mirror of
https://github.com/hanzoai/visor.git
synced 2026-09-24 17:04:14 +00:00
Tianyu Zhou
fc1afc9de5
SlimVM is a lightweight VMX-based platform for gVisor that accelerates sentry system calls by eliminating unnecessary user/kernel context switches. In a traditional gVisor + KVM setup, every guest syscall triggers a full VM exit to userspace, then re-enters the kernel via a regular syscall: gr0 -> hr0 (vm_exit) -> hr3 (ioctl return) -> hr0 (re-invoke syscall) SlimVM eliminates this overhead. When the guest executes a VMCALL, the VM exits into VMX root mode where SlimVM directly invokes the host kernel's syscall handler, then immediately resumes the guest: gr0 -> hr0 (vmcall + direct function call) SlimVM requires a companion kernel module that provides the /dev/slimvm device interface. See https://github.com/antgroup/slimvm for the kernel module and full documentation. A sentry->host getpid microbenchmark (Xeon 6982P-C, host kernel 5.10.134) shows the effect: │ ns/op │ vs base │ base 82.5 1.0x slimvm 420.9 5.1x kvm 3566.1 43.2x slimvm is ~8.5x faster than the kvm platform on this path. See the issue for details. Updates: https://github.com/google/gvisor/issues/13330 Co-developed-by: Chenggang <[email protected]> Co-developed-by: Robin Luk <[email protected]> Co-developed-by: Tiwei Bie <[email protected]> Co-developed-by: Jianfeng Tan <[email protected]> Co-developed-by: liushi.ls <[email protected]> Co-developed-by: Yong He <[email protected]> Co-developed-by: Aaron Lu <[email protected]> Co-developed-by: chris.zn <[email protected]> Co-developed-by: Lai Jiangshan <[email protected]> Co-developed-by: Lingfu <[email protected]> Co-developed-by: Min Le <[email protected]> Co-developed-by: Quan Xu <[email protected]> Co-developed-by: Zhang Haoyu <[email protected]> Co-developed-by: Dawei Shen <[email protected]> Signed-off-by: Tianyu Zhou <[email protected]>
109 lines
2.9 KiB
Go
109 lines
2.9 KiB
Go
// Copyright 2018 The gVisor Authors.
|
|
//
|
|
// Licensed under the Apache License, Version 2.0 (the "License");
|
|
// you may not use this file except in compliance with the License.
|
|
// You may obtain a copy of the License at
|
|
//
|
|
// http://www.apache.org/licenses/LICENSE-2.0
|
|
//
|
|
// Unless required by applicable law or agreed to in writing, software
|
|
// distributed under the License is distributed on an "AS IS" BASIS,
|
|
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
// See the License for the specific language governing permissions and
|
|
// limitations under the License.
|
|
|
|
package pagetables
|
|
|
|
import (
|
|
"gvisor.dev/gvisor/pkg/cpuid"
|
|
)
|
|
|
|
// Address constraints.
|
|
var (
|
|
lowerTop uintptr = 0x00007fffffffffff
|
|
upperBottom uintptr = 0xffff800000000000
|
|
pgdShift = 39
|
|
pgdMask uintptr = 0x1ff << pgdShift
|
|
pgdSize uintptr = 1 << pgdShift
|
|
|
|
// la57Enabled gates whether InitArch promotes new PageTables to a
|
|
// 5-level layout when the host CPU advertises LA57. Defaults to true
|
|
// so existing behavior is unchanged; platforms whose hardware-
|
|
// virtualization layer cannot walk a 5-level page table call
|
|
// DisableLA57 once at startup to force 4-level.
|
|
la57Enabled = true
|
|
)
|
|
|
|
// DisableLA57 forces all subsequently-created PageTables to use a 4-level
|
|
// layout regardless of host CPUID. Must be called before any PageTables is
|
|
// created.
|
|
func DisableLA57() {
|
|
la57Enabled = false
|
|
}
|
|
|
|
const (
|
|
pteShift = 12
|
|
pmdShift = 21
|
|
pudShift = 30
|
|
p4dShift = 39
|
|
|
|
pteMask = 0x1ff << pteShift
|
|
pmdMask = 0x1ff << pmdShift
|
|
pudMask = 0x1ff << pudShift
|
|
p4dMask = 0x1ff << p4dShift
|
|
|
|
pteSize = 1 << pteShift
|
|
pmdSize = 1 << pmdShift
|
|
pudSize = 1 << pudShift
|
|
p4dSize = 1 << p4dShift
|
|
|
|
executeDisable = 1 << 63
|
|
entriesPerPage = 512
|
|
)
|
|
|
|
// InitArch does some additional initialization related to the architecture.
|
|
//
|
|
// +checkescape:hard,stack
|
|
//
|
|
//go:nosplit
|
|
func (p *PageTables) InitArch(allocator Allocator) {
|
|
featureSet := cpuid.HostFeatureSet()
|
|
if la57Enabled && featureSet.HasFeature(cpuid.X86FeatureLA57) {
|
|
p.largeAddressesEnabled = true
|
|
lowerTop = 0x00FFFFFFFFFFFFFF
|
|
upperBottom = 0xFF00000000000000
|
|
pgdShift = 48
|
|
pgdMask = 0x1ff << pgdShift
|
|
pgdSize = 1 << pgdShift
|
|
}
|
|
|
|
if p.upperSharedPageTables != nil {
|
|
p.cloneUpperShared()
|
|
}
|
|
}
|
|
|
|
//go:nosplit
|
|
func pgdIndex(upperStart uintptr) uintptr {
|
|
if upperStart&(pgdSize-1) != 0 {
|
|
panic("upperStart should be pgd size aligned")
|
|
}
|
|
if upperStart >= upperBottom {
|
|
return entriesPerPage/2 + (upperStart-upperBottom)>>pgdShift
|
|
}
|
|
if upperStart < lowerTop {
|
|
return upperStart >> pgdShift
|
|
}
|
|
panic("upperStart should be in canonical range")
|
|
}
|
|
|
|
// cloneUpperShared clone the upper from the upper shared page tables.
|
|
//
|
|
//go:nosplit
|
|
func (p *PageTables) cloneUpperShared() {
|
|
start := pgdIndex(p.upperStart)
|
|
copy(p.root[start:entriesPerPage], p.upperSharedPageTables.root[start:entriesPerPage])
|
|
}
|
|
|
|
// PTEs is a collection of entries.
|
|
type PTEs [entriesPerPage]PTE
|