2019-06-04 16:11:32 +08:00
|
|
|
// SPDX-License-Identifier: GPL-2.0-only
|
2016-09-08 02:47:23 +08:00
|
|
|
/*
|
|
|
|
* Kernel-based Virtual Machine driver for Linux
|
|
|
|
*
|
|
|
|
* Copyright 2016 Red Hat, Inc. and/or its affiliates.
|
|
|
|
*/
|
|
|
|
#include <linux/kvm_host.h>
|
2016-09-16 22:27:36 +08:00
|
|
|
#include <linux/debugfs.h>
|
2019-05-20 16:18:07 +08:00
|
|
|
#include "lapic.h"
|
2021-07-31 06:04:52 +08:00
|
|
|
#include "mmu.h"
|
|
|
|
#include "mmu/mmu_internal.h"
|
2016-09-08 02:47:23 +08:00
|
|
|
|
2019-05-20 16:18:07 +08:00
|
|
|
static int vcpu_get_timer_advance_ns(void *data, u64 *val)
|
|
|
|
{
|
|
|
|
struct kvm_vcpu *vcpu = (struct kvm_vcpu *) data;
|
|
|
|
*val = vcpu->arch.apic->lapic_timer.timer_advance_ns;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
DEFINE_SIMPLE_ATTRIBUTE(vcpu_timer_advance_ns_fops, vcpu_get_timer_advance_ns, NULL, "%llu\n");
|
|
|
|
|
2021-06-10 02:03:39 +08:00
|
|
|
static int vcpu_get_guest_mode(void *data, u64 *val)
|
|
|
|
{
|
|
|
|
struct kvm_vcpu *vcpu = (struct kvm_vcpu *) data;
|
|
|
|
*val = vcpu->stat.guest_mode;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
DEFINE_SIMPLE_ATTRIBUTE(vcpu_guest_mode_fops, vcpu_get_guest_mode, NULL, "%lld\n");
|
|
|
|
|
2016-09-16 22:27:36 +08:00
|
|
|
static int vcpu_get_tsc_offset(void *data, u64 *val)
|
|
|
|
{
|
|
|
|
struct kvm_vcpu *vcpu = (struct kvm_vcpu *) data;
|
|
|
|
*val = vcpu->arch.tsc_offset;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
DEFINE_SIMPLE_ATTRIBUTE(vcpu_tsc_offset_fops, vcpu_get_tsc_offset, NULL, "%lld\n");
|
|
|
|
|
|
|
|
static int vcpu_get_tsc_scaling_ratio(void *data, u64 *val)
|
|
|
|
{
|
|
|
|
struct kvm_vcpu *vcpu = (struct kvm_vcpu *) data;
|
|
|
|
*val = vcpu->arch.tsc_scaling_ratio;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
DEFINE_SIMPLE_ATTRIBUTE(vcpu_tsc_scaling_fops, vcpu_get_tsc_scaling_ratio, NULL, "%llu\n");
|
|
|
|
|
|
|
|
static int vcpu_get_tsc_scaling_frac_bits(void *data, u64 *val)
|
|
|
|
{
|
2022-05-24 21:56:23 +08:00
|
|
|
*val = kvm_caps.tsc_scaling_ratio_frac_bits;
|
2016-09-16 22:27:36 +08:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
DEFINE_SIMPLE_ATTRIBUTE(vcpu_tsc_scaling_frac_fops, vcpu_get_tsc_scaling_frac_bits, NULL, "%llu\n");
|
|
|
|
|
2020-06-04 21:16:52 +08:00
|
|
|
void kvm_arch_create_vcpu_debugfs(struct kvm_vcpu *vcpu, struct dentry *debugfs_dentry)
|
2016-09-08 02:47:23 +08:00
|
|
|
{
|
2021-06-10 02:03:39 +08:00
|
|
|
debugfs_create_file("guest_mode", 0444, debugfs_dentry, vcpu,
|
|
|
|
&vcpu_guest_mode_fops);
|
2020-06-04 21:16:52 +08:00
|
|
|
debugfs_create_file("tsc-offset", 0444, debugfs_dentry, vcpu,
|
2019-08-01 02:56:20 +08:00
|
|
|
&vcpu_tsc_offset_fops);
|
2016-09-16 22:27:36 +08:00
|
|
|
|
2019-08-01 02:56:20 +08:00
|
|
|
if (lapic_in_kernel(vcpu))
|
|
|
|
debugfs_create_file("lapic_timer_advance_ns", 0444,
|
2020-06-04 21:16:52 +08:00
|
|
|
debugfs_dentry, vcpu,
|
2019-08-01 02:56:20 +08:00
|
|
|
&vcpu_timer_advance_ns_fops);
|
2019-05-20 16:18:07 +08:00
|
|
|
|
2022-05-24 21:56:23 +08:00
|
|
|
if (kvm_caps.has_tsc_control) {
|
2019-08-01 02:56:20 +08:00
|
|
|
debugfs_create_file("tsc-scaling-ratio", 0444,
|
2020-06-04 21:16:52 +08:00
|
|
|
debugfs_dentry, vcpu,
|
2019-08-01 02:56:20 +08:00
|
|
|
&vcpu_tsc_scaling_fops);
|
|
|
|
debugfs_create_file("tsc-scaling-ratio-frac-bits", 0444,
|
2020-06-04 21:16:52 +08:00
|
|
|
debugfs_dentry, vcpu,
|
2019-08-01 02:56:20 +08:00
|
|
|
&vcpu_tsc_scaling_frac_fops);
|
2016-09-16 22:27:36 +08:00
|
|
|
}
|
2016-09-08 02:47:23 +08:00
|
|
|
}
|
2021-07-31 06:04:52 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* This covers statistics <1024 (11=log(1024)+1), which should be enough to
|
|
|
|
* cover RMAP_RECYCLE_THRESHOLD.
