_SIZE || - (vma_pagesize == PUD_SIZE && kvm_stage2_has_pmd(kvm))) && - !force_pte) { + if (vma_pagesize == PMD_SIZE || + (vma_pagesize == PUD_SIZE && kvm_stage2_has_pmd(kvm))) gfn = (fault_ipa & huge_page_mask(hstate_vma(vma))) >> PAGE_SHIFT; - } up_read(¤t->mm->mmap_sem); /* We need minimum second+third level pages */ @@ -2353,7 +2372,7 @@ int kvm_arch_create_memslot(struct kvm *kvm, struct kvm_memory_slot *slot, return 0; } -void kvm_arch_memslots_updated(struct kvm *kvm, struct kvm_memslots *slots) +void kvm_arch_memslots_updated(struct kvm *kvm, u64 gen) { } diff --git a/virt/kvm/arm/trace.h b/virt/kvm/arm/trace.h index 3828beab93f2..204d210d01c2 100644 --- a/virt/kvm/arm/trace.h +++ b/virt/kvm/arm/trace.h @@ -2,6 +2,7 @@ #if !defined(_TRACE_KVM_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_KVM_H +#include <kvm/arm_arch_timer.h> #include <linux/tracepoint.h> #undef TRACE_SYSTEM @@ -262,10 +263,114 @@ TRACE_EVENT(kvm_timer_update_irq, __entry->vcpu_id, __entry->irq, __entry->level) ); +TRACE_EVENT(kvm_get_timer_map, + TP_PROTO(unsigned long vcpu_id, struct timer_map *map), + TP_ARGS(vcpu_id, map), + + TP_STRUCT__entry( + __field( unsigned long, vcpu_id ) + __field( int, direct_vtimer ) + __field( int, direct_ptimer ) + __field( int, emul_ptimer ) + ), + + TP_fast_assign( + __entry->vcpu_id = vcpu_id; + __entry->direct_vtimer = arch_timer_ctx_index(map->direct_vtimer); + __entry->direct_ptimer = + (map->direct_ptimer) ? arch_timer_ctx_index(map->direct_ptimer) : -1; + __entry->emul_ptimer = + (map->emul_ptimer) ? arch_timer_ctx_index(map->emul_ptimer) : -1; + ), + + TP_printk("VCPU: %ld, dv: %d, dp: %d, ep: %d", + __entry->vcpu_id, + __entry->direct_vtimer, + __entry->direct_ptimer, + __entry->emul_ptimer) +); + +TRACE_EVENT(kvm_timer_save_state, + TP_PROTO(struct arch_timer_context *ctx), + TP_ARGS(ctx), + + TP_STRUCT__entry( + __field( unsigned long, ctl ) + __field( unsigned long long, cval ) + __field( int, timer_idx ) + ), + + TP_fast_assign( + __entry->ctl = ctx->cnt_ctl; + __entry->cval = ctx->cnt_cval; + __entry->timer_idx = arch_timer_ctx_index(ctx); + ), + + TP_printk(" CTL: %#08lx CVAL: %#16llx arch_timer_ctx_index: %d", + __entry->ctl, + __entry->cval, + __entry->timer_idx) +); + +TRACE_EVENT(kvm_timer_restore_state, + TP_PROTO(struct arch_timer_context *ctx), + TP_ARGS(ctx), + + TP_STRUCT__entry( + __field( unsigned long, ctl ) + __field( unsigned long long, cval ) + __field( int, timer_idx ) + ), + + TP_fast_assign( + __entry->ctl = ctx->cnt_ctl; + __entry->cval = ctx->cnt_cval; + __entry->timer_idx = arch_timer_ctx_index(ctx); + ), + + TP_printk("CTL: %#08lx CVAL: %#16llx arch_timer_ctx_index: %d", + __entry->ctl, + __entry->cval, + __entry->timer_idx) +); + +TRACE_EVENT(kvm_timer_hrtimer_expire, + TP_PROTO(struct arch_timer_context *ctx), + TP_ARGS(ctx), + + TP_STRUCT__entry( + __field( int, timer_idx ) + ), + + TP_fast_assign( + __entry->timer_idx = arch_timer_ctx_index(ctx); + ), + + TP_printk("arch_timer_ctx_index: %d", __entry->timer_idx) +); + +TRACE_EVENT(kvm_timer_emulate, + TP_PROTO(struct