diff --git a/vendor/github.com/hashicorp/go-getter/.travis.yml b/vendor/github.com/hashicorp/go-getter/.travis.yml deleted file mode 100644 index 4fe9176aab8..00000000000 --- a/vendor/github.com/hashicorp/go-getter/.travis.yml +++ /dev/null @@ -1,24 +0,0 @@ -sudo: false - -addons: - apt: - sources: - - sourceline: 'ppa:git-core/ppa' - packages: - - git - -language: go - -os: - - linux - - osx - -go: - - "1.11.x" - -before_script: - - go build ./cmd/go-getter - -branches: - only: - - master diff --git a/vendor/github.com/hashicorp/go-getter/README.md b/vendor/github.com/hashicorp/go-getter/README.md index 3de23c70943..bbcd15de968 100644 --- a/vendor/github.com/hashicorp/go-getter/README.md +++ b/vendor/github.com/hashicorp/go-getter/README.md @@ -1,10 +1,10 @@ # go-getter -[![Build Status](http://img.shields.io/travis/hashicorp/go-getter.svg?style=flat-square)][travis] +[![CircleCI](https://circleci.com/gh/hashicorp/go-getter/tree/master.svg?style=svg)][circleci] [![Build status](https://ci.appveyor.com/api/projects/status/ulq3qr43n62croyq/branch/master?svg=true)][appveyor] [![Go Documentation](http://img.shields.io/badge/go-documentation-blue.svg?style=flat-square)][godocs] -[travis]: http://travis-ci.org/hashicorp/go-getter +[circleci]: https://circleci.com/gh/hashicorp/go-getter/tree/master [godocs]: http://godoc.org/github.com/hashicorp/go-getter [appveyor]: https://ci.appveyor.com/project/hashicorp/go-getter/branch/master @@ -356,3 +356,7 @@ In order to access to GCS, authentication credentials should be provided. More i - gcs::https://www.googleapis.com/storage/v1/bucket - gcs::https://www.googleapis.com/storage/v1/bucket/foo.zip - www.googleapis.com/storage/v1/bucket/foo + +#### GCS Testing + +The tests for `get_gcs.go` require you to have GCP credentials set in your environment. These credentials can have any level of permissions to any project, they just need to exist. This means setting `GOOGLE_APPLICATION_CREDENTIALS="~/path/to/credentials.json"` or `GOOGLE_CREDENTIALS="{stringified-credentials-json}"`. Due to this configuration, `get_gcs_test.go` will fail for external contributors in CircleCI. diff --git a/vendor/github.com/hashicorp/go-getter/client.go b/vendor/github.com/hashicorp/go-getter/client.go index 007a78ba7c8..38fb43b8f54 100644 --- a/vendor/github.com/hashicorp/go-getter/client.go +++ b/vendor/github.com/hashicorp/go-getter/client.go @@ -19,7 +19,7 @@ import ( // Using a client directly allows more fine-grained control over how downloading // is done, as well as customizing the protocols supported. type Client struct { - // Ctx for cancellation + // Ctx for cancellation Ctx context.Context // Src is the source URL to get. diff --git a/vendor/github.com/hashicorp/go-getter/get_git.go b/vendor/github.com/hashicorp/go-getter/get_git.go index bb1ec316d3e..1b9f4be8195 100644 --- a/vendor/github.com/hashicorp/go-getter/get_git.go +++ b/vendor/github.com/hashicorp/go-getter/get_git.go @@ -1,6 +1,7 @@ package getter import ( + "bytes" "context" "encoding/base64" "fmt" @@ -9,6 +10,7 @@ import ( "os" "os/exec" "path/filepath" + "regexp" "runtime" "strconv" "strings" @@ -24,6 +26,8 @@ type GitGetter struct { getter } +var defaultBranchRegexp = regexp.MustCompile(`\s->\sorigin/(.*)`) + func (g *GitGetter) ClientMode(_ *url.URL) (ClientMode, error) { return ClientModeDir, nil } @@ -182,10 +186,10 @@ func (g *GitGetter) update(ctx context.Context, dst, sshKeyFile, ref string, dep cmd.Dir = dst if getRunCommand(cmd) != nil { - // Not a branch, switch to master. This will also catch non-existent - // branches, in which case we want to switch to master and then - // checkout the proper branch later. - ref = "master" + // Not a branch, switch to default branch. This will also catch + // non-existent branches, in which case we want to switch to default + // and then checkout the proper branch later. + ref = findDefaultBranch(dst) } // We have to be on a branch to pull @@ -216,6 +220,22 @@ func (g *GitGetter) fetchSubmodules(ctx context.Context, dst, sshKeyFile string, return getRunCommand(cmd) } +// findDefaultBranch checks the repo's origin remote for its default branch +// (generally "master"). "master" is returned if an origin default branch +// can't be determined. +func findDefaultBranch(dst string) string { + var stdoutbuf bytes.Buffer + cmd := exec.Command("git", "branch", "-r", "--points-at", "refs/remotes/origin/HEAD") + cmd.Dir = dst + cmd.Stdout = &stdoutbuf + err := cmd.Run() + matches := defaultBranchRegexp.FindStringSubmatch(stdoutbuf.String()) + if err != nil || matches == nil { + return "master" + } + return matches[len(matches)-1] +} + // setupGitEnv sets up the environment for the given command. This is used to // pass configuration data to git and ssh and enables advanced cloning methods. func setupGitEnv(cmd *exec.Cmd, sshKeyFile string) { diff --git a/vendor/github.com/hashicorp/go-getter/get_http.go b/vendor/github.com/hashicorp/go-getter/get_http.go index 7c4541c6e95..9ffdba78a5f 100644 --- a/vendor/github.com/hashicorp/go-getter/get_http.go +++ b/vendor/github.com/hashicorp/go-getter/get_http.go @@ -9,7 +9,6 @@ import ( "net/url" "os" "path/filepath" - "strconv" "strings" safetemp "github.com/hashicorp/go-safetemp" @@ -88,7 +87,10 @@ func (g *HttpGetter) Get(dst string, u *url.URL) error { return err } - req.Header = g.Header + if g.Header != nil { + req.Header = g.Header + } + resp, err := g.Client.Do(req) if err != nil { return err @@ -128,6 +130,12 @@ func (g *HttpGetter) Get(dst string, u *url.URL) error { return g.getSubdir(ctx, dst, source, subDir) } +// GetFile fetches the file from src and stores it at dst. +// If the server supports Accept-Range, HttpGetter will attempt a range +// request. This means it is the caller's responsibility to ensure that an +// older version of the destination file does not exist, else it will be either +// falsely identified as being replaced, or corrupted with extra bytes +// appended. func (g *HttpGetter) GetFile(dst string, src *url.URL) error { ctx := g.Context() if g.Netrc { @@ -136,7 +144,6 @@ func (g *HttpGetter) GetFile(dst string, src *url.URL) error { return err } } - // Create all the parent directories if needed if err := os.MkdirAll(filepath.Dir(dst), 0755); err != nil { return err @@ -165,18 +172,17 @@ func (g *HttpGetter) GetFile(dst string, src *url.URL) error { req.Header = g.Header } headResp, err := g.Client.Do(req) - if err == nil && headResp != nil { + if err == nil { headResp.Body.Close() if headResp.StatusCode == 200 { // If the HEAD request succeeded, then attempt to set the range // query if we can. - if headResp.Header.Get("Accept-Ranges") == "bytes" { + if headResp.Header.Get("Accept-Ranges") == "bytes" && headResp.ContentLength >= 0 { if fi, err := f.Stat(); err == nil { - if _, err = f.Seek(0, os.SEEK_END); err == nil { - req.Header.Set("Range", fmt.Sprintf("bytes=%d-", fi.Size())) + if _, err = f.Seek(0, io.SeekEnd); err == nil { currentFileSize = fi.Size() - totalFileSize, _ := strconv.ParseInt(headResp.Header.Get("Content-Length"), 10, 64) - if currentFileSize >= totalFileSize { + req.Header.Set("Range", fmt.Sprintf("bytes=%d-", currentFileSize)) + if currentFileSize >= headResp.ContentLength { // file already present return nil } diff --git a/vendor/github.com/hashicorp/go-plugin/README.md b/vendor/github.com/hashicorp/go-plugin/README.md index fe305ad5989..46ee09fc0ca 100644 --- a/vendor/github.com/hashicorp/go-plugin/README.md +++ b/vendor/github.com/hashicorp/go-plugin/README.md @@ -141,11 +141,6 @@ This plugin system will give host processes a system for constraining versions. This is in addition to the protocol versioning already present which is more for larger underlying changes. -**Plugin fetching.** We will integrate with [go-getter](https://github.com/hashicorp/go-getter) -to support automatic download + install of plugins. Paired with cryptographically -secure plugins (above), we can make this a safe operation for an amazing -user experience. - ## What About Shared Libraries? When we started using plugins (late 2012, early 2013), plugins over RPC diff --git a/vendor/github.com/hashicorp/go-plugin/client.go b/vendor/github.com/hashicorp/go-plugin/client.go index f9e9effe7d8..780a3121da3 100644 --- a/vendor/github.com/hashicorp/go-plugin/client.go +++ b/vendor/github.com/hashicorp/go-plugin/client.go @@ -212,6 +212,12 @@ type ReattachConfig struct { Protocol Protocol Addr net.Addr Pid int + + // Test is set to true if this is reattaching to to a plugin in "test mode" + // (see ServeConfig.Test). In this mode, client.Kill will NOT kill the + // process and instead will rely on the plugin to terminate itself. This + // should not be used in non-test environments. + Test bool } // SecureConfig is used to configure a client to verify the integrity of an @@ -825,15 +831,21 @@ func (c *Client) reattach() (net.Addr, error) { c.exited = true }(p.Pid) - // Set the address and process + // Set the address and protocol c.address = c.config.Reattach.Addr - c.process = p c.protocol = c.config.Reattach.Protocol if c.protocol == "" { // Default the protocol to net/rpc for backwards compatibility c.protocol = ProtocolNetRPC } + // If we're in test mode, we do NOT set the process. This avoids the + // process being killed (the only purpose we have for c.process), since + // in test mode the process is responsible for exiting on its own. + if !c.config.Reattach.Test { + c.process = p + } + return c.address, nil } diff --git a/vendor/github.com/hashicorp/go-plugin/grpc_stdio.go b/vendor/github.com/hashicorp/go-plugin/grpc_stdio.go index 40977c0ac9e..6231a9fd625 100644 --- a/vendor/github.com/hashicorp/go-plugin/grpc_stdio.go +++ b/vendor/github.com/hashicorp/go-plugin/grpc_stdio.go @@ -7,12 +7,11 @@ import ( "io" empty "github.com/golang/protobuf/ptypes/empty" - "github.com/hashicorp/go-hclog" + hclog "github.com/hashicorp/go-hclog" + "github.com/hashicorp/go-plugin/internal/plugin" "google.golang.org/grpc" "google.golang.org/grpc/codes" "google.golang.org/grpc/status" - - "github.com/hashicorp/go-plugin/internal/plugin" ) // grpcStdioBuffer is the buffer size we try to fill when sending a chunk of @@ -101,10 +100,10 @@ func newGRPCStdioClient( // Connect immediately to the endpoint stdioClient, err := client.StreamStdio(ctx, &empty.Empty{}) - // If we get an Unavailable error, this means that the plugin isn't + // If we get an Unavailable or Unimplemented error, this means that the plugin isn't // updated and linking to the latest version of go-plugin that supports // this. We fall back to the previous behavior of just not syncing anything. - if status.Code(err) == codes.Unavailable { + if status.Code(err) == codes.Unavailable || status.Code(err) == codes.Unimplemented { log.Warn("stdio service not available, stdout/stderr syncing unavailable") stdioClient = nil err = nil @@ -135,6 +134,7 @@ func (c *grpcStdioClient) Run(stdout, stderr io.Writer) { if err == io.EOF || status.Code(err) == codes.Unavailable || status.Code(err) == codes.Canceled || + status.Code(err) == codes.Unimplemented || err == context.Canceled { c.log.Warn("received EOF, stopping recv loop", "err", err) return diff --git a/vendor/github.com/hashicorp/go-plugin/server.go b/vendor/github.com/hashicorp/go-plugin/server.go index e8b8c9778d3..002d6080d4f 100644 --- a/vendor/github.com/hashicorp/go-plugin/server.go +++ b/vendor/github.com/hashicorp/go-plugin/server.go @@ -1,11 +1,13 @@ package plugin import ( + "context" "crypto/tls" "crypto/x509" "encoding/base64" "errors" "fmt" + "io" "io/ioutil" "log" "net" @@ -83,6 +85,51 @@ type ServeConfig struct { // Logger is used to pass a logger into the server. If none is provided the // server will create a default logger. Logger hclog.Logger + + // Test, if non-nil, will put plugin serving into "test mode". This is + // meant to be used as part of `go test` within a plugin's codebase to + // launch the plugin in-process and output a ReattachConfig. + // + // This changes the behavior of the server in a number of ways to + // accomodate the expectation of running in-process: + // + // * The handshake cookie is not validated. + // * Stdout/stderr will receive plugin reads and writes + // * Connection information will not be sent to stdout + // + Test *ServeTestConfig +} + +// ServeTestConfig configures plugin serving for test mode. See ServeConfig.Test. +type ServeTestConfig struct { + // Context, if set, will force the plugin serving to end when cancelled. + // This is only a test configuration because the non-test configuration + // expects to take over the process and therefore end on an interrupt or + // kill signal. For tests, we need to kill the plugin serving routinely + // and this provides a way to do so. + // + // If you want to wait for the plugin process to close before moving on, + // you can wait on CloseCh. + Context context.Context + + // If this channel is non-nil, we will send the ReattachConfig via + // this channel. This can be encoded (via JSON recommended) to the + // plugin client to attach to this plugin. + ReattachConfigCh chan<- *ReattachConfig + + // CloseCh, if non-nil, will be closed when serving exits. This can be + // used along with Context to determine when the server is fully shut down. + // If this is not set, you can still use Context on its own, but note there + // may be a period of time between canceling the context and the plugin + // server being shut down. + CloseCh chan<- struct{} + + // SyncStdio, if true, will enable the client side "SyncStdout/Stderr" + // functionality to work. This defaults to false because the implementation + // of making this work within test environments is particularly messy + // and SyncStdio functionality is fairly rare, so we default to the simple + // scenario. + SyncStdio bool } // protocolVersion determines the protocol version and plugin set to be used by @@ -167,26 +214,46 @@ func protocolVersion(opts *ServeConfig) (int, Protocol, PluginSet) { // Serve serves the plugins given by ServeConfig. // // Serve doesn't return until the plugin is done being executed. Any -// errors will be outputted to os.Stderr. +// fixable errors will be output to os.Stderr and the process will +// exit with a status code of 1. Serve will panic for unexpected +// conditions where a user's fix is unknown. // // This is the method that plugins should call in their main() functions. func Serve(opts *ServeConfig) { - // Validate the handshake config - if opts.MagicCookieKey == "" || opts.MagicCookieValue == "" { - fmt.Fprintf(os.Stderr, - "Misconfigured ServeConfig given to serve this plugin: no magic cookie\n"+ - "key or value was set. Please notify the plugin author and report\n"+ - "this as a bug.