|
|
|
|
*/
|
|
|
|
#define RMAP_LOG_SIZE 11
|
|
|
|
|
|
|
|
static const char *kvm_lpage_str[KVM_NR_PAGE_SIZES] = { "4K", "2M", "1G" };
|
|
|
|
|
|
|
|
static int kvm_mmu_rmaps_stat_show(struct seq_file *m, void *v)
|
|
|
|
{
|
|
|
|
struct kvm_rmap_head *rmap;
|
|
|
|
struct kvm *kvm = m->private;
|
|
|
|
struct kvm_memory_slot *slot;
|
|
|
|
struct kvm_memslots *slots;
|
|
|
|
unsigned int lpage_size, index;
|
|
|
|
/* Still small enough to be on the stack */
|
|
|
|
unsigned int *log[KVM_NR_PAGE_SIZES], *cur;
|
|
|
|
int i, j, k, l, ret;
|
|
|
|
|
2022-01-05 12:03:37 +08:00
|
|
|
if (!kvm_memslots_have_rmaps(kvm))
|
|
|
|
return 0;
|
|
|
|
|
2021-07-31 06:04:52 +08:00
|
|
|
ret = -ENOMEM;
|
|
|
|
memset(log, 0, sizeof(log));
|
|
|
|
for (i = 0; i < KVM_NR_PAGE_SIZES; i++) {
|
|
|
|
log[i] = kcalloc(RMAP_LOG_SIZE, sizeof(unsigned int), GFP_KERNEL);
|
|
|
|
if (!log[i])
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
mutex_lock(&kvm->slots_lock);
|
|
|
|
write_lock(&kvm->mmu_lock);
|
|
|
|
|
|
|
|
for (i = 0; i < KVM_ADDRESS_SPACE_NUM; i++) {
|
KVM: Keep memslots in tree-based structures instead of array-based ones
The current memslot code uses a (reverse gfn-ordered) memslot array for
keeping track of them.
Because the memslot array that is currently in use cannot be modified
every memslot management operation (create, delete, move, change flags)
has to make a copy of the whole array so it has a scratch copy to work on.
Strictly speaking, however, it is only necessary to make copy of the
memslot that is being modified, copying all the memslots currently present
is just a limitation of the array-based memslot implementation.
Two memslot sets, however, are still needed so the VM continues to run
on the currently active set while the requested operation is being
performed on the second, currently inactive one.
In order to have two memslot sets, but only one copy of actual memslots
it is necessary to split out the memslot data from the memslot sets.
The memslots themselves should be also kept independent of each other
so they can be individually added or deleted.
These two memslot sets should normally point to the same set of
memslots. They can, however, be desynchronized when performing a
memslot management operation by replacing the memslot to be modified
by its copy. After the operation is complete, both memslot sets once
again point to the same, common set of memslot data.
This commit implements the aforementioned idea.
For tracking of gfns an ordinary rbtree is used since memslots cannot
overlap in the guest address space and so this data structure is
sufficient for ensuring that lookups are done quickly.
The "last used slot" mini-caches (both per-slot set one and per-vCPU one),
that keep track of the last found-by-gfn memslot, are still present in the
new code.
Co-developed-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Maciej S. Szmigiero <maciej.szmigiero@oracle.com>
Message-Id: <17c0cf3663b760a0d3753d4ac08c0753e941b811.1638817641.git.maciej.szmigiero@oracle.com>
2021-12-07 03:54:30 +08:00
|
|
|
int bkt;
|
|
|
|
|
2021-07-31 06:04:52 +08:00
|
|
|
slots = __kvm_memslots(kvm, i);
|
KVM: Keep memslots in tree-based structures instead of array-based ones
The current memslot code uses a (reverse gfn-ordered) memslot array for
keeping track of them.
Because the memslot array that is currently in use cannot be modified
every memslot management operation (create, delete, move, change flags)
has to make a copy of the whole array so it has a scratch copy to work on.
