Merge pull request #193 from containers/renovate/github.com-containers-storage-1.x

fix(deps): update module github.com/containers/storage to v1.58.0

.cirrus.yml

@@ -11,8 +11,8 @@ env:
     GOCACHE: "${HOME}/.cache/go-build"
 
     # VM Image built in containers/automation_images
-    _BUILT_IMAGE_SUFFIX: "c5878804328480768"
+    IMAGE_SUFFIX: "c20250324t111922z-f41f40d13"
-    FEDORA_CACHE_IMAGE_NAME: "fedora-${_BUILT_IMAGE_SUFFIX}"
+    FEDORA_CACHE_IMAGE_NAME: "fedora-${IMAGE_SUFFIX}"
 
     # Must be defined true when testing w/in containers
     CONTAINER: "false"
@@ -21,16 +21,6 @@ env:
 gcp_credentials: ENCRYPTED[2ba9cffb563741f8538eab6d4a8b2d4684c0de23693a8ade80aced34596669a87a6c01e45ce45fad7f7db1d995e8c777]
 
 
-# Default VM to use unless set or modified by task
-gce_instance:
-    image_project: "${IMAGE_PROJECT}"
-    zone: "us-central1-c"  # Required by Cirrus for the time being
-    cpu: 2
-    memory: "4Gb"
-    disk: 200  # Required for performance reasons
-    image_name: "${FEDORA_CACHE_IMAGE_NAME}"
-
-
 # Update metadata on VM images referenced by this repository state
 meta_task:
 
@@ -58,7 +48,7 @@ meta_task:
 vendor_task:
 
     container:
-      image: golang:1.18
+      image: golang:1.23
 
     script:
         - make vendor
@@ -68,7 +58,17 @@ vendor_task:
 build_and_test_task:
 
     gce_instance:
-        image_name: $FEDORA_CACHE_IMAGE_NAME
+        image_project: "${IMAGE_PROJECT}"
+        zone: "us-central1-c"  # Required by Cirrus for the time being
+        cpu: 2
+        memory: "4Gb"
+        disk: 200  # Required for performance reasons
+        image_name: "${FEDORA_CACHE_IMAGE_NAME}"
+
+    matrix:
+        - name: "Test on Fedora"
+          gce_instance:
+              image_name: "${FEDORA_CACHE_IMAGE_NAME}"
 
     # Avoid downloading this stuff every time
     gocache_cache:
@@ -100,6 +100,10 @@ build_and_test_task:
         - make test-unit
         - sudo make test-integration
 
+    # The hook runs in the background, errors will be in the journal.
+    always:
+        journal_script: journalctl -b
+
     binaries_artifacts:
         path: "bin/*"
         type: "application/octet-stream"

.github/renovate.json5

@@ -0,0 +1,54 @@
+/*
+   Renovate is a service similar to GitHub Dependabot, but with
+   (fantastically) more configuration options.  So many options
+   in fact, if you're new I recommend glossing over this cheat-sheet
+   prior to the official documentation:
+
+   https://www.augmentedmind.de/2021/07/25/renovate-bot-cheat-sheet
+
+   Configuration Update/Change Procedure:
+     1. Make changes
+     2. Manually validate changes (from repo-root):
+
+        podman run -it \
+            -v ./.github/renovate.json5:/usr/src/app/renovate.json5:z \
+            docker.io/renovate/renovate:latest \
+            renovate-config-validator
+     3. Commit.
+
+   Configuration Reference:
+   https://docs.renovatebot.com/configuration-options/
+
+   Monitoring Dashboard:
+   https://app.renovatebot.com/dashboard#github/containers
+
+   Note: The Renovate bot will create/manage it's business on
+         branches named 'renovate/*'.  Otherwise, and by
+         default, the only the copy of this file that matters
+         is the one on the `main` branch.  No other branches
+         will be monitored or touched in any way.
+*/
+
+{
+  "$schema": "https://docs.renovatebot.com/renovate-schema.json",
+
+  /*************************************************
+   ****** Global/general configuration options *****
+   *************************************************/
+
+  // Re-use predefined sets of configuration options to DRY
+  "extends": [
+    // https://github.com/containers/automation/blob/main/renovate/defaults.json5
+    "github>containers/automation//renovate/defaults.json5"
+  ],
+
+  // Permit automatic rebasing when base-branch changes by more than
+  // one commit.
+  "rebaseWhen": "behind-base-branch",
+
+  /*************************************************
+   *** Repository-specific configuration options ***
+   *************************************************/
+
+  "assignees": ["vrothberg", "rhatdan"],
+}

Makefile

@@ -6,6 +6,7 @@ GO_BUILD=$(GO) build
 ifeq ($(shell go help mod >/dev/null 2>&1 && echo true), true)
 	GO_BUILD=GO111MODULE=on $(GO) build -mod=vendor
 endif
+BUILDDIR ?= .
 DESTDIR ?=
 PREFIX ?= /usr/local
 SELINUXOPT ?= $(shell test -x /usr/sbin/selinuxenabled && selinuxenabled && echo -Z)
@@ -46,7 +47,7 @@ docs:
 
 .PHONY: binary
 binary:
-	$(GO_BUILD) -mod=vendor -o bin/oci-seccomp-bpf-hook -ldflags "-X main.version=$(OSBH_VERSION)" $(PROJECT)
+	$(GO_BUILD) -mod=vendor -o $(BUILDDIR)/bin/oci-seccomp-bpf-hook -ldflags "-X main.version=$(OSBH_VERSION)" $(PROJECT)
 
 .PHONY: validate
 validate:
@@ -73,7 +74,7 @@ test-unit:
 .PHONY: install.tools
 install.tools: .install.golangci-lint .install.md2man
 
-.install.golangci-lint: VERSION=v1.45.2
+.install.golangci-lint: VERSION=v1.60.3
 .install.golangci-lint:
 	curl -fsSL https://raw.githubusercontent.com/golangci/golangci-lint/$(VERSION)/install.sh | sh -s -- -b ./build $(VERSION)
 
@@ -94,7 +95,7 @@ install.docs:
 install-nobuild: install.docs-nobuild
 	install $(SELINUXOPT) -d -m 755 $(DESTDIR)$(HOOK_BIN_DIR)
 	install $(SELINUXOPT) -d -m 755 $(DESTDIR)$(HOOK_DIR)
-	install $(SELINUXOPT) -m 755 bin/oci-seccomp-bpf-hook $(DESTDIR)$(HOOK_BIN_DIR)
+	install $(SELINUXOPT) -m 755 $(BUILDDIR)/bin/oci-seccomp-bpf-hook $(DESTDIR)$(HOOK_BIN_DIR)
 	install $(SELINUXOPT) -m 644 oci-seccomp-bpf-hook.json $(DESTDIR)$(HOOK_DIR)
 	sed -i 's|HOOK_BIN_DIR|$(HOOK_BIN_DIR)|g' $(DESTDIR)$(HOOK_DIR)/oci-seccomp-bpf-hook.json
 

VERSION

@@ -1 +1 @@
-v1.2.7
+v1.2.12-dev

contrib/cirrus/Dockerfile

@@ -1,4 +1,4 @@
-FROM fedora:31
+FROM fedora:latest
 
 ENV GOPATH=/var/tmp/go
 ENV GOSRC=$GOPATH/src/github.com/containers/oci-seccomp-bpf-hook

contrib/cirrus/lib.sh

@@ -36,12 +36,6 @@ fi
 GOSRC="${GOSRC:-$(realpath $(dirname $0)/../../)}"
 SCRIPT_BASE="${SCRIPT_BASE:-./contrib/cirrus}"
 
-# GCE image-name compatible string representation of distribution name
-OS_RELEASE_ID="$(source /etc/os-release; echo $ID)"
-# GCE image-name compatible string representation of distribution _major_ version
-OS_RELEASE_VER="$(source /etc/os-release; echo $VERSION_ID | cut -d '.' -f 1)"
-# Combined to ease soe usage
-OS_REL_VER="${OS_RELEASE_ID}-${OS_RELEASE_VER}"
 # Set 'true' when operating inside a container
 CONTAINER="${CONTAINER:-false}"
 

ebpf.go

@@ -18,7 +18,6 @@ type event struct {
 // Complete documentation is available at
 // https://github.com/iovisor/bcc/blob/master/docs/reference_guide.md
 const source string = `
-#include <linux/bpf.h>
 #include <linux/nsproxy.h>
 #include <linux/pid_namespace.h>
 #include <linux/ns_common.h>
@@ -112,6 +111,13 @@ int enter_trace(struct tracepoint__raw_syscalls__sys_enter* args)
         // The syscall was already notified.
         if (seen > 0)
             return 0;
+
+        // if prctl was not seen, ignore the current syscall.
+        u64 prctl = __NR_prctl;
+        u64 *prctl_seen = seen_syscalls.lookup(&prctl);
+        if (prctl_seen == NULL) {
+	    return 0;
+        }
     }
 
     data.stopTracing = false;

go.mod

@@ -1,12 +1,28 @@
 module github.com/containers/oci-seccomp-bpf-hook
 
-go 1.14
+go 1.23.0
+
+toolchain go1.23.8
 
 require (
-	github.com/containers/common v0.48.0
+	github.com/containers/common v0.62.3
-	github.com/iovisor/gobpf v0.2.0
+	github.com/containers/storage v1.58.0
-	github.com/opencontainers/runtime-spec v1.0.3-0.20210326190908-1c3f411f0417
+	github.com/iovisor/gobpf v0.2.1-0.20221005153822-16120a1bf4d4
-	github.com/seccomp/libseccomp-golang v0.9.2-0.20210429002308-3879420cc921
+	github.com/opencontainers/runtime-spec v1.2.1
-	github.com/sirupsen/logrus v1.8.1
+	github.com/seccomp/libseccomp-golang v0.10.0
-	github.com/stretchr/testify v1.7.1
+	github.com/sirupsen/logrus v1.9.3
+	github.com/stretchr/testify v1.10.0
+)
+
+require (
+	github.com/davecgh/go-spew v1.1.2-0.20180830191138-d8f796af33cc // indirect
+	github.com/docker/go-units v0.5.0 // indirect
+	github.com/kr/pretty v0.3.1 // indirect
+	github.com/moby/sys/capability v0.4.0 // indirect
+	github.com/moby/sys/mountinfo v0.7.2 // indirect
+	github.com/moby/sys/user v0.4.0 // indirect
+	github.com/pmezard/go-difflib v1.0.1-0.20181226105442-5d4384ee4fb2 // indirect
+	golang.org/x/sys v0.32.0 // indirect
+	gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c // indirect
+	gopkg.in/yaml.v3 v3.0.1 // indirect
 )

oci-seccomp-bpf-hook.go

@@ -8,7 +8,6 @@ import (
 	"errors"
 	"flag"
 	"fmt"
-	"io/ioutil"
 	"log/syslog"
 	"os"
 	"os/exec"
@@ -23,6 +22,7 @@ import (
 	"time"
 
 	types "github.com/containers/common/pkg/seccomp"
+	"github.com/containers/storage/pkg/unshare"
 	"github.com/iovisor/gobpf/bcc"
 	spec "github.com/opencontainers/runtime-spec/specs-go"
 	seccomp "github.com/seccomp/libseccomp-golang"
@@ -57,6 +57,11 @@ func main() {
 		logrus.AddHook(hook)
 	}
 
+	if os.Getuid() != 0 || unshare.IsRootless() {
+		logrus.Errorf("running the hook requires root privileges")
+		os.Exit(1)
+	}
+
 	runBPF := flag.Int("r", 0, "Trace the specified PID")
 	outputFile := flag.String("o", "", "Path of the output file")
 	inputFile := flag.String("i", "", "Path of the input file")
@@ -223,6 +228,13 @@ func runBPFSource(pid int, profilePath string, inputFile string) (finalErr error
 	m := bcc.NewModule(src, []string{})
 	defer m.Close()
 
+	table := bcc.NewTable(m.TableId("events"), m)
+	channel := make(chan []byte)
+	perfMap, err := bcc.InitPerfMap(table, channel, nil)
+	if err != nil {
+		return fmt.Errorf("error initializing perf map: %v", err)
+	}
+
 	logrus.Info("Loading enter tracepoint")
 	enterTrace, err := m.LoadTracepoint("enter_trace")
 	if err != nil {
@@ -242,13 +254,6 @@ func runBPFSource(pid int, profilePath string, inputFile string) (finalErr error
 		return fmt.Errorf("error attaching to tracepoint: %v", err)
 	}
 
-	table := bcc.NewTable(m.TableId("events"), m)
-	channel := make(chan []byte)
-	perfMap, err := bcc.InitPerfMap(table, channel, nil)
-	if err != nil {
-		return fmt.Errorf("error initializing perf map: %v", err)
-	}
-
 	// Initialize the wait group used to wait for the tracing to be finished.
 	wg.Add(1)
 	var events []event
@@ -322,7 +327,7 @@ func generateProfile(syscalls map[string]int, profilePath string, inputFile stri
 	inputProfile := types.Seccomp{}
 
 	if inputFile != "" {
-		input, err := ioutil.ReadFile(inputFile)
+		input, err := os.ReadFile(inputFile)
 		if err != nil {
 			return fmt.Errorf("error reading input file: %v", err)
 		}
@@ -359,7 +364,7 @@ func generateProfile(syscalls map[string]int, profilePath string, inputFile stri
 	if err != nil {
 		return fmt.Errorf("error writing seccomp profile: %v", err)
 	}
-	if err := ioutil.WriteFile(profilePath, sJSON, 0644); err != nil {
+	if err := os.WriteFile(profilePath, sJSON, 0644); err != nil {
 		return fmt.Errorf("error writing seccomp profile: %v", err)
 	}
 	return nil
@@ -381,7 +386,7 @@ func parseAnnotation(annotation string) (outputFile string, inputFile string, er
 				return "", "", fmt.Errorf("%v: input file path must be absolute: %q", errInvalidAnnotation, inputFile)
 			}
 			inputProfile := types.Seccomp{}
-			input, err := ioutil.ReadFile(inputFile)
+			input, err := os.ReadFile(inputFile)
 			if err != nil {
 				return "", "", fmt.Errorf("%v: error reading input file: %q", errInvalidAnnotation, inputFile)
 			}

oci-seccomp-bpf-hook_test.go

@@ -2,7 +2,6 @@ package main
 
 import (
 	"encoding/json"
-	"io/ioutil"
 	"os"
 	"runtime"
 	"testing"
@@ -15,7 +14,7 @@ func TestParseAnnotation(t *testing.T) {
 	testProfile := types.Seccomp{}
 	testProfile.DefaultAction = types.ActErrno
 
-	tmpFile, err := ioutil.TempFile(os.TempDir(), "input-*.json")
+	tmpFile, err := os.CreateTemp(os.TempDir(), "input-*.json")
 	if err != nil {
 		t.Fatalf("cannot create temporary file")
 	}

vendor/github.com/containers/common/pkg/seccomp/conversion.go

@@ -7,7 +7,6 @@ import (
 	"fmt"
 
 	"github.com/opencontainers/runtime-spec/specs-go"
-	"github.com/pkg/errors"
 )
 
 var (
@@ -71,11 +70,12 @@ var (
 		// https://github.com/opencontainers/runtime-spec/pull/1064
 		// specs.ActKillProcess   ActKillProcess,
 		// specs.ActKillThread   ActKillThread,
-		specs.ActErrno: ActErrno,
+		specs.ActErrno:  ActErrno,
-		specs.ActTrap:  ActTrap,
+		specs.ActTrap:   ActTrap,
-		specs.ActAllow: ActAllow,
+		specs.ActAllow:  ActAllow,
-		specs.ActTrace: ActTrace,
+		specs.ActTrace:  ActTrace,
-		specs.ActLog:   ActLog,
+		specs.ActLog:    ActLog,
+		specs.ActNotify: ActNotify,
 	}
 	specOperatorToSeccompOperatorMap = map[specs.LinuxSeccompOperator]Operator{
 		specs.OpNotEqual:     OpNotEqual,
@@ -107,7 +107,7 @@ func specToSeccomp(spec *specs.LinuxSeccomp) (*Seccomp, error) {
 	for _, arch := range spec.Architectures {
 		newArch, err := specArchToSeccompArch(arch)
 		if err != nil {
-			return nil, errors.Wrap(err, "convert spec arch")
+			return nil, fmt.Errorf("convert spec arch: %w", err)
 		}
 		res.Architectures = append(res.Architectures, newArch)
 	}
@@ -115,7 +115,7 @@ func specToSeccomp(spec *specs.LinuxSeccomp) (*Seccomp, error) {
 	// Convert default action
 	newDefaultAction, err := specActionToSeccompAction(spec.DefaultAction)
 	if err != nil {
-		return nil, errors.Wrap(err, "convert default action")
+		return nil, fmt.Errorf("convert default action: %w", err)
 	}
 	res.DefaultAction = newDefaultAction
 	res.DefaultErrnoRet = spec.DefaultErrnoRet
@@ -124,7 +124,7 @@ func specToSeccomp(spec *specs.LinuxSeccomp) (*Seccomp, error) {
 	for _, call := range spec.Syscalls {
 		newAction, err := specActionToSeccompAction(call.Action)
 		if err != nil {
-			return nil, errors.Wrap(err, "convert action")
+			return nil, fmt.Errorf("convert action: %w", err)
 		}
 
 		for _, name := range call.Names {
@@ -139,7 +139,7 @@ func specToSeccomp(spec *specs.LinuxSeccomp) (*Seccomp, error) {
 			for _, arg := range call.Args {
 				newOp, err := specOperatorToSeccompOperator(arg.Op)
 				if err != nil {
-					return nil, errors.Wrap(err, "convert operator")
+					return nil, fmt.Errorf("convert operator: %w", err)
 				}
 
 				newArg := Arg{
@@ -163,7 +163,7 @@ func specArchToLibseccompArch(arch specs.Arch) (string, error) {
 	if res, ok := specArchToLibseccompArchMap[arch]; ok {
 		return res, nil
 	}
-	return "", errors.Errorf(
+	return "", fmt.Errorf(
 		"architecture %q is not valid for libseccomp", arch,
 	)
 }
@@ -173,7 +173,7 @@ func specArchToSeccompArch(arch specs.Arch) (Arch, error) {
 	if res, ok := specArchToSeccompArchMap[arch]; ok {
 		return res, nil
 	}
-	return "", errors.Errorf("architecture %q is not valid", arch)
+	return "", fmt.Errorf("architecture %q is not valid", arch)
 }
 
 // specActionToSeccompAction converts a spec action into a seccomp one.
@@ -181,7 +181,7 @@ func specActionToSeccompAction(action specs.LinuxSeccompAction) (Action, error)
 	if res, ok := specActionToSeccompActionMap[action]; ok {
 		return res, nil
 	}
-	return "", errors.Errorf(
+	return "", fmt.Errorf(
 		"spec action %q is not valid internal action", action,
 	)
 }
@@ -191,7 +191,7 @@ func specOperatorToSeccompOperator(operator specs.LinuxSeccompOperator) (Operato
 	if op, ok := specOperatorToSeccompOperatorMap[operator]; ok {
 		return op, nil
 	}
-	return "", errors.Errorf(
+	return "", fmt.Errorf(
 		"spec operator %q is not a valid internal operator", operator,
 	)
 }

vendor/github.com/containers/common/pkg/seccomp/default_linux.go

@@ -51,9 +51,16 @@ func DefaultProfile() *Seccomp {
 		{
 			Names: []string{
 				"bdflush",
+				"cachestat",
+				"futex_requeue",
+				"futex_wait",
+				"futex_waitv",
+				"futex_wake",
 				"io_pgetevents",
+				"io_pgetevents_time64",
 				"kexec_file_load",
 				"kexec_load",
+				"map_shadow_stack",
 				"migrate_pages",
 				"move_pages",
 				"nfsservctl",
@@ -68,9 +75,9 @@ func DefaultProfile() *Seccomp {
 				"pciconfig_write",
 				"sgetmask",
 				"ssetmask",
-				"swapcontext",
 				"swapoff",
 				"swapon",
+				"syscall",
 				"sysfs",
 				"uselib",
 				"userfaultfd",
@@ -142,6 +149,7 @@ func DefaultProfile() *Seccomp {
 				"fchdir",
 				"fchmod",
 				"fchmodat",
+				"fchmodat2",
 				"fchown",
 				"fchown32",
 				"fchownat",
@@ -221,6 +229,9 @@ func DefaultProfile() *Seccomp {
 				"ipc",
 				"keyctl",
 				"kill",
+				"landlock_add_rule",
+				"landlock_create_ruleset",
+				"landlock_restrict_self",
 				"lchown",
 				"lchown32",
 				"lgetxattr",
@@ -306,7 +317,6 @@ func DefaultProfile() *Seccomp {
 				"pwritev2",
 				"read",
 				"readahead",
-				"readdir",
 				"readlink",
 				"readlinkat",
 				"readv",
@@ -394,15 +404,12 @@ func DefaultProfile() *Seccomp {
 				"shmdt",
 				"shmget",
 				"shutdown",
-				"sigaction",
 				"sigaltstack",
 				"signal",
 				"signalfd",
 				"signalfd4",
-				"sigpending",
 				"sigprocmask",
 				"sigreturn",
-				"sigsuspend",
 				"socketcall",
 				"socketpair",
 				"splice",
@@ -416,7 +423,6 @@ func DefaultProfile() *Seccomp {
 				"sync",
 				"sync_file_range",
 				"syncfs",
-				"syscall",
 				"sysinfo",
 				"syslog",
 				"tee",
@@ -429,7 +435,6 @@ func DefaultProfile() *Seccomp {
 				"timer_gettime64",
 				"timer_settime",
 				"timer_settime64",
-				"timerfd",
 				"timerfd_create",
 				"timerfd_gettime",
 				"timerfd_gettime64",
@@ -519,6 +524,7 @@ func DefaultProfile() *Seccomp {
 		{
 			Names: []string{
 				"sync_file_range2",
+				"swapcontext",
 			},
 			Action: ActAllow,
 			Args:   []*Arg{},
@@ -573,6 +579,16 @@ func DefaultProfile() *Seccomp {
 				Arches: []string{"s390", "s390x"},
 			},
 		},
+		{
+			Names: []string{
+				"riscv_flush_icache",
+			},
+			Action: ActAllow,
+			Args:   []*Arg{},
+			Includes: Filter{
+				Arches: []string{"riscv64"},
+			},
+		},
 		{
 			Names: []string{
 				"open_by_handle_at",
@@ -600,8 +616,8 @@ func DefaultProfile() *Seccomp {
 				"bpf",
 				"fanotify_init",
 				"lookup_dcookie",
-				"perf_event_open",
 				"quotactl",
+				"quotactl_fd",
 				"setdomainname",
 				"sethostname",
 				"setns",
@@ -614,11 +630,11 @@ func DefaultProfile() *Seccomp {
 		},
 		{
 			Names: []string{
-				"bpf",
 				"fanotify_init",
 				"lookup_dcookie",
 				"perf_event_open",
 				"quotactl",
+				"quotactl_fd",
 				"setdomainname",
 				"sethostname",
 				"setns",
@@ -881,6 +897,50 @@ func DefaultProfile() *Seccomp {
 				Caps: []string{"CAP_AUDIT_WRITE"},
 			},
 		},
+		{
+			Names: []string{
+				"bpf",
+			},
+			Action:   ActErrno,
+			Errno:    "EPERM",
+			ErrnoRet: &eperm,
+			Args:     []*Arg{},
+			Excludes: Filter{
+				Caps: []string{"CAP_SYS_ADMIN", "CAP_BPF"},
+			},
+		},
+		{
+			Names: []string{
+				"bpf",
+			},
+			Action: ActAllow,
+			Args:   []*Arg{},
+			Includes: Filter{
+				Caps: []string{"CAP_BPF"},
+			},
+		},
+		{
+			Names: []string{
+				"perf_event_open",
+			},
+			Action:   ActErrno,
+			Errno:    "EPERM",
+			ErrnoRet: &eperm,
+			Args:     []*Arg{},
+			Excludes: Filter{
+				Caps: []string{"CAP_SYS_ADMIN", "CAP_BPF"},
+			},
+		},
+		{
+			Names: []string{
+				"perf_event_open",
+			},
+			Action: ActAllow,
+			Args:   []*Arg{},
+			Includes: Filter{
+				Caps: []string{"CAP_PERFMON"},
+			},
+		},
 	}
 