arch_timer_context *ctx, bool should_fire), + TP_ARGS(ctx, should_fire), + + TP_STRUCT__entry( + __field( int, timer_idx ) + __field( bool, should_fire ) + ), + + TP_fast_assign( + __entry->timer_idx = arch_timer_ctx_index(ctx); + __entry->should_fire = should_fire; + ), + + TP_printk("arch_timer_ctx_index: %d (should_fire: %d)", + __entry->timer_idx, __entry->should_fire) +); + #endif /* _TRACE_KVM_H */ #undef TRACE_INCLUDE_PATH -#define TRACE_INCLUDE_PATH ../../../virt/kvm/arm +#define TRACE_INCLUDE_PATH ../../virt/kvm/arm #undef TRACE_INCLUDE_FILE #define TRACE_INCLUDE_FILE trace diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c index ab3f47745d9c..44ceaccb18cf 100644 --- a/virt/kvm/arm/vgic/vgic-its.c +++ b/virt/kvm/arm/vgic/vgic-its.c @@ -754,8 +754,9 @@ static bool vgic_its_check_id(struct vgic_its *its, u64 baser, u32 id, u64 indirect_ptr, type = GITS_BASER_TYPE(baser); phys_addr_t base = GITS_BASER_ADDR_48_to_52(baser); int esz = GITS_BASER_ENTRY_SIZE(baser); - int index; + int index, idx; gfn_t gfn; + bool ret; switch (type) { case GITS_BASER_TYPE_DEVICE: @@ -782,7 +783,8 @@ static bool vgic_its_check_id(struct vgic_its *its, u64 baser, u32 id, if (eaddr) *eaddr = addr; - return kvm_is_visible_gfn(its->dev->kvm, gfn); + + goto out; } /* calculate and check the index into the 1st level */ @@ -812,7 +814,12 @@ static bool vgic_its_check_id(struct vgic_its *its, u64 baser, u32 id, if (eaddr) *eaddr = indirect_ptr; - return kvm_is_visible_gfn(its->dev->kvm, gfn); + +out: + idx = srcu_read_lock(&its->dev->kvm->srcu); + ret = kvm_is_visible_gfn(its->dev->kvm, gfn); + srcu_read_unlock(&its->dev->kvm->srcu, idx); + return ret; } static int vgic_its_alloc_collection(struct vgic_its *its, @@ -1729,8 +1736,8 @@ static void vgic_its_destroy(struct kvm_device *kvm_dev) kfree(its); } -int vgic_its_has_attr_regs(struct kvm_device *dev, - struct kvm_device_attr *attr) +static int vgic_its_has_attr_regs(struct kvm_device *dev, + struct kvm_device_attr *attr) { const struct vgic_register_region *region; gpa_t offset = attr->attr; @@ -1750,9 +1757,9 @@ int vgic_its_has_attr_regs(struct kvm_device *dev, return 0; } -int vgic_its_attr_regs_access(struct kvm_device *dev, - struct kvm_device_attr *attr, - u64 *reg, bool is_write) +static int vgic_its_attr_regs_access(struct kvm_device *dev, + struct kvm_device_attr *attr, + u64 *reg, bool is_write) { const struct vgic_register_region *region; struct vgic_its *its; @@ -1919,7 +1926,7 @@ static int vgic_its_save_ite(struct vgic_its *its, struct its_device *dev, ((u64)ite->irq->intid << KVM_ITS_ITE_PINTID_SHIFT) | ite->collection->collection_id; val = cpu_to_le64(val); - return kvm_write_guest(kvm, gpa, &val, ite_esz); + return kvm_write_guest_lock(kvm, gpa, &val, ite_esz); } /** @@ -2066,7 +2073,7 @@ static int vgic_its_save_dte(struct vgic_its *its, struct its_device *dev, (itt_addr_field << KVM_ITS_DTE_ITTADDR_SHIFT) | (dev->num_eventid_bits - 1)); val = cpu_to_le64(val); - return kvm_write_guest(kvm, ptr, &val, dte_esz); + return kvm_write_guest_lock(kvm, ptr, &val, dte_esz); } /** @@ -2246,7 +2253,7 @@ static int vgic_its_save_cte(struct vgic_its *its, ((u64)collection->target_addr << KVM_ITS_CTE_RDBASE_SHIFT) | collection->collection_id); val = cpu_to_le64(val); - return kvm_write_guest(its->dev->kvm, gpa, &val, esz); + return kvm_write_guest_lock(its->dev->kvm, gpa, &val, esz); } static int vgic_its_restore_cte(struct vgic_its *its, gpa_t gpa, int esz) @@ -2317,7 +2324,7 @@ static int vgic_its_save_collection_table(struct vgic_its *its) */ val = 0; BUG_ON(cte_esz > sizeof(val)); - ret = kvm_write_guest(its->dev->kvm, gpa, &val, cte_esz); + ret = kvm_write_guest_lock(its->dev->kvm, gpa, &val, cte_esz); return ret; } diff --git a/virt/kvm/arm/vgic/vgic-v3.c b/virt/kvm/arm/vgic/vgic-v3.c index 4ee0aeb9a905..9f87e58dbd4a 100644 --- a/virt/kvm/arm/vgic/vgic-v3.c +++ b/virt/kvm/arm/vgic/vgic-v3.c @@ -358,7 +358,7 @@ retry: if (status) { /* clear consumed data */ val &= ~(1 << bit_nr); - ret = kvm_write_guest(kvm, ptr, &val, 1); + ret = kvm_write_guest_lock(kvm, ptr, &val, 1); if (ret) return ret; } @@ -409,7 +409,7 @@ int vgic_v3_save_pending_tables(struct kvm *kvm) else val &= ~(1 << bit_nr); - ret = kvm_write_guest(kvm, ptr, &val, 1); + ret = kvm_write_guest_lock(kvm, ptr, &val, 1); if (ret) return ret; } @@ -589,7 +589,7 @@ early_param("kvm-arm.vgic_v4_enable", early_gicv4_enable); */ int vgic_v3_probe(const struct gic_kvm_info *info) { - u32 ich_vtr_el2 = kvm_call_hyp(__vgic_v3_get_ich_vtr_el2); + u32 ich_vtr_el2 = kvm_call_hyp_ret(__vgic_v3_get_ich_vtr_el2); int ret; /* @@ -679,7 +679,7 @@ void vgic_v3_put(struct kvm_vcpu *vcpu) struct vgic_v3_cpu_if *cpu_if = &vcpu->arch.vgic_cpu.vgic_v3; if (likely(cpu_if->vgic_sre)) - cpu_if->vgic_vmcr = kvm_call_hyp(__vgic_v3_read_vmcr); + cpu_if->vgic_vmcr = kvm_call_hyp_ret(__vgic_v3_read_vmcr); kvm_call_hyp(__vgic_v3_save_aprs, vcpu); diff --git a/virt/kvm/arm/vgic/vgic.c b/virt/kvm/arm/vgic/vgic.c index abd9c7352677..3af69f2a3866 100644 --- a/virt/kvm/arm/vgic/vgic.c +++ b/virt/kvm/arm/vgic/vgic.c @@ -867,15 +867,21 @@ void kvm_vgic_flush_hwstate(struct kvm_vcpu *vcpu) * either observe the new interrupt before or after doing this check, * and introducing additional synchronization mechanism doesn't change * this. + * + * Note that we still need to go through the whole thing if anything + * can be directly injected (GICv4). */ - if (list_empty(&vcpu->arch.vgic_cpu.ap_list_head)) + if (list_empty(&vcpu->arch.vgic_cpu.ap_list_head) && + !vgic_supports_direct_msis(vcpu->kvm)) return; DEBUG_SPINLOCK_BUG_ON(!irqs_disabled()); - raw_spin_lock(&vcpu->arch.vgic_cpu.ap_list_lock); - vgic_flush_lr_state(vcpu); - raw_spin_unlock(&vcpu->arch.vgic_cpu.ap_list_lock); + if (!list_empty(&vcpu->arch.vgic_cpu.ap_list_head)) { + raw_spin_lock(&vcpu->arch.vgic_cpu.ap_list_lock); + vgic_flush_lr_state(vcpu); + raw_spin_unlock(&vcpu->arch.vgic_cpu.ap_list_lock); + } if (can_access_vgic_from_kernel()) vgic_restore_state(vcpu); diff --git a/virt/kvm/coalesced_mmio.c b/virt/kvm/coalesced_mmio.c index 6855cce3e528..5294abb3f178 100644 --- a/virt/kvm/coalesced_mmio.c +++ b/virt/kvm/coalesced_mmio.c @@ -144,7 +144,8 @@ int kvm_vm_ioctl_register_coalesced_mmio(struct kvm *kvm, if (zone->pio != 1 && zone->pio != 0) return -EINVAL; - dev = kzalloc(sizeof(struct kvm_coalesced_mmio_dev), GFP_KERNEL); + dev = kzalloc(sizeof(struct kvm_coalesced_mmio_dev), + GFP_KERNEL_ACCOUNT); if (!dev) return -ENOMEM; diff --git a/virt/kvm/eventfd.c b/virt/kvm/eventfd.c index b20b751286fc..001aeda4c154 100644 --- a/virt/kvm/eventfd.c +++ b/virt/kvm/eventfd.c @@ -214,9 +214,9 @@ irqfd_wakeup(wait_queue_entry_t *wait, unsigned mode, int sync, void *key) if (flags & EPOLLHUP) { /* The eventfd is closing, detach from KVM */ - unsigned long flags; + unsigned long iflags; - spin_lock_irqsave(&kvm->irqfds.lock, flags); + spin_lock_irqsave(&kvm->irqfds.lock, iflags); /* * We must check if someone deactivated the irqfd before @@ -230,7 +230,7 @@ irqfd_wakeup(wait_queue_entry_t *wait, unsigned mode, int sync, void *key) if (irqfd_is_active(irqfd)) irqfd_deactivate(irqfd); - spin_unlock_irqrestore(&kvm->irqfds.lock, flags); + spin_unlock_irqrestore(&kvm->irqfds.lock, iflags); } return 0; @@ -297,7 +297,7 @@ kvm_irqfd_assign(struct kvm *kvm, struct kvm_irqfd *args) if (!kvm_arch_intc_initialized(kvm)) return -EAGAIN; - irqfd = kzalloc(sizeof(*irqfd), GFP_KERNEL); + irqfd = kzalloc(sizeof(*irqfd), GFP_KERNEL_ACCOUNT); if (!irqfd) return -ENOMEM; @@ -345,7 +345,8 @@ kvm_irqfd_assign(struct kvm *kvm, struct kvm_irqfd *args) } if (!irqfd->resampler) { - resampler = kzalloc(sizeof(*resampler), GFP_KERNEL); + resampler = kzalloc(sizeof(*resampler), + GFP_KERNEL_ACCOUNT); if (!resampler) { ret = -ENOMEM; mutex_unlock(&kvm->irqfds.resampler_lock); @@ -797,7 +798,7 @@ static int kvm_assign_ioeventfd_idx(struct kvm *kvm, if (IS_ERR(eventfd)) return PTR_ERR(eventfd); - p = kzalloc(sizeof(*p), GFP_KERNEL); + p = kzalloc(sizeof(*p), GFP_KERNEL_ACCOUNT); if (!p) { ret = -ENOMEM; goto fail; diff --git a/virt/kvm/irqchip.c b/virt/kvm/irqchip.c index b1286c4e0712..3547b0d8c91e 100644 --- a/virt/kvm/irqchip.c +++ b/virt/kvm/irqchip.c @@ -196,7 +196,7 @@ int kvm_set_irq_routing(struct kvm *kvm, nr_rt_entries += 1; new = kzalloc(sizeof(*new) + (nr_rt_entries * sizeof(struct hlist_head)), - GFP_KERNEL); + GFP_KERNEL_ACCOUNT); if (!new) return -ENOMEM; @@ -208,7 +208,7 @@ int kvm_set_irq_routing(struct kvm *kvm, for (i = 0; i < nr; ++i) { r = -ENOMEM; - e = kzalloc(sizeof(*e), GFP_KERNEL); + e = kzalloc(sizeof(*e), GFP_KERNEL_ACCOUNT); if (!e) goto out; diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index d237d3350a99..55fe8e20d8fd 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -81,6 +81,11 @@ unsigned int halt_poll_ns_grow = 2; module_param(halt_poll_ns_grow, uint, 0644); EXPORT_SYMBOL_GPL(halt_poll_ns_grow); +/* The start value to grow halt_poll_ns from */ +unsigned int halt_poll_ns_grow_start = 10000; /* 10us */ +module_param(halt_poll_ns_grow_start, uint, 0644); +EXPORT_SYMBOL_GPL(halt_poll_ns_grow_start); + /* Default resets per-vcpu halt_poll_ns . */ unsigned int halt_poll_ns_shrink; module_param(halt_poll_ns_shrink, uint, 0644); @@ -525,7 +530,7 @@ static struct kvm_memslots *kvm_alloc_memslots(void) int i; struct kvm_memslots *slots; - slots = kvzalloc(sizeof(struct kvm_memslots), GFP_KERNEL); + slots = kvzalloc(sizeof(struct kvm_memslots), GFP_KERNEL_ACCOUNT); if (!slots) return NULL; @@ -601,12 +606,12 @@ static int kvm_create_vm_debugfs(struct kvm *kvm, int fd) kvm->debugfs_stat_data = kcalloc(kvm_debugfs_num_entries, sizeof(*kvm->debugfs_stat_data), - GFP_KERNEL); + GFP_KERNEL_ACCOUNT); if (!kvm->debugfs_stat_data) return -ENOMEM; for (p = debugfs_entries; p->name; p++) { - stat_data = kzalloc(sizeof(*stat_data), GFP_KERNEL); + stat_data = kzalloc(sizeof(*stat_data), GFP_KERNEL_ACCOUNT); if (!stat_data) return -ENOMEM; @@ -656,12 +661,8 @@ static struct kvm *kvm_create_vm(unsigned long type) struct kvm_memslots *slots = kvm_alloc_memslots(); if (!slots) goto out_err_no_srcu; - /* - * Generations must be different for each address space. - * Init kvm generation close to the maximum to easily test the - * code of handling generation number wrap-around. - */ - slots->generation = i * 2 - 150; + /* Generations must be different for each address space. */ + slots->generation = i; rcu_assign_pointer(kvm->memslots[i], slots); } @@ -671,7 +672,7 @@ static struct kvm *kvm_create_vm(unsigned long type) goto out_err_no_irq_srcu; for (i = 0; i < KVM_NR_BUSES; i++) { rcu_assign_pointer(kvm->buses[i], - kzalloc(sizeof(struct kvm_io_bus), GFP_KERNEL)); + kzalloc(sizeof(struct kvm_io_bus), GFP_KERNEL_ACCOUNT)); if (!kvm->buses[i]) goto out_err; } @@ -789,7 +790,7 @@ static int kvm_create_dirty_bitmap(struct kvm_memory_slot *memslot) { unsigned long dirty_bytes = 2 * kvm_dirty_bitmap_bytes(memslot); - memslot->dirty_bitmap = kvzalloc(dirty_bytes, GFP_KERNEL); + memslot->dirty_bitmap = kvzalloc(dirty_bytes, GFP_KERNEL_ACCOUNT); if (!memslot->dirty_bitmap) return -ENOMEM; @@ -874,31 +875,34 @@ static struct kvm_memslots *install_new_memslots(struct kvm *kvm, int as_id, struct kvm_memslots *slots) { struct kvm_memslots *old_memslots = __kvm_memslots(kvm, as_id); + u64 gen = old_memslots->generation; - /* - * Set the low bit in the generation, which disables SPTE caching - * until the end of synchronize_srcu_expedited. - */ - WARN_ON(old_memslots->generation & 1); - slots->generation = old_memslots->generation + 1; + WARN_ON(gen & KVM_MEMSLOT_GEN_UPDATE_IN_PROGRESS); + slots->generation = gen | KVM_MEMSLOT_GEN_UPDATE_IN_PROGRESS; rcu_assign_pointer(kvm->memslots[as_id], slots); synchronize_srcu_expedited(&kvm->srcu); /* - * Increment the new memslot generation a second time. This prevents - * vm exits that race with memslot updates from caching a memslot - * generation that will (potentially) be valid forever. - * + * Increment the new memslot generation a second time, dropping the + * update in-progress flag and incrementing then generation based on + * the number of address spaces. This provides a unique and easily + * identifiable generation number while the memslots are in flux. + */ + gen = slots->generation & ~KVM_MEMSLOT_GEN_UPDATE_IN_PROGRESS; + + /* * Generations must be unique even across address