Strictly speaking, however, it is only necessary to make copy of the
memslot that is being modified, copying all the memslots currently present
is just a limitation of the array-based memslot implementation.
Two memslot sets, however, are still needed so the VM continues to run
on the currently active set while the requested operation is being
performed on the second, currently inactive one.
In order to have two memslot sets, but only one copy of actual memslots
it is necessary to split out the memslot data from the memslot sets.
The memslots themselves should be also kept independent of each other
so they can be individually added or deleted.
These two memslot sets should normally point to the same set of
memslots. They can, however, be desynchronized when performing a
memslot management operation by replacing the memslot to be modified
by its copy. After the operation is complete, both memslot sets once
again point to the same, common set of memslot data.
This commit implements the aforementioned idea.
For tracking of gfns an ordinary rbtree is used since memslots cannot
overlap in the guest address space and so this data structure is
sufficient for ensuring that lookups are done quickly.
The "last used slot" mini-caches (both per-slot set one and per-vCPU one),
that keep track of the last found-by-gfn memslot, are still present in the
new code.
Co-developed-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Maciej S. Szmigiero <maciej.szmigiero@oracle.com>
Message-Id: <17c0cf3663b760a0d3753d4ac08c0753e941b811.1638817641.git.maciej.szmigiero@oracle.com>
2021-12-07 03:54:30 +08:00
|
|
|
kvm_for_each_memslot(slot, bkt, slots)
|
2021-07-31 06:04:52 +08:00
|
|
|
for (k = 0; k < KVM_NR_PAGE_SIZES; k++) {
|
|
|
|
rmap = slot->arch.rmap[k];
|
|
|
|
lpage_size = kvm_mmu_slot_lpages(slot, k + 1);
|
|
|
|
cur = log[k];
|
|
|
|
for (l = 0; l < lpage_size; l++) {
|
|
|
|
index = ffs(pte_list_count(&rmap[l]));
|
|
|
|
if (WARN_ON_ONCE(index >= RMAP_LOG_SIZE))
|
|
|
|
index = RMAP_LOG_SIZE - 1;
|
|
|
|
cur[index]++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
write_unlock(&kvm->mmu_lock);
|
|
|
|
mutex_unlock(&kvm->slots_lock);
|
|
|
|
|
|
|
|
/* index=0 counts no rmap; index=1 counts 1 rmap */
|
|
|
|
seq_printf(m, "Rmap_Count:\t0\t1\t");
|
|
|
|
for (i = 2; i < RMAP_LOG_SIZE; i++) {
|
|
|
|
j = 1 << (i - 1);
|
|
|
|
k = (1 << i) - 1;
|
|
|
|
seq_printf(m, "%d-%d\t", j, k);
|
|
|
|
}
|
|
|
|
seq_printf(m, "\n");
|
|
|
|
|
|
|
|
for (i = 0; i < KVM_NR_PAGE_SIZES; i++) {
|
|
|
|
seq_printf(m, "Level=%s:\t", kvm_lpage_str[i]);
|
|
|
|
cur = log[i];
|
|
|
|
for (j = 0; j < RMAP_LOG_SIZE; j++)
|
|
|
|
seq_printf(m, "%d\t", cur[j]);
|
|
|
|
seq_printf(m, "\n");
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = 0;
|
|
|
|
out:
|
|
|
|
for (i = 0; i < KVM_NR_PAGE_SIZES; i++)
|
|
|
|
kfree(log[i]);
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int kvm_mmu_rmaps_stat_open(struct inode *inode, struct file *file)
|
|
|
|
{
|
|
|
|
struct kvm *kvm = inode->i_private;
|
|
|
|
|
|
|
|
if (!kvm_get_kvm_safe(kvm))
|
|
|
|
return -ENOENT;
|
|
|
|
|
|
|
|
return single_open(file, kvm_mmu_rmaps_stat_show, kvm);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int kvm_mmu_rmaps_stat_release(struct inode *inode, struct file *file)
|
|
|
|
{
|
|
|
|
struct kvm *kvm = inode->i_private;
|
|
|
|
|
|
|
|
kvm_put_kvm(kvm);
|
|
|
|
|
|
|
|
return single_release(inode, file);
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct file_operations mmu_rmaps_stat_fops = {
|
|
|
|
.open = kvm_mmu_rmaps_stat_open,
|
|
|
|
.read = seq_read,
|
|
|
|
.llseek = seq_lseek,
|
|
|
|
.release = kvm_mmu_rmaps_stat_release,
|
|
|
|
};
|
|
|
|
|
|
|
|
int kvm_arch_create_vm_debugfs(struct kvm *kvm)
|
|
|
|
{
|
|
|
|
debugfs_create_file("mmu_rmaps_stat", 0644, kvm->debugfs_dentry, kvm,
|
|
|
|
&mmu_rmaps_stat_fops);
|
|
|
|
return 0;
|
|
|
|
}
|