 	return &Seccomp{

vendor/github.com/containers/common/pkg/seccomp/errno_list.go

@@ -1,5 +1,4 @@
 //go:build linux && seccomp
-// +build linux,seccomp
 
 package seccomp
 

vendor/github.com/containers/common/pkg/seccomp/filter.go

@@ -1,5 +1,4 @@
 //go:build seccomp
-// +build seccomp
 
 // NOTE: this package has originally been copied from
 // github.com/opencontainers/runc and modified to work for other use cases
@@ -7,8 +6,10 @@
 package seccomp
 
 import (
+	"errors"
+	"fmt"
+
 	specs "github.com/opencontainers/runtime-spec/specs-go"
-	"github.com/pkg/errors"
 	libseccomp "github.com/seccomp/libseccomp-golang"
 	"golang.org/x/sys/unix"
 )
@@ -39,39 +40,39 @@ func BuildFilter(spec *specs.LinuxSeccomp) (*libseccomp.ScmpFilter, error) {
 
 	profile, err := specToSeccomp(spec)
 	if err != nil {
-		return nil, errors.Wrap(err, "convert spec to seccomp profile")
+		return nil, fmt.Errorf("convert spec to seccomp profile: %w", err)
 	}
 
 	defaultAction, err := toAction(profile.DefaultAction, profile.DefaultErrnoRet)
 	if err != nil {
-		return nil, errors.Wrapf(err, "convert default action %s", profile.DefaultAction)
+		return nil, fmt.Errorf("convert default action %s: %w", profile.DefaultAction, err)
 	}
 
 	filter, err := libseccomp.NewFilter(defaultAction)
 	if err != nil {
-		return nil, errors.Wrapf(err, "create filter for default action %s", defaultAction)
+		return nil, fmt.Errorf("create filter for default action %s: %w", defaultAction, err)
 	}
 
 	// Add extra architectures
 	for _, arch := range spec.Architectures {
 		libseccompArch, err := specArchToLibseccompArch(arch)
 		if err != nil {
-			return nil, errors.Wrap(err, "convert spec arch")
+			return nil, fmt.Errorf("convert spec arch: %w", err)
 		}
 
 		scmpArch, err := libseccomp.GetArchFromString(libseccompArch)
 		if err != nil {
-			return nil, errors.Wrapf(err, "validate Seccomp architecture %s", arch)
+			return nil, fmt.Errorf("validate Seccomp architecture %s: %w", arch, err)
 		}
 
 		if err := filter.AddArch(scmpArch); err != nil {
-			return nil, errors.Wrap(err, "add architecture to seccomp filter")
+			return nil, fmt.Errorf("add architecture to seccomp filter: %w", err)
 		}
 	}
 
 	// Unset no new privs bit
 	if err := filter.SetNoNewPrivsBit(false); err != nil {
-		return nil, errors.Wrap(err, "set no new privileges flag")
+		return nil, fmt.Errorf("set no new privileges flag: %w", err)
 	}
 
 	// Add a rule for each syscall
@@ -81,7 +82,7 @@ func BuildFilter(spec *specs.LinuxSeccomp) (*libseccomp.ScmpFilter, error) {
 		}
 
 		if err = matchSyscall(filter, call); err != nil {
-			return nil, errors.Wrap(err, "filter matches syscall")
+			return nil, fmt.Errorf("filter matches syscall: %w", err)
 		}
 	}
 
@@ -107,13 +108,13 @@ func matchSyscall(filter *libseccomp.ScmpFilter, call *Syscall) error {
 	// Convert the call's action to the libseccomp equivalent
 	callAct, err := toAction(call.Action, call.ErrnoRet)
 	if err != nil {
-		return errors.Wrapf(err, "convert action %s", call.Action)
+		return fmt.Errorf("convert action %s: %w", call.Action, err)
 	}
 
 	// Unconditional match - just add the rule
 	if len(call.Args) == 0 {
 		if err = filter.AddRule(callNum, callAct); err != nil {
-			return errors.Wrapf(err, "add seccomp filter rule for syscall %s", call.Name)
+			return fmt.Errorf("add seccomp filter rule for syscall %s: %w", call.Name, err)
 		}
 	} else {
 		// Linux system calls can have at most 6 arguments
@@ -127,10 +128,10 @@ func matchSyscall(filter *libseccomp.ScmpFilter, call *Syscall) error {
 		for _, cond := range call.Args {
 			newCond, err := toCondition(cond)
 			if err != nil {
-				return errors.Wrapf(err, "create seccomp syscall condition for syscall %s", call.Name)
+				return fmt.Errorf("create seccomp syscall condition for syscall %s: %w", call.Name, err)
 			}
 
-			argCounts[cond.Index] += 1
+			argCounts[cond.Index]++
 
 			conditions = append(conditions, newCond)
 		}
@@ -150,13 +151,13 @@ func matchSyscall(filter *libseccomp.ScmpFilter, call *Syscall) error {
 				condArr := []libseccomp.ScmpCondition{cond}
 
 				if err = filter.AddRuleConditional(callNum, callAct, condArr); err != nil {
-					return errors.Wrapf(err, "add seccomp rule for syscall %s", call.Name)
+					return fmt.Errorf("add seccomp rule for syscall %s: %w", call.Name, err)
 				}
 			}
 		} else if err = filter.AddRuleConditional(callNum, callAct, conditions); err != nil {
 			// No conditions share same argument
 			// Use new, proper behavior
-			return errors.Wrapf(err, "add seccomp rule for syscall %s", call.Name)
+			return fmt.Errorf("add seccomp rule for syscall %s: %w", call.Name, err)
 		}
 	}
 
@@ -168,7 +169,8 @@ func matchSyscall(filter *libseccomp.ScmpFilter, call *Syscall) error {
 func toAction(act Action, errnoRet *uint) (libseccomp.ScmpAction, error) {
 	switch act {
 	case ActKill:
-		return libseccomp.ActKill, nil
+		// lint was not passing until this was changed from ActKill to ActKilThread.
+		return libseccomp.ActKillThread, nil
 	case ActKillProcess:
 		return libseccomp.ActKillProcess, nil
 	case ActErrno:
@@ -188,7 +190,7 @@ func toAction(act Action, errnoRet *uint) (libseccomp.ScmpAction, error) {
 	case ActLog:
 		return libseccomp.ActLog, nil
 	default:
-		return libseccomp.ActInvalid, errors.Errorf("invalid action %s", act)
+		return libseccomp.ActInvalid, fmt.Errorf("invalid action %s", act)
 	}
 }
 
@@ -201,14 +203,14 @@ func toCondition(arg *Arg) (cond libseccomp.ScmpCondition, err error) {
 
 	op, err := toCompareOp(arg.Op)
 	if err != nil {
-		return cond, errors.Wrap(err, "convert compare operator")
+		return cond, fmt.Errorf("convert compare operator: %w", err)
 	}
 
 	condition, err := libseccomp.MakeCondition(
 		arg.Index, op, arg.Value, arg.ValueTwo,
 	)
 	if err != nil {
-		return cond, errors.Wrap(err, "make condition")
+		return cond, fmt.Errorf("make condition: %w", err)
 	}
 
 	return condition, nil
@@ -233,6 +235,6 @@ func toCompareOp(op Operator) (libseccomp.ScmpCompareOp, error) {
 	case OpMaskedEqual:
 		return libseccomp.CompareMaskedEqual, nil
 	default:
-		return libseccomp.CompareInvalid, errors.Errorf("invalid operator %s", op)
+		return libseccomp.CompareInvalid, fmt.Errorf("invalid operator %s", op)
 	}
 }

vendor/github.com/containers/common/pkg/seccomp/seccomp.json

@@ -55,9 +55,16 @@
 		{
 			"names": [
 				"bdflush",
+				"cachestat",
+				"futex_requeue",
+				"futex_wait",
+				"futex_waitv",
+				"futex_wake",
 				"io_pgetevents",
+				"io_pgetevents_time64",
 				"kexec_file_load",
 				"kexec_load",
+				"map_shadow_stack",
 				"migrate_pages",
 				"move_pages",
 				"nfsservctl",
@@ -72,9 +79,9 @@
 				"pciconfig_write",
 				"sgetmask",
 				"ssetmask",
-				"swapcontext",
 				"swapoff",
 				"swapon",
+				"syscall",
 				"sysfs",
 				"uselib",
 				"userfaultfd",
@@ -149,6 +156,7 @@
 				"fchdir",
 				"fchmod",
 				"fchmodat",
+				"fchmodat2",
 				"fchown",
 				"fchown32",
 				"fchownat",
@@ -228,6 +236,9 @@
 				"ipc",
 				"keyctl",
 				"kill",
+				"landlock_add_rule",
+				"landlock_create_ruleset",
+				"landlock_restrict_self",
 				"lchown",
 				"lchown32",
 				"lgetxattr",
@@ -313,7 +324,6 @@
 				"pwritev2",
 				"read",
 				"readahead",
-				"readdir",
 				"readlink",
 				"readlinkat",
 				"readv",
@@ -401,15 +411,12 @@
 				"shmdt",
 				"shmget",
 				"shutdown",
-				"sigaction",
 				"sigaltstack",
 				"signal",
 				"signalfd",
 				"signalfd4",
-				"sigpending",
 				"sigprocmask",
 				"sigreturn",
-				"sigsuspend",
 				"socketcall",
 				"socketpair",
 				"splice",
@@ -423,7 +430,6 @@
 				"sync",
 				"sync_file_range",
 				"syncfs",
-				"syscall",
 				"sysinfo",
 				"syslog",
 				"tee",
@@ -436,7 +442,6 @@
 				"timer_gettime64",
 				"timer_settime",
 				"timer_settime64",
-				"timerfd",
 				"timerfd_create",
 				"timerfd_gettime",
 				"timerfd_gettime64",
@@ -558,7 +563,8 @@
 		},
 		{
 			"names": [
-				"sync_file_range2"
+				"sync_file_range2",
+				"swapcontext"
 			],
 			"action": "SCMP_ACT_ALLOW",
 			"args": [],
@@ -638,6 +644,20 @@
 			},
 			"excludes": {}
 		},
+		{
+			"names": [
+				"riscv_flush_icache"
+			],
+			"action": "SCMP_ACT_ALLOW",
+			"args": [],
+			"comment": "",
+			"includes": {
+				"arches": [
+					"riscv64"
+				]
+			},
+			"excludes": {}
+		},
 		{
 			"names": [
 				"open_by_handle_at"
@@ -673,8 +693,8 @@
 				"bpf",
 				"fanotify_init",
 				"lookup_dcookie",
-				"perf_event_open",
 				"quotactl",
+				"quotactl_fd",
 				"setdomainname",
 				"sethostname",
 				"setns"
@@ -691,11 +711,11 @@
 		},
 		{
 			"names": [
-				"bpf",
 				"fanotify_init",
 				"lookup_dcookie",
 				"perf_event_open",
 				"quotactl",
+				"quotactl_fd",
 				"setdomainname",
 				"sethostname",
 				"setns"
@@ -1043,6 +1063,68 @@
 				]
 			},
 			"excludes": {}
+		},
+		{
+			"names": [
+				"bpf"
+			],
+			"action": "SCMP_ACT_ERRNO",
+			"args": [],
+			"comment": "",
+			"includes": {},
+			"excludes": {
+				"caps": [
+					"CAP_SYS_ADMIN",
+					"CAP_BPF"
+				]
+			},
+			"errnoRet": 1,
+			"errno": "EPERM"
+		},
+		{
+			"names": [
+				"bpf"
+			],
+			"action": "SCMP_ACT_ALLOW",
+			"args": [],
+			"comment": "",
+			"includes": {
+				"caps": [
+					"CAP_BPF"
+				]
+			},
+			"excludes": {}
+		},
+		{
+			"names": [
+				"perf_event_open"
+			],
+			"action": "SCMP_ACT_ERRNO",
+			"args": [],
+			"comment": "",
+			"includes": {},
+			"excludes": {
+				"caps": [
+					"CAP_SYS_ADMIN",
+					"CAP_BPF"
+				]
+			},
+			"errnoRet": 1,
+			"errno": "EPERM"
+		},
+		{
+			"names": [
+				"perf_event_open"
+			],
+			"action": "SCMP_ACT_ALLOW",
+			"args": [],
+			"comment": "",
+			"includes": {
+				"caps": [
+					"CAP_PERFMON"
+				]
+			},
+			"excludes": {}
 		}
 	]
 }

vendor/github.com/containers/common/pkg/seccomp/seccomp_linux.go

@@ -1,5 +1,4 @@
 //go:build seccomp
-// +build seccomp
 
 // SPDX-License-Identifier: Apache-2.0
 

vendor/github.com/containers/common/pkg/seccomp/seccomp_unsupported.go

@@ -1,5 +1,4 @@
 //go:build !linux || !seccomp
-// +build !linux !seccomp
 
 // SPDX-License-Identifier: Apache-2.0
 
@@ -15,12 +14,12 @@ import (
 
 var errNotSupported = errors.New("seccomp not enabled in this build")
 
-// LoadProfile returns an error on unsuppored systems
+// LoadProfile returns an error on unsupported systems
 func LoadProfile(body string, rs *specs.Spec) (*specs.LinuxSeccomp, error) {
 	return nil, errNotSupported
 }
 
-// GetDefaultProfile returns an error on unsuppored systems
+// GetDefaultProfile returns an error on unsupported systems
 func GetDefaultProfile(rs *specs.Spec) (*specs.LinuxSeccomp, error) {
 	return nil, errNotSupported
 }

vendor/github.com/containers/common/pkg/seccomp/supported.go

@@ -1,5 +1,4 @@
 //go:build linux && seccomp
-// +build linux,seccomp
 
 package seccomp
 

vendor/github.com/containers/common/pkg/seccomp/types.go

@@ -75,6 +75,7 @@ const (
 	ActTrace      Action = "SCMP_ACT_TRACE"
 	ActAllow      Action = "SCMP_ACT_ALLOW"
 	ActLog        Action = "SCMP_ACT_LOG"
+	ActNotify     Action = "SCMP_ACT_NOTIFY"
 )
 
 // Operator used to match syscall arguments in Seccomp

vendor/github.com/containers/common/pkg/seccomp/validate.go

@@ -1,12 +1,10 @@
 //go:build seccomp
-// +build seccomp
 
 package seccomp
 
 import (
 	"encoding/json"
-
+	"fmt"
-	"github.com/pkg/errors"
 )
 
 // ValidateProfile does a basic validation for the provided seccomp profile
@@ -14,16 +12,16 @@ import (
 func ValidateProfile(content string) error {
 	profile := &Seccomp{}
 	if err := json.Unmarshal([]byte(content), &profile); err != nil {
-		return errors.Wrap(err, "decoding seccomp profile")
+		return fmt.Errorf("decoding seccomp profile: %w", err)
 	}
 
 	spec, err := setupSeccomp(profile, nil)
 	if err != nil {
-		return errors.Wrap(err, "create seccomp spec")
+		return fmt.Errorf("create seccomp spec: %w", err)
 	}
 
 	if _, err := BuildFilter(spec); err != nil {
-		return errors.Wrap(err, "build seccomp filter")
+		return fmt.Errorf("build seccomp filter: %w", err)
 	}
 
 	return nil

vendor/github.com/containers/storage/LICENSE

@@ -0,0 +1,191 @@
+
+                                 Apache License
+                           Version 2.0, January 2004
+                        https://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   Copyright 2013-2016 Docker, Inc.
+
+   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
+
+       https://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.

vendor/github.com/containers/storage/NOTICE

@@ -0,0 +1,19 @@
+Docker
+Copyright 2012-2016 Docker, Inc.
+
+This product includes software developed at Docker, Inc. (https://www.docker.com).
+
+This product contains software (https://github.com/kr/pty) developed
+by Keith Rarick, licensed under the MIT License.
+
+The following is courtesy of our legal counsel:
+
+
+Use and transfer of Docker may be subject to certain restrictions by the
+United States and other governments.
+It is your responsibility to ensure that your use and/or transfer does not
+violate applicable laws.
+
+For more information, please see https://www.bis.doc.gov
+
+See also https://www.apache.org/dev/crypto.html and/or seek legal counsel.

vendor/github.com/containers/storage/pkg/fileutils/exists_freebsd.go

@@ -0,0 +1,38 @@
+package fileutils
+
+import (
+	"errors"
+	"os"
+	"syscall"
+
+	"golang.org/x/sys/unix"
+)
+
+// Exists checks whether a file or directory exists at the given path.
+// If the path is a symlink, the symlink is followed.
+func Exists(path string) error {
+	// It uses unix.Faccessat which is a faster operation compared to os.Stat for
+	// simply checking the existence of a file.
+	err := unix.Faccessat(unix.AT_FDCWD, path, unix.F_OK, 0)
+	if err != nil {
+		return &os.PathError{Op: "faccessat", Path: path, Err: err}
+	}
+	return nil
+}
+
+// Lexists checks whether a file or directory exists at the given path.
+// If the path is a symlink, the symlink itself is checked.
+func Lexists(path string) error {
+	// FreeBSD before 15.0 does not support the AT_SYMLINK_NOFOLLOW flag for
+	// faccessat. In this case, the call to faccessat will return EINVAL and
+	// we fall back to using Lstat.
+	err := unix.Faccessat(unix.AT_FDCWD, path, unix.F_OK, unix.AT_SYMLINK_NOFOLLOW)
+	if err != nil {
+		if errors.Is(err, syscall.EINVAL) {
+			_, err = os.Lstat(path)
+			return err
+		}
+		return &os.PathError{Op: "faccessat", Path: path, Err: err}
+	}
+	return nil
+}

vendor/github.com/containers/storage/pkg/fileutils/exists_unix.go

@@ -0,0 +1,33 @@
+//go:build !windows && !freebsd
+
+package fileutils
+
+import (
+	"os"
+
+	"golang.org/x/sys/unix"
+)
+
+// Exists checks whether a file or directory exists at the given path.
+// If the path is a symlink, the symlink is followed.
+func Exists(path string) error {
+	// It uses unix.Faccessat which is a faster operation compared to os.Stat for
+	// simply checking the existence of a file.
+	err := unix.Faccessat(unix.AT_FDCWD, path, unix.F_OK, unix.AT_EACCESS)
+	if err != nil {
+		return &os.PathError{Op: "faccessat", Path: path, Err: err}
+	}
+	return nil
+}
+
+// Lexists checks whether a file or directory exists at the given path.
+// If the path is a symlink, the symlink itself is checked.
+func Lexists(path string) error {
+	// It uses unix.Faccessat which is a faster operation compared to os.Stat for
+	// simply checking the existence of a file.
+	err := unix.Faccessat(unix.AT_FDCWD, path, unix.F_OK, unix.AT_SYMLINK_NOFOLLOW|unix.AT_EACCESS)
+	if err != nil {
+		return &os.PathError{Op: "faccessat", Path: path, Err: err}
+	}
+	return nil
+}

vendor/github.com/containers/storage/pkg/fileutils/exists_windows.go

@@ -0,0 +1,18 @@
+package fileutils
+
+import (
+	"os"
+)
+
+// Exists checks whether a file or directory exists at the given path.
+func Exists(path string) error {
+	_, err := os.Stat(path)
+	return err
+}
+
+// Lexists checks whether a file or directory exists at the given path, without
+// resolving symlinks
+func Lexists(path string) error {
+	_, err := os.Lstat(path)
+	return err
+}