\n") - os.Exit(1) - } + exitCode := -1 + // We use this to trigger an `os.Exit` so that we can execute our other + // deferred functions. In test mode, we just output the err to stderr + // and return. + defer func() { + if opts.Test == nil && exitCode >= 0 { + os.Exit(exitCode) + } - // First check the cookie - if os.Getenv(opts.MagicCookieKey) != opts.MagicCookieValue { - fmt.Fprintf(os.Stderr, - "This binary is a plugin. These are not meant to be executed directly.\n"+ - "Please execute the program that consumes these plugins, which will\n"+ - "load any plugins automatically\n") - os.Exit(1) + if opts.Test != nil && opts.Test.CloseCh != nil { + close(opts.Test.CloseCh) + } + }() + + if opts.Test == nil { + // Validate the handshake config + if opts.MagicCookieKey == "" || opts.MagicCookieValue == "" { + fmt.Fprintf(os.Stderr, + "Misconfigured ServeConfig given to serve this plugin: no magic cookie\n"+ + "key or value was set. Please notify the plugin author and report\n"+ + "this as a bug.\n") + exitCode = 1 + return + } + + // First check the cookie + if os.Getenv(opts.MagicCookieKey) != opts.MagicCookieValue { + fmt.Fprintf(os.Stderr, + "This binary is a plugin. These are not meant to be executed directly.\n"+ + "Please execute the program that consumes these plugins, which will\n"+ + "load any plugins automatically\n") + exitCode = 1 + return + } } // negotiate the version and plugins @@ -206,19 +273,6 @@ func Serve(opts *ServeConfig) { }) } - // Create our new stdout, stderr files. These will override our built-in - // stdout/stderr so that it works across the stream boundary. - stdout_r, stdout_w, err := os.Pipe() - if err != nil { - fmt.Fprintf(os.Stderr, "Error preparing plugin: %s\n", err) - os.Exit(1) - } - stderr_r, stderr_w, err := os.Pipe() - if err != nil { - fmt.Fprintf(os.Stderr, "Error preparing plugin: %s\n", err) - os.Exit(1) - } - // Register a listener so we can accept a connection listener, err := serverListener() if err != nil { @@ -279,6 +333,33 @@ func Serve(opts *ServeConfig) { // Create the channel to tell us when we're done doneCh := make(chan struct{}) + // Create our new stdout, stderr files. These will override our built-in + // stdout/stderr so that it works across the stream boundary. + var stdout_r, stderr_r io.Reader + stdout_r, stdout_w, err := os.Pipe() + if err != nil { + fmt.Fprintf(os.Stderr, "Error preparing plugin: %s\n", err) + os.Exit(1) + } + stderr_r, stderr_w, err := os.Pipe() + if err != nil { + fmt.Fprintf(os.Stderr, "Error preparing plugin: %s\n", err) + os.Exit(1) + } + + // If we're in test mode, we tee off the reader and write the data + // as-is to our normal Stdout and Stderr so that they continue working + // while stdio works. This is because in test mode, we assume we're running + // in `go test` or some equivalent and we want output to go to standard + // locations. + if opts.Test != nil { + // TODO(mitchellh): This isn't super ideal because a TeeReader + // only works if the reader side is actively read. If we never + // connect via a plugin client, the output still gets swallowed. + stdout_r = io.TeeReader(stdout_r, os.Stdout) + stderr_r = io.TeeReader(stderr_r, os.Stderr) + } + // Build the server type var server ServerProtocol switch protoType { @@ -321,35 +402,96 @@ func Serve(opts *ServeConfig) { logger.Debug("plugin address", "network", listener.Addr().Network(), "address", listener.Addr().String()) - // Output the address and service name to stdout so that the client can bring it up. - fmt.Printf("%d|%d|%s|%s|%s|%s\n", - CoreProtocolVersion, - protoVersion, - listener.Addr().Network(), - listener.Addr().String(), - protoType, - serverCert) - os.Stdout.Sync() - - // Eat the interrupts - ch := make(chan os.Signal, 1) - signal.Notify(ch, os.Interrupt) - go func() { - count := 0 - for { - <-ch - count++ - logger.Trace("plugin received interrupt signal, ignoring", "count", count) + // Output the address and service name to stdout so that the client can + // bring it up. In test mode, we don't do this because clients will + // attach via a reattach config. + if opts.Test == nil { + fmt.Printf("%d|%d|%s|%s|%s|%s\n", + CoreProtocolVersion, + protoVersion, + listener.Addr().Network(), + listener.Addr().String(), + protoType, + serverCert) + os.Stdout.Sync() + } else if ch := opts.Test.ReattachConfigCh; ch != nil { + // Send back the reattach config that can be used. This isn't + // quite ready if they connect immediately but the client should + // retry a few times. + ch <- &ReattachConfig{ + Protocol: protoType, + Addr: listener.Addr(), + Pid: os.Getpid(), + Test: true, } - }() + } + + // Eat the interrupts. In test mode we disable this so that go test + // can be cancelled properly. + if opts.Test == nil { + ch := make(chan os.Signal, 1) + signal.Notify(ch, os.Interrupt) + go func() { + count := 0 + for { + <-ch + count++ + logger.Trace("plugin received interrupt signal, ignoring", "count", count) + } + }() + } - // Set our new out, err - os.Stdout = stdout_w - os.Stderr = stderr_w + // Set our stdout, stderr to the stdio stream that clients can retrieve + // using ClientConfig.SyncStdout/err. We only do this for non-test mode + // or if the test mode explicitly requests it. + // + // In test mode, we use a multiwriter so that the data continues going + // to the normal stdout/stderr so output can show up in test logs. We + // also send to the stdio stream so that clients can continue working + // if they depend on that. + if opts.Test == nil || opts.Test.SyncStdio { + if opts.Test != nil { + // In test mode we need to maintain the original values so we can + // reset it. + defer func(out, err *os.File) { + os.Stdout = out + os.Stderr = err + }(os.Stdout, os.Stderr) + } + os.Stdout = stdout_w + os.Stderr = stderr_w + } // Accept connections and wait for completion go server.Serve(listener) - <-doneCh + + ctx := context.Background() + if opts.Test != nil && opts.Test.Context != nil { + ctx = opts.Test.Context + } + select { + case <-ctx.Done(): + // Cancellation. We can stop the server by closing the listener. + // This isn't graceful at all but this is currently only used by + // tests and its our only way to stop. + listener.Close() + + // If this is a grpc server, then we also ask the server itself to + // end which will kill all connections. There isn't an easy way to do + // this for net/rpc currently but net/rpc is more and more unused. + if s, ok := server.(*GRPCServer); ok { + s.Stop() + } + + // Wait for the server itself to shut down + <-doneCh + + case <-doneCh: + // Note that given the documentation of Serve we should probably be + // setting exitCode = 0 and using os.Exit here. That's how it used to + // work before extracting this library. However, for years we've done + // this so we'll keep this functionality. + } } func serverListener() (net.Listener, error) { @@ -388,7 +530,7 @@ func serverListener_tcp() (net.Listener, error) { } if minPort > maxPort { - return nil, fmt.Errorf("ENV_MIN_PORT value of %d is greater than PLUGIN_MAX_PORT value of %d", minPort, maxPort) + return nil, fmt.Errorf("PLUGIN_MIN_PORT value of %d is greater than PLUGIN_MAX_PORT value of %d", minPort, maxPort) } for port := minPort; port <= maxPort; port++ { diff --git a/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/load.go b/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/load.go index faa93ed6fb0..a070f76e048 100644 --- a/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/load.go +++ b/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/load.go @@ -52,12 +52,12 @@ func (m *Module) init(diags Diagnostics) { // case so callers can easily recognize it. for _, r := range m.ManagedResources { if _, exists := m.RequiredProviders[r.Provider.Name]; !exists { - m.RequiredProviders[r.Provider.Name] = []string{} + m.RequiredProviders[r.Provider.Name] = &ProviderRequirement{} } } for _, r := range m.DataResources { if _, exists := m.RequiredProviders[r.Provider.Name]; !exists { - m.RequiredProviders[r.Provider.Name] = []string{} + m.RequiredProviders[r.Provider.Name] = &ProviderRequirement{} } } diff --git a/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/load_hcl.go b/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/load_hcl.go index 9cb3aeef1bf..f83ac872673 100644 --- a/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/load_hcl.go +++ b/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/load_hcl.go @@ -7,8 +7,8 @@ import ( "github.com/hashicorp/hcl/v2/hclsyntax" - "github.com/hashicorp/hcl/v2/gohcl" "github.com/hashicorp/hcl/v2" + "github.com/hashicorp/hcl/v2/gohcl" "github.com/hashicorp/hcl/v2/hclparse" ctyjson "github.com/zclconf/go-cty/cty/json" ) @@ -51,18 +51,17 @@ func loadModule(dir string) (*Module, Diagnostics) { } } - for _, block := range content.Blocks { - // Our schema only allows required_providers here, so we - // assume that we'll only get that block type. - attrs, attrDiags := block.Body.JustAttributes() - diags = append(diags, attrDiags...) - - for name, attr := range attrs { - var version string - valDiags := gohcl.DecodeExpression(attr.Expr, nil, &version) - diags = append(diags, valDiags...) - if !valDiags.HasErrors() { - mod.RequiredProviders[name] = append(mod.RequiredProviders[name], version) + for _, innerBlock := range content.Blocks { + switch innerBlock.Type { + case "required_providers": + reqs, reqsDiags := decodeRequiredProvidersBlock(innerBlock) + diags = append(diags, reqsDiags...) + for name, req := range reqs { + if _, exists := mod.RequiredProviders[name]; !exists { + mod.RequiredProviders[name] = req + } else { + mod.RequiredProviders[name].VersionConstraints = append(mod.RequiredProviders[name].VersionConstraints, req.VersionConstraints...) + } } } } @@ -178,22 +177,20 @@ func loadModule(dir string) (*Module, Diagnostics) { diags = append(diags, contentDiags...) name := block.Labels[0] - + // Even if there isn't an explicit version required, we still + // need an entry in our map to signal the unversioned dependency. + if _, exists := mod.RequiredProviders[name]; !exists { + mod.RequiredProviders[name] = &ProviderRequirement{} + } if attr, defined := content.Attributes["version"]; defined { var version string valDiags := gohcl.DecodeExpression(attr.Expr, nil, &version) diags = append(diags, valDiags...) if !valDiags.HasErrors() { - mod.RequiredProviders[name] = append(mod.RequiredProviders[name], version) + mod.RequiredProviders[name].VersionConstraints = append(mod.RequiredProviders[name].VersionConstraints, version) } } - // Even if there wasn't an explicit version required, we still - // need an entry in our map to signal the unversioned dependency. - if _, exists := mod.RequiredProviders[name]; !exists { - mod.RequiredProviders[name] = []string{} - } - case "resource", "data": content, _, contentDiags := block.Body.PartialContent(resourceSchema) diff --git a/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/load_legacy.go b/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/load_legacy.go index 86ffdf11dd3..c79b033b6bd 100644 --- a/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/load_legacy.go +++ b/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/load_legacy.go @@ -267,17 +267,15 @@ func loadModuleLegacyHCL(dir string) (*Module, Diagnostics) { if err != nil { return nil, diagnosticsErrorf("invalid provider block at %s: %s", item.Pos(), err) } - - if block.Version != "" { - mod.RequiredProviders[name] = append(mod.RequiredProviders[name], block.Version) - } - // Even if there wasn't an explicit version required, we still // need an entry in our map to signal the unversioned dependency. if _, exists := mod.RequiredProviders[name]; !exists { - mod.RequiredProviders[name] = []string{} + mod.RequiredProviders[name] = &ProviderRequirement{} } + if block.Version != "" { + mod.RequiredProviders[name].VersionConstraints = append(mod.RequiredProviders[name].VersionConstraints, block.Version) + } } } } diff --git a/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/module.go b/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/module.go index 65ddb230735..63027d18412 100644 --- a/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/module.go +++ b/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/module.go @@ -9,8 +9,8 @@ type Module struct { Variables map[string]*Variable `json:"variables"` Outputs map[string]*Output `json:"outputs"` - RequiredCore []string `json:"required_core,omitempty"` - RequiredProviders map[string][]string `json:"required_providers"` + RequiredCore []string `json:"required_core,omitempty"` + RequiredProviders map[string]*ProviderRequirement `json:"required_providers"` ManagedResources map[string]*Resource `json:"managed_resources"` DataResources map[string]*Resource `json:"data_resources"` @@ -27,7 +27,7 @@ func newModule(path string) *Module { Path: path, Variables: make(map[string]*Variable), Outputs: make(map[string]*Output), - RequiredProviders: make(map[string][]string), + RequiredProviders: make(map[string]*ProviderRequirement), ManagedResources: make(map[string]*Resource), DataResources: make(map[string]*Resource), ModuleCalls: make(map[string]*ModuleCall), diff --git a/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/provider_ref.go b/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/provider_ref.go index d924837785f..157c8c2c154 100644 --- a/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/provider_ref.go +++ b/vendor/github.com/hashicorp/terraform-config-inspect/tfconfig/provider_ref.go @@ -1,5 +1,11 @@ package tfconfig +import ( + "github.com/hashicorp/hcl/v2" + "github.com/hashicorp/hcl/v2/gohcl" + "github.com/zclconf/go-cty/cty/gocty" +) + // ProviderRef is a reference to a provider configuration within a module. // It represents the contents of a "provider" argument in a resource, or // a value in the "providers" map for a module call. @@ -7,3 +13,73 @@ type ProviderRef struct { Name string `json:"name"` Alias string `json:"alias,omitempty"` // Empty if the default provider configuration is referenced } + +type ProviderRequirement struct { + Source string `json:"source,omitempty"` + VersionConstraints []string `json:"version_constraints,omitempty"` +} + +func decodeRequiredProvidersBlock(block *hcl.Block) (map[string]*ProviderRequirement, hcl.Diagnostics) { + attrs, diags := block.Body.JustAttributes() + reqs := make(map[string]*ProviderRequirement) + for name, attr := range attrs { + expr, err := attr.Expr.Value(nil) + if err != nil { + diags = append(diags, err...) + } + + switch { + case expr.Type().IsPrimitiveType(): + var version string + valDiags := gohcl.DecodeExpression(attr.Expr, nil, &version) + diags = append(diags, valDiags...) + if !valDiags.HasErrors() { + reqs[name] = &ProviderRequirement{ + VersionConstraints: []string{version}, + } + } + + case expr.Type().IsObjectType(): + var pr ProviderRequirement + if expr.Type().HasAttribute("version") { + var version string + err := gocty.FromCtyValue(expr.GetAttr("version"), &version) + if err == nil { + pr.VersionConstraints = append(pr.VersionConstraints, version) + } else { + diags = append(diags, &hcl.Diagnostic{ + Severity: hcl.DiagError, + Summary: "Unsuitable value type", + Detail: "Unsuitable value: string required", + Subject: attr.Expr.Range().Ptr(), + }) + } + } + if expr.Type().HasAttribute("source") { + var source string + err := gocty.FromCtyValue(expr.GetAttr("source"), &source) + if err == nil { + pr.Source = source + } else { + diags = append(diags, &hcl.Diagnostic{ + Severity: hcl.DiagError, + Summary: "Unsuitable value type", + Detail: "Unsuitable value: string required", + Subject: attr.Expr.Range().Ptr(), + }) + } + } + reqs[name] = &pr + + default: + diags = append(diags, &hcl.Diagnostic{ + Severity: hcl.DiagError, + Summary: "Unsuitable value type", + Detail: "Unsuitable value: string required", + Subject: attr.Expr.Range().Ptr(), + }) + } + } + + return reqs, diags +} diff --git a/vendor/golang.org/x/crypto/chacha20/chacha_arm64.go b/vendor/golang.org/x/crypto/chacha20/chacha_arm64.go new file mode 100644 index 00000000000..b799e440b4a --- /dev/null +++ b/vendor/golang.org/x/crypto/chacha20/chacha_arm64.go @@ -0,0 +1,16 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build go1.11,!gccgo,!purego + +package chacha20 + +const bufSize = 256 + +//go:noescape +func xorKeyStreamVX(dst, src []byte, key *[8]uint32, nonce *[3]uint32, counter *uint32) + +func (c *Cipher) xorKeyStreamBlocks(dst, src []byte) { + xorKeyStreamVX(dst, src, &c.key, &c.nonce, &c.counter) +} diff --git a/vendor/golang.org/x/crypto/internal/chacha20/asm_arm64.s b/vendor/golang.org/x/crypto/chacha20/chacha_arm64.s similarity index 99% rename from vendor/golang.org/x/crypto/internal/chacha20/asm_arm64.s rename to vendor/golang.org/x/crypto/chacha20/chacha_arm64.s index b3a16ef751a..891481539a1 100644 --- a/vendor/golang.org/x/crypto/internal/chacha20/asm_arm64.s +++ b/vendor/golang.org/x/crypto/chacha20/chacha_arm64.s @@ -2,8 +2,7 @@ // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. -// +build go1.11 -// +build !gccgo,!appengine +// +build go1.11,!gccgo,!purego #include "textflag.h" diff --git a/vendor/golang.org/x/crypto/chacha20/chacha_generic.go b/vendor/golang.org/x/crypto/chacha20/chacha_generic.go new file mode 100644 index 00000000000..a2ecf5c325b --- /dev/null +++ b/vendor/golang.org/x/crypto/chacha20/chacha_generic.go @@ -0,0 +1,398 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// Package chacha20 implements the ChaCha20 and XChaCha20 encryption algorithms +// as specified in RFC 8439 and draft-irtf-cfrg-xchacha-01. +package chacha20 + +import ( + "crypto/cipher" + "encoding/binary" + "errors" + "math/bits" + + "golang.org/x/crypto/internal/subtle" +) + +const ( + // KeySize is the size of the key used by this cipher, in bytes. + KeySize = 32 + + // NonceSize is the size of the nonce used with the standard variant of this + // cipher, in bytes. + // + // Note that this is too short to be safely generated at random if the same + // key is reused more than 2³² times. + NonceSize = 12 + + // NonceSizeX is the size of the nonce used with the XChaCha20 variant of + // this cipher, in bytes. + NonceSizeX = 24 +) + +// Cipher is a stateful instance of ChaCha20 or XChaCha20 using a particular key +// and nonce. A *Cipher implements the cipher.Stream interface. +type Cipher struct { + // The ChaCha20 state is 16 words: 4 constant, 8 of key, 1 of counter + // (incremented after each block), and 3 of nonce. + key [8]uint32 + counter uint32 + nonce [3]uint32 + + // The last len bytes of buf are leftover key stream bytes from the previous + // XORKeyStream invocation. The size of buf depends on how many blocks are + // computed at a time by xorKeyStreamBlocks. + buf [bufSize]byte + len int + + // overflow is set when the counter overflowed, no more blocks can be + // generated, and the next XORKeyStream call should panic. + overflow bool + + // The counter-independent results of the first round are cached after they + // are computed the first time. + precompDone bool + p1, p5, p9, p13 uint32 + p2, p6, p10, p14 uint32 + p3, p7, p11, p15 uint32 +} + +var _ cipher.Stream = (*Cipher)(nil) + +// NewUnauthenticatedCipher creates a new ChaCha20 stream cipher with the given +// 32 bytes key and a 12 or 24 bytes nonce. If a nonce of 24 bytes is provided, +// the XChaCha20 construction will be used. It returns an error if key or nonce +// have any other length. +// +// Note that ChaCha20, like all stream ciphers, is not authenticated and allows +// attackers to silently tamper with the plaintext. For this reason, it is more +// appropriate as a building block than as a standalone encryption mechanism. +// Instead, consider using package golang.org/x/crypto/chacha20poly1305. +func NewUnauthenticatedCipher(key, nonce []byte) (*Cipher, error) { + // This function is split into a wrapper so that the Cipher allocation will + // be inlined, and depending on how the caller uses the return value, won't + // escape to the heap. + c := &Cipher{} + return newUnauthenticatedCipher(c, key, nonce) +} + +func newUnauthenticatedCipher(c *Cipher, key, nonce []byte) (*Cipher, error) { + if len(key) != KeySize { + return nil, errors.New("chacha20: wrong key size") + } + if len(nonce) == NonceSizeX { + // XChaCha20 uses the ChaCha20 core to mix 16 bytes of the nonce into a + // derived key, allowing it to operate on a nonce of 24 bytes. See + // draft-irtf-cfrg-xchacha-01, Section 2.3. + key, _ = HChaCha20(key, nonce[0:16]) + cNonce := make([]byte, NonceSize) + copy(cNonce[4:12], nonce[16:24]) + nonce = cNonce + } else if len(nonce) != NonceSize { + return nil, errors.New("chacha20: wrong nonce size") + } + + key, nonce = key[:KeySize], nonce[:NonceSize] // bounds check elimination hint + c.key = [8]uint32{ + binary.LittleEndian.Uint32(key[0:4]), + binary.LittleEndian.Uint32(key[4:8]), + binary.LittleEndian.Uint32(key[8:12]), + binary.LittleEndian.Uint32(key[12:16]), + binary.LittleEndian.Uint32(key[16:20]), + binary.LittleEndian.Uint32(key[20:24]), + binary.LittleEndian.Uint32(key[24:28]), + binary.LittleEndian.Uint32(key[28:32]), + } + c.nonce = [3]uint32{ + binary.LittleEndian.Uint32(nonce[0:4]), + binary.LittleEndian.Uint32(nonce[4:8]), + binary.LittleEndian.Uint32(nonce[8:12]), + } + return c, nil +} + +// The constant first 4 words of the ChaCha20 state. +const ( + j0 uint32 = 0x61707865 // expa + j1 uint32 = 0x3320646e // nd 3 + j2 uint32 = 0x79622d32 // 2-by + j3 uint32 = 0x6b206574 // te k +) + +const blockSize = 64 + +// quarterRound is the core of ChaCha20. It shuffles the bits of 4 state words. +// It's executed 4 times for each of the 20 ChaCha20 rounds, operating on all 16 +// words each round, in columnar or diagonal groups of 4 at a time. +func quarterRound(a, b, c, d uint32) (uint32, uint32, uint32, uint32) { + a += b + d ^= a + d = bits.RotateLeft32(d, 16) + c += d + b ^= c + b = bits.RotateLeft32(b, 12) + a += b + d ^= a + d = bits.RotateLeft32(d, 8) + c += d + b ^= c + b = bits.RotateLeft32(b, 7) + return a, b, c, d +} + +// SetCounter sets the Cipher counter. The next invocation of XORKeyStream will +// behave as if (64 * counter) bytes had been encrypted so far. +// +// To prevent accidental counter reuse, SetCounter panics if counter is less +// than the current value. +// +// Note that the execution time of XORKeyStream is not independent of the +// counter value. +func (s *Cipher) SetCounter(counter uint32) { + // Internally, s may buffer multiple blocks, which complicates this + // implementation slightly. When checking whether the counter has rolled + // back, we must use both s.counter and s.len to determine how many blocks + // we have already output. + outputCounter := s.counter - uint32(s.len)/blockSize + if s.overflow || counter < outputCounter { + panic("chacha20: SetCounter attempted to rollback counter") + } + + // In the general case, we set the new counter value and reset s.len to 0, + // causing the next call to XORKeyStream to refill the buffer. However, if + // we're advancing within the existing buffer, we can save work by simply + // setting s.len. + if counter < s.counter { + s.len = int(s.counter-counter) * blockSize + } else { + s.counter = counter + s.len = 0 + } +} + +// XORKeyStream XORs each byte in the given slice with a byte from the +// cipher's key stream. Dst and src must overlap entirely or not at all. +// +// If len(dst) < len(src), XORKeyStream will panic. It is acceptable +// to pass a dst bigger than src, and in that case, XORKeyStream will +// only update dst[:len(src)] and will not touch the rest of dst. +// +// Multiple calls to XORKeyStream behave as if the concatenation of +// the src buffers was passed in a single run. That is, Cipher +// maintains state and does not reset at each XORKeyStream call. +func (s *Cipher) XORKeyStream(dst, src []byte) { + if len(src) == 0 { + return + } + if len(dst) < len(src) { + panic("chacha20: output smaller than input") + } + dst = dst[:len(src)] + if subtle.InexactOverlap(dst, src) { + panic("chacha20: invalid buffer overlap") + } + + // First, drain any remaining key stream from a previous XORKeyStream. + if s.len != 0 { + keyStream := s.buf[bufSize-s.len:] + if len(src) < len(keyStream) { + keyStream = keyStream[:len(src)] + } + _ = src[len(keyStream)-1] // bounds check elimination hint + for i, b := range keyStream { + dst[i] = src[i] ^ b + } + s.len -= len(keyStream) + dst, src = dst[len(keyStream):], src[len(keyStream):] + } + if len(src) == 0 { + return + } + + // If we'd need to let the counter overflow and keep generating output, + // panic immediately. If instead we'd only reach the last block, remember + // not to generate any more output after the buffer is drained. + numBlocks := (uint64(len(src)) + blockSize - 1) / blockSize + if s.overflow || uint64(s.counter)+numBlocks > 1<<32 { + panic("chacha20: counter overflow") + } else if uint64(s.counter)+numBlocks == 1<<32 { + s.overflow = true + } + + // xorKeyStreamBlocks implementations expect input lengths that are a + // multiple of bufSize. Platform-specific ones process multiple blocks at a + // time, so have bufSizes that are a multiple of blockSize. + + full := len(src) - len(src)%bufSize + if full > 0 { + s.xorKeyStreamBlocks(dst[:full], src[:full]) + } + dst, src = dst[full:], src[full:] + + // If using a multi-block xorKeyStreamBlocks would overflow, use the generic + // one that does one block at a time. + const blocksPerBuf = bufSize / blockSize + if uint64(s.counter)+blocksPerBuf > 1<<32 { + s.buf = [bufSize]byte{} + numBlocks := (len(src) + blockSize - 1) / blockSize + buf := s.buf[bufSize-numBlocks*blockSize:] + copy(buf, src) + s.xorKeyStreamBlocksGeneric(buf, buf) + s.len = len(buf) - copy(dst, buf) + return + } + + // If we have a partial (multi-)block, pad it for xorKeyStreamBlocks, and + // keep the leftover keystream for the next XORKeyStream invocation. + if len(src) > 0 { + s.buf = [bufSize]byte{} + copy(s.buf[:], src) + s.xorKeyStreamBlocks(s.buf[:], s.buf[:]) + s.len = bufSize - copy(dst, s.buf[:]) + } +} + +func (s *Cipher) xorKeyStreamBlocksGeneric(dst, src []byte) { + if len(dst) != len(src) || len(dst)%blockSize != 0 { + panic("chacha20: internal error: wrong dst and/or src length") + } + + // To generate each block of key stream, the initial cipher state + // (represented below) is passed through 20 rounds of shuffling, + // alternatively applying quarterRounds by columns (like 1, 5, 9, 13) + // or by diagonals (like 1, 6, 11, 12). + // + // 0:cccccccc 1:cccccccc 2:cccccccc 3:cccccccc + // 4:kkkkkkkk 5:kkkkkkkk 6:kkkkkkkk 7:kkkkkkkk + // 8:kkkkkkkk 9:kkkkkkkk 10:kkkkkkkk 11:kkkkkkkk + // 12:bbbbbbbb 13:nnnnnnnn 14:nnnnnnnn 15:nnnnnnnn + // + // c=constant k=key b=blockcount n=nonce + var ( + c0, c1, c2, c3 = j0, j1, j2, j3 + c4, c5, c6, c7 = s.key[0], s.key[1], s.key[2], s.key[3] + c8, c9, c10, c11 = s.key[4], s.key[5], s.key[6], s.key[7] + _, c13, c14, c15 = s.counter, s.nonce[0], s.nonce[1], s.nonce[2] + ) + + // Three quarters of the first round don't depend on the counter, so we can + // calculate them here, and reuse them for multiple blocks in the loop, and + // for future XORKeyStream invocations. + if !s.precompDone { + s.p1, s.p5, s.p9, s.p13 = quarterRound(c1, c5, c9, c13) + s.p2, s.p6, s.p10, s.p14 = quarterRound(c2, c6, c10, c14) + s.p3, s.p7, s.p11, s.p15 = quarterRound(c3, c7, c11, c15) + s.precompDone = true + } + + // A condition of len(src) > 0 would be sufficient, but this also + // acts as a bounds check elimination hint. + for len(src) >= 64 && len(dst) >= 64 { + // The remainder of the first column round. + fcr0, fcr4, fcr8, fcr12 := quarterRound(c0, c4, c8, s.counter) + + // The second diagonal round. + x0, x5, x10, x15 := quarterRound(fcr0, s.p5, s.p10, s.p15) + x1, x6, x11, x12 := quarterRound(s.p1, s.p6, s.p11, fcr12) + x2, x7, x8, x13 := quarterRound(s.p2, s.p7, fcr8, s.p13) + x3, x4, x9, x14 := quarterRound(s.p3, fcr4, s.p9, s.p14) + + // The remaining 18 rounds. + for i := 0; i < 9; i++ { + // Column round. + x0, x4, x8, x12 = quarterRound(x0, x4, x8, x12) + x1, x5, x9, x13 = quarterRound(x1, x5, x9, x13) + x2, x6, x10, x14 = quarterRound(x2, x6, x10, x14) + x3, x7, x11, x15 = quarterRound(x3, x7, x11, x15) + + // Diagonal round. + x0, x5, x10, x15 = quarterRound(x0, x5, x10, x15) + x1, x6, x11, x12 = quarterRound(x1, x6, x11, x12) + x2, x7, x8, x13 = quarterRound(x2, x7, x8, x13) + x3, x4, x9, x14 = quarterRound(x3, x4, x9, x14) + } + + // Add back the initial state to generate the key stream, then + // XOR the key stream with the source and write out the result. + addXor(dst[0:4], src[0:4], x0, c0) + addXor(dst[4:8], src[4:8], x1, c1) + addXor(dst[8:12], src[8:12], x2, c2) + addXor(dst[12:16], src[12:16], x3, c3) + addXor(dst[16:20], src[16:20], x4, c4) + addXor(dst[20:24], src[20:24], x5, c5) + addXor(dst[24:28], src[24:28], x6, c6) + addXor(dst[28:32], src[28:32], x7, c7) + addXor(dst[32:36], src[32:36], x8, c8) + addXor(dst[36:40], src[36:40], x9, c9) + addXor(dst[40:44], src[40:44], x10, c10) + addXor(dst[44:48], src[44:48], x11, c11) + addXor(dst[48:52], src[48:52], x12, s.counter) + addXor(dst[52:56], src[52:56], x13, c13) + addXor(dst[56:60], src[56:60], x14, c14) + addXor(dst[60:64], src[60:64], x15, c15) + + s.counter += 1 + + src, dst = src[blockSize:], dst[blockSize:] + } +} + +// HChaCha20 uses the ChaCha20 core to generate a derived key from a 32 bytes +// key and a 16 bytes nonce. It returns an error if key or nonce have any other +// length. It is used as part of the XChaCha20 construction. +func HChaCha20(key, nonce []byte) ([]byte, error) { + // This function is split into a wrapper so that the slice allocation will + // be inlined, and depending on how the caller uses the return value, won't + // escape to the heap. + out := make([]byte, 32) + return hChaCha20(out, key, nonce) +} + +func hChaCha20(out, key, nonce []byte) ([]byte, error) { + if len(key) != KeySize { + return nil, errors.New("chacha20: wrong HChaCha20 key size") + } + if len(nonce) != 16 { + return nil, errors.New("chacha20: wrong HChaCha20 nonce size") + } + + x0, x1, x2, x3 := j0, j1, j2, j3 + x4 := binary.LittleEndian.Uint32(key[0:4]) + x5 := binary.LittleEndian.Uint32(key[4:8]) + x6 := binary.LittleEndian.Uint32(key[8:12]) + x7 := binary.LittleEndian.Uint32(key[12:16]) + x8 := binary.LittleEndian.Uint32(key[16:20]) + x9 := binary.LittleEndian.Uint32(key[20:24]) + x10 := binary.LittleEndian.Uint32(key[24:28]) + x11 := binary.LittleEndian.Uint32(key[28:32]) + x12 := binary.LittleEndian.Uint32(nonce[0:4]) + x13 := binary.LittleEndian.Uint32(nonce[4:8]) + x14 := binary.LittleEndian.Uint32(nonce[8:12]) + x15 := binary.LittleEndian.Uint32(nonce[12:16]) + + for i := 0; i < 10; i++ { + // Diagonal round. + x0, x4, x8, x12 = quarterRound(x0, x4, x8, x12) + x1, x5, x9, x13 = quarterRound(x1, x5, x9, x13) + x2, x6, x10, x14 = quarterRound(x2, x6, x10, x14) + x3, x7, x11, x15 = quarterRound(x3, x7, x11, x15) + + // Column round. + x0, x5, x10, x15 = quarterRound(x0, x5, x10, x15) + x1, x6, x11, x12 = quarterRound(x1, x6, x11, x12) + x2, x7, x8, x13 = quarterRound(x2, x7, x8, x13) + x3, x4, x9, x14 = quarterRound(x3, x4, x9, x14) + } + + _ = out[31] // bounds check elimination hint + binary.LittleEndian.PutUint32(out[0:4], x0) + binary.LittleEndian.PutUint32(out[4:8], x1) + binary.LittleEndian.PutUint32(out[8:12], x2) + binary.LittleEndian.PutUint32(out[12:16], x3) + binary.LittleEndian.PutUint32(out[16:20], x12) + binary.LittleEndian.PutUint32(out[20:24], x13) + binary.LittleEndian.PutUint32(out[24:28], x14) + binary.LittleEndian.PutUint32(out[28:32], x15) + return out, nil +} diff --git a/vendor/golang.org/x/crypto/internal/chacha20/chacha_noasm.go b/vendor/golang.org/x/crypto/chacha20/chacha_noasm.go similarity index 50% rename from vendor/golang.org/x/crypto/internal/chacha20/chacha_noasm.go rename to vendor/golang.org/x/crypto/chacha20/chacha_noasm.go index fc2682528b5..4635307b8f2 100644 --- a/vendor/golang.org/x/crypto/internal/chacha20/chacha_noasm.go +++ b/vendor/golang.org/x/crypto/chacha20/chacha_noasm.go @@ -2,15 +2,12 @@ // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. -// +build !arm64,!s390x,!ppc64le arm64,!go1.11 gccgo appengine +// +build !arm64,!s390x,!ppc64le arm64,!go1.11 gccgo purego package chacha20 -const ( - bufSize = 64 - haveAsm = false -) +const bufSize = blockSize -func (*Cipher) xorKeyStreamAsm(dst, src []byte) { - panic("not implemented") +func (s *Cipher) xorKeyStreamBlocks(dst, src []byte) { + s.xorKeyStreamBlocksGeneric(dst, src) } diff --git a/vendor/golang.org/x/crypto/chacha20/chacha_ppc64le.go b/vendor/golang.org/x/crypto/chacha20/chacha_ppc64le.go new file mode 100644 index 00000000000..b7993303415 --- /dev/null +++ b/vendor/golang.org/x/crypto/chacha20/chacha_ppc64le.go @@ -0,0 +1,16 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !gccgo,!purego + +package chacha20 + +const bufSize = 256 + +//go:noescape +func chaCha20_ctr32_vsx(out, inp *byte, len int, key *[8]uint32, counter *uint32) + +func (c *Cipher) xorKeyStreamBlocks(dst, src []byte) { + chaCha20_ctr32_vsx(&dst[0], &src[0], len(src), &c.key, &c.counter) +} diff --git a/vendor/golang.org/x/crypto/internal/chacha20/asm_ppc64le.s b/vendor/golang.org/x/crypto/chacha20/chacha_ppc64le.s similarity index 95% rename from vendor/golang.org/x/crypto/internal/chacha20/asm_ppc64le.s rename to vendor/golang.org/x/crypto/chacha20/chacha_ppc64le.s index 54418522994..23c60216430 100644 --- a/vendor/golang.org/x/crypto/internal/chacha20/asm_ppc64le.s +++ b/vendor/golang.org/x/crypto/chacha20/chacha_ppc64le.s @@ -19,7 +19,7 @@ // The differences in this and the original implementation are // due to the calling conventions and initialization of constants. -// +build ppc64le,!gccgo,!appengine +// +build !gccgo,!purego #include "textflag.h" @@ -31,24 +31,7 @@ #define TMP R15 #define CONSTBASE R16 - -#define X0 R11 -#define X1 R12 -#define X2 R14 -#define X3 R15 -#define X4 R16 -#define X5 R17 -#define X6 R18 -#define X7 R19 -#define X8 R20 -#define X9 R21 -#define X10 R22 -#define X11 R23 -#define X12 R24 -#define X13 R25 -#define X14 R26 -#define X15 R27 - +#define BLOCKS R17 DATA consts<>+0x00(SB)/8, $0x3320646e61707865 DATA consts<>+0x08(SB)/8, $0x6b20657479622d32 @@ -72,13 +55,13 @@ DATA consts<>+0x90(SB)/8, $0x0000000100000000 DATA consts<>+0x98(SB)/8, $0x0000000300000002 GLOBL consts<>(SB), RODATA, $0xa0 -//func chaCha20_ctr32_vsx(out, inp []byte, len int, key *[32]byte, counter *[16]byte) +//func chaCha20_ctr32_vsx(out, inp *byte, len int, key *[8]uint32, counter *uint32) TEXT ·chaCha20_ctr32_vsx(SB),NOSPLIT,$64-40 MOVD out+0(FP), OUT MOVD inp+8(FP), INP MOVD len+16(FP), LEN MOVD key+24(FP), KEY - MOVD cnt+32(FP), CNT + MOVD counter+32(FP), CNT // Addressing for constants MOVD $consts<>+0x00(SB), CONSTBASE @@ -86,6 +69,7 @@ TEXT ·chaCha20_ctr32_vsx(SB),NOSPLIT,$64-40 MOVD $32, R9 MOVD $48, R10 MOVD $64, R11 + SRD $6, LEN, BLOCKS // V16 LXVW4X (CONSTBASE)(R0), VS48 ADD $80,CONSTBASE @@ -429,9 +413,9 @@ loop_vsx: BNE loop_outer_vsx done_vsx: - // Increment counter by 4 + // Increment counter by number of 64 byte blocks MOVD (CNT), R14 - ADD $4, R14 + ADD BLOCKS, R14 MOVD R14, (CNT) RET diff --git a/vendor/golang.org/x/crypto/chacha20/chacha_s390x.go b/vendor/golang.org/x/crypto/chacha20/chacha_s390x.go new file mode 100644 index 00000000000..a9244bdf4db --- /dev/null +++ b/vendor/golang.org/x/crypto/chacha20/chacha_s390x.go @@ -0,0 +1,26 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !gccgo,!purego + +package chacha20 + +import "golang.org/x/sys/cpu" + +var haveAsm = cpu.S390X.HasVX + +const bufSize = 256 + +// xorKeyStreamVX is an assembly implementation of XORKeyStream. It must only +// be called when the vector facility is available. Implementation in asm_s390x.s. +//go:noescape +func xorKeyStreamVX(dst, src []byte, key *[8]uint32, nonce *[3]uint32, counter *uint32) + +func (c *Cipher) xorKeyStreamBlocks(dst, src []byte) { + if cpu.S390X.HasVX { + xorKeyStreamVX(dst, src, &c.key, &c.nonce, &c.counter) + } else { + c.xorKeyStreamBlocksGeneric(dst, src) + } +} diff --git a/vendor/golang.org/x/crypto/internal/chacha20/chacha_s390x.s b/vendor/golang.org/x/crypto/chacha20/chacha_s390x.s similarity index 87% rename from vendor/golang.org/x/crypto/internal/chacha20/chacha_s390x.s rename to vendor/golang.org/x/crypto/chacha20/chacha_s390x.s index 57df404465c..89c658c410b 100644 --- a/vendor/golang.org/x/crypto/internal/chacha20/chacha_s390x.s +++ b/vendor/golang.org/x/crypto/chacha20/chacha_s390x.s @@ -2,7 +2,7 @@ // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. -// +build s390x,!gccgo,!appengine +// +build !gccgo,!purego #include "go_asm.h" #include "textflag.h" @@ -24,15 +24,6 @@ DATA ·constants<>+0x14(SB)/4, $0x3320646e DATA ·constants<>+0x18(SB)/4, $0x79622d32 DATA ·constants<>+0x1c(SB)/4, $0x6b206574 -// EXRL targets: -TEXT ·mvcSrcToBuf(SB), NOFRAME|NOSPLIT, $0 - MVC $1, (R1), (R8) - RET - -TEXT ·mvcBufToDst(SB), NOFRAME|NOSPLIT, $0 - MVC $1, (R8), (R9) - RET - #define BSWAP V5 #define J0 V6 #define KEY0 V7 @@ -144,7 +135,7 @@ TEXT ·mvcBufToDst(SB), NOFRAME|NOSPLIT, $0 VMRHF v, w, c \ // c = {a[2], b[2], c[2], d[2]} VMRLF v, w, d // d = {a[3], b[3], c[3], d[3]} -// func xorKeyStreamVX(dst, src []byte, key *[8]uint32, nonce *[3]uint32, counter *uint32, buf *[256]byte, len *int) +// func xorKeyStreamVX(dst, src []byte, key *[8]uint32, nonce *[3]uint32, counter *uint32) TEXT ·xorKeyStreamVX(SB), NOSPLIT, $0 MOVD $·constants<>(SB), R1 MOVD dst+0(FP), R2 // R2=&dst[0] @@ -152,25 +143,10 @@ TEXT ·xorKeyStreamVX(SB), NOSPLIT, $0 MOVD key+48(FP), R5 // R5=key MOVD nonce+56(FP), R6 // R6=nonce MOVD counter+64(FP), R7 // R7=counter - MOVD buf+72(FP), R8 // R8=buf - MOVD len+80(FP), R9 // R9=len // load BSWAP and J0 VLM (R1), BSWAP, J0 - // set up tail buffer - ADD $-1, R4, R12 - MOVBZ R12, R12 - CMPUBEQ R12, $255, aligned - MOVD R4, R1 - AND $~255, R1 - MOVD $(R3)(R1*1), R1 - EXRL $·mvcSrcToBuf(SB), R12 - MOVD $255, R0 - SUB R12, R0 - MOVD R0, (R9) // update len - -aligned: // setup MOVD $95, R0 VLM (R5), KEY0, KEY1 @@ -217,9 +193,7 @@ loop: // decrement length ADD $-256, R4 - BLT tail -continue: // rearrange vectors SHUFFLE(X0, X1, X2, X3, M0, M1, M2, M3) ADDV(J0, X0, X1, X2, X3) @@ -245,16 +219,6 @@ continue: MOVD $256(R3), R3 CMPBNE R4, $0, chacha - CMPUBEQ R12, $255, return - EXRL $·mvcBufToDst(SB), R12 // len was updated during setup -return: VSTEF $0, CTR, (R7) RET - -tail: - MOVD R2, R9 - MOVD R8, R2 - MOVD R8, R3 - MOVD $0, R4 - JMP continue diff --git a/vendor/golang.org/x/crypto/internal/chacha20/xor.go b/vendor/golang.org/x/crypto/chacha20/xor.go similarity index 73% rename from vendor/golang.org/x/crypto/internal/chacha20/xor.go rename to vendor/golang.org/x/crypto/chacha20/xor.go index 9c5ba0b33ae..c2d04851e0d 100644 --- a/vendor/golang.org/x/crypto/internal/chacha20/xor.go +++ b/vendor/golang.org/x/crypto/chacha20/xor.go @@ -4,9 +4,7 @@ package chacha20 -import ( - "runtime" -) +import "runtime" // Platforms that have fast unaligned 32-bit little endian accesses. const unaligned = runtime.GOARCH == "386" || @@ -15,10 +13,10 @@ const unaligned = runtime.GOARCH == "386" || runtime.GOARCH == "ppc64le" || runtime.GOARCH == "s390x" -// xor reads a little endian uint32 from src, XORs it with u and +// addXor reads a little endian uint32 from src, XORs it with (a + b) and // places the result in little endian byte order in dst. -func xor(dst, src []byte, u uint32) { - _, _ = src[3], dst[3] // eliminate bounds checks +func addXor(dst, src []byte, a, b uint32) { + _, _ = src[3], dst[3] // bounds check elimination hint if unaligned { // The compiler should optimize this code into // 32-bit unaligned little endian loads and stores. @@ -29,15 +27,16 @@ func xor(dst, src []byte, u uint32) { v |= uint32(src[1]) << 8 v |= uint32(src[2]) << 16 v |= uint32(src[3]) << 24 - v ^= u + v ^= a + b dst[0] = byte(v) dst[1] = byte(v >> 8) dst[2] = byte(v >> 16) dst[3] = byte(v >> 24) } else { - dst[0] = src[0] ^ byte(u) - dst[1] = src[1] ^ byte(u>>8) - dst[2] = src[2] ^ byte(u>>16) - dst[3] = src[3] ^ byte(u>>24) + a += b + dst[0] = src[0] ^ byte(a) + dst[1] = src[1] ^ byte(a>>8) + dst[2] = src[2] ^ byte(a>>16) + dst[3] = src[3] ^ byte(a>>24) } } diff --git a/vendor/golang.org/x/crypto/curve25519/const_amd64.h b/vendor/golang.org/x/crypto/curve25519/const_amd64.h deleted file mode 100644 index b3f74162f60..00000000000 --- a/vendor/golang.org/x/crypto/curve25519/const_amd64.h +++ /dev/null @@ -1,8 +0,0 @@ -// Copyright 2012 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// This code was translated into a form compatible with 6a from the public -// domain sources in SUPERCOP: https://bench.cr.yp.to/supercop.html - -#define REDMASK51 0x0007FFFFFFFFFFFF diff --git a/vendor/golang.org/x/crypto/curve25519/const_amd64.s b/vendor/golang.org/x/crypto/curve25519/const_amd64.s deleted file mode 100644 index ee7b4bd5f8e..00000000000 --- a/vendor/golang.org/x/crypto/curve25519/const_amd64.s +++ /dev/null @@ -1,20 +0,0 @@ -// Copyright 2012 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// This code was translated into a form compatible with 6a from the public -// domain sources in SUPERCOP: https://bench.cr.yp.to/supercop.html - -// +build amd64,!gccgo,!appengine - -// These constants cannot be encoded in non-MOVQ immediates. -// We access them directly from memory instead. - -DATA ·_121666_213(SB)/8, $996687872 -GLOBL ·_121666_213(SB), 8, $8 - -DATA ·_2P0(SB)/8, $0xFFFFFFFFFFFDA -GLOBL ·_2P0(SB), 8, $8 - -DATA ·_2P1234(SB)/8, $0xFFFFFFFFFFFFE -GLOBL ·_2P1234(SB), 8, $8 diff --git a/vendor/golang.org/x/crypto/curve25519/cswap_amd64.s b/vendor/golang.org/x/crypto/curve25519/cswap_amd64.s deleted file mode 100644 index cd793a5b5f2..00000000000 --- a/vendor/golang.org/x/crypto/curve25519/cswap_amd64.s +++ /dev/null @@ -1,65 +0,0 @@ -// Copyright 2012 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// +build amd64,!gccgo,!appengine - -// func cswap(inout *[4][5]uint64, v uint64) -TEXT ·cswap(SB),7,$0 - MOVQ inout+0(FP),DI - MOVQ v+8(FP),SI - - SUBQ $1, SI - NOTQ SI - MOVQ SI, X15 - PSHUFD $0x44, X15, X15 - - MOVOU 0(DI), X0 - MOVOU 16(DI), X2 - MOVOU 32(DI), X4 - MOVOU 48(DI), X6 - MOVOU 64(DI), X8 - MOVOU 80(DI), X1 - MOVOU 96(DI), X3 - MOVOU 112(DI), X5 - MOVOU 128(DI), X7 - MOVOU 144(DI), X9 - - MOVO X1, X10 - MOVO X3, X11 - MOVO X5, X12 - MOVO X7, X13 - MOVO X9, X14 - - PXOR X0, X10 - PXOR X2, X11 - PXOR X4, X12 - PXOR X6, X13 - PXOR X8, X14 - PAND X15, X10 - PAND X15, X11 - PAND X15, X12 - PAND X15, X13 - PAND X15, X14 - PXOR X10, X0 - PXOR X10, X1 - PXOR X11, X2 - PXOR X11, X3 - PXOR X12, X4 - PXOR X12, X5 - PXOR X13, X6 - PXOR X13, X7 - PXOR X14, X8 - PXOR X14, X9 - - MOVOU X0, 0(DI) - MOVOU X2, 16(DI) - MOVOU X4, 32(DI) - MOVOU X6, 48(DI) - MOVOU X8, 64(DI) - MOVOU X1, 80(DI) - MOVOU X3, 96(DI) - MOVOU X5, 112(DI) - MOVOU X7, 128(DI) - MOVOU X9, 144(DI) - RET diff --git a/vendor/golang.org/x/crypto/curve25519/curve25519.go b/vendor/golang.org/x/crypto/curve25519/curve25519.go index 75f24babb69..4b9a655d1b5 100644 --- a/vendor/golang.org/x/crypto/curve25519/curve25519.go +++ b/vendor/golang.org/x/crypto/curve25519/curve25519.go @@ -1,834 +1,95 @@ -// Copyright 2013 The Go Authors. All rights reserved. +// Copyright 2019 The Go Authors. All rights reserved. // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. -// We have an implementation in amd64 assembly so this code is only run on -// non-amd64 platforms. The amd64 assembly does not support gccgo. -// +build !amd64 gccgo appengine - -package curve25519 +// Package curve25519 provides an implementation of the X25519 function, which +// performs scalar multiplication on the elliptic curve known as Curve25519. +// See RFC 7748. +package curve25519 // import "golang.org/x/crypto/curve25519" import ( - "encoding/binary" + "crypto/subtle" + "fmt" ) -// This code is a port of the public domain, "ref10" implementation of -// curve25519 from SUPERCOP 20130419 by D. J. Bernstein. - -// fieldElement represents an element of the field GF(2^255 - 19). An element -// t, entries t[0]...t[9], represents the integer t[0]+2^26 t[1]+2^51 t[2]+2^77 -// t[3]+2^102 t[4]+...