^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) // SPDX-License-Identifier: GPL-2.0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) * KVM coalesced MMIO
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) * Copyright (c) 2008 Bull S.A.S.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6) * Copyright 2009 Red Hat, Inc. and/or its affiliates.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8) * Author: Laurent Vivier <Laurent.Vivier@bull.net>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) #include <kvm/iodev.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) #include <linux/kvm_host.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) #include <linux/slab.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16) #include <linux/kvm.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) #include "coalesced_mmio.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) static inline struct kvm_coalesced_mmio_dev *to_mmio(struct kvm_io_device *dev)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) return container_of(dev, struct kvm_coalesced_mmio_dev, dev);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) static int coalesced_mmio_in_range(struct kvm_coalesced_mmio_dev *dev,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) gpa_t addr, int len)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) /* is it in a batchable area ?
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) * (addr,len) is fully included in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) * (zone->addr, zone->size)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) if (len < 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) if (addr + len < addr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) if (addr < dev->zone.addr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) if (addr + len > dev->zone.addr + dev->zone.size)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40) return 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) static int coalesced_mmio_has_room(struct kvm_coalesced_mmio_dev *dev, u32 last)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) struct kvm_coalesced_mmio_ring *ring;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46) unsigned avail;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) /* Are we able to batch it ? */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50) /* last is the first free entry
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) * check if we don't meet the first used entry
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) * there is always one unused entry in the buffer
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) ring = dev->kvm->coalesced_mmio_ring;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) avail = (ring->first - last - 1) % KVM_COALESCED_MMIO_MAX;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) if (avail == 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57) /* full */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) return 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) static int coalesced_mmio_write(struct kvm_vcpu *vcpu,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) struct kvm_io_device *this, gpa_t addr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66) int len, const void *val)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) struct kvm_coalesced_mmio_dev *dev = to_mmio(this);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) struct kvm_coalesced_mmio_ring *ring = dev->kvm->coalesced_mmio_ring;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) __u32 insert;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) if (!coalesced_mmio_in_range(dev, addr, len))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) return -EOPNOTSUPP;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) spin_lock(&dev->kvm->ring_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) insert = READ_ONCE(ring->last);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) if (!coalesced_mmio_has_room(dev, insert) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) insert >= KVM_COALESCED_MMIO_MAX) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80) spin_unlock(&dev->kvm->ring_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81) return -EOPNOTSUPP;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84) /* copy data in first free entry of the ring */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86) ring->coalesced_mmio[insert].phys_addr = addr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87) ring->coalesced_mmio[insert].len = len;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88) memcpy(ring->coalesced_mmio[insert].data, val, len);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) ring->coalesced_mmio[insert].pio = dev->zone.pio;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90) smp_wmb();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91) ring->last = (insert + 1) % KVM_COALESCED_MMIO_MAX;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92) spin_unlock(&dev->kvm->ring_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96) static void coalesced_mmio_destructor(struct kvm_io_device *this)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98) struct kvm_coalesced_mmio_dev *dev = to_mmio(this);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) list_del(&dev->list);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102) kfree(dev);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) static const struct kvm_io_device_ops coalesced_mmio_ops = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106) .write = coalesced_mmio_write,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) .destructor = coalesced_mmio_destructor,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) int kvm_coalesced_mmio_init(struct kvm *kvm)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) struct page *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) page = alloc_page(GFP_KERNEL | __GFP_ZERO);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115) if (!page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116) return -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) kvm->coalesced_mmio_ring = page_address(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121) * We're using this spinlock to sync access to the coalesced ring.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122) * The list doesn't need its own lock since device registration and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) * unregistration should only happen when kvm->slots_lock is held.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) spin_lock_init(&kvm->ring_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) INIT_LIST_HEAD(&kvm->coalesced_zones);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) void kvm_coalesced_mmio_free(struct kvm *kvm)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133) if (kvm->coalesced_mmio_ring)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) free_page((unsigned long)kvm->coalesced_mmio_ring);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) int kvm_vm_ioctl_register_coalesced_mmio(struct kvm *kvm,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) struct kvm_coalesced_mmio_zone *zone)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) struct kvm_coalesced_mmio_dev *dev;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143) if (zone->pio != 1 && zone->pio != 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) dev = kzalloc(sizeof(struct kvm_coalesced_mmio_dev),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147) GFP_KERNEL_ACCOUNT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) if (!dev)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149) return -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151) kvm_iodevice_init(&dev->dev, &coalesced_mmio_ops);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) dev->kvm = kvm;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) dev->zone = *zone;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) mutex_lock(&kvm->slots_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156) ret = kvm_io_bus_register_dev(kvm,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) zone->pio ? KVM_PIO_BUS : KVM_MMIO_BUS,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) zone->addr, zone->size, &dev->dev);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159) if (ret < 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) goto out_free_dev;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) list_add_tail(&dev->list, &kvm->coalesced_zones);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) mutex_unlock(&kvm->slots_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) out_free_dev:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) mutex_unlock(&kvm->slots_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) kfree(dev);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) int kvm_vm_ioctl_unregister_coalesced_mmio(struct kvm *kvm,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) struct kvm_coalesced_mmio_zone *zone)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) struct kvm_coalesced_mmio_dev *dev, *tmp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) int r;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) if (zone->pio != 1 && zone->pio != 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182) mutex_lock(&kvm->slots_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) list_for_each_entry_safe(dev, tmp, &kvm->coalesced_zones, list) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185) if (zone->pio == dev->zone.pio &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) coalesced_mmio_in_range(dev, zone->addr, zone->size)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187) r = kvm_io_bus_unregister_dev(kvm,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188) zone->pio ? KVM_PIO_BUS : KVM_MMIO_BUS, &dev->dev);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191) * On failure, unregister destroys all devices on the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192) * bus _except_ the target device, i.e. coalesced_zones
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193) * has been modified. No need to restart the walk as
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194) * there aren't any zones left.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196) if (r)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198) kvm_iodevice_destructor(&dev->dev);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202) mutex_unlock(&kvm->slots_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205) * Ignore the result of kvm_io_bus_unregister_dev(), from userspace's
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206) * perspective, the coalesced MMIO is most definitely unregistered.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209) }