vendor/github.com/containers/storage/pkg/fileutils/fileutils.go

@@ -0,0 +1,371 @@
+package fileutils
+
+import (
+	"errors"
+	"fmt"
+	"io"
+	"os"
+	"path/filepath"
+	"regexp"
+	"strings"
+	"text/scanner"
+
+	"github.com/sirupsen/logrus"
+)
+
+// PatternMatcher allows checking paths against a list of patterns
+type PatternMatcher struct {
+	patterns   []*Pattern
+	exclusions bool
+}
+
+// NewPatternMatcher creates a new matcher object for specific patterns that can
+// be used later to match against patterns against paths
+func NewPatternMatcher(patterns []string) (*PatternMatcher, error) {
+	pm := &PatternMatcher{
+		patterns: make([]*Pattern, 0, len(patterns)),
+	}
+	for _, p := range patterns {
+		// Eliminate leading and trailing whitespace.
+		p = strings.TrimSpace(p)
+		if p == "" {
+			continue
+		}
+		p = filepath.Clean(p)
+		newp := &Pattern{}
+		if p[0] == '!' {
+			if len(p) == 1 {
+				return nil, errors.New("illegal exclusion pattern: \"!\"")
+			}
+			newp.exclusion = true
+			p = strings.TrimPrefix(filepath.Clean(p[1:]), "/")
+			pm.exclusions = true
+		}
+		// Do some syntax checking on the pattern.
+		// filepath's Match() has some really weird rules that are inconsistent
+		// so instead of trying to dup their logic, just call Match() for its
+		// error state and if there is an error in the pattern return it.
+		// If this becomes an issue we can remove this since its really only
+		// needed in the error (syntax) case - which isn't really critical.
+		if _, err := filepath.Match(p, "."); err != nil {
+			return nil, err
+		}
+		newp.cleanedPattern = p
+		newp.dirs = strings.Split(p, string(os.PathSeparator))
+		pm.patterns = append(pm.patterns, newp)
+	}
+	return pm, nil
+}
+
+// Deprecated: Please use the `MatchesResult` method instead.
+// Matches matches path against all the patterns. Matches is not safe to be
+// called concurrently
+func (pm *PatternMatcher) Matches(file string) (bool, error) {
+	matched := false
+	file = filepath.FromSlash(file)
+
+	for _, pattern := range pm.patterns {
+		negative := false
+
+		if pattern.exclusion {
+			negative = true
+		}
+
+		match, err := pattern.match(file)
+		if err != nil {
+			return false, err
+		}
+
+		if match {
+			matched = !negative
+		}
+	}
+
+	if matched {
+		logrus.Debugf("Skipping excluded path: %s", file)
+	}
+
+	return matched, nil
+}
+
+type MatchResult struct {
+	isMatched         bool
+	matches, excludes uint
+}
+
+// Excludes returns true if the overall result is matched
+func (m *MatchResult) IsMatched() bool {
+	return m.isMatched
+}
+
+// Excludes returns the amount of matches of an MatchResult
+func (m *MatchResult) Matches() uint {
+	return m.matches
+}
+
+// Excludes returns the amount of excludes of an MatchResult
+func (m *MatchResult) Excludes() uint {
+	return m.excludes
+}
+
+// MatchesResult verifies the provided filepath against all patterns.
+// It returns the `*MatchResult` result for the patterns on success, otherwise
+// an error. This method is not safe to be called concurrently.
+func (pm *PatternMatcher) MatchesResult(file string) (res *MatchResult, err error) {
+	file = filepath.FromSlash(file)
+	res = &MatchResult{false, 0, 0}
+
+	for _, pattern := range pm.patterns {
+		negative := false
+
+		if pattern.exclusion {
+			negative = true
+		}
+
+		match, err := pattern.match(file)
+		if err != nil {
+			return nil, err
+		}
+
+		if match {
+			res.isMatched = !negative
+			if negative {
+				res.excludes++
+			} else {
+				res.matches++
+			}
+		}
+	}
+
+	if res.matches > 0 {
+		logrus.Debugf("Skipping excluded path: %s", file)
+	}
+
+	return res, nil
+}
+
+// IsMatch verifies the provided filepath against all patterns and returns true
+// if it matches. A match is valid if the last match is a positive one.
+// It returns an error on failure and is not safe to be called concurrently.
+func (pm *PatternMatcher) IsMatch(file string) (matched bool, err error) {
+	res, err := pm.MatchesResult(file)
+	if err != nil {
+		return false, err
+	}
+	return res.isMatched, nil
+}
+
+// Exclusions returns true if any of the patterns define exclusions
+func (pm *PatternMatcher) Exclusions() bool {
+	return pm.exclusions
+}
+
+// Patterns returns array of active patterns
+func (pm *PatternMatcher) Patterns() []*Pattern {
+	return pm.patterns
+}
+
+// Pattern defines a single regexp used to filter file paths.
+type Pattern struct {
+	cleanedPattern string
+	dirs           []string
+	regexp         *regexp.Regexp
+	exclusion      bool
+}
+
+func (p *Pattern) String() string {
+	return p.cleanedPattern
+}
+
+// Exclusion returns true if this pattern defines exclusion
+func (p *Pattern) Exclusion() bool {
+	return p.exclusion
+}
+
+func (p *Pattern) match(path string) (bool, error) {
+	if p.regexp == nil {
+		if err := p.compile(); err != nil {
+			return false, filepath.ErrBadPattern
+		}
+	}
+
+	b := p.regexp.MatchString(path)
+
+	return b, nil
+}
+
+func (p *Pattern) compile() error {
+	regStr := "^"
+	pattern := p.cleanedPattern
+	// Go through the pattern and convert it to a regexp.
+	// We use a scanner so we can support utf-8 chars.
+	var scan scanner.Scanner
+	scan.Init(strings.NewReader(pattern))
+
+	sl := string(os.PathSeparator)
+	escSL := sl
+	const bs = `\`
+	if sl == bs {
+		escSL += bs
+	}
+
+	for scan.Peek() != scanner.EOF {
+		ch := scan.Next()
+
+		if ch == '*' {
+			if scan.Peek() == '*' {
+				// is some flavor of "**"
+				scan.Next()
+
+				// Treat **/ as ** so eat the "/"
+				if string(scan.Peek()) == sl {
+					scan.Next()
+				}
+
+				if scan.Peek() == scanner.EOF {
+					// is "**EOF" - to align with .gitignore just accept all
+					regStr += ".*"
+				} else {
+					// is "**"
+					// Note that this allows for any # of /'s (even 0) because
+					// the .* will eat everything, even /'s
+					regStr += "(.*" + escSL + ")?"
+				}
+			} else {
+				// is "*" so map it to anything but "/"
+				regStr += "[^" + escSL + "]*"
+			}
+		} else if ch == '?' {
+			// "?" is any char except "/"
+			regStr += "[^" + escSL + "]"
+		} else if ch == '.' || ch == '$' {
+			// Escape some regexp special chars that have no meaning
+			// in golang's filepath.Match
+			regStr += bs + string(ch)
+		} else if ch == '\\' {
+			// escape next char.
+			if sl == bs {
+				// On windows map "\" to "\\", meaning an escaped backslash,
+				// and then just continue because filepath.Match on
+				// Windows doesn't allow escaping at all
+				regStr += escSL
+				continue
+			}
+			if scan.Peek() != scanner.EOF {
+				regStr += bs + string(scan.Next())
+			} else {
+				return filepath.ErrBadPattern
+			}
+		} else {
+			regStr += string(ch)
+		}
+	}
+
+	regStr += "(" + escSL + ".*)?$"
+
+	re, err := regexp.Compile(regStr)
+	if err != nil {
+		return err
+	}
+
+	p.regexp = re
+	return nil
+}
+
+// Matches returns true if file matches any of the patterns
+// and isn't excluded by any of the subsequent patterns.
+func Matches(file string, patterns []string) (bool, error) {
+	pm, err := NewPatternMatcher(patterns)
+	if err != nil {
+		return false, err
+	}
+	file = filepath.Clean(file)
+
+	if file == "." {
+		// Don't let them exclude everything, kind of silly.
+		return false, nil
+	}
+
+	return pm.IsMatch(file)
+}
+
+// CopyFile copies from src to dst until either EOF is reached
+// on src or an error occurs. It verifies src exists and removes
+// the dst if it exists.
+func CopyFile(src, dst string) (int64, error) {
+	cleanSrc := filepath.Clean(src)
+	cleanDst := filepath.Clean(dst)
+	if cleanSrc == cleanDst {
+		return 0, nil
+	}
+	sf, err := os.Open(cleanSrc)
+	if err != nil {
+		return 0, err
+	}
+	defer sf.Close()
+	if err := os.Remove(cleanDst); err != nil && !os.IsNotExist(err) {
+		return 0, err
+	}
+	df, err := os.Create(cleanDst)
+	if err != nil {
+		return 0, err
+	}
+	defer df.Close()
+	return io.Copy(df, sf)
+}
+
+// ReadSymlinkedDirectory returns the target directory of a symlink.
+// The target of the symbolic link may not be a file.
+func ReadSymlinkedDirectory(path string) (string, error) {
+	var realPath string
+	var err error
+	if realPath, err = filepath.Abs(path); err != nil {
+		return "", fmt.Errorf("unable to get absolute path for %s: %w", path, err)
+	}
+	if realPath, err = filepath.EvalSymlinks(realPath); err != nil {
+		return "", fmt.Errorf("failed to canonicalise path for %s: %w", path, err)
+	}
+	realPathInfo, err := os.Stat(realPath)
+	if err != nil {
+		return "", fmt.Errorf("failed to stat target '%s' of '%s': %w", realPath, path, err)
+	}
+	if !realPathInfo.Mode().IsDir() {
+		return "", fmt.Errorf("canonical path points to a file '%s'", realPath)
+	}
+	return realPath, nil
+}
+
+// ReadSymlinkedPath returns the target directory of a symlink.
+// The target of the symbolic link can be a file and a directory.
+func ReadSymlinkedPath(path string) (realPath string, err error) {
+	if realPath, err = filepath.Abs(path); err != nil {
+		return "", fmt.Errorf("unable to get absolute path for %q: %w", path, err)
+	}
+	if realPath, err = filepath.EvalSymlinks(realPath); err != nil {
+		return "", fmt.Errorf("failed to canonicalise path for %q: %w", path, err)
+	}
+	if err := Exists(realPath); err != nil {
+		return "", fmt.Errorf("failed to stat target %q of %q: %w", realPath, path, err)
+	}
+	return realPath, nil
+}
+
+// CreateIfNotExists creates a file or a directory only if it does not already exist.
+func CreateIfNotExists(path string, isDir bool) error {
+	if err := Exists(path); err != nil {
+		if os.IsNotExist(err) {
+			if isDir {
+				return os.MkdirAll(path, 0o755)
+			}
+			if err := os.MkdirAll(filepath.Dir(path), 0o755); err != nil {
+				return err
+			}
+			f, err := os.OpenFile(path, os.O_CREATE, 0o755)
+			if err != nil {
+				return err
+			}
+			f.Close()
+		}
+	}
+	return nil
+}

vendor/github.com/containers/storage/pkg/fileutils/fileutils_darwin.go

@@ -0,0 +1,27 @@
+package fileutils
+
+import (
+	"os"
+	"os/exec"
+	"strconv"
+	"strings"
+)
+
+// GetTotalUsedFds returns the number of used File Descriptors by
+// executing `lsof -p PID`
+func GetTotalUsedFds() int {
+	pid := os.Getpid()
+
+	cmd := exec.Command("lsof", "-p", strconv.Itoa(pid))
+
+	output, err := cmd.CombinedOutput()
+	if err != nil {
+		return -1
+	}
+
+	outputStr := strings.TrimSpace(string(output))
+
+	fds := strings.Split(outputStr, "\n")
+
+	return len(fds) - 1
+}

vendor/github.com/containers/storage/pkg/fileutils/fileutils_solaris.go

@@ -0,0 +1,7 @@
+package fileutils
+
+// GetTotalUsedFds Returns the number of used File Descriptors.
+// On Solaris these limits are per process and not systemwide
+func GetTotalUsedFds() int {
+	return -1
+}

vendor/github.com/containers/storage/pkg/fileutils/fileutils_unix.go

@@ -0,0 +1,21 @@
+//go:build linux || freebsd
+
+package fileutils
+
+import (
+	"fmt"
+	"os"
+
+	"github.com/sirupsen/logrus"
+)
+
+// GetTotalUsedFds Returns the number of used File Descriptors by
+// reading it via /proc filesystem.
+func GetTotalUsedFds() int {
+	if fds, err := os.ReadDir(fmt.Sprintf("/proc/%d/fd", os.Getpid())); err != nil {
+		logrus.Errorf("%v", err)
+	} else {
+		return len(fds)
+	}
+	return -1
+}

vendor/github.com/containers/storage/pkg/fileutils/fileutils_windows.go

@@ -0,0 +1,7 @@
+package fileutils
+
+// GetTotalUsedFds Returns the number of used File Descriptors. Not supported
+// on Windows.
+func GetTotalUsedFds() int {
+	return -1
+}

vendor/github.com/containers/storage/pkg/fileutils/reflink_linux.go

@@ -0,0 +1,20 @@
+package fileutils
+
+import (
+	"io"
+	"os"
+
+	"golang.org/x/sys/unix"
+)
+
+// ReflinkOrCopy attempts to reflink the source to the destination fd.
+// If reflinking fails or is unsupported, it falls back to io.Copy().
+func ReflinkOrCopy(src, dst *os.File) error {
+	err := unix.IoctlFileClone(int(dst.Fd()), int(src.Fd()))
+	if err == nil {
+		return nil
+	}
+
+	_, err = io.Copy(dst, src)
+	return err
+}

vendor/github.com/containers/storage/pkg/fileutils/reflink_unsupported.go

@@ -0,0 +1,15 @@
+//go:build !linux
+
+package fileutils
+
+import (
+	"io"
+	"os"
+)
+
+// ReflinkOrCopy attempts to reflink the source to the destination fd.
+// If reflinking fails or is unsupported, it falls back to io.Copy().
+func ReflinkOrCopy(src, dst *os.File) error {
+	_, err := io.Copy(dst, src)
+	return err
+}