+2^230 t[9]. Bounds on each t[i] vary depending on -// context. -type fieldElement [10]int32 - -func feZero(fe *fieldElement) { - for i := range fe { - fe[i] = 0 - } -} - -func feOne(fe *fieldElement) { - feZero(fe) - fe[0] = 1 -} - -func feAdd(dst, a, b *fieldElement) { - for i := range dst { - dst[i] = a[i] + b[i] - } -} - -func feSub(dst, a, b *fieldElement) { - for i := range dst { - dst[i] = a[i] - b[i] - } -} - -func feCopy(dst, src *fieldElement) { - for i := range dst { - dst[i] = src[i] - } -} - -// feCSwap replaces (f,g) with (g,f) if b == 1; replaces (f,g) with (f,g) if b == 0. -// -// Preconditions: b in {0,1}. -func feCSwap(f, g *fieldElement, b int32) { - b = -b - for i := range f { - t := b & (f[i] ^ g[i]) - f[i] ^= t - g[i] ^= t - } -} - -// load3 reads a 24-bit, little-endian value from in. -func load3(in []byte) int64 { - var r int64 - r = int64(in[0]) - r |= int64(in[1]) << 8 - r |= int64(in[2]) << 16 - return r -} - -// load4 reads a 32-bit, little-endian value from in. -func load4(in []byte) int64 { - return int64(binary.LittleEndian.Uint32(in)) -} - -func feFromBytes(dst *fieldElement, src *[32]byte) { - h0 := load4(src[:]) - h1 := load3(src[4:]) << 6 - h2 := load3(src[7:]) << 5 - h3 := load3(src[10:]) << 3 - h4 := load3(src[13:]) << 2 - h5 := load4(src[16:]) - h6 := load3(src[20:]) << 7 - h7 := load3(src[23:]) << 5 - h8 := load3(src[26:]) << 4 - h9 := (load3(src[29:]) & 0x7fffff) << 2 - - var carry [10]int64 - carry[9] = (h9 + 1<<24) >> 25 - h0 += carry[9] * 19 - h9 -= carry[9] << 25 - carry[1] = (h1 + 1<<24) >> 25 - h2 += carry[1] - h1 -= carry[1] << 25 - carry[3] = (h3 + 1<<24) >> 25 - h4 += carry[3] - h3 -= carry[3] << 25 - carry[5] = (h5 + 1<<24) >> 25 - h6 += carry[5] - h5 -= carry[5] << 25 - carry[7] = (h7 + 1<<24) >> 25 - h8 += carry[7] - h7 -= carry[7] << 25 - - carry[0] = (h0 + 1<<25) >> 26 - h1 += carry[0] - h0 -= carry[0] << 26 - carry[2] = (h2 + 1<<25) >> 26 - h3 += carry[2] - h2 -= carry[2] << 26 - carry[4] = (h4 + 1<<25) >> 26 - h5 += carry[4] - h4 -= carry[4] << 26 - carry[6] = (h6 + 1<<25) >> 26 - h7 += carry[6] - h6 -= carry[6] << 26 - carry[8] = (h8 + 1<<25) >> 26 - h9 += carry[8] - h8 -= carry[8] << 26 - - dst[0] = int32(h0) - dst[1] = int32(h1) - dst[2] = int32(h2) - dst[3] = int32(h3) - dst[4] = int32(h4) - dst[5] = int32(h5) - dst[6] = int32(h6) - dst[7] = int32(h7) - dst[8] = int32(h8) - dst[9] = int32(h9) -} - -// feToBytes marshals h to s. -// Preconditions: -// |h| bounded by 1.1*2^25,1.1*2^24,1.1*2^25,1.1*2^24,etc. -// -// Write p=2^255-19; q=floor(h/p). -// Basic claim: q = floor(2^(-255)(h + 19 2^(-25)h9 + 2^(-1))). -// -// Proof: -// Have |h|<=p so |q|<=1 so |19^2 2^(-255) q|<1/4. -// Also have |h-2^230 h9|<2^230 so |19 2^(-255)(h-2^230 h9)|<1/4. -// -// Write y=2^(-1)-19^2 2^(-255)q-19 2^(-255)(h-2^230 h9). -// Then 0> 25 - q = (h[0] + q) >> 26 - q = (h[1] + q) >> 25 - q = (h[2] + q) >> 26 - q = (h[3] + q) >> 25 - q = (h[4] + q) >> 26 - q = (h[5] + q) >> 25 - q = (h[6] + q) >> 26 - q = (h[7] + q) >> 25 - q = (h[8] + q) >> 26 - q = (h[9] + q) >> 25 - - // Goal: Output h-(2^255-19)q, which is between 0 and 2^255-20. - h[0] += 19 * q - // Goal: Output h-2^255 q, which is between 0 and 2^255-20. - - carry[0] = h[0] >> 26 - h[1] += carry[0] - h[0] -= carry[0] << 26 - carry[1] = h[1] >> 25 - h[2] += carry[1] - h[1] -= carry[1] << 25 - carry[2] = h[2] >> 26 - h[3] += carry[2] - h[2] -= carry[2] << 26 - carry[3] = h[3] >> 25 - h[4] += carry[3] - h[3] -= carry[3] << 25 - carry[4] = h[4] >> 26 - h[5] += carry[4] - h[4] -= carry[4] << 26 - carry[5] = h[5] >> 25 - h[6] += carry[5] - h[5] -= carry[5] << 25 - carry[6] = h[6] >> 26 - h[7] += carry[6] - h[6] -= carry[6] << 26 - carry[7] = h[7] >> 25 - h[8] += carry[7] - h[7] -= carry[7] << 25 - carry[8] = h[8] >> 26 - h[9] += carry[8] - h[8] -= carry[8] << 26 - carry[9] = h[9] >> 25 - h[9] -= carry[9] << 25 - // h10 = carry9 - - // Goal: Output h[0]+...+2^255 h10-2^255 q, which is between 0 and 2^255-20. - // Have h[0]+...+2^230 h[9] between 0 and 2^255-1; - // evidently 2^255 h10-2^255 q = 0. - // Goal: Output h[0]+...+2^230 h[9]. - - s[0] = byte(h[0] >> 0) - s[1] = byte(h[0] >> 8) - s[2] = byte(h[0] >> 16) - s[3] = byte((h[0] >> 24) | (h[1] << 2)) - s[4] = byte(h[1] >> 6) - s[5] = byte(h[1] >> 14) - s[6] = byte((h[1] >> 22) | (h[2] << 3)) - s[7] = byte(h[2] >> 5) - s[8] = byte(h[2] >> 13) - s[9] = byte((h[2] >> 21) | (h[3] << 5)) - s[10] = byte(h[3] >> 3) - s[11] = byte(h[3] >> 11) - s[12] = byte((h[3] >> 19) | (h[4] << 6)) - s[13] = byte(h[4] >> 2) - s[14] = byte(h[4] >> 10) - s[15] = byte(h[4] >> 18) - s[16] = byte(h[5] >> 0) - s[17] = byte(h[5] >> 8) - s[18] = byte(h[5] >> 16) - s[19] = byte((h[5] >> 24) | (h[6] << 1)) - s[20] = byte(h[6] >> 7) - s[21] = byte(h[6] >> 15) - s[22] = byte((h[6] >> 23) | (h[7] << 3)) - s[23] = byte(h[7] >> 5) - s[24] = byte(h[7] >> 13) - s[25] = byte((h[7] >> 21) | (h[8] << 4)) - s[26] = byte(h[8] >> 4) - s[27] = byte(h[8] >> 12) - s[28] = byte((h[8] >> 20) | (h[9] << 6)) - s[29] = byte(h[9] >> 2) - s[30] = byte(h[9] >> 10) - s[31] = byte(h[9] >> 18) +// Deprecated: when provided a low-order point, ScalarMult will set dst to all +// zeroes, irrespective of the scalar. Instead, use the X25519 function, which +// will return an error. +func ScalarMult(dst, scalar, point *[32]byte) { + scalarMult(dst, scalar, point) } -// feMul calculates h = f * g -// Can overlap h with f or g. -// -// Preconditions: -// |f| bounded by 1.1*2^26,1.1*2^25,1.1*2^26,1.1*2^25,etc. -// |g| bounded by 1.1*2^26,1.1*2^25,1.1*2^26,1.1*2^25,etc. -// -// Postconditions: -// |h| bounded by 1.1*2^25,1.1*2^24,1.1*2^25,1.1*2^24,etc. -// -// Notes on implementation strategy: -// -// Using schoolbook multiplication. -// Karatsuba would save a little in some cost models. +// ScalarBaseMult sets dst to the product scalar * base where base is the +// standard generator. // -// Most multiplications by 2 and 19 are 32-bit precomputations; -// cheaper than 64-bit postcomputations. -// -// There is one remaining multiplication by 19 in the carry chain; -// one *19 precomputation can be merged into this, -// but the resulting data flow is considerably less clean. -// -// There are 12 carries below. -// 10 of them are 2-way parallelizable and vectorizable. -// Can get away with 11 carries, but then data flow is much deeper. -// -// With tighter constraints on inputs can squeeze carries into int32. -func feMul(h, f, g *fieldElement) { - f0 := f[0] - f1 := f[1] - f2 := f[2] - f3 := f[3] - f4 := f[4] - f5 := f[5] - f6 := f[6] - f7 := f[7] - f8 := f[8] - f9 := f[9] - g0 := g[0] - g1 := g[1] - g2 := g[2] - g3 := g[3] - g4 := g[4] - g5 := g[5] - g6 := g[6] - g7 := g[7] - g8 := g[8] - g9 := g[9] - g1_19 := 19 * g1 // 1.4*2^29 - g2_19 := 19 * g2 // 1.4*2^30; still ok - g3_19 := 19 * g3 - g4_19 := 19 * g4 - g5_19 := 19 * g5 - g6_19 := 19 * g6 - g7_19 := 19 * g7 - g8_19 := 19 * g8 - g9_19 := 19 * g9 - f1_2 := 2 * f1 - f3_2 := 2 * f3 - f5_2 := 2 * f5 - f7_2 := 2 * f7 - f9_2 := 2 * f9 - f0g0 := int64(f0) * int64(g0) - f0g1 := int64(f0) * int64(g1) - f0g2 := int64(f0) * int64(g2) - f0g3 := int64(f0) * int64(g3) - f0g4 := int64(f0) * int64(g4) - f0g5 := int64(f0) * int64(g5) - f0g6 := int64(f0) * int64(g6) - f0g7 := int64(f0) * int64(g7) - f0g8 := int64(f0) * int64(g8) - f0g9 := int64(f0) * int64(g9) - f1g0 := int64(f1) * int64(g0) - f1g1_2 := int64(f1_2) * int64(g1) - f1g2 := int64(f1) * int64(g2) - f1g3_2 := int64(f1_2) * int64(g3) - f1g4 := int64(f1) * int64(g4) - f1g5_2 := int64(f1_2) * int64(g5) - f1g6 := int64(f1) * int64(g6) - f1g7_2 := int64(f1_2) * int64(g7) - f1g8 := int64(f1) * int64(g8) - f1g9_38 := int64(f1_2) * int64(g9_19) - f2g0 := int64(f2) * int64(g0) - f2g1 := int64(f2) * int64(g1) - f2g2 := int64(f2) * int64(g2) - f2g3 := int64(f2) * int64(g3) - f2g4 := int64(f2) * int64(g4) - f2g5 := int64(f2) * int64(g5) - f2g6 := int64(f2) * int64(g6) - f2g7 := int64(f2) * int64(g7) - f2g8_19 := int64(f2) * int64(g8_19) - f2g9_19 := int64(f2) * int64(g9_19) - f3g0 := int64(f3) * int64(g0) - f3g1_2 := int64(f3_2) * int64(g1) - f3g2 := int64(f3) * int64(g2) - f3g3_2 := int64(f3_2) * int64(g3) - f3g4 := int64(f3) * int64(g4) - f3g5_2 := int64(f3_2) * int64(g5) - f3g6 := int64(f3) * int64(g6) - f3g7_38 := int64(f3_2) * int64(g7_19) - f3g8_19 := int64(f3) * int64(g8_19) - f3g9_38 := int64(f3_2) * int64(g9_19) - f4g0 := int64(f4) * int64(g0) - f4g1 := int64(f4) * int64(g1) - f4g2 := int64(f4) * int64(g2) - f4g3 := int64(f4) * int64(g3) - f4g4 := int64(f4) * int64(g4) - f4g5 := int64(f4) * int64(g5) - f4g6_19 := int64(f4) * int64(g6_19) - f4g7_19 := int64(f4) * int64(g7_19) - f4g8_19 := int64(f4) * int64(g8_19) - f4g9_19 := int64(f4) * int64(g9_19) - f5g0 := int64(f5) * int64(g0) - f5g1_2 := int64(f5_2) * int64(g1) - f5g2 := int64(f5) * int64(g2) - f5g3_2 := int64(f5_2) * int64(g3) - f5g4 := int64(f5) * int64(g4) - f5g5_38 := int64(f5_2) * int64(g5_19) - f5g6_19 := int64(f5) * int64(g6_19) - f5g7_38 := int64(f5_2) * int64(g7_19) - f5g8_19 := int64(f5) * int64(g8_19) - f5g9_38 := int64(f5_2) * int64(g9_19) - f6g0 := int64(f6) * int64(g0) - f6g1 := int64(f6) * int64(g1) - f6g2 := int64(f6) * int64(g2) - f6g3 := int64(f6) * int64(g3) - f6g4_19 := int64(f6) * int64(g4_19) - f6g5_19 := int64(f6) * int64(g5_19) - f6g6_19 := int64(f6) * int64(g6_19) - f6g7_19 := int64(f6) * int64(g7_19) - f6g8_19 := int64(f6) * int64(g8_19) - f6g9_19 := int64(f6) * int64(g9_19) - f7g0 := int64(f7) * int64(g0) - f7g1_2 := int64(f7_2) * int64(g1) - f7g2 := int64(f7) * int64(g2) - f7g3_38 := int64(f7_2) * int64(g3_19) - f7g4_19 := int64(f7) * int64(g4_19) - f7g5_38 := int64(f7_2) * int64(g5_19) - f7g6_19 := int64(f7) * int64(g6_19) - f7g7_38 := int64(f7_2) * int64(g7_19) - f7g8_19 := int64(f7) * int64(g8_19) - f7g9_38 := int64(f7_2) * int64(g9_19) - f8g0 := int64(f8) * int64(g0) - f8g1 := int64(f8) * int64(g1) - f8g2_19 := int64(f8) * int64(g2_19) - f8g3_19 := int64(f8) * int64(g3_19) - f8g4_19 := int64(f8) * int64(g4_19) - f8g5_19 := int64(f8) * int64(g5_19) - f8g6_19 := int64(f8) * int64(g6_19) - f8g7_19 := int64(f8) * int64(g7_19) - f8g8_19 := int64(f8) * int64(g8_19) - f8g9_19 := int64(f8) * int64(g9_19) - f9g0 := int64(f9) * int64(g0) - f9g1_38 := int64(f9_2) * int64(g1_19) - f9g2_19 := int64(f9) * int64(g2_19) - f9g3_38 := int64(f9_2) * int64(g3_19) - f9g4_19 := int64(f9) * int64(g4_19) - f9g5_38 := int64(f9_2) * int64(g5_19) - f9g6_19 := int64(f9) * int64(g6_19) - f9g7_38 := int64(f9_2) * int64(g7_19) - f9g8_19 := int64(f9) * int64(g8_19) - f9g9_38 := int64(f9_2) * int64(g9_19) - h0 := f0g0 + f1g9_38 + f2g8_19 + f3g7_38 + f4g6_19 + f5g5_38 + f6g4_19 + f7g3_38 + f8g2_19 + f9g1_38 - h1 := f0g1 + f1g0 + f2g9_19 + f3g8_19 + f4g7_19 + f5g6_19 + f6g5_19 + f7g4_19 + f8g3_19 + f9g2_19 - h2 := f0g2 + f1g1_2 + f2g0 + f3g9_38 + f4g8_19 + f5g7_38 + f6g6_19 + f7g5_38 + f8g4_19 + f9g3_38 - h3 := f0g3 + f1g2 + f2g1 + f3g0 + f4g9_19 + f5g8_19 + f6g7_19 + f7g6_19 + f8g5_19 + f9g4_19 - h4 := f0g4 + f1g3_2 + f2g2 + f3g1_2 + f4g0 + f5g9_38 + f6g8_19 + f7g7_38 + f8g6_19 + f9g5_38 - h5 := f0g5 + f1g4 + f2g3 + f3g2 + f4g1 + f5g0 + f6g9_19 + f7g8_19 + f8g7_19 + f9g6_19 - h6 := f0g6 + f1g5_2 + f2g4 + f3g3_2 + f4g2 + f5g1_2 + f6g0 + f7g9_38 + f8g8_19 + f9g7_38 - h7 := f0g7 + f1g6 + f2g5 + f3g4 + f4g3 + f5g2 + f6g1 + f7g0 + f8g9_19 + f9g8_19 - h8 := f0g8 + f1g7_2 + f2g6 + f3g5_2 + f4g4 + f5g3_2 + f6g2 + f7g1_2 + f8g0 + f9g9_38 - h9 := f0g9 + f1g8 + f2g7 + f3g6 + f4g5 + f5g4 + f6g3 + f7g2 + f8g1 + f9g0 - var carry [10]int64 - - // |h0| <= (1.1*1.1*2^52*(1+19+19+19+19)+1.1*1.1*2^50*(38+38+38+38+38)) - // i.e. |h0| <= 1.2*2^59; narrower ranges for h2, h4, h6, h8 - // |h1| <= (1.1*1.1*2^51*(1+1+19+19+19+19+19+19+19+19)) - // i.e. |h1| <= 1.5*2^58; narrower ranges for h3, h5, h7, h9 - - carry[0] = (h0 + (1 << 25)) >> 26 - h1 += carry[0] - h0 -= carry[0] << 26 - carry[4] = (h4 + (1 << 25)) >> 26 - h5 += carry[4] - h4 -= carry[4] << 26 - // |h0| <= 2^25 - // |h4| <= 2^25 - // |h1| <= 1.51*2^58 - // |h5| <= 1.51*2^58 - - carry[1] = (h1 + (1 << 24)) >> 25 - h2 += carry[1] - h1 -= carry[1] << 25 - carry[5] = (h5 + (1 << 24)) >> 25 - h6 += carry[5] - h5 -= carry[5] << 25 - // |h1| <= 2^24; from now on fits into int32 - // |h5| <= 2^24; from now on fits into int32 - // |h2| <= 1.21*2^59 - // |h6| <= 1.21*2^59 - - carry[2] = (h2 + (1 << 25)) >> 26 - h3 += carry[2] - h2 -= carry[2] << 26 - carry[6] = (h6 + (1 << 25)) >> 26 - h7 += carry[6] - h6 -= carry[6] << 26 - // |h2| <= 2^25; from now on fits into int32 unchanged - // |h6| <= 2^25; from now on fits into int32 unchanged - // |h3| <= 1.51*2^58 - // |h7| <= 1.51*2^58 - - carry[3] = (h3 + (1 << 24)) >> 25 - h4 += carry[3] - h3 -= carry[3] << 25 - carry[7] = (h7 + (1 << 24)) >> 25 - h8 += carry[7] - h7 -= carry[7] << 25 - // |h3| <= 2^24; from now on fits into int32 unchanged - // |h7| <= 2^24; from now on fits into int32 unchanged - // |h4| <= 1.52*2^33 - // |h8| <= 1.52*2^33 - - carry[4] = (h4 + (1 << 25)) >> 26 - h5 += carry[4] - h4 -= carry[4] << 26 - carry[8] = (h8 + (1 << 25)) >> 26 - h9 += carry[8] - h8 -= carry[8] << 26 - // |h4| <= 2^25; from now on fits into int32 unchanged - // |h8| <= 2^25; from now on fits into int32 unchanged - // |h5| <= 1.01*2^24 - // |h9| <= 1.51*2^58 - - carry[9] = (h9 + (1 << 24)) >> 25 - h0 += carry[9] * 19 - h9 -= carry[9] << 25 - // |h9| <= 2^24; from now on fits into int32 unchanged - // |h0| <= 1.8*2^37 - - carry[0] = (h0 + (1 << 25)) >> 26 - h1 += carry[0] - h0 -= carry[0] << 26 - // |h0| <= 2^25; from now on fits into int32 unchanged - // |h1| <= 1.01*2^24 - - h[0] = int32(h0) - h[1] = int32(h1) - h[2] = int32(h2) - h[3] = int32(h3) - h[4] = int32(h4) - h[5] = int32(h5) - h[6] = int32(h6) - h[7] = int32(h7) - h[8] = int32(h8) - h[9] = int32(h9) +// It is recommended to use the X25519 function with Basepoint instead, as +// copying into fixed size arrays can lead to unexpected bugs. +func ScalarBaseMult(dst, scalar *[32]byte) { + ScalarMult(dst, scalar, &basePoint) } -// feSquare calculates h = f*f. Can overlap h with f. -// -// Preconditions: -// |f| bounded by 1.1*2^26,1.1*2^25,1.1*2^26,1.1*2^25,etc. -// -// Postconditions: -// |h| bounded