spaces. We do not need * a global counter for that, instead the generation space is evenly split * across address spaces. For example, with two address spaces, address - * space 0 will use generations 0, 4, 8, ... while * address space 1 will - * use generations 2, 6, 10, 14, ... + * space 0 will use generations 0, 2, 4, ... while address space 1 will + * use generations 1, 3, 5, ... */ - slots->generation += KVM_ADDRESS_SPACE_NUM * 2 - 1; + gen += KVM_ADDRESS_SPACE_NUM; - kvm_arch_memslots_updated(kvm, slots); + kvm_arch_memslots_updated(kvm, gen); + + slots->generation = gen; return old_memslots; } @@ -1018,7 +1022,7 @@ int __kvm_set_memory_region(struct kvm *kvm, goto out_free; } - slots = kvzalloc(sizeof(struct kvm_memslots), GFP_KERNEL); + slots = kvzalloc(sizeof(struct kvm_memslots), GFP_KERNEL_ACCOUNT); if (!slots) goto out_free; memcpy(slots, __kvm_memslots(kvm, as_id), sizeof(struct kvm_memslots)); @@ -1201,11 +1205,9 @@ int kvm_get_dirty_log_protect(struct kvm *kvm, mask = xchg(&dirty_bitmap[i], 0); dirty_bitmap_buffer[i] = mask; - if (mask) { - offset = i * BITS_PER_LONG; - kvm_arch_mmu_enable_log_dirty_pt_masked(kvm, memslot, - offset, mask); - } + offset = i * BITS_PER_LONG; + kvm_arch_mmu_enable_log_dirty_pt_masked(kvm, memslot, + offset, mask); } spin_unlock(&kvm->mmu_lock); } @@ -2185,20 +2187,23 @@ void kvm_sigset_deactivate(struct kvm_vcpu *vcpu) static void grow_halt_poll_ns(struct kvm_vcpu *vcpu) { - unsigned int old, val, grow; + unsigned int old, val, grow, grow_start; old = val = vcpu->halt_poll_ns; + grow_start = READ_ONCE(halt_poll_ns_grow_start); grow = READ_ONCE(halt_poll_ns_grow); - /* 10us base */ - if (val == 0 && grow) - val = 10000; - else - val *= grow; + if (!grow) + goto out; + + val *= grow; + if (val < grow_start) + val = grow_start; if (val > halt_poll_ns) val = halt_poll_ns; vcpu->halt_poll_ns = val; +out: trace_kvm_halt_poll_ns_grow(vcpu->vcpu_id, val, old); } @@ -2683,7 +2688,7 @@ static long kvm_vcpu_ioctl(struct file *filp, struct kvm_regs *kvm_regs; r = -ENOMEM; - kvm_regs = kzalloc(sizeof(struct kvm_regs), GFP_KERNEL); + kvm_regs = kzalloc(sizeof(struct kvm_regs), GFP_KERNEL_ACCOUNT); if (!kvm_regs) goto out; r = kvm_arch_vcpu_ioctl_get_regs(vcpu, kvm_regs); @@ -2711,7 +2716,8 @@ out_free1: break; } case KVM_GET_SREGS: { - kvm_sregs = kzalloc(sizeof(struct kvm_sregs), GFP_KERNEL); + kvm_sregs = kzalloc(sizeof(struct kvm_sregs), + GFP_KERNEL_ACCOUNT); r = -ENOMEM; if (!kvm_sregs) goto out; @@ -2803,7 +2809,7 @@ out_free1: break; } case KVM_GET_FPU: { - fpu = kzalloc(sizeof(struct kvm_fpu), GFP_KERNEL); + fpu = kzalloc(sizeof(struct kvm_fpu), GFP_KERNEL_ACCOUNT); r = -ENOMEM; if (!fpu) goto out; @@ -2899,6 +2905,9 @@ static long kvm_device_ioctl(struct file *filp, unsigned int ioctl, { struct kvm_device *dev = filp->private_data; + if (dev->kvm->mm != current->mm) + return -EIO; + switch (ioctl) { case KVM_SET_DEVICE_ATTR: return kvm_device_ioctl_attr(dev, dev->ops->set_attr, arg); @@ -2980,7 +2989,7 @@ static int kvm_ioctl_create_device(struct kvm *kvm, if (test) return 0; - dev = kzalloc(sizeof(*dev), GFP_KERNEL); + dev = kzalloc(sizeof(*dev), GFP_KERNEL_ACCOUNT); if (!dev) return -ENOMEM; @@ -3625,6 +3634,7 @@ int kvm_io_bus_write(struct kvm_vcpu *vcpu, enum kvm_bus bus_idx, gpa_t addr, r = __kvm_io_bus_write(vcpu, bus, &range, val); return r < 0 ? r : 0; } +EXPORT_SYMBOL_GPL(kvm_io_bus_write); /* kvm_io_bus_write_cookie - called under kvm->slots_lock */ int kvm_io_bus_write_cookie(struct kvm_vcpu *vcpu, enum kvm_bus bus_idx, @@ -3675,7 +3685,6 @@ static int __kvm_io_bus_read(struct kvm_vcpu *vcpu, struct kvm_io_bus *bus, return -EOPNOTSUPP; } -EXPORT_SYMBOL_GPL(kvm_io_bus_write); /* kvm_io_bus_read - called under kvm->slots_lock */ int kvm_io_bus_read(struct kvm_vcpu *vcpu, enum kvm_bus bus_idx, gpa_t addr, @@ -3697,7 +3706,6 @@ int kvm_io_bus_read(struct kvm_vcpu *vcpu, enum kvm_bus bus_idx, gpa_t addr, return r < 0 ? r : 0; } - /* Caller must hold slots_lock. */ int kvm_io_bus_register_dev(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr, int len, struct kvm_io_device *dev) @@ -3714,8 +3722,8 @@ int kvm_io_bus_register_dev(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr, if (bus->dev_count - bus->ioeventfd_count > NR_IOBUS_DEVS - 1) return -ENOSPC; - new_bus = kmalloc(sizeof(*bus) + ((bus->dev_count + 1) * - sizeof(struct kvm_io_range)), GFP_KERNEL); + new_bus = kmalloc(struct_size(bus, range, bus->dev_count + 1), + GFP_KERNEL_ACCOUNT); if (!new_bus) return -ENOMEM; @@ -3760,8 +3768,8 @@ void kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx, if (i == bus->dev_count) return; - new_bus = kmalloc(sizeof(*bus) + ((bus->dev_count - 1) * - sizeof(struct kvm_io_range)), GFP_KERNEL); + new_bus = kmalloc(struct_size(bus, range, bus->dev_count - 1), + GFP_KERNEL_ACCOUNT); if (!new_bus) { pr_err("kvm: failed to shrink bus, removing it completely\n"); goto broken; @@ -4029,7 +4037,7 @@ static void kvm_uevent_notify_change(unsigned int type, struct kvm *kvm) active = kvm_active_vms; spin_unlock(&kvm_lock); - env = kzalloc(sizeof(*env), GFP_KERNEL); + env = kzalloc(sizeof(*env), GFP_KERNEL_ACCOUNT); if (!env) return; @@ -4045,7 +4053,7 @@ static void kvm_uevent_notify_change(unsigned int type, struct kvm *kvm) add_uevent_var(env, "PID=%d", kvm->userspace_pid); if (!IS_ERR_OR_NULL(kvm->debugfs_dentry)) { - char *tmp, *p = kmalloc(PATH_MAX, GFP_KERNEL); + char *tmp, *p = kmalloc(PATH_MAX, GFP_KERNEL_ACCOUNT); if (p) { tmp = dentry_path_raw(kvm->debugfs_dentry, p, PATH_MAX); diff --git a/virt/kvm/vfio.c b/virt/kvm/vfio.c index d99850c462a1..524cbd20379f 100644 --- a/virt/kvm/vfio.c +++ b/virt/kvm/vfio.c @@ -219,7 +219,7 @@ static int kvm_vfio_set_group(struct kvm_device *dev, long attr, u64 arg) } } - kvg = kzalloc(sizeof(*kvg), GFP_KERNEL); + kvg = kzalloc(sizeof(*kvg), GFP_KERNEL_ACCOUNT); if (!kvg) { mutex_unlock(&kv->lock); kvm_vfio_group_put_external_user(vfio_group); @@ -405,7 +405,7 @@ static int kvm_vfio_create(struct kvm_device *dev, u32 type) if (tmp->ops == &kvm_vfio_ops) return -EBUSY; - kv = kzalloc(sizeof(*kv), GFP_KERNEL); + kv = kzalloc(sizeof(*kv), GFP_KERNEL_ACCOUNT); if (!kv) return -ENOMEM; |