vendor/github.com/containers/storage/pkg/idtools/idtools.go

@@ -0,0 +1,620 @@
+package idtools
+
+import (
+	"bufio"
+	"errors"
+	"fmt"
+	"io/fs"
+	"os"
+	"os/user"
+	"runtime"
+	"sort"
+	"strconv"
+	"strings"
+	"sync"
+	"syscall"
+
+	"github.com/containers/storage/pkg/system"
+	"github.com/sirupsen/logrus"
+)
+
+// IDMap contains a single entry for user namespace range remapping. An array
+// of IDMap entries represents the structure that will be provided to the Linux
+// kernel for creating a user namespace.
+type IDMap struct {
+	ContainerID int `json:"container_id"`
+	HostID      int `json:"host_id"`
+	Size        int `json:"size"`
+}
+
+type subIDRange struct {
+	Start  int
+	Length int
+}
+
+type ranges []subIDRange
+
+func (e ranges) Len() int           { return len(e) }
+func (e ranges) Swap(i, j int)      { e[i], e[j] = e[j], e[i] }
+func (e ranges) Less(i, j int) bool { return e[i].Start < e[j].Start }
+
+const (
+	subuidFileName          string = "/etc/subuid"
+	subgidFileName          string = "/etc/subgid"
+	ContainersOverrideXattr        = "user.containers.override_stat"
+)
+
+// MkdirAllAs creates a directory (include any along the path) and then modifies
+// ownership to the requested uid/gid.  If the directory already exists, this
+// function will still change ownership to the requested uid/gid pair.
+// Deprecated: Use MkdirAllAndChown
+func MkdirAllAs(path string, mode os.FileMode, ownerUID, ownerGID int) error {
+	return mkdirAs(path, mode, ownerUID, ownerGID, true, true)
+}
+
+// MkdirAs creates a directory and then modifies ownership to the requested uid/gid.
+// If the directory already exists, this function still changes ownership
+// Deprecated: Use MkdirAndChown with a IDPair
+func MkdirAs(path string, mode os.FileMode, ownerUID, ownerGID int) error {
+	return mkdirAs(path, mode, ownerUID, ownerGID, false, true)
+}
+
+// MkdirAllAndChown creates a directory (include any along the path) and then modifies
+// ownership to the requested uid/gid.  If the directory already exists, this
+// function will still change ownership to the requested uid/gid pair.
+func MkdirAllAndChown(path string, mode os.FileMode, ids IDPair) error {
+	return mkdirAs(path, mode, ids.UID, ids.GID, true, true)
+}
+
+// MkdirAndChown creates a directory and then modifies ownership to the requested uid/gid.
+// If the directory already exists, this function still changes ownership
+func MkdirAndChown(path string, mode os.FileMode, ids IDPair) error {
+	return mkdirAs(path, mode, ids.UID, ids.GID, false, true)
+}
+
+// MkdirAllAndChownNew creates a directory (include any along the path) and then modifies
+// ownership ONLY of newly created directories to the requested uid/gid. If the
+// directories along the path exist, no change of ownership will be performed
+func MkdirAllAndChownNew(path string, mode os.FileMode, ids IDPair) error {
+	return mkdirAs(path, mode, ids.UID, ids.GID, true, false)
+}
+
+// GetRootUIDGID retrieves the remapped root uid/gid pair from the set of maps.
+// If the maps are empty, then the root uid/gid will default to "real" 0/0
+func GetRootUIDGID(uidMap, gidMap []IDMap) (int, int, error) {
+	var uid, gid int
+	var err error
+	if len(uidMap) == 1 && uidMap[0].Size == 1 {
+		uid = uidMap[0].HostID
+	} else {
+		uid, err = RawToHost(0, uidMap)
+		if err != nil {
+			return -1, -1, err
+		}
+	}
+	if len(gidMap) == 1 && gidMap[0].Size == 1 {
+		gid = gidMap[0].HostID
+	} else {
+		gid, err = RawToHost(0, gidMap)
+		if err != nil {
+			return -1, -1, err
+		}
+	}
+	return uid, gid, nil
+}
+
+// RawToContainer takes an id mapping, and uses it to translate a host ID to
+// the remapped ID. If no map is provided, then the translation assumes a
+// 1-to-1 mapping and returns the passed in id.
+//
+// If you wish to map a (uid,gid) combination you should use the corresponding
+// IDMappings methods, which ensure that you are mapping the correct ID against
+// the correct mapping.
+func RawToContainer(hostID int, idMap []IDMap) (int, error) {
+	if idMap == nil {
+		return hostID, nil
+	}
+	for _, m := range idMap {
+		if (hostID >= m.HostID) && (hostID <= (m.HostID + m.Size - 1)) {
+			contID := m.ContainerID + (hostID - m.HostID)
+			return contID, nil
+		}
+	}
+	return -1, fmt.Errorf("host ID %d cannot be mapped to a container ID", hostID)
+}
+
+// RawToHost takes an id mapping and a remapped ID, and translates the ID to
+// the mapped host ID. If no map is provided, then the translation assumes a
+// 1-to-1 mapping and returns the passed in id.
+//
+// If you wish to map a (uid,gid) combination you should use the corresponding
+// IDMappings methods, which ensure that you are mapping the correct ID against
+// the correct mapping.
+func RawToHost(contID int, idMap []IDMap) (int, error) {
+	if idMap == nil {
+		return contID, nil
+	}
+	for _, m := range idMap {
+		if (contID >= m.ContainerID) && (contID <= (m.ContainerID + m.Size - 1)) {
+			hostID := m.HostID + (contID - m.ContainerID)
+			return hostID, nil
+		}
+	}
+	return -1, fmt.Errorf("container ID %d cannot be mapped to a host ID", contID)
+}
+
+// IDPair is a UID and GID pair
+type IDPair struct {
+	UID int
+	GID int
+}
+
+// IDMappings contains a mappings of UIDs and GIDs
+type IDMappings struct {
+	uids []IDMap
+	gids []IDMap
+}
+
+// NewIDMappings takes a requested user and group name and
+// using the data from /etc/sub{uid,gid} ranges, creates the
+// proper uid and gid remapping ranges for that user/group pair
+func NewIDMappings(username, groupname string) (*IDMappings, error) {
+	subuidRanges, err := readSubuid(username)
+	if err != nil {
+		return nil, err
+	}
+	subgidRanges, err := readSubgid(groupname)
+	if err != nil {
+		return nil, err
+	}
+	if len(subuidRanges) == 0 {
+		return nil, fmt.Errorf("no subuid ranges found for user %q in %s", username, subuidFileName)
+	}
+	if len(subgidRanges) == 0 {
+		return nil, fmt.Errorf("no subgid ranges found for group %q in %s", groupname, subgidFileName)
+	}
+
+	return &IDMappings{
+		uids: createIDMap(subuidRanges),
+		gids: createIDMap(subgidRanges),
+	}, nil
+}
+
+// NewIDMappingsFromMaps creates a new mapping from two slices
+// Deprecated: this is a temporary shim while transitioning to IDMapping
+func NewIDMappingsFromMaps(uids []IDMap, gids []IDMap) *IDMappings {
+	return &IDMappings{uids: uids, gids: gids}
+}
+
+// RootPair returns a uid and gid pair for the root user. The error is ignored
+// because a root user always exists, and the defaults are correct when the uid
+// and gid maps are empty.
+func (i *IDMappings) RootPair() IDPair {
+	uid, gid, _ := GetRootUIDGID(i.uids, i.gids)
+	return IDPair{UID: uid, GID: gid}
+}
+
+// ToHost returns the host UID and GID for the container uid, gid.
+func (i *IDMappings) ToHost(pair IDPair) (IDPair, error) {
+	var err error
+	var target IDPair
+
+	target.UID, err = RawToHost(pair.UID, i.uids)
+	if err != nil {
+		return target, err
+	}
+
+	target.GID, err = RawToHost(pair.GID, i.gids)
+	return target, err
+}
+
+var (
+	overflowUIDOnce sync.Once
+	overflowGIDOnce sync.Once
+	overflowUID     int
+	overflowGID     int
+)
+
+// getOverflowUID returns the UID mapped to the overflow user
+func getOverflowUID() int {
+	overflowUIDOnce.Do(func() {
+		// 65534 is the value on older kernels where /proc/sys/kernel/overflowuid is not present
+		overflowUID = 65534
+		if content, err := os.ReadFile("/proc/sys/kernel/overflowuid"); err == nil {
+			if tmp, err := strconv.Atoi(string(content)); err == nil {
+				overflowUID = tmp
+			}
+		}
+	})
+	return overflowUID
+}
+
+// getOverflowGID returns the GID mapped to the overflow user
+func getOverflowGID() int {
+	overflowGIDOnce.Do(func() {
+		// 65534 is the value on older kernels where /proc/sys/kernel/overflowgid is not present
+		overflowGID = 65534
+		if content, err := os.ReadFile("/proc/sys/kernel/overflowgid"); err == nil {
+			if tmp, err := strconv.Atoi(string(content)); err == nil {
+				overflowGID = tmp
+			}
+		}
+	})
+	return overflowGID
+}
+
+// ToHost returns the host UID and GID for the container uid, gid.
+// Remapping is only performed if the ids aren't already the remapped root ids
+// If the mapping is not possible because the target ID is not mapped into
+// the namespace, then the overflow ID is used.
+func (i *IDMappings) ToHostOverflow(pair IDPair) (IDPair, error) {
+	var err error
+	target := i.RootPair()
+
+	if pair.UID != target.UID {
+		target.UID, err = RawToHost(pair.UID, i.uids)
+		if err != nil {
+			target.UID = getOverflowUID()
+			logrus.Debugf("Failed to map UID %v to the target mapping, using the overflow ID %v", pair.UID, target.UID)
+		}
+	}
+
+	if pair.GID != target.GID {
+		target.GID, err = RawToHost(pair.GID, i.gids)
+		if err != nil {
+			target.GID = getOverflowGID()
+			logrus.Debugf("Failed to map GID %v to the target mapping, using the overflow ID %v", pair.GID, target.GID)
+		}
+	}
+	return target, nil
+}
+
+// ToContainer returns the container UID and GID for the host uid and gid
+func (i *IDMappings) ToContainer(pair IDPair) (int, int, error) {
+	uid, err := RawToContainer(pair.UID, i.uids)
+	if err != nil {
+		return -1, -1, err
+	}
+	gid, err := RawToContainer(pair.GID, i.gids)
+	return uid, gid, err
+}
+
+// Empty returns true if there are no id mappings
+func (i *IDMappings) Empty() bool {
+	return len(i.uids) == 0 && len(i.gids) == 0
+}
+
+// UIDs return the UID mapping
+// TODO: remove this once everything has been refactored to use pairs
+func (i *IDMappings) UIDs() []IDMap {
+	return i.uids
+}
+
+// GIDs return the UID mapping
+// TODO: remove this once everything has been refactored to use pairs
+func (i *IDMappings) GIDs() []IDMap {
+	return i.gids
+}
+
+func createIDMap(subidRanges ranges) []IDMap {
+	idMap := []IDMap{}
+
+	// sort the ranges by lowest ID first
+	sort.Sort(subidRanges)
+	containerID := 0
+	for _, idrange := range subidRanges {
+		idMap = append(idMap, IDMap{
+			ContainerID: containerID,
+			HostID:      idrange.Start,
+			Size:        idrange.Length,
+		})
+		containerID = containerID + idrange.Length
+	}
+	return idMap
+}
+
+// parseSubidFile will read the appropriate file (/etc/subuid or /etc/subgid)
+// and return all found ranges for a specified username. If the special value
+// "ALL" is supplied for username, then all ranges in the file will be returned
+func parseSubidFile(path, username string) (ranges, error) {
+	var (
+		rangeList ranges
+		uidstr    string
+	)
+	if u, err := user.Lookup(username); err == nil {
+		uidstr = u.Uid
+	}
+
+	subidFile, err := os.Open(path)
+	if err != nil {
+		return rangeList, err
+	}
+	defer subidFile.Close()
+
+	s := bufio.NewScanner(subidFile)
+	for s.Scan() {
+		if err := s.Err(); err != nil {
+			return rangeList, err
+		}
+
+		text := strings.TrimSpace(s.Text())
+		if text == "" || strings.HasPrefix(text, "#") {
+			continue
+		}
+		parts := strings.Split(text, ":")
+		if len(parts) != 3 {
+			return rangeList, fmt.Errorf("cannot parse subuid/gid information: Format not correct for %s file", path)
+		}
+		if parts[0] == username || username == "ALL" || (parts[0] == uidstr && parts[0] != "") {
+			startid, err := strconv.Atoi(parts[1])
+			if err != nil {
+				return rangeList, fmt.Errorf("string to int conversion failed during subuid/gid parsing of %s: %w", path, err)
+			}
+			length, err := strconv.Atoi(parts[2])
+			if err != nil {
+				return rangeList, fmt.Errorf("string to int conversion failed during subuid/gid parsing of %s: %w", path, err)
+			}
+			rangeList = append(rangeList, subIDRange{startid, length})
+		}
+	}
+	return rangeList, nil
+}
+
+func checkChownErr(err error, name string, uid, gid int) error {
+	var e *os.PathError
+	if errors.As(err, &e) && e.Err == syscall.EINVAL {
+		return fmt.Errorf(`potentially insufficient UIDs or GIDs available in user namespace (requested %d:%d for %s): Check /etc/subuid and /etc/subgid if configured locally and run "podman system migrate": %w`, uid, gid, name, err)
+	}
+	return err
+}
+
+// Stat contains file states that can be overridden with ContainersOverrideXattr.
+type Stat struct {
+	IDs   IDPair
+	Mode  os.FileMode
+	Major int
+	Minor int
+}
+
+// FormatContainersOverrideXattr will format the given uid, gid, and mode into a string
+// that can be used as the value for the ContainersOverrideXattr xattr.
+func FormatContainersOverrideXattr(uid, gid, mode int) string {
+	return FormatContainersOverrideXattrDevice(uid, gid, fs.FileMode(mode), 0, 0)
+}
+
+// FormatContainersOverrideXattrDevice will format the given uid, gid, and mode into a string
+// that can be used as the value for the ContainersOverrideXattr xattr.  For devices, it also
+// needs the major and minor numbers.
+func FormatContainersOverrideXattrDevice(uid, gid int, mode fs.FileMode, major, minor int) string {
+	typ := ""
+	switch mode & os.ModeType {
+	case os.ModeDir:
+		typ = "dir"
+	case os.ModeSymlink:
+		typ = "symlink"
+	case os.ModeNamedPipe:
+		typ = "pipe"
+	case os.ModeSocket:
+		typ = "socket"
+	case os.ModeDevice:
+		typ = fmt.Sprintf("block-%d-%d", major, minor)
+	case os.ModeDevice | os.ModeCharDevice:
+		typ = fmt.Sprintf("char-%d-%d", major, minor)
+	default:
+		typ = "file"
+	}
+	unixMode := mode & os.ModePerm
+	if mode&os.ModeSetuid != 0 {
+		unixMode |= 0o4000
+	}
+	if mode&os.ModeSetgid != 0 {
+		unixMode |= 0o2000
+	}
+	if mode&os.ModeSticky != 0 {
+		unixMode |= 0o1000
+	}
+	return fmt.Sprintf("%d:%d:%04o:%s", uid, gid, unixMode, typ)
+}
+
+// GetContainersOverrideXattr will get and decode ContainersOverrideXattr.
+func GetContainersOverrideXattr(path string) (Stat, error) {
+	xstat, err := system.Lgetxattr(path, ContainersOverrideXattr)
+	if err != nil {
+		return Stat{}, err
+	}
+	return parseOverrideXattr(xstat) // This will fail if (xstat, err) == (nil, nil), i.e. the xattr does not exist.
+}
+
+func parseOverrideXattr(xstat []byte) (Stat, error) {
+	var stat Stat
+	attrs := strings.Split(string(xstat), ":")
+	if len(attrs) < 3 {
+		return stat, fmt.Errorf("the number of parts in %s is less than 3",
+			ContainersOverrideXattr)
+	}
+
+	value, err := strconv.ParseUint(attrs[0], 10, 32)
+	if err != nil {
+		return stat, fmt.Errorf("failed to parse UID: %w", err)
+	}
+	stat.IDs.UID = int(value)
+
+	value, err = strconv.ParseUint(attrs[1], 10, 32)
+	if err != nil {
+		return stat, fmt.Errorf("failed to parse GID: %w", err)
+	}
+	stat.IDs.GID = int(value)
+
+	value, err = strconv.ParseUint(attrs[2], 8, 32)
+	if err != nil {
+		return stat, fmt.Errorf("failed to parse mode: %w", err)
+	}
+	stat.Mode = os.FileMode(value) & os.ModePerm
+	if value&0o1000 != 0 {
+		stat.Mode |= os.ModeSticky
+	}
+	if value&0o2000 != 0 {
+		stat.Mode |= os.ModeSetgid
+	}
+	if value&0o4000 != 0 {
+		stat.Mode |= os.ModeSetuid
+	}
+
+	if len(attrs) > 3 {
+		typ := attrs[3]
+		if strings.HasPrefix(typ, "file") {
+		} else if strings.HasPrefix(typ, "dir") {
+			stat.Mode |= os.ModeDir
+		} else if strings.HasPrefix(typ, "symlink") {
+			stat.Mode |= os.ModeSymlink
+		} else if strings.HasPrefix(typ, "pipe") {
+			stat.Mode |= os.ModeNamedPipe
+		} else if strings.HasPrefix(typ, "socket") {
+			stat.Mode |= os.ModeSocket
+		} else if strings.HasPrefix(typ, "block") {
+			stat.Mode |= os.ModeDevice
+			stat.Major, stat.Minor, err = parseDevice(typ)
+			if err != nil {
+				return stat, err
+			}
+		} else if strings.HasPrefix(typ, "char") {
+			stat.Mode |= os.ModeDevice | os.ModeCharDevice
+			stat.Major, stat.Minor, err = parseDevice(typ)
+			if err != nil {
+				return stat, err
+			}
+		} else {
+			return stat, fmt.Errorf("invalid file type %s", typ)
+		}
+	}
+	return stat, nil
+}
+
+func parseDevice(typ string) (int, int, error) {
+	parts := strings.Split(typ, "-")
+	// If there are more than 3 parts, just ignore them to be forward compatible
+	if len(parts) < 3 {
+		return 0, 0, fmt.Errorf("invalid device type %s", typ)
+	}
+	if parts[0] != "block" && parts[0] != "char" {
+		return 0, 0, fmt.Errorf("invalid device type %s", typ)
+	}
+	major, err := strconv.Atoi(parts[1])
+	if err != nil {
+		return 0, 0, fmt.Errorf("failed to parse major number: %w", err)
+	}
+	minor, err := strconv.Atoi(parts[2])
+	if err != nil {
+		return 0, 0, fmt.Errorf("failed to parse minor number: %w", err)
+	}
+	return major, minor, nil
+}
+
+// SetContainersOverrideXattr will encode and set ContainersOverrideXattr.
+func SetContainersOverrideXattr(path string, stat Stat) error {
+	value := FormatContainersOverrideXattrDevice(stat.IDs.UID, stat.IDs.GID, stat.Mode, stat.Major, stat.Minor)
+	return system.Lsetxattr(path, ContainersOverrideXattr, []byte(value), 0)
+}
+
+func SafeChown(name string, uid, gid int) error {
+	if runtime.GOOS == "darwin" {
+		stat := Stat{
+			Mode: os.FileMode(0o0700),
+		}
+		xstat, err := system.Lgetxattr(name, ContainersOverrideXattr)
+		if err == nil && xstat != nil {
+			stat, err = parseOverrideXattr(xstat)
+			if err != nil {
+				return err
+			}
+		} else {
+			st, err := os.Stat(name) // Ideally we would share this with system.Stat below, but then we would need to convert Mode.
+			if err != nil {
+				return err
+			}
+			stat.Mode = st.Mode()
+		}
+		stat.IDs = IDPair{UID: uid, GID: gid}
+		if err = SetContainersOverrideXattr(name, stat); err != nil {
+			return err
+		}
+		uid = os.Getuid()
+		gid = os.Getgid()
+	}
+	if stat, statErr := system.Stat(name); statErr == nil {
+		if stat.UID() == uint32(uid) && stat.GID() == uint32(gid) {
+			return nil
+		}
+	}
+	return checkChownErr(os.Chown(name, uid, gid), name, uid, gid)
+}
+
+func SafeLchown(name string, uid, gid int) error {
+	if runtime.GOOS == "darwin" {
+		stat := Stat{
+			Mode: os.FileMode(0o0700),
+		}
+		xstat, err := system.Lgetxattr(name, ContainersOverrideXattr)
+		if err == nil && xstat != nil {
+			stat, err = parseOverrideXattr(xstat)
+			if err != nil {
+				return err
+			}
+		} else {
+			st, err := os.Lstat(name) // Ideally we would share this with system.Stat below, but then we would need to convert Mode.
+			if err != nil {
+				return err
+			}
+			stat.Mode = st.Mode()
+		}
+		stat.IDs = IDPair{UID: uid, GID: gid}
+		if err = SetContainersOverrideXattr(name, stat); err != nil {
+			return err
+		}
+		uid = os.Getuid()
+		gid = os.Getgid()
+	}
+	if stat, statErr := system.Lstat(name); statErr == nil {
+		if stat.UID() == uint32(uid) && stat.GID() == uint32(gid) {
+			return nil
+		}
+	}
+	return checkChownErr(os.Lchown(name, uid, gid), name, uid, gid)
+}
+
+type sortByHostID []IDMap
+
+func (e sortByHostID) Len() int           { return len(e) }
+func (e sortByHostID) Swap(i, j int)      { e[i], e[j] = e[j], e[i] }
+func (e sortByHostID) Less(i, j int) bool { return e[i].HostID < e[j].HostID }
+
+type sortByContainerID []IDMap
+
+func (e sortByContainerID) Len() int           { return len(e) }
+func (e sortByContainerID) Swap(i, j int)      { e[i], e[j] = e[j], e[i] }
+func (e sortByContainerID) Less(i, j int) bool { return e[i].ContainerID < e[j].ContainerID }
+
+// IsContiguous checks if the specified mapping is contiguous and doesn't
+// have any hole.
+func IsContiguous(mappings []IDMap) bool {
+	if len(mappings) < 2 {
+		return true
+	}
+
+	var mh sortByHostID = mappings[:]
+	sort.Sort(mh)
+	for i := 1; i < len(mh); i++ {
+		if mh[i].HostID != mh[i-1].HostID+mh[i-1].Size {
+			return false
+		}
+	}
+
+	var mc sortByContainerID = mappings[:]
+	sort.Sort(mc)
+	for i := 1; i < len(mc); i++ {
+		if mc[i].ContainerID != mc[i-1].ContainerID+mc[i-1].Size {
+			return false
+		}
+	}
+	return true
+}

vendor/github.com/containers/storage/pkg/idtools/idtools_supported.go

@@ -0,0 +1,91 @@
+//go:build linux && cgo && libsubid
+
+package idtools
+
+import (
+	"errors"
+	"os/user"
+	"sync"
+	"unsafe"
+)
+
+/*
+#cgo LDFLAGS: -l subid
+#include <shadow/subid.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+struct subid_range get_range(struct subid_range *ranges, int i)
+{
+    return ranges[i];
+}
+
+#if !defined(SUBID_ABI_MAJOR) || (SUBID_ABI_MAJOR < 4)
+# define subid_init libsubid_init
+# define subid_get_uid_ranges get_subuid_ranges
+# define subid_get_gid_ranges get_subgid_ranges
+#endif
+
+*/
+import "C"
+
+var onceInit sync.Once
+
+func readSubid(username string, isUser bool) (ranges, error) {
+	var ret ranges
+	uidstr := ""
+
+	if username == "ALL" {
+		return nil, errors.New("username ALL not supported")
+	}
+
+	if u, err := user.Lookup(username); err == nil {
+		uidstr = u.Uid
+	}
+
+	onceInit.Do(func() {
+		C.subid_init(C.CString("storage"), C.stderr)
+	})
+
+	cUsername := C.CString(username)
+	defer C.free(unsafe.Pointer(cUsername))
+
+	cuidstr := C.CString(uidstr)
+	defer C.free(unsafe.Pointer(cuidstr))
+
+	var nRanges C.int
+	var cRanges *C.struct_subid_range
+	if isUser {
+		nRanges = C.subid_get_uid_ranges(cUsername, &cRanges)
+		if nRanges <= 0 {
+			nRanges = C.subid_get_uid_ranges(cuidstr, &cRanges)
+		}
+	} else {
+		nRanges = C.subid_get_gid_ranges(cUsername, &cRanges)
+		if nRanges <= 0 {
+			nRanges = C.subid_get_gid_ranges(cuidstr, &cRanges)
+		}
+	}
+	if nRanges < 0 {
+		return nil, errors.New("cannot read subids")
+	}
+	defer C.free(unsafe.Pointer(cRanges))
+
+	for i := 0; i < int(nRanges); i++ {
+		r := C.get_range(cRanges, C.int(i))
+		newRange := subIDRange{
+			Start:  int(r.start),
+			Length: int(r.count),
+		}
+		ret = append(ret, newRange)
+	}
+	return ret, nil
+}
+
+func readSubuid(username string) (ranges, error) {
+	return readSubid(username, true)
+}
+
+func readSubgid(username string) (ranges, error) {
+	return readSubid(username, false)
+}

vendor/github.com/containers/storage/pkg/idtools/idtools_unix.go

@@ -0,0 +1,214 @@
+//go:build !windows
+
+package idtools
+
+import (
+	"bytes"
+	"fmt"
+	"io"
+	"os"
+	"path/filepath"
+	"strings"
+	"sync"
+	"syscall"
+
+	"github.com/containers/storage/pkg/fileutils"
+	"github.com/containers/storage/pkg/system"
+	"github.com/moby/sys/user"
+)
+
+var (
+	entOnce   sync.Once
+	getentCmd string
+)
+
+func mkdirAs(path string, mode os.FileMode, ownerUID, ownerGID int, mkAll, chownExisting bool) error {
+	// make an array containing the original path asked for, plus (for mkAll == true)
+	// all path components leading up to the complete path that don't exist before we MkdirAll
+	// so that we can chown all of them properly at the end.  If chownExisting is false, we won't
+	// chown the full directory path if it exists
+	var paths []string
+	st, err := os.Stat(path)
+	if err != nil && os.IsNotExist(err) {
+		paths = []string{path}
+	} else if err == nil {
+		if !st.IsDir() {
+			return &os.PathError{Op: "mkdir", Path: path, Err: syscall.ENOTDIR}
+		}
+		if chownExisting {
+			// short-circuit--we were called with an existing directory and chown was requested
+			return SafeChown(path, ownerUID, ownerGID)
+		}
+		// nothing to do; directory exists and chown was NOT requested
+		return nil
+	}
+
+	if mkAll {
+		// walk back to "/" looking for directories which do not exist
+		// and add them to the paths array for chown after creation
+		dirPath := path
+		if !filepath.IsAbs(dirPath) {
+			return fmt.Errorf("path: %s should be absolute", dirPath)
+		}
+		for {
+			dirPath = filepath.Dir(dirPath)
+			if dirPath == "/" {
+				break
+			}
+			if err := fileutils.Exists(dirPath); err != nil && os.IsNotExist(err) {
+				paths = append(paths, dirPath)
+			}
+		}
+		if err := os.MkdirAll(path, mode); err != nil {
+			return err
+		}
+	} else {
+		if err := os.Mkdir(path, mode); err != nil && !os.IsExist(err) {
+			return err
+		}
+	}
+	// even if it existed, we will chown the requested path + any subpaths that
+	// didn't exist when we called MkdirAll
+	for _, pathComponent := range paths {
+		if err := SafeChown(pathComponent, ownerUID, ownerGID); err != nil {
+			return err
+		}
+	}
+	return nil
+}
+
+// CanAccess takes a valid (existing) directory and a uid, gid pair and determines
+// if that uid, gid pair has access (execute bit) to the directory
+func CanAccess(path string, pair IDPair) bool {
+	statInfo, err := system.Stat(path)
+	if err != nil {
+		return false
+	}
+	fileMode := os.FileMode(statInfo.Mode())
+	permBits := fileMode.Perm()
+	return accessible(statInfo.UID() == uint32(pair.UID),
+		statInfo.GID() == uint32(pair.GID), permBits)
+}
+
+func accessible(isOwner, isGroup bool, perms os.FileMode) bool {
+	if isOwner && (perms&0o100 == 0o100) {
+		return true
+	}
+	if isGroup && (perms&0o010 == 0o010) {
+		return true
+	}
+	if perms&0o001 == 0o001 {
+		return true
+	}
+	return false
+}
+
+// LookupUser uses traditional local system files lookup (from libcontainer/user) on a username,
+// followed by a call to `getent` for supporting host configured non-files passwd and group dbs
+func LookupUser(username string) (user.User, error) {
+	// first try a local system files lookup using existing capabilities
+	usr, err := user.LookupUser(username)
+	if err == nil {
+		return usr, nil
+	}
+	// local files lookup failed; attempt to call `getent` to query configured passwd dbs
+	usr, err = getentUser(fmt.Sprintf("%s %s", "passwd", username))
+	if err != nil {
+		return user.User{}, err
+	}
+	return usr, nil
+}
+
+// LookupUID uses traditional local system files lookup (from libcontainer/user) on a uid,
+// followed by a call to `getent` for supporting host configured non-files passwd and group dbs
+func LookupUID(uid int) (user.User, error) {
+	// first try a local system files lookup using existing capabilities
+	usr, err := user.LookupUid(uid)
+	if err == nil {
+		return usr, nil
+	}
+	// local files lookup failed; attempt to call `getent` to query configured passwd dbs
+	return getentUser(fmt.Sprintf("%s %d", "passwd", uid))
+}
+
+func getentUser(args string) (user.User, error) {
+	reader, err := callGetent(args)
+	if err != nil {
+		return user.User{}, err
+	}
+	users, err := user.ParsePasswd(reader)
+	if err != nil {
+		return user.User{}, err
+	}
+	if len(users) == 0 {
+		return user.User{}, fmt.Errorf("getent failed to find passwd entry for %q", strings.Split(args, " ")[1])
+	}
+	return users[0], nil
+}
+
+// LookupGroup uses traditional local system files lookup (from libcontainer/user) on a group name,
+// followed by a call to `getent` for supporting host configured non-files passwd and group dbs
+func LookupGroup(groupname string) (user.Group, error) {
+	// first try a local system files lookup using existing capabilities
+	group, err := user.LookupGroup(groupname)
+	if err == nil {
+		return group, nil
+	}
+	// local files lookup failed; attempt to call `getent` to query configured group dbs
+	return getentGroup(fmt.Sprintf("%s %s", "group", groupname))
+}
+
+// LookupGID uses traditional local system files lookup (from libcontainer/user) on a group ID,
+// followed by a call to `getent` for supporting host configured non-files passwd and group dbs
+func LookupGID(gid int) (user.Group, error) {
+	// first try a local system files lookup using existing capabilities
+	group, err := user.LookupGid(gid)
+	if err == nil {
+		return group, nil
+	}
+	// local files lookup failed; attempt to call `getent` to query configured group dbs
+	return getentGroup(fmt.Sprintf("%s %d", "group", gid))
+}
+
+func getentGroup(args string) (user.Group, error) {
+	reader, err := callGetent(args)
+	if err != nil {
+		return user.Group{}, err
+	}
+	groups, err := user.ParseGroup(reader)
+	if err != nil {
+		return user.Group{}, err
+	}
+	if len(groups) == 0 {
+		return user.Group{}, fmt.Errorf("getent failed to find groups entry for %q", strings.Split(args, " ")[1])
+	}
+	return groups[0], nil
+}
+
+func callGetent(args string) (io.Reader, error) {
+	entOnce.Do(func() { getentCmd, _ = resolveBinary("getent") })
+	// if no `getent` command on host, can't do anything else
+	if getentCmd == "" {
+		return nil, fmt.Errorf("")
+	}
+	out, err := execCmd(getentCmd, args)
+	if err != nil {
+		exitCode, errC := system.GetExitCode(err)
+		if errC != nil {
+			return nil, err
+		}
+		switch exitCode {
+		case 1:
+			return nil, fmt.Errorf("getent reported invalid parameters/database unknown")
+		case 2:
+			terms := strings.Split(args, " ")
+			return nil, fmt.Errorf("getent unable to find entry %q in %s database", terms[1], terms[0])
+		case 3:
+			return nil, fmt.Errorf("getent database doesn't support enumeration")
+		default:
+			return nil, err
+		}
+
+	}
+	return bytes.NewReader(out), nil
+}

vendor/github.com/containers/storage/pkg/idtools/idtools_unsupported.go

@@ -0,0 +1,11 @@
+//go:build !linux || !libsubid || !cgo
+
+package idtools
+
+func readSubuid(username string) (ranges, error) {
+	return parseSubidFile(subuidFileName, username)
+}
+
+func readSubgid(username string) (ranges, error) {
+	return parseSubidFile(subgidFileName, username)
+}

vendor/github.com/containers/storage/pkg/idtools/idtools_windows.go

@@ -0,0 +1,23 @@
+//go:build windows
+
+package idtools
+
+import (
+	"os"
+)
+
+// Platforms such as Windows do not support the UID/GID concept. So make this
+// just a wrapper around system.MkdirAll.
+func mkdirAs(path string, mode os.FileMode, ownerUID, ownerGID int, mkAll, chownExisting bool) error {
+	if err := os.MkdirAll(path, mode); err != nil {
+		return err
+	}
+	return nil
+}
+
+// CanAccess takes a valid (existing) directory and a uid, gid pair and determines
+// if that uid, gid pair has access (execute bit) to the directory
+// Windows does not require/support this function, so always return true
+func CanAccess(path string, pair IDPair) bool {
+	return true
+}

vendor/github.com/containers/storage/pkg/idtools/parser.go

@@ -0,0 +1,59 @@
+package idtools
+
+import (
+	"fmt"
+	"math"
+	"math/bits"
+	"strconv"
+	"strings"
+)
+
+func parseTriple(spec []string) (container, host, size uint32, err error) {
+	cid, err := strconv.ParseUint(spec[0], 10, 32)
+	if err != nil {
+		return 0, 0, 0, fmt.Errorf("parsing id map value %q: %w", spec[0], err)
+	}
+	hid, err := strconv.ParseUint(spec[1], 10, 32)
+	if err != nil {
+		return 0, 0, 0, fmt.Errorf("parsing id map value %q: %w", spec[1], err)
+	}
+	sz, err := strconv.ParseUint(spec[2], 10, 32)
+	if err != nil {
+		return 0, 0, 0, fmt.Errorf("parsing id map value %q: %w", spec[2], err)
+	}
+	return uint32(cid), uint32(hid), uint32(sz), nil
+}
+
+// ParseIDMap parses idmap triples from string.
+func ParseIDMap(mapSpec []string, mapSetting string) (idmap []IDMap, err error) {
+	stdErr := fmt.Errorf("initializing ID mappings: %s setting is malformed expected [\"uint32:uint32:uint32\"]: %q", mapSetting, mapSpec)
+	for _, idMapSpec := range mapSpec {
+		if idMapSpec == "" {
+			continue
+		}
+		idSpec := strings.Split(idMapSpec, ":")
+		if len(idSpec)%3 != 0 {
+			return nil, stdErr
+		}
+		for i := range idSpec {
+			if i%3 != 0 {
+				continue
+			}
+			cid, hid, size, err := parseTriple(idSpec[i : i+3])
+			if err != nil {
+				return nil, stdErr
+			}
+			// Avoid possible integer overflow on 32bit builds
+			if bits.UintSize == 32 && (cid > math.MaxInt32 || hid > math.MaxInt32 || size > math.MaxInt32) {
+				return nil, stdErr
+			}
+			mapping := IDMap{
+				ContainerID: int(cid),
+				HostID:      int(hid),
+				Size:        int(size),
+			}
+			idmap = append(idmap, mapping)
+		}
+	}
+	return idmap, nil
+}

vendor/github.com/containers/storage/pkg/idtools/usergroupadd_linux.go

@@ -0,0 +1,164 @@
+package idtools
+
+import (
+	"fmt"
+	"sort"
+	"strconv"
+	"strings"
+	"sync"
+
+	"github.com/containers/storage/pkg/regexp"
+)
+
+// add a user and/or group to Linux /etc/passwd, /etc/group using standard
+// Linux distribution commands:
+// adduser --system --shell /bin/false --disabled-login --disabled-password --no-create-home --group <username>
+// useradd -r -s /bin/false <username>
+
+var (
+	once        sync.Once
+	userCommand string
+
+	cmdTemplates = map[string]string{
+		"adduser": "--system --shell /bin/false --no-create-home --disabled-login --disabled-password --group %s",
+		"useradd": "-r -s /bin/false %s",
+		"usermod": "-%s %d-%d %s",
+	}
+
+	idOutRegexp = regexp.Delayed(`uid=([0-9]+).*gid=([0-9]+)`)
+	// default length for a UID/GID subordinate range
+	defaultRangeLen   = 65536
+	defaultRangeStart = 100000
+	userMod           = "usermod"
+)
+
+// AddNamespaceRangesUser takes a username and uses the standard system
+// utility to create a system user/group pair used to hold the
+// /etc/sub{uid,gid} ranges which will be used for user namespace
+// mapping ranges in containers.
+func AddNamespaceRangesUser(name string) (int, int, error) {
+	if err := addUser(name); err != nil {
+		return -1, -1, fmt.Errorf("adding user %q: %w", name, err)
+	}
+
+	// Query the system for the created uid and gid pair
+	out, err := execCmd("id", name)
+	if err != nil {
+		return -1, -1, fmt.Errorf("trying to find uid/gid for new user %q: %w", name, err)
+	}
+	matches := idOutRegexp.FindStringSubmatch(strings.TrimSpace(string(out)))
+	if len(matches) != 3 {
+		return -1, -1, fmt.Errorf("can't find uid, gid from `id` output: %q", string(out))
+	}
+	uid, err := strconv.Atoi(matches[1])
+	if err != nil {
+		return -1, -1, fmt.Errorf("can't convert found uid (%s) to int: %w", matches[1], err)
+	}
+	gid, err := strconv.Atoi(matches[2])
+	if err != nil {
+		return -1, -1, fmt.Errorf("can't convert found gid (%s) to int: %w", matches[2], err)
+	}
+
+	// Now we need to create the subuid/subgid ranges for our new user/group (system users
+	// do not get auto-created ranges in subuid/subgid)
+
+	if err := createSubordinateRanges(name); err != nil {
+		return -1, -1, fmt.Errorf("couldn't create subordinate ID ranges: %w", err)
+	}
+	return uid, gid, nil
+}
+
+func addUser(userName string) error {
+	once.Do(func() {
+		// set up which commands are used for adding users/groups dependent on distro
+		if _, err := resolveBinary("adduser"); err == nil {
+			userCommand = "adduser"
+		} else if _, err := resolveBinary("useradd"); err == nil {
+			userCommand = "useradd"
+		}
+	})
+	if userCommand == "" {
+		return fmt.Errorf("cannot add user; no useradd/adduser binary found")
+	}
+	args := fmt.Sprintf(cmdTemplates[userCommand], userName)
+	out, err := execCmd(userCommand, args)
+	if err != nil {
+		return fmt.Errorf("failed to add user with error: %w; output: %q", err, string(out))
+	}
+	return nil
+}
+
+func createSubordinateRanges(name string) error {
+	// first, we should verify that ranges weren't automatically created
+	// by the distro tooling
+	ranges, err := readSubuid(name)
+	if err != nil {
+		return fmt.Errorf("while looking for subuid ranges for user %q: %w", name, err)
+	}
+	if len(ranges) == 0 {
+		// no UID ranges; let's create one
+		startID, err := findNextUIDRange()
+		if err != nil {
+			return fmt.Errorf("can't find available subuid range: %w", err)
+		}
+		out, err := execCmd(userMod, fmt.Sprintf(cmdTemplates[userMod], "v", startID, startID+defaultRangeLen-1, name))
+		if err != nil {
+			return fmt.Errorf("unable to add subuid range to user: %q; output: %s, err: %w", name, out, err)
+		}
+	}
+
+	ranges, err = readSubgid(name)
+	if err != nil {
+		return fmt.Errorf("while looking for subgid ranges for user %q: %w", name, err)
+	}
+	if len(ranges) == 0 {
+		// no GID ranges; let's create one
+		startID, err := findNextGIDRange()
+		if err != nil {
+			return fmt.Errorf("can't find available subgid range: %w", err)
+		}
+		out, err := execCmd(userMod, fmt.Sprintf(cmdTemplates[userMod], "w", startID, startID+defaultRangeLen-1, name))
+		if err != nil {
+			return fmt.Errorf("unable to add subgid range to user: %q; output: %s, err: %w", name, out, err)
+		}
+	}
+	return nil
+}
+
+func findNextUIDRange() (int, error) {
+	ranges, err := readSubuid("ALL")
+	if err != nil {
+		return -1, fmt.Errorf("couldn't parse all ranges in /etc/subuid file: %w", err)
+	}
+	sort.Sort(ranges)
+	return findNextRangeStart(ranges)
+}
+
+func findNextGIDRange() (int, error) {
+	ranges, err := readSubgid("ALL")
+	if err != nil {
+		return -1, fmt.Errorf("couldn't parse all ranges in /etc/subgid file: %w", err)
+	}
+	sort.Sort(ranges)
+	return findNextRangeStart(ranges)
+}
+
+func findNextRangeStart(rangeList ranges) (int, error) {
+	startID := defaultRangeStart
+	for _, arange := range rangeList {
+		if wouldOverlap(arange, startID) {
+			startID = arange.Start + arange.Length
+		}
+	}
+	return startID, nil
+}
+
+func wouldOverlap(arange subIDRange, ID int) bool {
+	low := ID
+	high := ID + defaultRangeLen
+	if (low >= arange.Start && low <= arange.Start+arange.Length) ||
+		(high <= arange.Start+arange.Length && high >= arange.Start) {
+		return true
+	}
+	return false
+}

vendor/github.com/containers/storage/pkg/idtools/usergroupadd_unsupported.go

@@ -0,0 +1,12 @@
+//go:build !linux
+
+package idtools
+
+import "fmt"
+
+// AddNamespaceRangesUser takes a name and finds an unused uid, gid pair
+// and calls the appropriate helper function to add the group and then
+// the user to the group in /etc/group and /etc/passwd respectively.
+func AddNamespaceRangesUser(name string) (int, int, error) {
+	return -1, -1, fmt.Errorf("no support for adding users or groups on this OS")
+}

vendor/github.com/containers/storage/pkg/idtools/utils_unix.go

@@ -0,0 +1,32 @@
+//go:build !windows
+
+package idtools
+
+import (
+	"fmt"
+	"os/exec"
+	"path/filepath"
+	"strings"
+)
+
+func resolveBinary(binname string) (string, error) {
+	binaryPath, err := exec.LookPath(binname)
+	if err != nil {
+		return "", err
+	}
+	resolvedPath, err := filepath.EvalSymlinks(binaryPath)
+	if err != nil {
+		return "", err
+	}
+	// only return no error if the final resolved binary basename
+	// matches what was searched for
+	if filepath.Base(resolvedPath) == binname {
+		return resolvedPath, nil
+	}
+	return "", fmt.Errorf("binary %q does not resolve to a binary of that name in $PATH (%q)", binname, resolvedPath)
+}
+
+func execCmd(cmd, args string) ([]byte, error) {
+	execCmd := exec.Command(cmd, strings.Split(args, " ")...)
+	return execCmd.CombinedOutput()
+}

vendor/github.com/containers/storage/pkg/mount/flags.go

@@ -0,0 +1,149 @@
+package mount
+
+import (
+	"fmt"
+	"strings"
+)
+
+var flags = map[string]struct {
+	clear bool
+	flag  int
+}{
+	"defaults":      {false, 0},
+	"ro":            {false, RDONLY},
+	"rw":            {true, RDONLY},
+	"suid":          {true, NOSUID},
+	"nosuid":        {false, NOSUID},
+	"dev":           {true, NODEV},
+	"nodev":         {false, NODEV},
+	"exec":          {true, NOEXEC},
+	"noexec":        {false, NOEXEC},
+	"sync":          {false, SYNCHRONOUS},
+	"async":         {true, SYNCHRONOUS},
+	"dirsync":       {false, DIRSYNC},
+	"remount":       {false, REMOUNT},
+	"mand":          {false, MANDLOCK},
+	"nomand":        {true, MANDLOCK},
+	"atime":         {true, NOATIME},
+	"noatime":       {false, NOATIME},
+	"diratime":      {true, NODIRATIME},
+	"nodiratime":    {false, NODIRATIME},
+	"bind":          {false, BIND},
+	"rbind":         {false, RBIND},
+	"unbindable":    {false, UNBINDABLE},
+	"runbindable":   {false, RUNBINDABLE},
+	"private":       {false, PRIVATE},
+	"rprivate":      {false, RPRIVATE},
+	"shared":        {false, SHARED},
+	"rshared":       {false, RSHARED},
+	"slave":         {false, SLAVE},
+	"rslave":        {false, RSLAVE},
+	"relatime":      {false, RELATIME},
+	"norelatime":    {true, RELATIME},
+	"strictatime":   {false, STRICTATIME},
+	"nostrictatime": {true, STRICTATIME},
+}
+
+var validFlags = map[string]bool{
+	"":          true,
+	"size":      true,
+	"mode":      true,
+	"uid":       true,
+	"gid":       true,
+	"nr_inodes": true,
+	"nr_blocks": true,
+	"mpol":      true,
+}
+
+var propagationFlags = map[string]bool{
+	"bind":        true,
+	"rbind":       true,
+	"unbindable":  true,
+	"runbindable": true,
+	"private":     true,
+	"rprivate":    true,
+	"shared":      true,
+	"rshared":     true,
+	"slave":       true,
+	"rslave":      true,
+}
+
+// MergeTmpfsOptions merge mount options to make sure there is no duplicate.
+func MergeTmpfsOptions(options []string) ([]string, error) {
+	// We use collisions maps to remove duplicates.
+	// For flag, the key is the flag value (the key for propagation flag is -1)
+	// For data=value, the key is the data
+	flagCollisions := map[int]bool{}
+	dataCollisions := map[string]bool{}
+
+	var newOptions []string
+	// We process in reverse order
+	for i := len(options) - 1; i >= 0; i-- {
+		option := options[i]
+		if option == "defaults" {
+			continue
+		}
+		if f, ok := flags[option]; ok && f.flag != 0 {
+			// There is only one propagation mode
+			key := f.flag
+			if propagationFlags[option] {
+				key = -1
+			}
+			// Check to see if there is collision for flag
+			if !flagCollisions[key] {
+				// We prepend the option and add to collision map
+				newOptions = append([]string{option}, newOptions...)
+				flagCollisions[key] = true
+			}
+			continue
+		}
+		opt, _, ok := strings.Cut(option, "=")
+		if !ok || !validFlags[opt] {
+			return nil, fmt.Errorf("invalid tmpfs option %q", opt)
+		}
+		if !dataCollisions[opt] {
+			// We prepend the option and add to collision map
+			newOptions = append([]string{option}, newOptions...)
+			dataCollisions[opt] = true
+		}
+	}
+
+	return newOptions, nil
+}
+
+// ParseOptions parses fstab type mount options into mount() flags
+// and device specific data
+func ParseOptions(options string) (int, string) {
+	var (
+		flag int
+		data []string
+	)
+
+	for _, o := range strings.Split(options, ",") {
+		// If the option does not exist in the flags table or the flag
+		// is not supported on the platform,
+		// then it is a data value for a specific fs type
+		if f, exists := flags[o]; exists && f.flag != 0 {
+			if f.clear {
+				flag &= ^f.flag
+			} else {
+				flag |= f.flag
+			}
+		} else {
+			data = append(data, o)
+		}
+	}
+	return flag, strings.Join(data, ",")
+}
+
+// ParseTmpfsOptions parse fstab type mount options into flags and data
+func ParseTmpfsOptions(options string) (int, string, error) {
+	flags, data := ParseOptions(options)
+	for _, o := range strings.Split(data, ",") {
+		opt, _, _ := strings.Cut(o, "=")
+		if !validFlags[opt] {
+			return 0, "", fmt.Errorf("invalid tmpfs option %q", opt)
+		}
+	}
+	return flags, data, nil
+}

vendor/github.com/containers/storage/pkg/mount/flags_freebsd.go

@@ -0,0 +1,48 @@
+package mount
+
+import (
+	"golang.org/x/sys/unix"
+)
+
+const (
+	// RDONLY will mount the file system read-only.
+	RDONLY = unix.MNT_RDONLY
+
+	// NOSUID will not allow set-user-identifier or set-group-identifier bits to
+	// take effect.
+	NOSUID = unix.MNT_NOSUID
+
+	// NOEXEC will not allow execution of any binaries on the mounted file system.
+	NOEXEC = unix.MNT_NOEXEC
+
+	// SYNCHRONOUS will allow I/O to the file system to be done synchronously.
+	SYNCHRONOUS = unix.MNT_SYNCHRONOUS
+
+	// REMOUNT will attempt to remount an already-mounted file system. This is
+	// commonly used to change the mount flags for a file system, especially to
+	// make a readonly file system writeable. It does not change device or mount
+	// point.
+	REMOUNT = unix.MNT_UPDATE
+
+	// NOATIME will not update the file access time when reading from a file.
+	NOATIME = unix.MNT_NOATIME
+
+	mntDetach = unix.MNT_FORCE
+
+	NODIRATIME  = 0
+	NODEV       = 0
+	DIRSYNC     = 0
+	MANDLOCK    = 0
+	BIND        = 0
+	RBIND       = 0
+	UNBINDABLE  = 0
+	RUNBINDABLE = 0
+	PRIVATE     = 0
+	RPRIVATE    = 0
+	SLAVE       = 0
+	RSLAVE      = 0
+	SHARED      = 0
+	RSHARED     = 0
+	RELATIME    = 0
+	STRICTATIME = 0
+)

vendor/github.com/containers/storage/pkg/mount/flags_linux.go

@@ -0,0 +1,87 @@
+package mount
+
+import (
+	"golang.org/x/sys/unix"
+)
+
+const (
+	// RDONLY will mount the file system read-only.
+	RDONLY = unix.MS_RDONLY
+
+	// NOSUID will not allow set-user-identifier or set-group-identifier bits to
+	// take effect.
+	NOSUID = unix.MS_NOSUID
+
+	// NODEV will not interpret character or block special devices on the file
+	// system.
+	NODEV = unix.MS_NODEV
+
+	// NOEXEC will not allow execution of any binaries on the mounted file system.
+	NOEXEC = unix.MS_NOEXEC
+
+	// SYNCHRONOUS will allow I/O to the file system to be done synchronously.
+	SYNCHRONOUS = unix.MS_SYNCHRONOUS
+
+	// DIRSYNC will force all directory updates within the file system to be done
+	// synchronously. This affects the following system calls: create, link,
+	// unlink, symlink, mkdir, rmdir, mknod and rename.
+	DIRSYNC = unix.MS_DIRSYNC
+
+	// REMOUNT will attempt to remount an already-mounted file system. This is
+	// commonly used to change the mount flags for a file system, especially to
+	// make a readonly file system writeable. It does not change device or mount
+	// point.
+	REMOUNT = unix.MS_REMOUNT
+
+	// MANDLOCK will force mandatory locks on a filesystem.
+	MANDLOCK = unix.MS_MANDLOCK
+
+	// NOATIME will not update the file access time when reading from a file.
+	NOATIME = unix.MS_NOATIME
+
+	// NODIRATIME will not update the directory access time.
+	NODIRATIME = unix.MS_NODIRATIME
+
+	// BIND remounts a subtree somewhere else.
+	BIND = unix.MS_BIND
+
+	// RBIND remounts a subtree and all possible submounts somewhere else.
+	RBIND = unix.MS_BIND | unix.MS_REC
+
+	// UNBINDABLE creates a mount which cannot be cloned through a bind operation.
+	UNBINDABLE = unix.MS_UNBINDABLE
+
+	// RUNBINDABLE marks the entire mount tree as UNBINDABLE.
+	RUNBINDABLE = unix.MS_UNBINDABLE | unix.MS_REC
+
+	// PRIVATE creates a mount which carries no propagation abilities.
+	PRIVATE = unix.MS_PRIVATE
+
+	// RPRIVATE marks the entire mount tree as PRIVATE.
+	RPRIVATE = unix.MS_PRIVATE | unix.MS_REC
+
+	// SLAVE creates a mount which receives propagation from its master, but not
+	// vice versa.
+	SLAVE = unix.MS_SLAVE
+
+	// RSLAVE marks the entire mount tree as SLAVE.
+	RSLAVE = unix.MS_SLAVE | unix.MS_REC
+
+	// SHARED creates a mount which provides the ability to create mirrors of
+	// that mount such that mounts and unmounts within any of the mirrors
+	// propagate to the other mirrors.
+	SHARED = unix.MS_SHARED
+
+	// RSHARED marks the entire mount tree as SHARED.
+	RSHARED = unix.MS_SHARED | unix.MS_REC
+
+	// RELATIME updates inode access times relative to modify or change time.
+	RELATIME = unix.MS_RELATIME
+
+	// STRICTATIME allows to explicitly request full atime updates.  This makes
+	// it possible for the kernel to default to relatime or noatime but still
+	// allow userspace to override it.
+	STRICTATIME = unix.MS_STRICTATIME
+
+	mntDetach = unix.MNT_DETACH
+)

vendor/github.com/containers/storage/pkg/mount/flags_unsupported.go

@@ -0,0 +1,31 @@
+//go:build !linux && !freebsd
+
+package mount
+
+// These flags are unsupported.
+const (
+	BIND        = 0
+	DIRSYNC     = 0
+	MANDLOCK    = 0
+	NOATIME     = 0
+	NODEV       = 0
+	NODIRATIME  = 0
+	NOEXEC      = 0
+	NOSUID      = 0
+	UNBINDABLE  = 0
+	RUNBINDABLE = 0
+	PRIVATE     = 0
+	RPRIVATE    = 0
+	SHARED      = 0
+	RSHARED     = 0
+	SLAVE       = 0
+	RSLAVE      = 0
+	RBIND       = 0
+	RELATIME    = 0
+	RELATIVE    = 0
+	REMOUNT     = 0
+	STRICTATIME = 0
+	SYNCHRONOUS = 0
+	RDONLY      = 0
+	mntDetach   = 0
+)

vendor/github.com/containers/storage/pkg/mount/mount.go

@@ -0,0 +1,110 @@
+package mount
+
+import (
+	"sort"
+	"strconv"
+	"strings"
+)
+
+// mountError holds an error from a mount or unmount operation
+type mountError struct {
+	op             string
+	source, target string
+	flags          uintptr
+	data           string
+	err            error
+}
+
+// Error returns a string representation of mountError
+func (e *mountError) Error() string {
+	out := e.op + " "
+
+	if e.source != "" {
+		out += e.source + ":" + e.target
+	} else {
+		out += e.target
+	}
+
+	if e.flags != uintptr(0) {
+		out += ", flags: 0x" + strconv.FormatUint(uint64(e.flags), 16)
+	}
+	if e.data != "" {
+		out += ", data: " + e.data
+	}
+
+	out += ": " + e.err.Error()
+	return out
+}
+
+// Cause returns the underlying cause of the error
+func (e *mountError) Cause() error {
+	return e.err
+}
+
+// Unwrap returns the underlying cause of the error
+func (e *mountError) Unwrap() error {
+	return e.err
+}
+
+// Mount will mount filesystem according to the specified configuration, on the
+// condition that the target path is *not* already mounted. Options must be
+// specified like the mount or fstab unix commands: "opt1=val1,opt2=val2". See
+// flags.go for supported option flags.
+func Mount(device, target, mType, options string) error {
+	flag, data := ParseOptions(options)
+	if flag&REMOUNT != REMOUNT {
+		if mounted, err := Mounted(target); err != nil || mounted {
+			return err
+		}
+	}
+	return mount(device, target, mType, uintptr(flag), data)
+}
+
+// ForceMount will mount a filesystem according to the specified configuration,
+// *regardless* if the target path is not already mounted. Options must be
+// specified like the mount or fstab unix commands: "opt1=val1,opt2=val2". See
+// flags.go for supported option flags.
+func ForceMount(device, target, mType, options string) error {
+	flag, data := ParseOptions(options)
+	return mount(device, target, mType, uintptr(flag), data)
+}
+
+// Unmount lazily unmounts a filesystem on supported platforms, otherwise
+// does a normal unmount.
+func Unmount(target string) error {
+	return unmount(target, mntDetach)
+}
+
+// RecursiveUnmount unmounts the target and all mounts underneath, starting with
+// the deepest mount first.
+func RecursiveUnmount(target string) error {
+	mounts, err := GetMounts()
+	if err != nil {
+		return err
+	}
+
+	// Make the deepest mount be first
+	sort.Slice(mounts, func(i, j int) bool {
+		return len(mounts[i].Mountpoint) > len(mounts[j].Mountpoint)
+	})
+
+	for i, m := range mounts {
+		if !strings.HasPrefix(m.Mountpoint, target) {
+			continue
+		}
+		if err := Unmount(m.Mountpoint); err != nil && i == len(mounts)-1 {
+			return err
+			// Ignore errors for submounts and continue trying to unmount others
+			// The final unmount should fail if there are any submounts remaining
+		}
+	}
+	return nil
+}
+
+// ForceUnmount lazily unmounts a filesystem on supported platforms,
+// otherwise does a normal unmount.
+//
+// Deprecated: please use Unmount instead, it is identical.
+func ForceUnmount(target string) error {
+	return unmount(target, mntDetach)
+}

vendor/github.com/containers/storage/pkg/mount/mounter_freebsd.go

@@ -0,0 +1,63 @@
+//go:build freebsd && cgo
+
+package mount
+
+/*
+#include <errno.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/_iovec.h>
+#include <sys/mount.h>
+#include <sys/param.h>
+*/
+import "C"
+
+import (
+	"fmt"
+	"strings"
+	"unsafe"
+)
+
+func allocateIOVecs(options []string) []C.struct_iovec {
+	out := make([]C.struct_iovec, len(options))
+	for i, option := range options {
+		out[i].iov_base = unsafe.Pointer(C.CString(option))
+		out[i].iov_len = C.size_t(len(option) + 1)
+	}
+	return out
+}
+
+func mount(device, target, mType string, flag uintptr, data string) error {
+	isNullFS := false
+
+	options := []string{"fspath", target}
+
+	if data != "" {
+		xs := strings.Split(data, ",")
+		for _, x := range xs {
+			if x == "bind" {
+				isNullFS = true
+				continue
+			}
+			name, val, _ := strings.Cut(x, "=")
+			options = append(options, name)
+			options = append(options, val)
+		}
+	}
+
+	if isNullFS {
+		options = append(options, "fstype", "nullfs", "target", device)
+	} else {
+		options = append(options, "fstype", mType, "from", device)
+	}
+	rawOptions := allocateIOVecs(options)
+	for _, rawOption := range rawOptions {
+		defer C.free(rawOption.iov_base)
+	}
+
+	if errno := C.nmount(&rawOptions[0], C.uint(len(options)), C.int(flag)); errno != 0 {
+		reason := C.GoString(C.strerror(*C.__error()))
+		return fmt.Errorf("failed to call nmount: %s", reason)
+	}
+	return nil
+}

vendor/github.com/containers/storage/pkg/mount/mounter_linux.go

@@ -0,0 +1,74 @@
+package mount
+
+import (
+	"golang.org/x/sys/unix"
+)
+
+const (
+	// ptypes is the set propagation types.
+	ptypes = unix.MS_SHARED | unix.MS_PRIVATE | unix.MS_SLAVE | unix.MS_UNBINDABLE
+
+	// pflags is the full set valid flags for a change propagation call.
+	pflags = ptypes | unix.MS_REC | unix.MS_SILENT
+
+	// broflags is the combination of bind and read only
+	broflags = unix.MS_BIND | unix.MS_RDONLY
+
+	none = "none"
+)
+
+// isremount returns true if either device name or flags identify a remount request, false otherwise.
+func isremount(device string, flags uintptr) bool {
+	switch {
+	// We treat device "" and "none" as a remount request to provide compatibility with
+	// requests that don't explicitly set MS_REMOUNT such as those manipulating bind mounts.
+	case flags&unix.MS_REMOUNT != 0, device == "", device == none:
+		return true
+	default:
+		return false
+	}
+}
+
+func mount(device, target, mType string, flags uintptr, data string) error {
+	oflags := flags &^ ptypes
+	if !isremount(device, flags) || data != "" {
+		// Initial call applying all non-propagation flags for mount
+		// or remount with changed data
+		if err := unix.Mount(device, target, mType, oflags, data); err != nil {
+			return &mountError{
+				op:     "mount",
+				source: device,
+				target: target,
+				flags:  oflags,
+				data:   data,
+				err:    err,
+			}
+		}
+	}
+
+	if flags&ptypes != 0 {
+		// Change the propagation type.
+		if err := unix.Mount("", target, "", flags&pflags, ""); err != nil {
+			return &mountError{
+				op:     "remount",
+				target: target,
+				flags:  flags & pflags,
+				err:    err,
+			}
+		}
+	}
+
+	if oflags&broflags == broflags {
+		// Remount the bind to apply read only.
+		if err := unix.Mount("", target, "", oflags|unix.MS_REMOUNT, ""); err != nil {
+			return &mountError{
+				op:     "remount-ro",
+				target: target,
+				flags:  oflags | unix.MS_REMOUNT,
+				err:    err,
+			}
+		}
+	}
+
+	return nil
+}

vendor/github.com/containers/storage/pkg/mount/mounter_unsupported.go

@@ -0,0 +1,7 @@
+//go:build !linux && !(freebsd && cgo)
+
+package mount
+
+func mount(device, target, mType string, flag uintptr, data string) error {
+	panic("Not implemented")
+}

vendor/github.com/containers/storage/pkg/mount/mountinfo.go

@@ -0,0 +1,13 @@
+package mount
+
+import (
+	"github.com/moby/sys/mountinfo"
+)
+
+type Info = mountinfo.Info
+
+var Mounted = mountinfo.Mounted
+
+func GetMounts() ([]*Info, error) {
+	return mountinfo.GetMounts(nil)
+}

vendor/github.com/containers/storage/pkg/mount/mountinfo_linux.go

@@ -0,0 +1,18 @@
+package mount
+
+import (
+	"fmt"
+	"os"
+
+	"github.com/moby/sys/mountinfo"
+)
+
+func PidMountInfo(pid int) ([]*Info, error) {
+	f, err := os.Open(fmt.Sprintf("/proc/%d/mountinfo", pid))
+	if err != nil {
+		return nil, err
+	}
+	defer f.Close()
+
+	return mountinfo.GetMountsFromReader(f, nil)
+}

vendor/github.com/containers/storage/pkg/mount/sharedsubtree_linux.go

@@ -0,0 +1,64 @@
+package mount
+
+// MakeShared ensures a mounted filesystem has the SHARED mount option enabled.
+// See the supported options in flags.go for further reference.
+func MakeShared(mountPoint string) error {
+	return ensureMountedAs(mountPoint, SHARED)
+}
+
+// MakeRShared ensures a mounted filesystem has the RSHARED mount option enabled.
+// See the supported options in flags.go for further reference.
+func MakeRShared(mountPoint string) error {
+	return ensureMountedAs(mountPoint, RSHARED)
+}
+
+// MakePrivate ensures a mounted filesystem has the PRIVATE mount option enabled.
+// See the supported options in flags.go for further reference.
+func MakePrivate(mountPoint string) error {
+	return ensureMountedAs(mountPoint, PRIVATE)
+}
+
+// MakeRPrivate ensures a mounted filesystem has the RPRIVATE mount option
+// enabled. See the supported options in flags.go for further reference.
+func MakeRPrivate(mountPoint string) error {
+	return ensureMountedAs(mountPoint, RPRIVATE)
+}
+
+// MakeSlave ensures a mounted filesystem has the SLAVE mount option enabled.
+// See the supported options in flags.go for further reference.
+func MakeSlave(mountPoint string) error {
+	return ensureMountedAs(mountPoint, SLAVE)
+}
+
+// MakeRSlave ensures a mounted filesystem has the RSLAVE mount option enabled.
+// See the supported options in flags.go for further reference.
+func MakeRSlave(mountPoint string) error {
+	return ensureMountedAs(mountPoint, RSLAVE)
+}
+
+// MakeUnbindable ensures a mounted filesystem has the UNBINDABLE mount option
+// enabled. See the supported options in flags.go for further reference.
+func MakeUnbindable(mountPoint string) error {
+	return ensureMountedAs(mountPoint, UNBINDABLE)
+}
+
+// MakeRUnbindable ensures a mounted filesystem has the RUNBINDABLE mount
+// option enabled. See the supported options in flags.go for further reference.
+func MakeRUnbindable(mountPoint string) error {
+	return ensureMountedAs(mountPoint, RUNBINDABLE)
+}
+
+func ensureMountedAs(mnt string, flags int) error {
+	mounted, err := Mounted(mnt)
+	if err != nil {
+		return err
+	}
+
+	if !mounted {
+		if err := mount(mnt, mnt, "none", uintptr(BIND), ""); err != nil {
+			return err
+		}
+	}
+
+	return mount("", mnt, "none", uintptr(flags), "")
+}

vendor/github.com/containers/storage/pkg/mount/unmount_unix.go

@@ -0,0 +1,34 @@
+//go:build !windows
+
+package mount
+
+import (
+	"time"
+
+	"golang.org/x/sys/unix"
+)
+
+func unmount(target string, flags int) error {
+	var err error
+	for range 50 {
+		err = unix.Unmount(target, flags)
+		switch err {
+		case unix.EBUSY:
+			time.Sleep(50 * time.Millisecond)
+			continue
+		case unix.EINVAL, nil:
+			// Ignore "not mounted" error here. Note the same error
+			// can be returned if flags are invalid, so this code
+			// assumes that the flags value is always correct.
+			return nil
+		}
+		break
+	}
+
+	return &mountError{
+		op:     "umount",
+		target: target,
+		flags:  uintptr(flags),
+		err:    err,
+	}
+}

vendor/github.com/containers/storage/pkg/mount/unmount_unsupported.go

@@ -0,0 +1,7 @@
+//go:build windows
+
+package mount
+
+func unmount(target string, flag int) error {
+	panic("Not implemented")
+}

vendor/github.com/containers/storage/pkg/reexec/README.md

@@ -0,0 +1,5 @@
+# reexec
+
+The `reexec` package facilitates the busybox style reexec of the docker binary that we require because 
+of the forking limitations of using Go.  Handlers can be registered with a name and the argv 0 of 
+the exec of the binary will be used to find and execute custom init paths.

vendor/github.com/containers/storage/pkg/reexec/command_freebsd.go

@@ -0,0 +1,37 @@
+//go:build freebsd
+
+package reexec
+
+import (
+	"context"
+	"os"
+	"os/exec"
+
+	"golang.org/x/sys/unix"
+)
+
+// Self returns the path to the current process's binary.
+// Uses sysctl.
+func Self() string {
+	path, err := unix.SysctlArgs("kern.proc.pathname", -1)
+	if err == nil {
+		return path
+	}
+	return os.Args[0]
+}
+
+// Command returns *exec.Cmd which has Path as current binary.
+// For example if current binary is "docker" at "/usr/bin/", then cmd.Path will
+// be set to "/usr/bin/docker".
+func Command(args ...string) *exec.Cmd {
+	cmd := exec.Command(Self())
+	cmd.Args = args
+	return cmd
+}
+
+// CommandContext returns *exec.Cmd which has Path as current binary.
+func CommandContext(ctx context.Context, args ...string) *exec.Cmd {
+	cmd := exec.CommandContext(ctx, Self())
+	cmd.Args = args
+	return cmd
+}

vendor/github.com/containers/storage/pkg/reexec/command_linux.go

@@ -0,0 +1,34 @@
+//go:build linux
+
+package reexec
+
+import (
+	"context"
+	"os/exec"
+)
+
+// Self returns the path to the current process's binary.
+// Returns "/proc/self/exe".
+func Self() string {
+	return "/proc/self/exe"
+}
+
+// Command returns *exec.Cmd which has Path as current binary.
+// This will use the in-memory version (/proc/self/exe) of the current binary,
+// it is thus safe to delete or replace the on-disk binary (os.Args[0]).
+func Command(args ...string) *exec.Cmd {
+	panicIfNotInitialized()
+	cmd := exec.Command(Self())
+	cmd.Args = args
+	return cmd
+}
+
+// CommandContext returns *exec.Cmd which has Path as current binary.
+// This will use the in-memory version (/proc/self/exe) of the current binary,
+// it is thus safe to delete or replace the on-disk binary (os.Args[0]).
+func CommandContext(ctx context.Context, args ...string) *exec.Cmd {
+	panicIfNotInitialized()
+	cmd := exec.CommandContext(ctx, Self())
+	cmd.Args = args
+	return cmd
+}

vendor/github.com/containers/storage/pkg/reexec/command_unix.go

@@ -0,0 +1,32 @@
+//go:build solaris || darwin
+
+package reexec
+
+import (
+	"context"
+	"os/exec"
+)
+
+// Self returns the path to the current process's binary.
+// Uses os.Args[0].
+func Self() string {
+	return naiveSelf()
+}
+
+// Command returns *exec.Cmd which has Path as current binary.
+// For example if current binary is "docker" at "/usr/bin/", then cmd.Path will
+// be set to "/usr/bin/docker".
+func Command(args ...string) *exec.Cmd {
+	panicIfNotInitialized()
+	cmd := exec.Command(Self())
+	cmd.Args = args
+	return cmd
+}
+
+// CommandContext returns *exec.Cmd which has Path as current binary.
+func CommandContext(ctx context.Context, args ...string) *exec.Cmd {
+	panicIfNotInitialized()
+	cmd := exec.CommandContext(ctx, Self())
+	cmd.Args = args
+	return cmd
+}

vendor/github.com/containers/storage/pkg/reexec/command_unsupported.go

@@ -0,0 +1,20 @@
+//go:build !linux && !windows && !freebsd && !solaris && !darwin
+
+package reexec
+
+import (
+	"context"
+	"os/exec"
+)
+
+// Command is unsupported on operating systems apart from Linux, Windows, Solaris and Darwin.
+func Command(args ...string) *exec.Cmd {
+	panicIfNotInitialized()
+	return nil
+}
+
+// CommandContext is unsupported on operating systems apart from Linux, Windows, Solaris and Darwin.
+func CommandContext(ctx context.Context, args ...string) *exec.Cmd {
+	panicIfNotInitialized()
+	return nil
+}

vendor/github.com/containers/storage/pkg/reexec/command_windows.go

@@ -0,0 +1,34 @@
+//go:build windows
+
+package reexec
+
+import (
+	"context"
+	"os/exec"
+)
+
+// Self returns the path to the current process's binary.
+// Uses os.Args[0].
+func Self() string {
+	return naiveSelf()
+}
+
+// Command returns *exec.Cmd which has Path as current binary.
+// For example if current binary is "docker.exe" at "C:\", then cmd.Path will
+// be set to "C:\docker.exe".
+func Command(args ...string) *exec.Cmd {
+	panicIfNotInitialized()
+	cmd := exec.Command(Self())
+	cmd.Args = args
+	return cmd
+}
+
+// Command returns *exec.Cmd which has Path as current binary.
+// For example if current binary is "docker.exe" at "C:\", then cmd.Path will
+// be set to "C:\docker.exe".
+func CommandContext(ctx context.Context, args ...string) *exec.Cmd {
+	panicIfNotInitialized()
+	cmd := exec.CommandContext(ctx, Self())
+	cmd.Args = args
+	return cmd
+}

vendor/github.com/containers/storage/pkg/reexec/reexec.go

@@ -0,0 +1,66 @@
+package reexec
+
+import (
+	"fmt"
+	"os"
+	"os/exec"
+	"path/filepath"
+)
+
+var (
+	registeredInitializers = make(map[string]func())
+	initWasCalled          = false
+)
+
+// Register adds an initialization func under the specified name
+func Register(name string, initializer func()) {
+	if _, exists := registeredInitializers[name]; exists {
+		panic(fmt.Sprintf("reexec func already registered under name %q", name))
+	}
+
+	registeredInitializers[name] = initializer
+}
+
+// Init is called as the first part of the exec process and returns true if an
+// initialization function was called.
+func Init() bool {
+	initializer, exists := registeredInitializers[os.Args[0]]
+	initWasCalled = true
+	if exists {
+		initializer()
+
+		return true
+	}
+	return false
+}
+
+func panicIfNotInitialized() {
+	if !initWasCalled {
+		// The reexec package is used to run subroutines in
+		// subprocesses which would otherwise have unacceptable side
+		// effects on the main thread.  If you found this error, then
+		// your program uses a package which needs to do this.  In
+		// order for that to work, main() should start with this
+		// boilerplate, or an equivalent:
+		//     if reexec.Init() {
+		//         return
+		//     }
+		panic("a library subroutine needed to run a subprocess, but reexec.Init() was not called in main()")
+	}
+}
+
+func naiveSelf() string {
+	name := os.Args[0]
+	if filepath.Base(name) == name {
+		if lp, err := exec.LookPath(name); err == nil {
+			return lp
+		}
+	}
+	// handle conversion of relative paths to absolute
+	if absName, err := filepath.Abs(name); err == nil {
+		return absName
+	}
+	// if we couldn't get absolute name, return original
+	// (NOTE: Go only errors on Abs() if os.Getwd fails)
+	return name
+}

vendor/github.com/containers/storage/pkg/regexp/regexp.go

@@ -0,0 +1,234 @@
+package regexp
+
+import (
+	"io"
+	"regexp"
+	"sync"
+)
+
+// Regexp is a wrapper struct used for wrapping MustCompile regex expressions
+// used as global variables. Using this structure helps speed the startup time
+// of apps that want to use global regex variables. This library initializes them on
+// first use as opposed to the start of the executable.
+type Regexp struct {
+	*regexpStruct
+}
+
+type regexpStruct struct {
+	_      noCopy
+	once   sync.Once
+	regexp *regexp.Regexp
+	val    string
+}
+
+func Delayed(val string) Regexp {
+	re := &regexpStruct{
+		val: val,
+	}
+	if precompile {
+		re.regexp = regexp.MustCompile(re.val)
+	}
+	return Regexp{re}
+}
+
+func (re *regexpStruct) compile() {
+	if precompile {
+		return
+	}
+	re.once.Do(func() {
+		re.regexp = regexp.MustCompile(re.val)
+	})
+}
+
+func (re *regexpStruct) Expand(dst []byte, template []byte, src []byte, match []int) []byte {
+	re.compile()
+	return re.regexp.Expand(dst, template, src, match)
+}
+
+func (re *regexpStruct) ExpandString(dst []byte, template string, src string, match []int) []byte {
+	re.compile()
+	return re.regexp.ExpandString(dst, template, src, match)
+}
+
+func (re *regexpStruct) Find(b []byte) []byte {
+	re.compile()
+	return re.regexp.Find(b)
+}
+
+func (re *regexpStruct) FindAll(b []byte, n int) [][]byte {
+	re.compile()
+	return re.regexp.FindAll(b, n)
+}
+
+func (re *regexpStruct) FindAllIndex(b []byte, n int) [][]int {
+	re.compile()
+	return re.regexp.FindAllIndex(b, n)
+}
+
+func (re *regexpStruct) FindAllString(s string, n int) []string {
+	re.compile()
+	return re.regexp.FindAllString(s, n)
+}
+
+func (re *regexpStruct) FindAllStringIndex(s string, n int) [][]int {
+	re.compile()
+	return re.regexp.FindAllStringIndex(s, n)
+}
+
+func (re *regexpStruct) FindAllStringSubmatch(s string, n int) [][]string {
+	re.compile()
+	return re.regexp.FindAllStringSubmatch(s, n)
+}
+
+func (re *regexpStruct) FindAllStringSubmatchIndex(s string, n int) [][]int {
+	re.compile()
+	return re.regexp.FindAllStringSubmatchIndex(s, n)
+}
+
+func (re *regexpStruct) FindAllSubmatch(b []byte, n int) [][][]byte {
+	re.compile()
+	return re.regexp.FindAllSubmatch(b, n)
+}
+
+func (re *regexpStruct) FindAllSubmatchIndex(b []byte, n int) [][]int {
+	re.compile()
+	return re.regexp.FindAllSubmatchIndex(b, n)
+}
+
+func (re *regexpStruct) FindIndex(b []byte) (loc []int) {
+	re.compile()
+	return re.regexp.FindIndex(b)
+}
+
+func (re *regexpStruct) FindReaderIndex(r io.RuneReader) (loc []int) {
+	re.compile()
+	return re.regexp.FindReaderIndex(r)
+}
+
+func (re *regexpStruct) FindReaderSubmatchIndex(r io.RuneReader) []int {
+	re.compile()
+	return re.regexp.FindReaderSubmatchIndex(r)
+}
+
+func (re *regexpStruct) FindString(s string) string {
+	re.compile()
+	return re.regexp.FindString(s)
+}
+
+func (re *regexpStruct) FindStringIndex(s string) (loc []int) {
+	re.compile()
+	return re.regexp.FindStringIndex(s)
+}
+
+func (re *regexpStruct) FindStringSubmatch(s string) []string {
+	re.compile()
+	return re.regexp.FindStringSubmatch(s)
+}
+
+func (re *regexpStruct) FindStringSubmatchIndex(s string) []int {
+	re.compile()
+	return re.regexp.FindStringSubmatchIndex(s)
+}
+
+func (re *regexpStruct) FindSubmatch(b []byte) [][]byte {
+	re.compile()
+	return re.regexp.FindSubmatch(b)
+}
+
+func (re *regexpStruct) FindSubmatchIndex(b []byte) []int {
+	re.compile()
+	return re.regexp.FindSubmatchIndex(b)
+}
+
+func (re *regexpStruct) LiteralPrefix() (prefix string, complete bool) {
+	re.compile()
+	return re.regexp.LiteralPrefix()
+}
+
+func (re *regexpStruct) Longest() {
+	re.compile()
+	re.regexp.Longest()
+}
+
+func (re *regexpStruct) Match(b []byte) bool {
+	re.compile()
+	return re.regexp.Match(b)
+}
+
+func (re *regexpStruct) MatchReader(r io.RuneReader) bool {
+	re.compile()
+	return re.regexp.MatchReader(r)
+}
+
+func (re *regexpStruct) MatchString(s string) bool {
+	re.compile()
+	return re.regexp.MatchString(s)
+}
+
+func (re *regexpStruct) NumSubexp() int {
+	re.compile()
+	return re.regexp.NumSubexp()
+}
+
+func (re *regexpStruct) ReplaceAll(src, repl []byte) []byte {
+	re.compile()
+	return re.regexp.ReplaceAll(src, repl)
+}
+
+func (re *regexpStruct) ReplaceAllFunc(src []byte, repl func([]byte) []byte) []byte {
+	re.compile()
+	return re.regexp.ReplaceAllFunc(src, repl)
+}
+
+func (re *regexpStruct) ReplaceAllLiteral(src, repl []byte) []byte {
+	re.compile()
+	return re.regexp.ReplaceAllLiteral(src, repl)
+}
+
+func (re *regexpStruct) ReplaceAllLiteralString(src, repl string) string {
+	re.compile()
+	return re.regexp.ReplaceAllLiteralString(src, repl)
+}
+
+func (re *regexpStruct) ReplaceAllString(src, repl string) string {
+	re.compile()
+	return re.regexp.ReplaceAllString(src, repl)
+}
+
+func (re *regexpStruct) ReplaceAllStringFunc(src string, repl func(string) string) string {
+	re.compile()
+	return re.regexp.ReplaceAllStringFunc(src, repl)
+}
+
+func (re *regexpStruct) Split(s string, n int) []string {
+	re.compile()
+	return re.regexp.Split(s, n)
+}
+
+func (re *regexpStruct) String() string {
+	re.compile()
+	return re.regexp.String()
+}
+
+func (re *regexpStruct) SubexpIndex(name string) int {
+	re.compile()
+	return re.regexp.SubexpIndex(name)
+}
+
+func (re *regexpStruct) SubexpNames() []string {
+	re.compile()
+	return re.regexp.SubexpNames()
+}
+
+// noCopy may be added to structs which must not be copied
+// after the first use.
+//
+// See https://golang.org/issues/8005#issuecomment-190753527
+// for details.
+//
+// Note that it must not be embedded, due to the Lock and Unlock methods.
+type noCopy struct{}
+
+// Lock is a no-op used by -copylocks checker from `go vet`.
+func (*noCopy) Lock()   {}
+func (*noCopy) Unlock() {}

vendor/github.com/containers/storage/pkg/regexp/regexp_dontprecompile.go

@@ -0,0 +1,5 @@
+//go:build !regexp_precompile
+
+package regexp
+
+const precompile = false

vendor/github.com/containers/storage/pkg/regexp/regexp_precompile.go

@@ -0,0 +1,5 @@
+//go:build regexp_precompile
+
+package regexp
+
+const precompile = true

vendor/github.com/containers/storage/pkg/system/chmod.go

@@ -0,0 +1,17 @@
+package system
+
+import (
+	"errors"
+	"os"
+	"syscall"
+)
+
+func Chmod(name string, mode os.FileMode) error {
+	err := os.Chmod(name, mode)
+
+	for err != nil && errors.Is(err, syscall.EINTR) {
+		err = os.Chmod(name, mode)
+	}
+
+	return err
+}

vendor/github.com/containers/storage/pkg/system/chtimes.go

@@ -0,0 +1,35 @@
+package system
+
+import (
+	"os"
+	"time"
+)
+
+// Chtimes changes the access time and modified time of a file at the given path
+func Chtimes(name string, atime time.Time, mtime time.Time) error {
+	unixMinTime := time.Unix(0, 0)
+	unixMaxTime := maxTime
+
+	// If the modified time is prior to the Unix Epoch, or after the
+	// end of Unix Time, os.Chtimes has undefined behavior
+	// default to Unix Epoch in this case, just in case
+
+	if atime.Before(unixMinTime) || atime.After(unixMaxTime) {
+		atime = unixMinTime
+	}
+
+	if mtime.Before(unixMinTime) || mtime.After(unixMaxTime) {
+		mtime = unixMinTime
+	}
+
+	if err := os.Chtimes(name, atime, mtime); err != nil {
+		return err
+	}
+
+	// Take platform specific action for setting create time.
+	if err := setCTime(name, mtime); err != nil {
+		return err
+	}
+
+	return nil
+}

vendor/github.com/containers/storage/pkg/system/chtimes_unix.go

@@ -0,0 +1,14 @@
+//go:build !windows
+
+package system
+
+import (
+	"time"
+)
+
+// setCTime will set the create time on a file. On Unix, the create
+// time is updated as a side effect of setting the modified time, so
+// no action is required.
+func setCTime(path string, ctime time.Time) error {
+	return nil
+}

vendor/github.com/containers/storage/pkg/system/chtimes_windows.go

@@ -0,0 +1,28 @@
+//go:build windows
+
+package system
+
+import (
+	"time"
+
+	"golang.org/x/sys/windows"
+)
+
+// setCTime will set the create time on a file. On Windows, this requires
+// calling SetFileTime and explicitly including the create time.
+func setCTime(path string, ctime time.Time) error {
+	ctimespec := windows.NsecToTimespec(ctime.UnixNano())
+	pathp, e := windows.UTF16PtrFromString(path)
+	if e != nil {
+		return e
+	}
+	h, e := windows.CreateFile(pathp,
+		windows.FILE_WRITE_ATTRIBUTES, windows.FILE_SHARE_WRITE, nil,
+		windows.OPEN_EXISTING, windows.FILE_FLAG_BACKUP_SEMANTICS, 0)
+	if e != nil {
+		return e
+	}
+	defer windows.Close(h)
+	c := windows.NsecToFiletime(windows.TimespecToNsec(ctimespec))
+	return windows.SetFileTime(h, &c, nil, nil)
+}

vendor/github.com/containers/storage/pkg/system/errors.go

@@ -0,0 +1,8 @@
+package system
+
+import (
+	"errors"
+)
+
+// ErrNotSupportedPlatform means the platform is not supported.
+var ErrNotSupportedPlatform = errors.New("platform and architecture is not supported")

vendor/github.com/containers/storage/pkg/system/exitcode.go

@@ -0,0 +1,33 @@
+package system
+
+import (
+	"fmt"
+	"os/exec"
+	"syscall"
+)
+
+// GetExitCode returns the ExitStatus of the specified error if its type is
+// exec.ExitError, returns 0 and an error otherwise.
+func GetExitCode(err error) (int, error) {
+	exitCode := 0
+	if exiterr, ok := err.(*exec.ExitError); ok {
+		if procExit, ok := exiterr.Sys().(syscall.WaitStatus); ok {
+			return procExit.ExitStatus(), nil
+		}
+	}
+	return exitCode, fmt.Errorf("failed to get exit code")
+}
+
+// ProcessExitCode process the specified error and returns the exit status code
+// if the error was of type exec.ExitError, returns nothing otherwise.
+func ProcessExitCode(err error) (exitCode int) {
+	if err != nil {
+		var exiterr error
+		if exitCode, exiterr = GetExitCode(err); exiterr != nil {
+			// TODO: Fix this so we check the error's text.
+			// we've failed to retrieve exit code, so we set it to 127
+			exitCode = 127
+		}
+	}
+	return
+}

vendor/github.com/containers/storage/pkg/system/extattr_freebsd.go

@@ -0,0 +1,93 @@
+//go:build freebsd
+
+package system
+
+import (
+	"os"
+	"unsafe"
+
+	"golang.org/x/sys/unix"
+)
+
+const (
+	EXTATTR_NAMESPACE_EMPTY  = unix.EXTATTR_NAMESPACE_EMPTY
+	EXTATTR_NAMESPACE_USER   = unix.EXTATTR_NAMESPACE_USER
+	EXTATTR_NAMESPACE_SYSTEM = unix.EXTATTR_NAMESPACE_SYSTEM
+)
+
+// ExtattrGetLink retrieves the value of the extended attribute identified by attrname
+// in the given namespace and associated with the given path in the file system.
+// If the path is a symbolic link, the extended attribute is retrieved from the link itself.
+// Returns a []byte slice if the extattr is set and nil otherwise.
+func ExtattrGetLink(path string, attrnamespace int, attrname string) ([]byte, error) {
+	size, errno := unix.ExtattrGetLink(path, attrnamespace, attrname,
+		uintptr(unsafe.Pointer(nil)), 0)
+	if errno != nil {
+		if errno == unix.ENOATTR {
+			return nil, nil
+		}
+		return nil, &os.PathError{Op: "extattr_get_link", Path: path, Err: errno}
+	}
+	if size == 0 {
+		return []byte{}, nil
+	}
+
+	dest := make([]byte, size)
+	size, errno = unix.ExtattrGetLink(path, attrnamespace, attrname,
+		uintptr(unsafe.Pointer(&dest[0])), size)
+	if errno != nil {
+		return nil, &os.PathError{Op: "extattr_get_link", Path: path, Err: errno}
+	}
+
+	return dest[:size], nil
+}
+
+// ExtattrSetLink sets the value of extended attribute identified by attrname
+// in the given namespace and associated with the given path in the file system.
+// If the path is a symbolic link, the extended attribute is set on the link itself.
+func ExtattrSetLink(path string, attrnamespace int, attrname string, data []byte) error {
+	if len(data) == 0 {
+		data = []byte{} // ensure non-nil for empty data
+	}
+	if _, errno := unix.ExtattrSetLink(path, attrnamespace, attrname,
+		uintptr(unsafe.Pointer(&data[0])), len(data)); errno != nil {
+		return &os.PathError{Op: "extattr_set_link", Path: path, Err: errno}
+	}
+
+	return nil
+}
+
+// ExtattrListLink lists extended attributes associated with the given path
+// in the specified namespace. If the path is a symbolic link, the attributes
+// are listed from the link itself.
+func ExtattrListLink(path string, attrnamespace int) ([]string, error) {
+	size, errno := unix.ExtattrListLink(path, attrnamespace,
+		uintptr(unsafe.Pointer(nil)), 0)
+	if errno != nil {
+		return nil, &os.PathError{Op: "extattr_list_link", Path: path, Err: errno}
+	}
+	if size == 0 {
+		return []string{}, nil
+	}
+
+	dest := make([]byte, size)
+	size, errno = unix.ExtattrListLink(path, attrnamespace,
+		uintptr(unsafe.Pointer(&dest[0])), size)
+	if errno != nil {
+		return nil, &os.PathError{Op: "extattr_list_link", Path: path, Err: errno}
+	}
+
+	var attrs []string
+	for i := 0; i < size; {
+		// Each attribute is preceded by a single byte length
+		length := int(dest[i])
+		i++
+		if i+length > size {
+			break
+		}
+		attrs = append(attrs, string(dest[i:i+length]))
+		i += length
+	}
+
+	return attrs, nil
+}

vendor/github.com/containers/storage/pkg/system/extattr_unsupported.go

@@ -0,0 +1,24 @@
+//go:build !freebsd
+
+package system
+
+const (
+	EXTATTR_NAMESPACE_EMPTY  = 0
+	EXTATTR_NAMESPACE_USER   = 0
+	EXTATTR_NAMESPACE_SYSTEM = 0
+)
+
+// ExtattrGetLink is not supported on platforms other than FreeBSD.
+func ExtattrGetLink(path string, attrnamespace int, attrname string) ([]byte, error) {
+	return nil, ErrNotSupportedPlatform
+}
+
+// ExtattrSetLink is not supported on platforms other than FreeBSD.
+func ExtattrSetLink(path string, attrnamespace int, attrname string, data []byte) error {
+	return ErrNotSupportedPlatform
+}
+
+// ExtattrListLink is not supported on platforms other than FreeBSD.
+func ExtattrListLink(path string, attrnamespace int) ([]string, error) {
+	return nil, ErrNotSupportedPlatform
+}

vendor/github.com/containers/storage/pkg/system/init.go

@@ -0,0 +1,22 @@
+package system
+
+import (
+	"syscall"
+	"time"
+	"unsafe"
+)
+
+// maxTime is used by chtimes.
+var maxTime time.Time
+
+func init() {
+	// chtimes initialization
+	if unsafe.Sizeof(syscall.Timespec{}.Nsec) == 8 {
+		// This is a 64 bit timespec
+		// os.Chtimes limits time to the following
+		maxTime = time.Unix(0, 1<<63-1)
+	} else {
+		// This is a 32 bit timespec
+		maxTime = time.Unix(1<<31-1, 0)
+	}
+}

vendor/github.com/containers/storage/pkg/system/init_windows.go

@@ -0,0 +1,16 @@
+package system
+
+import "os"
+
+// LCOWSupported determines if Linux Containers on Windows are supported.
+// Note: This feature is in development (06/17) and enabled through an
+// environment variable. At a future time, it will be enabled based
+// on build number. @jhowardmsft
+var lcowSupported = false
+
+func init() {
+	// LCOW initialization
+	if os.Getenv("LCOW_SUPPORTED") != "" {
+		lcowSupported = true
+	}
+}

vendor/github.com/containers/storage/pkg/system/lchflags_bsd.go

@@ -0,0 +1,55 @@
+//go:build freebsd
+
+package system
+
+import (
+	"unsafe"
+
+	"golang.org/x/sys/unix"
+)
+
+// Flag values from <sys/stat.h>
+const (
+	/*
+	 * Definitions of flags stored in file flags word.
+	 *
+	 * Super-user and owner changeable flags.
+	 */
+	UF_SETTABLE  uint32 = 0x0000ffff /* mask of owner changeable flags */
+	UF_NODUMP    uint32 = 0x00000001 /* do not dump file */
+	UF_IMMUTABLE uint32 = 0x00000002 /* file may not be changed */
+	UF_APPEND    uint32 = 0x00000004 /* writes to file may only append */
+	UF_OPAQUE    uint32 = 0x00000008 /* directory is opaque wrt. union */
+	UF_NOUNLINK  uint32 = 0x00000010 /* file may not be removed or renamed */
+
+	UF_SYSTEM   uint32 = 0x00000080 /* Windows system file bit */
+	UF_SPARSE   uint32 = 0x00000100 /* sparse file */
+	UF_OFFLINE  uint32 = 0x00000200 /* file is offline */
+	UF_REPARSE  uint32 = 0x00000400 /* Windows reparse point file bit */
+	UF_ARCHIVE  uint32 = 0x00000800 /* file needs to be archived */
+	UF_READONLY uint32 = 0x00001000 /* Windows readonly file bit */
+	/* This is the same as the MacOS X definition of UF_HIDDEN. */
+	UF_HIDDEN uint32 = 0x00008000 /* file is hidden */
+
+	/*
+	 * Super-user changeable flags.
+	 */
+	SF_SETTABLE  uint32 = 0xffff0000 /* mask of superuser changeable flags */
+	SF_ARCHIVED  uint32 = 0x00010000 /* file is archived */
+	SF_IMMUTABLE uint32 = 0x00020000 /* file may not be changed */
+	SF_APPEND    uint32 = 0x00040000 /* writes to file may only append */
+	SF_NOUNLINK  uint32 = 0x00100000 /* file may not be removed or renamed */
+	SF_SNAPSHOT  uint32 = 0x00200000 /* snapshot inode */
+)
+
+func Lchflags(path string, flags uint32) error {
+	p, err := unix.BytePtrFromString(path)
+	if err != nil {
+		return err
+	}
+	_, _, e1 := unix.Syscall(unix.SYS_LCHFLAGS, uintptr(unsafe.Pointer(p)), uintptr(flags), 0)
+	if e1 != 0 {
+		return e1
+	}
+	return nil
+}

vendor/github.com/containers/storage/pkg/system/lchown.go

@@ -0,0 +1,20 @@
+package system
+
+import (
+	"os"
+	"syscall"
+)
+
+func Lchown(name string, uid, gid int) error {
+	err := syscall.Lchown(name, uid, gid)
+
+	for err == syscall.EINTR {
+		err = syscall.Lchown(name, uid, gid)
+	}
+
+	if err != nil {
+		return &os.PathError{Op: "lchown", Path: name, Err: err}
+	}
+
+	return nil
+}

vendor/github.com/containers/storage/pkg/system/lcow_unix.go

@@ -0,0 +1,8 @@
+//go:build !windows
+
+package system
+
+// LCOWSupported returns true if Linux containers on Windows are supported.
+func LCOWSupported() bool {
+	return false
+}

vendor/github.com/containers/storage/pkg/system/lcow_windows.go

@@ -0,0 +1,6 @@
+package system
+
+// LCOWSupported returns true if Linux containers on Windows are supported.
+func LCOWSupported() bool {
+	return lcowSupported
+}

vendor/github.com/containers/storage/pkg/system/lstat_unix.go

@@ -0,0 +1,20 @@
+//go:build !windows
+
+package system
+
+import (
+	"os"
+	"syscall"
+)
+
+// Lstat takes a path to a file and returns
+// a system.StatT type pertaining to that file.
+//
+// Throws an error if the file does not exist
+func Lstat(path string) (*StatT, error) {
+	s := &syscall.Stat_t{}
+	if err := syscall.Lstat(path, s); err != nil {
+		return nil, &os.PathError{Op: "Lstat", Path: path, Err: err}
+	}
+	return fromStatT(s)
+}

vendor/github.com/containers/storage/pkg/system/lstat_windows.go

@@ -0,0 +1,14 @@
+package system
+
+import "os"
+
+// Lstat calls os.Lstat to get a fileinfo interface back.
+// This is then copied into our own locally defined structure.
+func Lstat(path string) (*StatT, error) {
+	fi, err := os.Lstat(path)
+	if err != nil {
+		return nil, err
+	}
+
+	return fromStatT(&fi)
+}

vendor/github.com/containers/storage/pkg/system/meminfo.go

@@ -0,0 +1,17 @@
+package system
+
+// MemInfo contains memory statistics of the host system.
+type MemInfo struct {
+	// Total usable RAM (i.e. physical RAM minus a few reserved bits and the
+	// kernel binary code).
+	MemTotal int64
+
+	// Amount of free memory.
+	MemFree int64
+
+	// Total amount of swap space available.
+	SwapTotal int64
+
+	// Amount of swap space that is currently unused.
+	SwapFree int64
+}

vendor/github.com/containers/storage/pkg/system/meminfo_freebsd.go

@@ -0,0 +1,85 @@
+//go:build freebsd && cgo
+
+package system
+
+import (
+	"errors"
+	"fmt"
+	"unsafe"
+
+	"golang.org/x/sys/unix"
+)
+
+// #include <unistd.h>
+// #include <sys/vmmeter.h>
+// #include <sys/sysctl.h>
+// #include <vm/vm_param.h>
+import "C"
+
+func getMemInfo() (int64, int64, error) {
+	data, err := unix.SysctlRaw("vm.vmtotal")
+	if err != nil {
+		return -1, -1, fmt.Errorf("can't get kernel info: %w", err)
+	}
+	if len(data) != C.sizeof_struct_vmtotal {
+		return -1, -1, fmt.Errorf("unexpected vmtotal size %d", len(data))
+	}
+
+	total := (*C.struct_vmtotal)(unsafe.Pointer(&data[0]))
+
+	pagesize := int64(C.sysconf(C._SC_PAGESIZE))
+	npages := int64(C.sysconf(C._SC_PHYS_PAGES))
+	return pagesize * npages, pagesize * int64(total.t_free), nil
+}
+
+func getSwapInfo() (int64, int64, error) {
+	var (
+		total int64 = 0
+		used  int64 = 0
+	)
+	swapCount, err := unix.SysctlUint32("vm.nswapdev")
+	if err != nil {
+		return -1, -1, fmt.Errorf("reading vm.nswapdev: %w", err)
+	}
+	for i := 0; i < int(swapCount); i++ {
+		data, err := unix.SysctlRaw("vm.swap_info", i)
+		if err != nil {
+			return -1, -1, fmt.Errorf("reading vm.swap_info.%d: %w", i, err)
+		}
+		if len(data) != C.sizeof_struct_xswdev {
+			return -1, -1, fmt.Errorf("unexpected swap_info size %d", len(data))
+		}
+		xsw := (*C.struct_xswdev)(unsafe.Pointer(&data[0]))
+		total += int64(xsw.xsw_nblks)
+		used += int64(xsw.xsw_used)
+	}
+	pagesize := int64(C.sysconf(C._SC_PAGESIZE))
+	return pagesize * total, pagesize * (total - used), nil
+}
+
+// ReadMemInfo retrieves memory statistics of the host system and returns a
+//
+//	MemInfo type.
+func ReadMemInfo() (*MemInfo, error) {
+	MemTotal, MemFree, err := getMemInfo()
+	if err != nil {
+		return nil, fmt.Errorf("getting memory totals %w", err)
+	}
+	SwapTotal, SwapFree, err := getSwapInfo()
+	if err != nil {
+		return nil, fmt.Errorf("getting swap totals %w", err)
+	}
+
+	if MemTotal < 0 || MemFree < 0 || SwapTotal < 0 || SwapFree < 0 {
+		return nil, errors.New("getting system memory info")
+	}
+
+	meminfo := &MemInfo{}
+	// Total memory is total physical memory less than memory locked by kernel
+	meminfo.MemTotal = MemTotal
+	meminfo.MemFree = MemFree
+	meminfo.SwapTotal = SwapTotal
+	meminfo.SwapFree = SwapFree
+
+	return meminfo, nil
+}

vendor/github.com/containers/storage/pkg/system/meminfo_linux.go

@@ -0,0 +1,65 @@
+package system
+
+import (
+	"bufio"
+	"io"
+	"os"
+	"strconv"
+	"strings"
+
+	"github.com/docker/go-units"
+)
+
+// ReadMemInfo retrieves memory statistics of the host system and returns a
+// MemInfo type.
+func ReadMemInfo() (*MemInfo, error) {
+	file, err := os.Open("/proc/meminfo")
+	if err != nil {
+		return nil, err
+	}
+	defer file.Close()
+	return parseMemInfo(file)
+}
+
+// parseMemInfo parses the /proc/meminfo file into
+// a MemInfo object given an io.Reader to the file.
+// Throws error if there are problems reading from the file
+func parseMemInfo(reader io.Reader) (*MemInfo, error) {
+	meminfo := &MemInfo{}
+	scanner := bufio.NewScanner(reader)
+	for scanner.Scan() {
+		// Expected format: ["MemTotal:", "1234", "kB"]
+		parts := strings.Fields(scanner.Text())
+
+		// Sanity checks: Skip malformed entries.
+		if len(parts) < 3 || parts[2] != "kB" {
+			continue
+		}
+
+		// Convert to bytes.
+		size, err := strconv.Atoi(parts[1])
+		if err != nil {
+			continue
+		}
+		bytes := int64(size) * units.KiB
+
+		switch parts[0] {
+		case "MemTotal:":
+			meminfo.MemTotal = bytes
+		case "MemFree:":
+			meminfo.MemFree = bytes
+		case "SwapTotal:":
+			meminfo.SwapTotal = bytes
+		case "SwapFree:":
+			meminfo.SwapFree = bytes
+		}
+
+	}
+
+	// Handle errors that may have occurred during the reading of the file.
+	if err := scanner.Err(); err != nil {
+		return nil, err
+	}
+
+	return meminfo, nil
+}

vendor/github.com/containers/storage/pkg/system/meminfo_solaris.go

@@ -0,0 +1,129 @@
+//go:build solaris && cgo
+
+package system
+
+import (
+	"fmt"
+	"unsafe"
+)
+
+// #cgo CFLAGS: -std=c99
+// #cgo LDFLAGS: -lkstat
+// #include <unistd.h>
+// #include <stdlib.h>
+// #include <stdio.h>
+// #include <kstat.h>
+// #include <sys/swap.h>
+// #include <sys/param.h>
+// struct swaptable *allocSwaptable(int num) {
+//	struct swaptable *st;
+//	struct swapent *swapent;
+// 	st = (struct swaptable *)malloc(num * sizeof(swapent_t) + sizeof (int));
+//	swapent = st->swt_ent;
+//	for (int i = 0; i < num; i++,swapent++) {
+//		swapent->ste_path = (char *)malloc(MAXPATHLEN * sizeof (char));
+//	}
+//	st->swt_n = num;
+//	return st;
+//}
+// void freeSwaptable (struct swaptable *st) {
+//	struct swapent *swapent = st->swt_ent;
+//	for (int i = 0; i < st->swt_n; i++,swapent++) {
+//		free(swapent->ste_path);
+//	}
+//	free(st);
+// }
+// swapent_t getSwapEnt(swapent_t *ent, int i) {
+//	return ent[i];
+// }
+// int64_t getPpKernel() {
+//	int64_t pp_kernel = 0;
+//	kstat_ctl_t *ksc;
+//	kstat_t *ks;
+//	kstat_named_t *knp;
+//	kid_t kid;
+//
+//	if ((ksc = kstat_open()) == NULL) {
+//		return -1;
+//	}
+//	if ((ks = kstat_lookup(ksc, "unix", 0, "system_pages")) == NULL) {
+//		return -1;
+//	}
+//	if (((kid = kstat_read(ksc, ks, NULL)) == -1) ||
+//	    ((knp = kstat_data_lookup(ks, "pp_kernel")) == NULL)) {
+//		return -1;
+//	}
+//	switch (knp->data_type) {
+//	case KSTAT_DATA_UINT64:
+//		pp_kernel = knp->value.ui64;
+//		break;
+//	case KSTAT_DATA_UINT32:
+//		pp_kernel = knp->value.ui32;
+//		break;
+//	}
+//	pp_kernel *= sysconf(_SC_PAGESIZE);
+//	return (pp_kernel > 0 ? pp_kernel : -1);
+// }
+import "C"
+
+// Get the system memory info using sysconf same as prtconf
+func getTotalMem() int64 {
+	pagesize := C.sysconf(C._SC_PAGESIZE)
+	npages := C.sysconf(C._SC_PHYS_PAGES)
+	return int64(pagesize * npages)
+}
+
+func getFreeMem() int64 {
+	pagesize := C.sysconf(C._SC_PAGESIZE)
+	npages := C.sysconf(C._SC_AVPHYS_PAGES)
+	return int64(pagesize * npages)
+}
+
+// ReadMemInfo retrieves memory statistics of the host system and returns a
+//
+//	MemInfo type.
+func ReadMemInfo() (*MemInfo, error) {
+	ppKernel := C.getPpKernel()
+	MemTotal := getTotalMem()
+	MemFree := getFreeMem()
+	SwapTotal, SwapFree, err := getSysSwap()
+
+	if ppKernel < 0 || MemTotal < 0 || MemFree < 0 || SwapTotal < 0 ||
+		SwapFree < 0 {
+		return nil, fmt.Errorf("getting system memory info %w", err)
+	}
+
+	meminfo := &MemInfo{}
+	// Total memory is total physical memory less than memory locked by kernel
+	meminfo.MemTotal = MemTotal - int64(ppKernel)
+	meminfo.MemFree = MemFree
+	meminfo.SwapTotal = SwapTotal
+	meminfo.SwapFree = SwapFree
+
+	return meminfo, nil
+}
+
+func getSysSwap() (int64, int64, error) {
+	var tSwap int64
+	var fSwap int64
+	var diskblksPerPage int64
+	num, err := C.swapctl(C.SC_GETNSWP, nil)
+	if err != nil {
+		return -1, -1, err
+	}
+	st := C.allocSwaptable(num)
+	_, err = C.swapctl(C.SC_LIST, unsafe.Pointer(st))
+	if err != nil {
+		C.freeSwaptable(st)
+		return -1, -1, err
+	}
+
+	diskblksPerPage = int64(C.sysconf(C._SC_PAGESIZE) >> C.DEV_BSHIFT)
+	for i := 0; i < int(num); i++ {
+		swapent := C.getSwapEnt(&st.swt_ent[0], C.int(i))
+		tSwap += int64(swapent.ste_pages) * diskblksPerPage
+		fSwap += int64(swapent.ste_free) * diskblksPerPage
+	}
+	C.freeSwaptable(st)
+	return tSwap, fSwap, nil
+}

vendor/github.com/containers/storage/pkg/system/meminfo_unsupported.go

@@ -0,0 +1,8 @@
+//go:build !linux && !windows && !solaris && !(freebsd && cgo)
+
+package system
+
+// ReadMemInfo is not supported on platforms other than linux and windows.
+func ReadMemInfo() (*MemInfo, error) {
+	return nil, ErrNotSupportedPlatform
+}

vendor/github.com/containers/storage/pkg/system/meminfo_windows.go

@@ -0,0 +1,46 @@
+package system
+
+import (
+	"unsafe"
+
+	"golang.org/x/sys/windows"
+)
+
+var (
+	modkernel32 = windows.NewLazySystemDLL("kernel32.dll")
+
+	procGlobalMemoryStatusEx = modkernel32.NewProc("GlobalMemoryStatusEx")
+)
+
+// https://msdn.microsoft.com/en-us/library/windows/desktop/aa366589(v=vs.85).aspx
+// https://msdn.microsoft.com/en-us/library/windows/desktop/aa366770(v=vs.85).aspx
+type memorystatusex struct {
+	dwLength                uint32
+	dwMemoryLoad            uint32
+	ullTotalPhys            uint64
+	ullAvailPhys            uint64
+	ullTotalPageFile        uint64
+	ullAvailPageFile        uint64
+	ullTotalVirtual         uint64
+	ullAvailVirtual         uint64
+	ullAvailExtendedVirtual uint64
+}
+
+// ReadMemInfo retrieves memory statistics of the host system and returns a
+//
+//	MemInfo type.
+func ReadMemInfo() (*MemInfo, error) {
+	msi := &memorystatusex{
+		dwLength: 64,
+	}
+	r1, _, _ := procGlobalMemoryStatusEx.Call(uintptr(unsafe.Pointer(msi)))
+	if r1 == 0 {
+		return &MemInfo{}, nil
+	}
+	return &MemInfo{
+		MemTotal:  int64(msi.ullTotalPhys),
+		MemFree:   int64(msi.ullAvailPhys),
+		SwapTotal: int64(msi.ullTotalPageFile),
+		SwapFree:  int64(msi.ullAvailPageFile),
+	}, nil
+}

vendor/github.com/containers/storage/pkg/system/mknod.go

@@ -0,0 +1,22 @@
+//go:build !windows && !freebsd
+
+package system
+
+import (
+	"golang.org/x/sys/unix"
+)
+
+// Mknod creates a filesystem node (file, device special file or named pipe) named path
+// with attributes specified by mode and dev.
+func Mknod(path string, mode uint32, dev uint32) error {
+	return unix.Mknod(path, mode, int(dev))
+}
+
+// Mkdev is used to build the value of linux devices (in /dev/) which specifies major
+// and minor number of the newly created device special file.
+// Linux device nodes are a bit weird due to backwards compat with 16 bit device nodes.
+// They are, from low to high: the lower 8 bits of the minor, then 12 bits of the major,
+// then the top 12 bits of the minor.
+func Mkdev(major int64, minor int64) uint32 {
+	return uint32(((minor & 0xfff00) << 12) | ((major & 0xfff) << 8) | (minor & 0xff))
+}

vendor/github.com/containers/storage/pkg/system/mknod_freebsd.go

@@ -0,0 +1,22 @@
+//go:build freebsd
+
+package system
+
+import (
+	"golang.org/x/sys/unix"
+)
+
+// Mknod creates a filesystem node (file, device special file or named pipe) named path
+// with attributes specified by mode and dev.
+func Mknod(path string, mode uint32, dev uint64) error {
+	return unix.Mknod(path, mode, dev)
+}
+
+// Mkdev is used to build the value of linux devices (in /dev/) which specifies major
+// and minor number of the newly created device special file.
+// Linux device nodes are a bit weird due to backwards compat with 16 bit device nodes.
+// They are, from low to high: the lower 8 bits of the minor, then 12 bits of the major,
+// then the top 12 bits of the minor.
+func Mkdev(major int64, minor int64) uint64 {
+	return uint64(((minor & 0xfff00) << 12) | ((major & 0xfff) << 8) | (minor & 0xff))
+}

vendor/github.com/containers/storage/pkg/system/mknod_windows.go

@@ -0,0 +1,13 @@
+//go:build windows
+
+package system
+
+// Mknod is not implemented on Windows.
+func Mknod(path string, mode uint32, dev int) error {
+	return ErrNotSupportedPlatform
+}
+
+// Mkdev is not implemented on Windows.
+func Mkdev(major int64, minor int64) uint32 {
+	panic("Mkdev not implemented on Windows.")
+}

vendor/github.com/containers/storage/pkg/system/path.go

@@ -0,0 +1,20 @@
+package system
+
+import "runtime"
+
+const defaultUnixPathEnv = "/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin"
+
+// DefaultPathEnv is unix style list of directories to search for
+// executables. Each directory is separated from the next by a colon
+// ':' character .
+func DefaultPathEnv(platform string) string {
+	if runtime.GOOS == "windows" {
+		if platform != runtime.GOOS && LCOWSupported() {
+			return defaultUnixPathEnv
+		}
+		// Deliberately empty on Windows containers on Windows as the default path will be set by
+		// the container. Docker has no context of what the default path should be.
+		return ""
+	}
+	return defaultUnixPathEnv
+}

vendor/github.com/containers/storage/pkg/system/path_unix.go

@@ -0,0 +1,9 @@
+//go:build !windows
+
+package system
+
+// CheckSystemDriveAndRemoveDriveLetter verifies that a path, if it includes a drive letter,
+// is the system drive. This is a no-op on Linux.
+func CheckSystemDriveAndRemoveDriveLetter(path string) (string, error) {
+	return path, nil
+}

vendor/github.com/containers/storage/pkg/system/path_windows.go

@@ -0,0 +1,33 @@
+//go:build windows
+
+package system
+
+import (
+	"fmt"
+	"path/filepath"
+	"strings"
+)
+
+// CheckSystemDriveAndRemoveDriveLetter verifies and manipulates a Windows path.
+// This is used, for example, when validating a user provided path in docker cp.
+// If a drive letter is supplied, it must be the system drive. The drive letter
+// is always removed. Also, it translates it to OS semantics (IOW / to \). We
+// need the path in this syntax so that it can ultimately be concatenated with
+// a Windows long-path which doesn't support drive-letters. Examples:
+// C:			--> Fail
+// C:\			--> \
+// a			--> a
+// /a			--> \a
+// d:\			--> Fail
+func CheckSystemDriveAndRemoveDriveLetter(path string) (string, error) {
+	if len(path) == 2 && string(path[1]) == ":" {
+		return "", fmt.Errorf("relative path not specified in %q", path)
+	}
+	if !filepath.IsAbs(path) || len(path) < 2 {
+		return filepath.FromSlash(path), nil
+	}
+	if string(path[1]) == ":" && !strings.EqualFold(string(path[0]), "c") {
+		return "", fmt.Errorf("specified path is not on the system drive (C:)")
+	}
+	return filepath.FromSlash(path[2:]), nil
+}

vendor/github.com/containers/storage/pkg/system/process_unix.go

@@ -0,0 +1,24 @@
+//go:build linux || freebsd || solaris || darwin
+
+package system
+
+import (
+	"syscall"
+
+	"golang.org/x/sys/unix"
+)
+
+// IsProcessAlive returns true if process with a given pid is running.
+func IsProcessAlive(pid int) bool {
+	err := unix.Kill(pid, syscall.Signal(0))
+	if err == nil || err == unix.EPERM {
+		return true
+	}
+
+	return false
+}
+
+// KillProcess force-stops a process.
+func KillProcess(pid int) {
+	_ = unix.Kill(pid, unix.SIGKILL)
+}

vendor/github.com/containers/storage/pkg/system/rm.go

@@ -0,0 +1,99 @@
+package system
+
+import (
+	"errors"
+	"fmt"
+	"os"
+	"syscall"
+	"time"
+
+	"github.com/containers/storage/pkg/mount"
+	"github.com/sirupsen/logrus"
+)
+
+// EnsureRemoveAll wraps `os.RemoveAll` to check for specific errors that can
+// often be remedied.
+// Only use `EnsureRemoveAll` if you really want to make every effort to remove
+// a directory.
+//
+// Because of the way `os.Remove` (and by extension `os.RemoveAll`) works, there
+// can be a race between reading directory entries and then actually attempting
+// to remove everything in the directory.
+// These types of errors do not need to be returned since it's ok for the dir to
+// be gone we can just retry the remove operation.
+//
+// This should not return a `os.ErrNotExist` kind of error under any circumstances
+func EnsureRemoveAll(dir string) error {
+	notExistErr := make(map[string]bool)
+
+	// track retries
+	exitOnErr := make(map[string]int)
+	maxRetry := 1000
+
+	// Attempt a simple remove all first, this avoids the more expensive
+	// RecursiveUnmount call if not needed.
+	if err := os.RemoveAll(dir); err == nil {
+		return nil
+	}
+
+	// Attempt to unmount anything beneath this dir first
+	if err := mount.RecursiveUnmount(dir); err != nil {
+		logrus.Debugf("RecursiveUnmount on %s failed: %v", dir, err)
+	}
+
+	for {
+		err := os.RemoveAll(dir)
+		if err == nil {
+			return nil
+		}
+
+		// If the RemoveAll fails with a permission error, we
+		// may have immutable files so try to remove the
+		// immutable flag and redo the RemoveAll.
+		if errors.Is(err, syscall.EPERM) {
+			if err = resetFileFlags(dir); err != nil {
+				return fmt.Errorf("resetting file flags: %w", err)
+			}
+			err = os.RemoveAll(dir)
+			if err == nil {
+				return nil
+			}
+		}
+
+		pe, ok := err.(*os.PathError)
+		if !ok {
+			return err
+		}
+
+		if os.IsNotExist(err) {
+			if notExistErr[pe.Path] {
+				return err
+			}
+			notExistErr[pe.Path] = true
+
+			// There is a race where some subdir can be removed but after the parent
+			//   dir entries have been read.
+			// So the path could be from `os.Remove(subdir)`
+			// If the reported non-existent path is not the passed in `dir` we
+			// should just retry, but otherwise return with no error.
+			if pe.Path == dir {
+				return nil
+			}
+			continue
+		}
+
+		if !IsEBUSY(pe.Err) {
+			return err
+		}
+
+		if e := mount.Unmount(pe.Path); e != nil {
+			return fmt.Errorf("while removing %s: %w", dir, e)
+		}
+
+		if exitOnErr[pe.Path] == maxRetry {
+			return err
+		}
+		exitOnErr[pe.Path]++
+		time.Sleep(10 * time.Millisecond)
+	}
+}

vendor/github.com/containers/storage/pkg/system/rm_common.go

@@ -0,0 +1,9 @@
+//go:build !freebsd
+
+package system
+
+// Reset file flags in a directory tree. This allows EnsureRemoveAll
+// to delete trees which have the immutable flag set.
+func resetFileFlags(dir string) error {
+	return nil
+}

vendor/github.com/containers/storage/pkg/system/rm_freebsd.go

@@ -0,0 +1,17 @@
+package system
+
+import (
+	"io/fs"
+	"path/filepath"
+)
+
+// Reset file flags in a directory tree. This allows EnsureRemoveAll
+// to delete trees which have the immutable flag set.
+func resetFileFlags(dir string) error {
+	return filepath.WalkDir(dir, func(path string, d fs.DirEntry, err error) error {
+		if err := Lchflags(path, 0); err != nil {
+			return err
+		}
+		return nil
+	})
+}

vendor/github.com/containers/storage/pkg/system/stat_common.go

@@ -0,0 +1,11 @@
+//go:build !freebsd
+
+package system
+
+type platformStatT struct{}
+
+// Flags return file flags if supported or zero otherwise
+func (s StatT) Flags() uint32 {
+	_ = s.platformStatT // Silence warnings that StatT.platformStatT is unused (on these platforms)
+	return 0
+}

vendor/github.com/containers/storage/pkg/system/stat_darwin.go

@@ -0,0 +1,15 @@
+package system
+
+import "syscall"
+
+// fromStatT converts a syscall.Stat_t type to a system.Stat_t type
+func fromStatT(s *syscall.Stat_t) (*StatT, error) {
+	return &StatT{
+		size: s.Size,
+		mode: uint32(s.Mode),
+		uid:  s.Uid,
+		gid:  s.Gid,
+		rdev: uint64(s.Rdev),
+		mtim: s.Mtimespec,
+	}, nil
+}