by 1.1*2^25,1.1*2^24,1.1*2^25,1.1*2^24,etc. -func feSquare(h, f *fieldElement) { - f0 := f[0] - f1 := f[1] - f2 := f[2] - f3 := f[3] - f4 := f[4] - f5 := f[5] - f6 := f[6] - f7 := f[7] - f8 := f[8] - f9 := f[9] - f0_2 := 2 * f0 - f1_2 := 2 * f1 - f2_2 := 2 * f2 - f3_2 := 2 * f3 - f4_2 := 2 * f4 - f5_2 := 2 * f5 - f6_2 := 2 * f6 - f7_2 := 2 * f7 - f5_38 := 38 * f5 // 1.31*2^30 - f6_19 := 19 * f6 // 1.31*2^30 - f7_38 := 38 * f7 // 1.31*2^30 - f8_19 := 19 * f8 // 1.31*2^30 - f9_38 := 38 * f9 // 1.31*2^30 - f0f0 := int64(f0) * int64(f0) - f0f1_2 := int64(f0_2) * int64(f1) - f0f2_2 := int64(f0_2) * int64(f2) - f0f3_2 := int64(f0_2) * int64(f3) - f0f4_2 := int64(f0_2) * int64(f4) - f0f5_2 := int64(f0_2) * int64(f5) - f0f6_2 := int64(f0_2) * int64(f6) - f0f7_2 := int64(f0_2) * int64(f7) - f0f8_2 := int64(f0_2) * int64(f8) - f0f9_2 := int64(f0_2) * int64(f9) - f1f1_2 := int64(f1_2) * int64(f1) - f1f2_2 := int64(f1_2) * int64(f2) - f1f3_4 := int64(f1_2) * int64(f3_2) - f1f4_2 := int64(f1_2) * int64(f4) - f1f5_4 := int64(f1_2) * int64(f5_2) - f1f6_2 := int64(f1_2) * int64(f6) - f1f7_4 := int64(f1_2) * int64(f7_2) - f1f8_2 := int64(f1_2) * int64(f8) - f1f9_76 := int64(f1_2) * int64(f9_38) - f2f2 := int64(f2) * int64(f2) - f2f3_2 := int64(f2_2) * int64(f3) - f2f4_2 := int64(f2_2) * int64(f4) - f2f5_2 := int64(f2_2) * int64(f5) - f2f6_2 := int64(f2_2) * int64(f6) - f2f7_2 := int64(f2_2) * int64(f7) - f2f8_38 := int64(f2_2) * int64(f8_19) - f2f9_38 := int64(f2) * int64(f9_38) - f3f3_2 := int64(f3_2) * int64(f3) - f3f4_2 := int64(f3_2) * int64(f4) - f3f5_4 := int64(f3_2) * int64(f5_2) - f3f6_2 := int64(f3_2) * int64(f6) - f3f7_76 := int64(f3_2) * int64(f7_38) - f3f8_38 := int64(f3_2) * int64(f8_19) - f3f9_76 := int64(f3_2) * int64(f9_38) - f4f4 := int64(f4) * int64(f4) - f4f5_2 := int64(f4_2) * int64(f5) - f4f6_38 := int64(f4_2) * int64(f6_19) - f4f7_38 := int64(f4) * int64(f7_38) - f4f8_38 := int64(f4_2) * int64(f8_19) - f4f9_38 := int64(f4) * int64(f9_38) - f5f5_38 := int64(f5) * int64(f5_38) - f5f6_38 := int64(f5_2) * int64(f6_19) - f5f7_76 := int64(f5_2) * int64(f7_38) - f5f8_38 := int64(f5_2) * int64(f8_19) - f5f9_76 := int64(f5_2) * int64(f9_38) - f6f6_19 := int64(f6) * int64(f6_19) - f6f7_38 := int64(f6) * int64(f7_38) - f6f8_38 := int64(f6_2) * int64(f8_19) - f6f9_38 := int64(f6) * int64(f9_38) - f7f7_38 := int64(f7) * int64(f7_38) - f7f8_38 := int64(f7_2) * int64(f8_19) - f7f9_76 := int64(f7_2) * int64(f9_38) - f8f8_19 := int64(f8) * int64(f8_19) - f8f9_38 := int64(f8) * int64(f9_38) - f9f9_38 := int64(f9) * int64(f9_38) - h0 := f0f0 + f1f9_76 + f2f8_38 + f3f7_76 + f4f6_38 + f5f5_38 - h1 := f0f1_2 + f2f9_38 + f3f8_38 + f4f7_38 + f5f6_38 - h2 := f0f2_2 + f1f1_2 + f3f9_76 + f4f8_38 + f5f7_76 + f6f6_19 - h3 := f0f3_2 + f1f2_2 + f4f9_38 + f5f8_38 + f6f7_38 - h4 := f0f4_2 + f1f3_4 + f2f2 + f5f9_76 + f6f8_38 + f7f7_38 - h5 := f0f5_2 + f1f4_2 + f2f3_2 + f6f9_38 + f7f8_38 - h6 := f0f6_2 + f1f5_4 + f2f4_2 + f3f3_2 + f7f9_76 + f8f8_19 - h7 := f0f7_2 + f1f6_2 + f2f5_2 + f3f4_2 + f8f9_38 - h8 := f0f8_2 + f1f7_4 + f2f6_2 + f3f5_4 + f4f4 + f9f9_38 - h9 := f0f9_2 + f1f8_2 + f2f7_2 + f3f6_2 + f4f5_2 - var carry [10]int64 - - carry[0] = (h0 + (1 << 25)) >> 26 - h1 += carry[0] - h0 -= carry[0] << 26 - carry[4] = (h4 + (1 << 25)) >> 26 - h5 += carry[4] - h4 -= carry[4] << 26 - - carry[1] = (h1 + (1 << 24)) >> 25 - h2 += carry[1] - h1 -= carry[1] << 25 - carry[5] = (h5 + (1 << 24)) >> 25 - h6 += carry[5] - h5 -= carry[5] << 25 - - carry[2] = (h2 + (1 << 25)) >> 26 - h3 += carry[2] - h2 -= carry[2] << 26 - carry[6] = (h6 + (1 << 25)) >> 26 - h7 += carry[6] - h6 -= carry[6] << 26 - - carry[3] = (h3 + (1 << 24)) >> 25 - h4 += carry[3] - h3 -= carry[3] << 25 - carry[7] = (h7 + (1 << 24)) >> 25 - h8 += carry[7] - h7 -= carry[7] << 25 +const ( + // ScalarSize is the size of the scalar input to X25519. + ScalarSize = 32 + // PointSize is the size of the point input to X25519. + PointSize = 32 +) - carry[4] = (h4 + (1 << 25)) >> 26 - h5 += carry[4] - h4 -= carry[4] << 26 - carry[8] = (h8 + (1 << 25)) >> 26 - h9 += carry[8] - h8 -= carry[8] << 26 +// Basepoint is the canonical Curve25519 generator. +var Basepoint []byte - carry[9] = (h9 + (1 << 24)) >> 25 - h0 += carry[9] * 19 - h9 -= carry[9] << 25 +var basePoint = [32]byte{9, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0} - carry[0] = (h0 + (1 << 25)) >> 26 - h1 += carry[0] - h0 -= carry[0] << 26 +func init() { Basepoint = basePoint[:] } - h[0] = int32(h0) - h[1] = int32(h1) - h[2] = int32(h2) - h[3] = int32(h3) - h[4] = int32(h4) - h[5] = int32(h5) - h[6] = int32(h6) - h[7] = int32(h7) - h[8] = int32(h8) - h[9] = int32(h9) +func checkBasepoint() { + if subtle.ConstantTimeCompare(Basepoint, []byte{ + 0x09, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + }) != 1 { + panic("curve25519: global Basepoint value was modified") + } } -// feMul121666 calculates h = f * 121666. Can overlap h with f. +// X25519 returns the result of the scalar multiplication (scalar * point), +// according to RFC 7748, Section 5. scalar, point and the return value are +// slices of 32 bytes. // -// Preconditions: -// |f| bounded by 1.1*2^26,1.1*2^25,1.1*2^26,1.1*2^25,etc. +// scalar can be generated at random, for example with crypto/rand. point should +// be either Basepoint or the output of another X25519 call. // -// Postconditions: -// |h| bounded by 1.1*2^25,1.1*2^24,1.1*2^25,1.1*2^24,etc. -func feMul121666(h, f *fieldElement) { - h0 := int64(f[0]) * 121666 - h1 := int64(f[1]) * 121666 - h2 := int64(f[2]) * 121666 - h3 := int64(f[3]) * 121666 - h4 := int64(f[4]) * 121666 - h5 := int64(f[5]) * 121666 - h6 := int64(f[6]) * 121666 - h7 := int64(f[7]) * 121666 - h8 := int64(f[8]) * 121666 - h9 := int64(f[9]) * 121666 - var carry [10]int64 - - carry[9] = (h9 + (1 << 24)) >> 25 - h0 += carry[9] * 19 - h9 -= carry[9] << 25 - carry[1] = (h1 + (1 << 24)) >> 25 - h2 += carry[1] - h1 -= carry[1] << 25 - carry[3] = (h3 + (1 << 24)) >> 25 - h4 += carry[3] - h3 -= carry[3] << 25 - carry[5] = (h5 + (1 << 24)) >> 25 - h6 += carry[5] - h5 -= carry[5] << 25 - carry[7] = (h7 + (1 << 24)) >> 25 - h8 += carry[7] - h7 -= carry[7] << 25 - - carry[0] = (h0 + (1 << 25)) >> 26 - h1 += carry[0] - h0 -= carry[0] << 26 - carry[2] = (h2 + (1 << 25)) >> 26 - h3 += carry[2] - h2 -= carry[2] << 26 - carry[4] = (h4 + (1 << 25)) >> 26 - h5 += carry[4] - h4 -= carry[4] << 26 - carry[6] = (h6 + (1 << 25)) >> 26 - h7 += carry[6] - h6 -= carry[6] << 26 - carry[8] = (h8 + (1 << 25)) >> 26 - h9 += carry[8] - h8 -= carry[8] << 26 - - h[0] = int32(h0) - h[1] = int32(h1) - h[2] = int32(h2) - h[3] = int32(h3) - h[4] = int32(h4) - h[5] = int32(h5) - h[6] = int32(h6) - h[7] = int32(h7) - h[8] = int32(h8) - h[9] = int32(h9) -} - -// feInvert sets out = z^-1. -func feInvert(out, z *fieldElement) { - var t0, t1, t2, t3 fieldElement - var i int - - feSquare(&t0, z) - for i = 1; i < 1; i++ { - feSquare(&t0, &t0) - } - feSquare(&t1, &t0) - for i = 1; i < 2; i++ { - feSquare(&t1, &t1) - } - feMul(&t1, z, &t1) - feMul(&t0, &t0, &t1) - feSquare(&t2, &t0) - for i = 1; i < 1; i++ { - feSquare(&t2, &t2) - } - feMul(&t1, &t1, &t2) - feSquare(&t2, &t1) - for i = 1; i < 5; i++ { - feSquare(&t2, &t2) - } - feMul(&t1, &t2, &t1) - feSquare(&t2, &t1) - for i = 1; i < 10; i++ { - feSquare(&t2, &t2) - } - feMul(&t2, &t2, &t1) - feSquare(&t3, &t2) - for i = 1; i < 20; i++ { - feSquare(&t3, &t3) - } - feMul(&t2, &t3, &t2) - feSquare(&t2, &t2) - for i = 1; i < 10; i++ { - feSquare(&t2, &t2) - } - feMul(&t1, &t2, &t1) - feSquare(&t2, &t1) - for i = 1; i < 50; i++ { - feSquare(&t2, &t2) - } - feMul(&t2, &t2, &t1) - feSquare(&t3, &t2) - for i = 1; i < 100; i++ { - feSquare(&t3, &t3) - } - feMul(&t2, &t3, &t2) - feSquare(&t2, &t2) - for i = 1; i < 50; i++ { - feSquare(&t2, &t2) - } - feMul(&t1, &t2, &t1) - feSquare(&t1, &t1) - for i = 1; i < 5; i++ { - feSquare(&t1, &t1) - } - feMul(out, &t1, &t0) +// If point is Basepoint (but not if it's a different slice with the same +// contents) a precomputed implementation might be used for performance. +func X25519(scalar, point []byte) ([]byte, error) { + // Outline the body of function, to let the allocation be inlined in the + // caller, and possibly avoid escaping to the heap. + var dst [32]byte + return x25519(&dst, scalar, point) } -func scalarMult(out, in, base *[32]byte) { - var e [32]byte - - copy(e[:], in[:]) - e[0] &= 248 - e[31] &= 127 - e[31] |= 64 - - var x1, x2, z2, x3, z3, tmp0, tmp1 fieldElement - feFromBytes(&x1, base) - feOne(&x2) - feCopy(&x3, &x1) - feOne(&z3) - - swap := int32(0) - for pos := 254; pos >= 0; pos-- { - b := e[pos/8] >> uint(pos&7) - b &= 1 - swap ^= int32(b) - feCSwap(&x2, &x3, swap) - feCSwap(&z2, &z3, swap) - swap = int32(b) - - feSub(&tmp0, &x3, &z3) - feSub(&tmp1, &x2, &z2) - feAdd(&x2, &x2, &z2) - feAdd(&z2, &x3, &z3) - feMul(&z3, &tmp0, &x2) - feMul(&z2, &z2, &tmp1) - feSquare(&tmp0, &tmp1) - feSquare(&tmp1, &x2) - feAdd(&x3, &z3, &z2) - feSub(&z2, &z3, &z2) - feMul(&x2, &tmp1, &tmp0) - feSub(&tmp1, &tmp1, &tmp0) - feSquare(&z2, &z2) - feMul121666(&z3, &tmp1) - feSquare(&x3, &x3) - feAdd(&tmp0, &tmp0, &z3) - feMul(&z3, &x1, &z2) - feMul(&z2, &tmp1, &tmp0) - } - - feCSwap(&x2, &x3, swap) - feCSwap(&z2, &z3, swap) - - feInvert(&z2, &z2) - feMul(&x2, &x2, &z2) - feToBytes(out, &x2) +func x25519(dst *[32]byte, scalar, point []byte) ([]byte, error) { + var in [32]byte + if l := len(scalar); l != 32 { + return nil, fmt.Errorf("bad scalar length: %d, expected %d", l, 32) + } + if l := len(point); l != 32 { + return nil, fmt.Errorf("bad point length: %d, expected %d", l, 32) + } + copy(in[:], scalar) + if &point[0] == &Basepoint[0] { + checkBasepoint() + ScalarBaseMult(dst, &in) + } else { + var base, zero [32]byte + copy(base[:], point) + ScalarMult(dst, &in, &base) + if subtle.ConstantTimeCompare(dst[:], zero[:]) == 1 { + return nil, fmt.Errorf("bad input point: low order point") + } + } + return dst[:], nil } diff --git a/vendor/golang.org/x/crypto/curve25519/mont25519_amd64.go b/vendor/golang.org/x/crypto/curve25519/curve25519_amd64.go similarity index 99% rename from vendor/golang.org/x/crypto/curve25519/mont25519_amd64.go rename to vendor/golang.org/x/crypto/curve25519/curve25519_amd64.go index 5822bd53383..5120b779b9b 100644 --- a/vendor/golang.org/x/crypto/curve25519/mont25519_amd64.go +++ b/vendor/golang.org/x/crypto/curve25519/curve25519_amd64.go @@ -2,7 +2,7 @@ // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. -// +build amd64,!gccgo,!appengine +// +build amd64,!gccgo,!appengine,!purego package curve25519 diff --git a/vendor/golang.org/x/crypto/curve25519/ladderstep_amd64.s b/vendor/golang.org/x/crypto/curve25519/curve25519_amd64.s similarity index 76% rename from vendor/golang.org/x/crypto/curve25519/ladderstep_amd64.s rename to vendor/golang.org/x/crypto/curve25519/curve25519_amd64.s index e0ac30c70f1..0250c888592 100644 --- a/vendor/golang.org/x/crypto/curve25519/ladderstep_amd64.s +++ b/vendor/golang.org/x/crypto/curve25519/curve25519_amd64.s @@ -5,9 +5,84 @@ // This code was translated into a form compatible with 6a from the public // domain sources in SUPERCOP: https://bench.cr.yp.to/supercop.html -// +build amd64,!gccgo,!appengine +// +build amd64,!gccgo,!appengine,!purego -#include "const_amd64.h" +#define REDMASK51 0x0007FFFFFFFFFFFF + +// These constants cannot be encoded in non-MOVQ immediates. +// We access them directly from memory instead. + +DATA ·_121666_213(SB)/8, $996687872 +GLOBL ·_121666_213(SB), 8, $8 + +DATA ·_2P0(SB)/8, $0xFFFFFFFFFFFDA +GLOBL ·_2P0(SB), 8, $8 + +DATA ·_2P1234(SB)/8, $0xFFFFFFFFFFFFE +GLOBL ·_2P1234(SB), 8, $8 + +// func freeze(inout *[5]uint64) +TEXT ·freeze(SB),7,$0-8 + MOVQ inout+0(FP), DI + + MOVQ 0(DI),SI + MOVQ 8(DI),DX + MOVQ 16(DI),CX + MOVQ 24(DI),R8 + MOVQ 32(DI),R9 + MOVQ $REDMASK51,AX + MOVQ AX,R10 + SUBQ $18,R10 + MOVQ $3,R11 +REDUCELOOP: + MOVQ SI,R12 + SHRQ $51,R12 + ANDQ AX,SI + ADDQ R12,DX + MOVQ DX,R12 + SHRQ $51,R12 + ANDQ AX,DX + ADDQ R12,CX + MOVQ CX,R12 + SHRQ $51,R12 + ANDQ AX,CX + ADDQ R12,R8 + MOVQ R8,R12 + SHRQ $51,R12 + ANDQ AX,R8 + ADDQ R12,R9 + MOVQ R9,R12 + SHRQ $51,R12 + ANDQ AX,R9 + IMUL3Q $19,R12,R12 + ADDQ R12,SI + SUBQ $1,R11 + JA REDUCELOOP + MOVQ $1,R12 + CMPQ R10,SI + CMOVQLT R11,R12 + CMPQ AX,DX + CMOVQNE R11,R12 + CMPQ AX,CX + CMOVQNE R11,R12 + CMPQ AX,R8 + CMOVQNE R11,R12 + CMPQ AX,R9 + CMOVQNE R11,R12 + NEGQ R12 + ANDQ R12,AX + ANDQ R12,R10 + SUBQ R10,SI + SUBQ AX,DX + SUBQ AX,CX + SUBQ AX,R8 + SUBQ AX,R9 + MOVQ SI,0(DI) + MOVQ DX,8(DI) + MOVQ CX,16(DI) + MOVQ R8,24(DI) + MOVQ R9,32(DI) + RET // func ladderstep(inout *[5][5]uint64) TEXT ·ladderstep(SB),0,$296-8 @@ -1375,3 +1450,344 @@ TEXT ·ladderstep(SB),0,$296-8 MOVQ AX,104(DI) MOVQ R10,112(DI) RET + +// func cswap(inout *[4][5]uint64, v uint64) +TEXT ·cswap(SB),7,$0 + MOVQ inout+0(FP),DI + MOVQ v+8(FP),SI + + SUBQ $1, SI + NOTQ SI + MOVQ SI, X15 + PSHUFD $0x44, X15, X15 + + MOVOU 0(DI), X0 + MOVOU 16(DI), X2 + MOVOU 32(DI), X4 + MOVOU 48(DI), X6 + MOVOU 64(DI), X8 + MOVOU 80(DI), X1 + MOVOU 96(DI), X3 + MOVOU 112(DI), X5 + MOVOU 128(DI), X7 + MOVOU 144(DI), X9 + + MOVO X1, X10 + MOVO X3, X11 + MOVO X5, X12 + MOVO X7, X13 + MOVO X9, X14 + + PXOR X0, X10 + PXOR X2, X11 + PXOR X4, X12 + PXOR X6, X13 + PXOR X8, X14 + PAND X15, X10 + PAND X15, X11 + PAND X15, X12 + PAND X15, X13 + PAND X15, X14 + PXOR X10, X0 + PXOR X10, X1 + PXOR X11, X2 + PXOR X11, X3 + PXOR X12, X4 + PXOR X12, X5 + PXOR X13, X6 + PXOR X13, X7 + PXOR X14, X8 + PXOR X14, X9 + + MOVOU X0, 0(DI) + MOVOU X2, 16(DI) + MOVOU X4, 32(DI) + MOVOU X6, 48(DI) + MOVOU X8, 64(DI) + MOVOU X1, 80(DI) + MOVOU X3, 96(DI) + MOVOU X5, 112(DI) + MOVOU X7, 128(DI) + MOVOU X9, 144(DI) + RET + +// func mul(dest, a, b *[5]uint64) +TEXT ·mul(SB),0,$16-24 + MOVQ dest+0(FP), DI + MOVQ a+8(FP), SI + MOVQ b+16(FP), DX + + MOVQ DX,CX + MOVQ 24(SI),DX + IMUL3Q $19,DX,AX + MOVQ AX,0(SP) + MULQ 16(CX) + MOVQ AX,R8 + MOVQ DX,R9 + MOVQ 32(SI),DX + IMUL3Q $19,DX,AX + MOVQ AX,8(SP) + MULQ 8(CX) + ADDQ AX,R8 + ADCQ DX,R9 + MOVQ 0(SI),AX + MULQ 0(CX) + ADDQ AX,R8 + ADCQ DX,R9 + MOVQ 0(SI),AX + MULQ 8(CX) + MOVQ AX,R10 + MOVQ DX,R11 + MOVQ 0(SI),AX + MULQ 16(CX) + MOVQ AX,R12 + MOVQ DX,R13 + MOVQ 0(SI),AX + MULQ 24(CX) + MOVQ AX,R14 + MOVQ DX,R15 + MOVQ 0(SI),AX + MULQ 32(CX) + MOVQ AX,BX + MOVQ DX,BP + MOVQ 8(SI),AX + MULQ 0(CX) + ADDQ AX,R10 + ADCQ DX,R11 + MOVQ 8(SI),AX + MULQ 8(CX) + ADDQ AX,R12 + ADCQ DX,R13 + MOVQ 8(SI),AX + MULQ 16(CX) + ADDQ AX,R14 + ADCQ DX,R15 + MOVQ 8(SI),AX + MULQ 24(CX) + ADDQ AX,BX + ADCQ DX,BP + MOVQ 8(SI),DX + IMUL3Q $19,DX,AX + MULQ 32(CX) + ADDQ AX,R8 + ADCQ DX,R9 + MOVQ 16(SI),AX + MULQ 0(CX) + ADDQ AX,R12 + ADCQ DX,R13 + MOVQ 16(SI),AX + MULQ 8(CX) + ADDQ AX,R14 + ADCQ DX,R15 + MOVQ 16(SI),AX + MULQ 16(CX) + ADDQ AX,BX + ADCQ DX,BP + MOVQ 16(SI),DX + IMUL3Q $19,DX,AX + MULQ 24(CX) + ADDQ AX,R8 + ADCQ DX,R9 + MOVQ 16(SI),DX + IMUL3Q $19,DX,AX + MULQ 32(CX) + ADDQ AX,R10 + ADCQ DX,R11 + MOVQ 24(SI),AX + MULQ 0(CX) + ADDQ AX,R14 + ADCQ DX,R15 + MOVQ 24(SI),AX + MULQ 8(CX) + ADDQ AX,BX + ADCQ DX,BP + MOVQ 0(SP),AX + MULQ 24(CX) + ADDQ AX,R10 + ADCQ DX,R11 + MOVQ 0(SP),AX + MULQ 32(CX) + ADDQ AX,R12 + ADCQ DX,R13 + MOVQ 32(SI),AX + MULQ 0(CX) + ADDQ AX,BX + ADCQ DX,BP + MOVQ 8(SP),AX + MULQ 16(CX) + ADDQ AX,R10 + ADCQ DX,R11 + MOVQ 8(SP),AX + MULQ 24(CX) + ADDQ AX,R12 + ADCQ DX,R13 + MOVQ 8(SP),AX + MULQ 32(CX) + ADDQ AX,R14 + ADCQ DX,R15 + MOVQ $REDMASK51,SI + SHLQ $13,R8,R9 + ANDQ SI,R8 + SHLQ $13,R10,R11 + ANDQ SI,R10 + ADDQ R9,R10 + SHLQ $13,R12,R13 + ANDQ SI,R12 + ADDQ R11,R12 + SHLQ $13,R14,R15 + ANDQ SI,R14 + ADDQ R13,R14 + SHLQ $13,BX,BP + ANDQ SI,BX + ADDQ R15,BX + IMUL3Q $19,BP,DX + ADDQ DX,R8 + MOVQ R8,DX + SHRQ $51,DX + ADDQ R10,DX + MOVQ DX,CX + SHRQ $51,DX + ANDQ SI,R8 + ADDQ R12,DX + MOVQ DX,R9 + SHRQ $51,DX + ANDQ SI,CX + ADDQ R14,DX + MOVQ DX,AX + SHRQ $51,DX + ANDQ SI,R9 + ADDQ BX,DX + MOVQ DX,R10 + SHRQ $51,DX + ANDQ SI,AX + IMUL3Q $19,DX,DX + ADDQ DX,R8 + ANDQ SI,R10 + MOVQ R8,0(DI) + MOVQ CX,8(DI) + MOVQ R9,16(DI) + MOVQ AX,24(DI) + MOVQ R10,32(DI) + RET + +// func square(out, in *[5]uint64) +TEXT ·square(SB),7,$0-16 + MOVQ out+0(FP), DI + MOVQ in+8(FP), SI + + MOVQ 0(SI),AX + MULQ 0(SI) + MOVQ AX,CX + MOVQ DX,R8 + MOVQ 0(SI),AX + SHLQ $1,AX + MULQ 8(SI) + MOVQ AX,R9 + MOVQ DX,R10 + MOVQ 0(SI),AX + SHLQ $1,AX + MULQ 16(SI) + MOVQ AX,R11 + MOVQ DX,R12 + MOVQ 0(SI),AX + SHLQ $1,AX + MULQ 24(SI) + MOVQ AX,R13 + MOVQ DX,R14 + MOVQ 0(SI),AX + SHLQ $1,AX + MULQ 32(SI) + MOVQ AX,R15 + MOVQ DX,BX + MOVQ 8(SI),AX + MULQ 8(SI) + ADDQ AX,R11 + ADCQ DX,R12 + MOVQ 8(SI),AX + SHLQ $1,AX + MULQ 16(SI) + ADDQ AX,R13 + ADCQ DX,R14 + MOVQ 8(SI),AX + SHLQ $1,AX + MULQ 24(SI) + ADDQ AX,R15 + ADCQ DX,BX + MOVQ 8(SI),DX + IMUL3Q $38,DX,AX + MULQ 32(SI) + ADDQ AX,CX + ADCQ DX,R8 + MOVQ 16(SI),AX + MULQ 16(SI) + ADDQ AX,R15 + ADCQ DX,BX + MOVQ 16(SI),DX + IMUL3Q $38,DX,AX + MULQ 24(SI) + ADDQ AX,CX + ADCQ DX,R8 + MOVQ 16(SI),DX + IMUL3Q $38,DX,AX + MULQ 32(SI) + ADDQ AX,R9 + ADCQ DX,R10 + MOVQ 24(SI),DX + IMUL3Q $19,DX,AX + MULQ 24(SI) + ADDQ AX,R9 + ADCQ DX,R10 + MOVQ 24(SI),DX + IMUL3Q $38,DX,AX + MULQ 32(SI) + ADDQ AX,R11 + ADCQ DX,R12 + MOVQ 32(SI),DX + IMUL3Q $19,DX,AX + MULQ 32(SI) + ADDQ AX,R13 + ADCQ DX,R14 + MOVQ $REDMASK51,SI + SHLQ $13,CX,R8 + ANDQ SI,CX + SHLQ $13,R9,R10 + ANDQ SI,R9 + ADDQ R8,R9 + SHLQ $13,R11,R12 + ANDQ SI,R11 + ADDQ R10,R11 + SHLQ $13,R13,R14 + ANDQ SI,R13 + ADDQ R12,R13 + SHLQ $13,R15,BX + ANDQ SI,R15 + ADDQ R14,R15 + IMUL3Q $19,BX,DX + ADDQ DX,CX + MOVQ CX,DX + SHRQ $51,DX + ADDQ R9,DX + ANDQ SI,CX + MOVQ DX,R8 + SHRQ $51,DX + ADDQ R11,DX + ANDQ SI,R8 + MOVQ DX,R9 + SHRQ $51,DX + ADDQ R13,DX + ANDQ SI,R9 + MOVQ DX,AX + SHRQ $51,DX + ADDQ R15,DX + ANDQ SI,AX + MOVQ DX,R10 + SHRQ $51,DX + IMUL3Q $19,DX,DX + ADDQ DX,CX + ANDQ SI,R10 + MOVQ CX,0(DI) + MOVQ R8,8(DI) + MOVQ R9,16(DI) + MOVQ AX,24(DI) + MOVQ R10,32(DI) + RET diff --git a/vendor/golang.org/x/crypto/curve25519/curve25519_generic.go b/vendor/golang.org/x/crypto/curve25519/curve25519_generic.go new file mode 100644 index 00000000000..c43b13fc83e --- /dev/null +++ b/vendor/golang.org/x/crypto/curve25519/curve25519_generic.go @@ -0,0 +1,828 @@ +// Copyright 2013 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package curve25519 + +import "encoding/binary" + +// This code is a port of the public domain, "ref10" implementation of +// curve25519 from SUPERCOP 20130419 by D. J. Bernstein. + +// fieldElement represents an element of the field GF(2^255 - 19). An element +// t, entries t[0]...t[9], represents the integer t[0]+2^26 t[1]+2^51 t[2]+2^77 +// t[3]+2^102 t[4]+...+2^230 t[9]. Bounds on each t[i] vary depending on +// context. +type fieldElement [10]int32 + +func feZero(fe *fieldElement) { + for i := range fe { + fe[i] = 0 + } +} + +func feOne(fe *fieldElement) { + feZero(fe) + fe[0] = 1 +} + +func feAdd(dst, a, b *fieldElement) { + for i := range dst { + dst[i] = a[i] + b[i] + } +} + +func feSub(dst, a, b *fieldElement) { + for i := range dst { + dst[i] = a[i] - b[i] + } +} + +func feCopy(dst, src *fieldElement) { + for i := range dst { + dst[i] = src[i] + } +} + +// feCSwap replaces (f,g) with (g,f) if b == 1; replaces (f,g) with (f,g) if b == 0. +// +// Preconditions: b in {0,1}. +func feCSwap(f, g *fieldElement, b int32) { + b = -b + for i := range f { + t := b & (f[i] ^ g[i]) + f[i] ^= t + g[i] ^= t + } +} + +// load3 reads a 24-bit, little-endian value from in. +func load3(in []byte) int64 { + var r int64 + r = int64(in[0]) + r |= int64(in[1]) << 8 + r |= int64(in[2]) << 16 + return r +} + +// load4 reads a 32-bit, little-endian value from in. +func load4(in []byte) int64 { + return int64(binary.LittleEndian.Uint32(in)) +} + +func feFromBytes(dst *fieldElement, src *[32]byte) { + h0 := load4(src[:]) + h1 := load3(src[4:]) << 6 + h2 := load3(src[7:]) << 5 + h3 := load3(src[10:]) << 3 + h4 := load3(src[13:]) << 2 + h5 := load4(src[16:]) + h6 := load3(src[20:]) << 7 + h7 := load3(src[23:]) << 5 + h8 := load3(src[26:]) << 4 + h9 := (load3(src[29:]) & 0x7fffff) << 2 + + var carry [10]int64 + carry[9] = (h9 + 1<<24) >> 25 + h0 += carry[9] * 19 + h9 -= carry[9] << 25 + carry[1] = (h1 + 1<<24) >> 25 + h2 += carry[1] + h1 -= carry[1] << 25 + carry[3] = (h3 + 1<<24) >> 25 + h4 += carry[3] + h3 -= carry[3] << 25 + carry[5] = (h5 + 1<<24) >> 25 + h6 += carry[5] + h5 -= carry[5] << 25 + carry[7] = (h7 + 1<<24) >> 25 + h8 += carry[7] + h7 -= carry[7] << 25 + + carry[0] = (h0 + 1<<25) >> 26 + h1 += carry[0] + h0 -= carry[0] << 26 + carry[2] = (h2 + 1<<25) >> 26 + h3 += carry[2] + h2 -= carry[2] << 26 + carry[4] = (h4 + 1<<25) >> 26 + h5 += carry[4] + h4 -= carry[4] << 26 + carry[6] = (h6 + 1<<25) >> 26 + h7 += carry[6] + h6 -= carry[6] << 26 + carry[8] = (h8 + 1<<25) >> 26 + h9 += carry[8] + h8 -= carry[8] << 26 + + dst[0] = int32(h0) + dst[1] = int32(h1) + dst[2] = int32(h2) + dst[3] = int32(h3) + dst[4] = int32(h4) + dst[5] = int32(h5) + dst[6] = int32(h6) + dst[7] = int32(h7) + dst[8] = int32(h8) + dst[9] = int32(h9) +} + +// feToBytes marshals h to s. +// Preconditions: +// |h| bounded by 1.1*2^25,1.1*2^24,1.1*2^25,1.1*2^24,etc. +// +// Write p=2^255-19; q=floor(h/p). +// Basic claim: q = floor(2^(-255)(h + 19 2^(-25)h9 + 2^(-1))). +// +// Proof: +// Have |h|<=p so |q|<=1 so |19^2 2^(-255) q|<1/4. +// Also have |h-2^230 h9|<2^230 so |19 2^(-255)(h-2^230 h9)|<1/4. +// +// Write y=2^(-1)-19^2 2^(-255)q-19 2^(-255)(h-2^230 h9). +// Then 0> 25 + q = (h[0] + q) >> 26 + q = (h[1] + q) >> 25 + q = (h[2] + q) >> 26 + q = (h[3] + q) >> 25 + q = (h[4] + q) >> 26 + q = (h[5] + q) >> 25 + q = (h[6] + q) >> 26 + q = (h[7] + q) >> 25 + q = (h[8] + q) >> 26 + q = (h[9] + q) >> 25 + + // Goal: Output h-(2^255-19)q, which is between 0 and 2^255-20. + h[0] += 19 * q + // Goal: Output h-2^255 q, which is between 0 and 2^255-20. + + carry[0] = h[0] >> 26 + h[1] += carry[0] + h[0] -= carry[0] << 26 + carry[1] = h[1] >> 25 + h[2] += carry[1] + h[1] -= carry[1] << 25 + carry[2] = h[2] >> 26 + h[3] += carry[2] + h[2] -= carry[2] << 26 + carry[3] = h[3] >> 25 + h[4] += carry[3] + h[3] -= carry[3] << 25 + carry[4] = h[4] >> 26 + h[5] += carry[4] + h[4] -= carry[4] << 26 + carry[5] = h[5] >> 25 + h[6] += carry[5] + h[5] -= carry[5] << 25 + carry[6] = h[6] >> 26 + h[7] += carry[6] + h[6] -= carry[6] << 26 + carry[7] = h[7] >> 25 + h[8] += carry[7] + h[7] -= carry[7] << 25 + carry[8] = h[8] >> 26 + h[9] += carry[8] + h[8] -= carry[8] << 26 + carry[9] = h[9] >> 25 + h[9] -= carry[9] << 25 + // h10 = carry9 + + // Goal: Output h[0]+...+2^255 h10-2^255 q, which is between 0 and 2^255-20. + // Have h[0]+...+2^230 h[9] between 0 and 2^255-1; + // evidently 2^255 h10-2^255 q = 0. + // Goal: Output h[0]+...+2^230 h[9]. + + s[0] = byte(h[0] >> 0) + s[1] = byte(h[0] >> 8) + s[2] = byte(h[0] >> 16) + s[3] = byte((h[0] >> 24) | (h[1] << 2)) + s[4] = byte(h[1] >> 6) + s[5] = byte(h[1] >> 14) + s[6] = byte((h[1] >> 22) | (h[2] << 3)) + s[7] = byte(h[2] >> 5) + s[8] = byte(h[2] >> 13) + s[9] = byte((h[2] >> 21) | (h[3] << 5)) + s[10] = byte(h[3] >> 3) + s[11] = byte(h[3] >> 11) + s[12] = byte((h[3] >> 19) | (h[4] << 6)) + s[13] = byte(h[4] >> 2) + s[14] = byte(h[4] >> 10) + s[15] = byte(h[4] >> 18) + s[16] = byte(h[5] >> 0) + s[17] = byte(h[5] >> 8) + s[18] = byte(h[5] >> 16) + s[19] = byte((h[5] >> 24) | (h[6] << 1)) + s[20] = byte(h[6] >> 7) + s[21] = byte(h[6] >> 15) + s[22] = byte((h[6] >> 23) | (h[7] << 3)) + s[23] = byte(h[7] >> 5) + s[24] = byte(h[7] >> 13) + s[25] = byte((h[7] >> 21) | (h[8] << 4)) + s[26] = byte(h[8] >> 4) + s[27] = byte(h[8] >> 12) + s[28] = byte((h[8] >> 20) | (h[9] << 6)) + s[29] = byte(h[9] >> 2) + s[30] = byte(h[9] >> 10) + s[31] = byte(h[9] >> 18) +} + +// feMul calculates h = f * g +// Can overlap h with f or g. +// +// Preconditions: +// |f| bounded by 1.1*2^26,1.1*2^25,1.1*2^26,1.1*2^25,etc. +// |g| bounded by 1.1*2^26,1.1*2^25,1.1*2^26,1.1*2^25,etc. +// +// Postconditions: +// |h| bounded by 1.1*2^25,1.1*2^24,1.1*2^25,1.1*2^24,etc. +// +// Notes on implementation strategy: +// +// Using schoolbook multiplication. +// Karatsuba would save a little in some cost models. +// +// Most multiplications by 2 and 19 are 32-bit precomputations; +// cheaper than 64-bit postcomputations. +// +// There is one remaining multiplication by 19 in the carry chain; +// one *19 precomputation can be merged into this, +// but the resulting data flow is considerably less clean. +// +// There are 12 carries below. +// 10 of them are 2-way parallelizable and vectorizable. +// Can get away with 11 carries, but then data flow is much deeper. +// +// With tighter constraints on inputs can squeeze carries into int32. +func feMul(h, f, g *fieldElement) { + f0 := f[0] + f1 := f[1] + f2 := f[2] + f3 := f[3] + f4 := f[4] + f5 := f[5] + f6 := f[6] + f7 := f[7] + f8 := f[8] + f9 := f[9] + g0 := g[0] + g1 := g[1] + g2 := g[2] + g3 := g[3] + g4 := g[4] + g5 := g[5] + g6 := g[6] + g7 := g[7] + g8 := g[8] + g9 := g[9] + g1_19 := 19 * g1 // 1.4*2^29 + g2_19 := 19 * g2 // 1.4*2^30; still ok + g3_19 := 19 * g3 + g4_19 := 19 * g4 + g5_19 := 19 * g5 + g6_19 := 19 * g6 + g7_19 := 19 * g7 + g8_19 := 19 * g8 + g9_19 := 19 * g9 + f1_2 := 2 * f1 + f3_2 := 2 * f3 + f5_2 := 2 * f5 + f7_2 := 2 * f7 + f9_2 := 2 * f9 + f0g0 := int64(f0) * int64(g0) + f0g1 := int64(f0) * int64(g1) + f0g2 := int64(f0) * int64(g2) + f0g3 := int64(f0) * int64(g3) + f0g4 := int64(f0) * int64(g4) + f0g5 := int64(f0) * int64(g5) + f0g6 := int64(f0) * int64(g6) + f0g7 := int64(f0) * int64(g7) + f0g8 := int64(f0) * int64(g8) + f0g9 := int64(f0) * int64(g9) + f1g0 := int64(f1) * int64(g0) + f1g1_2 := int64(f1_2) * int64(g1) + f1g2 := int64(f1) * int64(g2) + f1g3_2 := int64(f1_2) * int64(g3) + f1g4 := int64(f1) * int64(g4) + f1g5_2 := int64(f1_2) * int64(g5) + f1g6 := int64(f1) * int64(g6) + f1g7_2 := int64(f1_2) * int64(g7) + f1g8 := int64(f1) * int64(g8) + f1g9_38 := int64(f1_2) * int64(g9_19) + f2g0 := int64(f2) * int64(g0) + f2g1 := int64(f2) * int64(g1) + f2g2 := int64(f2) * int64(g2) + f2g3 := int64(f2) * int64(g3) + f2g4 := int64(f2) * int64(g4) + f2g5 := int64(f2) * int64(g5) + f2g6 := int64(f2) * int64(g6) + f2g7 := int64(f2) * int64(g7) + f2g8_19 := int64(f2) * int64(g8_19) + f2g9_19 := int64(f2) * int64(g9_19) + f3g0 := int64(f3) * int64(g0) + f3g1_2 := int64(f3_2) * int64(g1) + f3g2 := int64(f3) * int64(g2) + f3g3_2 := int64(f3_2) * int64(g3) + f3g4 := int64(f3) * int64(g4) + f3g5_2 := int64(f3_2) * int64(g5) + f3g6 := int64(f3) * int64(g6) + f3g7_38 := int64(f3_2) * int64(g7_19) + f3g8_19 := int64(f3) * int64(g8_19) + f3g9_38 := int64(f3_2) * int64(g9_19) + f4g0 := int64(f4) * int64(g0) + f4g1 := int64(f4) * int64(g1) + f4g2 := int64(f4) * int64(g2) + f4g3 := int64(f4) * int64(g3) + f4g4 := int64(f4) * int64(g4) + f4g5 := int64(f4) * int64(g5) + f4g6_19 := int64(f4) * int64(g6_19) + f4g7_19 := int64(f4) * int64(g7_19) + f4g8_19 := int64(f4) * int64(g8_19) + f4g9_19 := int64(f4) * int64(g9_19) + f5g0 := int64(f5) * int64(g0) + f5g1_2 := int64(f5_2) * int64(g1) + f5g2 := int64(f5) * int64(g2) + f5g3_2 := int64(f5_2) * int64(g3) + f5g4 := int64(f5) * int64(g4) + f5g5_38 := int64(f5_2) * int64(g5_19) + f5g6_19 := int64(f5) * int64(g6_19) + f5g7_38 := int64(f5_2) * int64(g7_19) + f5g8_19 := int64(f5) * int64(g8_19) + f5g9_38 := int64(f5_2) * int64(g9_19) + f6g0 := int64(f6) * int64(g0) + f6g1 := int64(f6) * int64(g1) + f6g2 := int64(f6) * int64(g2) + f6g3 := int64(f6) * int64(g3) + f6g4_19 := int64(f6) * int64(g4_19) + f6g5_19 := int64(f6) * int64(g5_19) + f6g6_19 := int64(f6) * int64(g6_19) + f6g7_19 := int64(f6) * int64(g7_19) + f6g8_19 := int64(f6) * int64(g8_19) + f6g9_19 := int64(f6) * int64(g9_19) + f7g0 := int64(f7) * int64(g0) + f7g1_2 := int64(f7_2) * int64(g1) + f7g2 := int64(f7) * int64(g2) + f7g3_38 := int64(f7_2) * int64(g3_19) + f7g4_19 := int64(f7) * int64(g4_19) + f7g5_38 := int64(f7_2) * int64(g5_19) + f7g6_19 := int64(f7) * int64(g6_19) + f7g7_38 := int64(f7_2) * int64(g7_19) + f7g8_19 := int64(f7) * int64(g8_19) + f7g9_38 := int64(f7_2) * int64(g9_19) + f8g0 := int64(f8) * int64(g0) + f8g1 := int64(f8) * int64(g1) + f8g2_19 := int64(f8) * int64(g2_19) + f8g3_19 := int64(f8) * int64(g3_19) + f8g4_19 := int64(f8) * int64(g4_19) + f8g5_19 := int64(f8) * int64(g5_19) + f8g6_19 := int64(f8) * int64(g6_19) + f8g7_19 := int64(f8) * int64(g7_19) + f8g8_19 := int64(f8) * int64(g8_19) + f8g9_19 := int64(f8) * int64(g9_19) + f9g0 := int64(f9) * int64(g0) + f9g1_38 := int64(f9_2) * int64(g1_19) + f9g2_19 := int64(f9) * int64(g2_19) + f9g3_38 := int64(f9_2) * int64(g3_19) + f9g4_19 := int64(f9) * int64(g4_19) + f9g5_38 := int64(f9_2) * int64(g5_19) + f9g6_19 := int64(f9) * int64(g6_19) + f9g7_38 := int64(f9_2) * int64(g7_19) + f9g8_19 := int64(f9) * int64(g8_19) + f9g9_38 := int64(f9_2) * int64(g9_19) + h0 := f0g0 + f1g9_38 + f2g8_19 + f3g7_38 + f4g6_19 + f5g5_38 + f6g4_19 + f7g3_38 + f8g2_19 + f9g1_38 + h1 := f0g1 + f1g0 + f2g9_19 + f3g8_19 + f4g7_19 + f5g6_19 + f6g5_19 + f7g4_19 + f8g3_19 + f9g2_19 + h2 := f0g2 + f1g1_2 + f2g0 + f3g9_38 + f4g8_19 + f5g7_38 + f6g6_19 + f7g5_38 + f8g4_19 + f9g3_38 + h3 := f0g3 + f1g2 + f2g1 + f3g0 + f4g9_19 + f5g8_19 + f6g7_19 + f7g6_19 + f8g5_19 + f9g4_19 + h4 := f0g4 + f1g3_2 + f2g2 + f3g1_2 + f4g0 + f5g9_38 + f6g8_19 + f7g7_38 + f8g6_19 + f9g5_38 + h5 := f0g5 + f1g4 + f2g3 + f3g2 + f4g1 + f5g0 + f6g9_19 + f7g8_19 + f8g7_19 + f9g6_19 + h6 := f0g6 + f1g5_2 + f2g4 + f3g3_2 + f4g2 + f5g1_2 + f6g0 + f7g9_38 + f8g8_19 + f9g7_38 + h7 := f0g7 + f1g6 + f2g5 + f3g4 + f4g3 + f5g2 + f6g1 + f7g0 + f8g9_19 + f9g8_19 + h8 := f0g8 + f1g7_2 + f2g6 + f3g5_2 + f4g4 + f5g3_2 + f6g2 + f7g1_2 + f8g0 + f9g9_38 + h9 := f0g9 + f1g8 + f2g7 + f3g6 + f4g5 + f5g4 + f6g3 + f7g2 + f8g1 + f9g0 + var carry [10]int64 + + // |h0| <= (1.1*1.1*2^52*(1+19+19+19+19)+1.1*1.1*2^50*(38+38+38+38+38)) + // i.e. |h0| <= 1.2*2^59; narrower ranges for h2, h4, h6, h8 + // |h1| <= (1.1*1.1*2^51*(1+1+19+19+19+19+19+19+19+19)) + // i.e. |h1| <= 1.5*2^58; narrower ranges for h3, h5, h7, h9 + + carry[0] = (h0 + (1 << 25)) >> 26 + h1 += carry[0] + h0 -= carry[0] << 26 + carry[4] = (h4 + (1 << 25)) >> 26 + h5 += carry[4] + h4 -= carry[4] << 26 + // |h0| <= 2^25 + // |h4| <= 2^25 + // |h1| <= 1.51*2^58 + // |h5| <= 1.51*2^58 + + carry[1] = (h1 + (1 << 24)) >> 25 + h2 += carry[1] + h1 -= carry[1] << 25 + carry[5] = (h5 + (1 << 24)) >> 25 + h6 += carry[5] + h5 -= carry[5] << 25 + // |h1| <= 2^24; from now on fits into int32 + // |h5| <= 2^24; from now on fits into int32 + // |h2| <= 1.21*2^59 + // |h6| <= 1.21*2^59 + + carry[2] = (h2 + (1 << 25)) >> 26 + h3 += carry[2] + h2 -= carry[2] << 26 + carry[6] = (h6 + (1 << 25)) >> 26 + h7 += carry[6] + h6 -= carry[6] << 26 + // |h2| <= 2^25; from now on fits into int32 unchanged + // |h6| <= 2^25; from now on fits into int32 unchanged + // |h3| <= 1.51*2^58 + // |h7| <= 1.51*2^58 + + carry[3] = (h3 + (1 << 24)) >> 25 + h4 += carry[3] + h3 -= carry[3] << 25 + carry[7] = (h7 + (1 << 24)) >> 25 + h8 += carry[7] + h7 -= carry[7] << 25 + // |h3| <= 2^24; from now on fits into int32 unchanged + // |h7| <= 2^24; from now on fits into int32 unchanged + // |h4| <= 1.52*2^33 + // |h8| <= 1.52*2^33 + + carry[4] = (h4 + (1 << 25)) >> 26 + h5 += carry[4] + h4 -= carry[4] << 26 + carry[8] = (h8 + (1 << 25)) >> 26 + h9 += carry[8] + h8 -= carry[8] << 26 + // |h4| <= 2^25; from now on fits into int32 unchanged + // |h8| <= 2^25; from now on fits into int32 unchanged + // |h5| <= 1.01*2^24 + // |h9| <= 1.51*2^58 + + carry[9] = (h9 + (1 << 24)) >> 25 + h0 += carry[9] * 19 + h9 -= carry[9] << 25 + // |h9| <= 2^24; from now on fits into int32 unchanged + // |h0| <= 1.8*2^37 + + carry[0] = (h0 + (1 << 25)) >> 26 + h1 += carry[0] + h0 -= carry[0] << 26 + // |h0| <= 2^25; from now on fits into int32 unchanged + // |h1| <= 1.01*2^24 + + h[0] = int32(h0) + h[1] = int32(h1) + h[2] = int32(h2) + h[3] = int32(h3) + h[4] = int32(h4) + h[5] = int32(h5) + h[6] = int32(h6) + h[7] = int32(h7) + h[8] = int32(h8) + h[9] = int32(h9) +} + +// feSquare calculates h = f*f. Can overlap h with f. +// +// Preconditions: +// |f| bounded by 1.1*2^26,1.1*2^25,1.1*2^26,1.1*2^25,etc. +// +// Postconditions: +// |h| bounded by 1.1*2^25,1.1*2^24,1.1*2^25,1.1*2^24,etc. +func feSquare(h, f *fieldElement) { + f0 := f[0] + f1 := f[1] + f2 := f[2] + f3 := f[3] + f4 := f[4] + f5 := f[5] + f6 := f[6] + f7 := f[7] + f8 := f[8] + f9 := f[9] + f0_2 := 2 * f0 + f1_2 := 2 * f1 + f2_2 := 2 * f2 + f3_2 := 2 * f3 + f4_2 := 2 * f4 + f5_2 := 2 * f5 + f6_2 := 2 * f6 + f7_2 := 2 * f7 + f5_38 := 38 * f5 // 1.31*2^30 + f6_19 := 19 * f6 // 1.31*2^30 + f7_38 := 38 * f7 // 1.31*2^30 + f8_19 := 19 * f8 // 1.31*2^30 + f9_38 := 38 * f9 // 1.31*2^30 + f0f0 := int64(f0) * int64(f0) + f0f1_2 := int64(f0_2) * int64(f1) + f0f2_2 := int64(f0_2) * int64(f2) + f0f3_2 := int64(f0_2) * int64(f3) + f0f4_2 := int64(f0_2) * int64(f4) + f0f5_2 := int64(f0_2) * int64(f5) + f0f6_2 := int64(f0_2) * int64(f6) + f0f7_2 := int64(f0_2) * int64(f7) + f0f8_2 := int64(f0_2) * int64(f8) + f0f9_2 := int64(f0_2) * int64(f9) + f1f1_2 := int64(f1_2) * int64(f1) + f1f2_2 := int64(f1_2) * int64(f2) + f1f3_4 := int64(f1_2) * int64(f3_2) + f1f4_2 := int64(f1_2) * int64(f4) + f1f5_4 := int64(f1_2) * int64(f5_2) + f1f6_2 := int64(f1_2) * int64(f6) + f1f7_4 := int64(f1_2) * int64(f7_2) + f1f8_2 := int64(f1_2) * int64(f8) + f1f9_76 := int64(f1_2) * int64(f9_38) + f2f2 := int64(f2) * int64(f2) + f2f3_2 := int64(f2_2) * int64(f3) + f2f4_2 := int64(f2_2) * int64(f4) + f2f5_2 := int64(f2_2) * int64(f5) + f2f6_2 := int64(f2_2) * int64(f6) + f2f7_2 := int64(f2_2) * int64(f7) + f2f8_38 := int64(f2_2) * int64(f8_19) + f2f9_38 := int64(f2) * int64(f9_38) + f3f3_2 := int64(f3_2) * int64(f3) + f3f4_2 := int64(f3_2) * int64(f4) + f3f5_4 := int64(f3_2) * int64(f5_2) + f3f6_2 := int64(f3_2) * int64(f6) + f3f7_76 := int64(f3_2) * int64(f7_38) + f3f8_38 := int64(f3_2) * int64(f8_19) + f3f9_76 := int64(f3_2) * int64(f9_38) + f4f4 := int64(f4) * int64(f4) + f4f5_2 := int64(f4_2) * int64(f5) + f4f6_38 := int64(f4_2) * int64(f6_19) + f4f7_38 := int64(f4) * int64(f7_38) + f4f8_38 := int64(f4_2) * int64(f8_19) + f4f9_38 := int64(f4) * int64(f9_38) + f5f5_38 := int64(f5) * int64(f5_38) + f5f6_38 := int64(f5_2) * int64(f6_19) + f5f7_76 := int64(f5_2) * int64(f7_38) + f5f8_38 := int64(f5_2) * int64(f8_19) + f5f9_76 := int64(f5_2) * int64(f9_38) + f6f6_19 := int64(f6) * int64(f6_19) + f6f7_38 := int64(f6) * int64(f7_38) + f6f8_38 := int64(f6_2) * int64(f8_19) + f6f9_38 := int64(f6) * int64(f9_38) + f7f7_38 := int64(f7) * int64(f7_38) + f7f8_38 := int64(f7_2) * int64(f8_19) + f7f9_76 := int64(f7_2) * int64(f9_38) + f8f8_19 := int64(f8) * int64(f8_19) + f8f9_38 := int64(f8) * int64(f9_38) + f9f9_38 := int64(f9) * int64(f9_38) + h0 := f0f0 + f1f9_76 + f2f8_38 + f3f7_76 + f4f6_38 + f5f5_38 + h1 := f0f1_2 + f2f9_38 + f3f8_38 + f4f7_38 + f5f6_38 + h2 := f0f2_2 + f1f1_2 + f3f9_76 + f4f8_38 + f5f7_76 + f6f6_19 + h3 := f0f3_2 + f1f2_2 + f4f9_38 + f5f8_38 + f6f7_38 + h4 := f0f4_2 + f1f3_4 + f2f2 + f5f9_76 + f6f8_38 + f7f7_38 + h5 := f0f5_2 + f1f4_2 + f2f3_2 + f6f9_38 + f7f8_38 + h6 := f0f6_2 + f1f5_4 + f2f4_2 + f3f3_2 + f7f9_76 + f8f8_19 + h7 := f0f7_2 + f1f6_2 + f2f5_2 + f3f4_2 + f8f9_38 + h8 := f0f8_2 + f1f7_4 + f2f6_2 + f3f5_4 + f4f4 + f9f9_38 + h9 := f0f9_2 + f1f8_2 + f2f7_2 + f3f6_2 + f4f5_2 + var carry [10]int64 + + carry[0] = (h0 + (1 << 25)) >> 26 + h1 += carry[0] + h0 -= carry[0] << 26 + carry[4] = (h4 + (1 << 25)) >> 26 + h5 += carry[4] + h4 -= carry[4] << 26 + + carry[1] = (h1 + (1 << 24)) >> 25 + h2 += carry[1] + h1 -= carry[1] << 25 + carry[5] = (h5 + (1 << 24)) >> 25 + h6 += carry[5] + h5 -= carry[5] << 25 + + carry[2] = (h2 + (1 << 25)) >> 26 + h3 += carry[2] + h2 -= carry[2] << 26 + carry[6] = (h6 + (1 << 25)) >> 26 + h7 += carry[6] + h6 -= carry[6] << 26 + + carry[3] = (h3 + (1 << 24)) >> 25 + h4 += carry[3] + h3 -= carry[3] << 25 + carry[7] = (h7 + (1 << 24)) >> 25 + h8 += carry[7] + h7 -= carry[7] << 25 + + carry[4] = (h4 + (1 << 25)) >> 26 + h5 += carry[4] + h4 -= carry[4] << 26 + carry[8] = (h8 + (1 << 25)) >> 26 + h9 += carry[8] + h8 -= carry[8] << 26 + + carry[9] = (h9 + (1 << 24)) >> 25 + h0 += carry[9] * 19 + h9 -= carry[9] << 25 + + carry[0] = (h0 + (1 << 25)) >> 26 + h1 += carry[0] + h0 -= carry[0] << 26 + + h[0] = int32(h0) + h[1] = int32(h1) + h[2] = int32(h2) + h[3] = int32(h3) + h[4] = int32(h4) + h[5] = int32(h5) + h[6] = int32(h6) + h[7] = int32(h7) + h[8] = int32(h8) + h[9] = int32(h9) +} + +// feMul121666 calculates h = f * 121666. Can overlap h with f. +// +// Preconditions: +// |f| bounded by 1.1*2^26,1.1*2^25,1.1*2^26,1.1*2^25,etc. +// +// Postconditions: +// |h| bounded by 1.1*2^25,1.1*2^24,1.1*2^25,1.1*2^24,etc. +func feMul121666(h, f *fieldElement) { + h0 := int64(f[0]) * 121666 + h1 := int64(f[1]) * 121666 + h2 := int64(f[2]) * 121666 + h3 := int64(f[3]) * 121666 + h4 := int64(f[4]) * 121666 + h5 := int64(f[5]) * 121666 + h6 := int64(f[6]) * 121666 + h7 := int64(f[7]) * 121666 + h8 := int64(f[8]) * 121666 + h9 := int64(f[9]) * 121666 + var carry [10]int64 + + carry[9] = (h9 + (1 << 24)) >> 25 + h0 += carry[9] * 19 + h9 -= carry[9] << 25 + carry[1] = (h1 + (1 << 24)) >> 25 + h2 += carry[1] + h1 -= carry[1] << 25 + carry[3] = (h3 + (1 << 24)) >> 25 + h4 += carry[3] + h3 -= carry[3] << 25 + carry[5] = (h5 + (1 << 24)) >> 25 + h6 += carry[5] + h5 -= carry[5] << 25 + carry[7] = (h7 + (1 << 24)) >> 25 + h8 += carry[7] + h7 -= carry[7] << 25 + + carry[0] = (h0 + (1 << 25)) >> 26 + h1 += carry[0] + h0 -= carry[0] << 26 + carry[2] = (h2 + (1 << 25)) >> 26 + h3 += carry[2] + h2 -= carry[2] << 26 + carry[4] = (h4 + (1 << 25)) >> 26 + h5 += carry[4] + h4 -= carry[4] << 26 + carry[6] = (h6 + (1 << 25)) >> 26 + h7 += carry[6] + h6 -= carry[6] << 26 + carry[8] = (h8 + (1 << 25)) >> 26 + h9 += carry[8] + h8 -= carry[8] << 26 + + h[0] = int32(h0) + h[1] = int32(h1) + h[2] = int32(h2) + h[3] = int32(h3) + h[4] = int32(h4) + h[5] = int32(h5) + h[6] = int32(h6) + h[7] = int32(h7) + h[8] = int32(h8) + h[9] = int32(h9) +} + +// feInvert sets out = z^-1. +func feInvert(out, z *fieldElement) { + var t0, t1, t2, t3 fieldElement + var i int + + feSquare(&t0, z) + for i = 1; i < 1; i++ { + feSquare(&t0, &t0) + } + feSquare(&t1, &t0) + for i = 1; i < 2; i++ { + feSquare(&t1, &t1) + } + feMul(&t1, z, &t1) + feMul(&t0, &t0, &t1) + feSquare(&t2, &t0) + for i = 1; i < 1; i++ { + feSquare(&t2, &t2) + } + feMul(&t1, &t1, &t2) + feSquare(&t2, &t1) + for i = 1; i < 5; i++ { + feSquare(&t2, &t2) + } + feMul(&t1, &t2, &t1) + feSquare(&t2, &t1) + for i = 1; i < 10; i++ { + feSquare(&t2, &t2) + } + feMul(&t2, &t2, &t1) + feSquare(&t3, &t2) + for i = 1; i < 20; i++ { + feSquare(&t3, &t3) + } + feMul(&t2, &t3, &t2) + feSquare(&t2, &t2) + for i = 1; i < 10; i++ { + feSquare(&t2, &t2) + } + feMul(&t1, &t2, &t1) + feSquare(&t2, &t1) + for i = 1; i < 50; i++ { + feSquare(&t2, &t2) + } + feMul(&t2, &t2, &t1) + feSquare(&t3, &t2) + for i = 1; i < 100; i++ { + feSquare(&t3, &t3) + } + feMul(&t2, &t3, &t2) + feSquare(&t2, &t2) + for i = 1; i < 50; i++ { + feSquare(&t2, &t2) + } + feMul(&t1, &t2, &t1) + feSquare(&t1, &t1) + for i = 1; i < 5; i++ { + feSquare(&t1, &t1) + } + feMul(out, &t1, &t0) +} + +func scalarMultGeneric(out, in, base *[32]byte) { + var e [32]byte + + copy(e[:], in[:]) + e[0] &= 248 + e[31] &= 127 + e[31] |= 64 + + var x1, x2, z2, x3, z3, tmp0, tmp1 fieldElement + feFromBytes(&x1, base) + feOne(&x2) + feCopy(&x3, &x1) + feOne(&z3) + + swap := int32(0) + for pos := 254; pos >= 0; pos-- { + b := e[pos/8] >> uint(pos&7) + b &= 1 + swap ^= int32(b) + feCSwap(&x2, &x3, swap) + feCSwap(&z2, &z3, swap) + swap = int32(b) + + feSub(&tmp0, &x3, &z3) + feSub(&tmp1, &x2, &z2) + feAdd(&x2, &x2, &z2) + feAdd(&z2, &x3, &z3) + feMul(&z3, &tmp0, &x2) + feMul(&z2, &z2, &tmp1) + feSquare(&tmp0, &tmp1) + feSquare(&tmp1, &x2) + feAdd(&x3, &z3, &z2) + feSub(&z2, &z3, &z2) + feMul(&x2, &tmp1, &tmp0) + feSub(&tmp1, &tmp1, &tmp0) + feSquare(&z2, &z2) + feMul121666(&z3, &tmp1) + feSquare(&x3, &x3) + feAdd(&tmp0, &tmp0, &z3) + feMul(&z3, &x1, &z2) + feMul(&z2, &tmp1, &tmp0) + } + + feCSwap(&x2, &x3, swap) + feCSwap(&z2, &z3, swap) + + feInvert(&z2, &z2) + feMul(&x2, &x2, &z2) + feToBytes(out, &x2) +} diff --git a/vendor/golang.org/x/crypto/curve25519/curve25519_noasm.go b/vendor/golang.org/x/crypto/curve25519/curve25519_noasm.go new file mode 100644 index 00000000000..047d49afc27 --- /dev/null +++ b/vendor/golang.org/x/crypto/curve25519/curve25519_noasm.go @@ -0,0 +1,11 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !amd64 gccgo appengine purego + +package curve25519 + +func scalarMult(out, in, base *[32]byte) { + scalarMultGeneric(out, in, base) +} diff --git a/vendor/golang.org/x/crypto/curve25519/doc.go b/vendor/golang.org/x/crypto/curve25519/doc.go deleted file mode 100644 index da9b10d9c1f..00000000000 --- a/vendor/golang.org/x/crypto/curve25519/doc.go +++ /dev/null @@ -1,23 +0,0 @@ -// Copyright 2012 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// Package curve25519 provides an implementation of scalar multiplication on -// the elliptic curve known as curve25519. See https://cr.yp.to/ecdh.html -package curve25519 // import "golang.org/x/crypto/curve25519" - -// basePoint is the x coordinate of the generator of the curve. -var basePoint = [32]byte{9, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0} - -// ScalarMult sets dst to the product in*base where dst and base are the x -// coordinates of group points and all values are in little-endian form. -func ScalarMult(dst, in, base *[32]byte) { - scalarMult(dst, in, base) -} - -// ScalarBaseMult sets dst to the product in*base where dst and base are the x -// coordinates of group points, base is the standard generator and all values -// are in little-endian form. -func ScalarBaseMult(dst, in *[32]byte) { - ScalarMult(dst, in, &basePoint) -} diff --git a/vendor/golang.org/x/crypto/curve25519/freeze_amd64.s b/vendor/golang.org/x/crypto/curve25519/freeze_amd64.s deleted file mode 100644 index 390816106ee..00000000000 --- a/vendor/golang.org/x/crypto/curve25519/freeze_amd64.s +++ /dev/null @@ -1,73 +0,0 @@ -// Copyright 2012 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// This code was translated into a form compatible with 6a from the public -// domain sources in SUPERCOP: https://bench.cr.yp.to/supercop.html - -// +build amd64,!gccgo,!appengine - -#include "const_amd64.h" - -// func freeze(inout *[5]uint64) -TEXT ·freeze(SB),7,$0-8 - MOVQ inout+0(FP), DI - - MOVQ 0(DI),SI - MOVQ 8(DI),DX - MOVQ 16(DI),CX - MOVQ 24(DI),R8 - MOVQ 32(DI),R9 - MOVQ $REDMASK51,AX - MOVQ AX,R10 - SUBQ $18,R10 - MOVQ $3,R11 -REDUCELOOP: - MOVQ SI,R12 - SHRQ $51,R12 - ANDQ AX,SI - ADDQ R12,DX - MOVQ DX,R12 - SHRQ $51,R12 - ANDQ AX,DX - ADDQ R12,CX - MOVQ CX,R12 - SHRQ $51,R12 - ANDQ AX,CX - ADDQ R12,R8 - MOVQ R8,R12 - SHRQ $51,R12 - ANDQ AX,R8 - ADDQ R12,R9 - MOVQ R9,R12 - SHRQ $51,R12 - ANDQ AX,R9 - IMUL3Q $19,R12,R12 - ADDQ R12,SI - SUBQ $1,R11 - JA REDUCELOOP - MOVQ $1,R12 - CMPQ R10,SI - CMOVQLT R11,R12 - CMPQ AX,DX - CMOVQNE R11,R12 - CMPQ AX,CX - CMOVQNE R11,R12 - CMPQ AX,R8 - CMOVQNE R11,R12 - CMPQ AX,R9 - CMOVQNE R11,R12 - NEGQ R12 - ANDQ R12,AX - ANDQ R12,R10 - SUBQ R10,SI - SUBQ AX,DX - SUBQ AX,CX - SUBQ AX,R8 - SUBQ AX,R9 - MOVQ SI,0(DI) - MOVQ DX,8(DI) - MOVQ CX,16(DI) - MOVQ R8,24(DI) - MOVQ R9,32(DI) - RET diff --git a/vendor/golang.org/x/crypto/curve25519/mul_amd64.s b/vendor/golang.org/x/crypto/curve25519/mul_amd64.s deleted file mode 100644 index 1f76d1a3f59..00000000000 --- a/vendor/golang.org/x/crypto/curve25519/mul_amd64.s +++ /dev/null @@ -1,169 +0,0 @@ -// Copyright 2012 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// This code was translated into a form compatible with 6a from the public -// domain sources in SUPERCOP: https://bench.cr.yp.to/supercop.html - -// +build amd64,!gccgo,!appengine - -#include "const_amd64.h" - -// func mul(dest, a, b *[5]uint64) -TEXT ·mul(SB),0,$16-24 - MOVQ dest+0(FP), DI - MOVQ a+8(FP), SI - MOVQ b+16(FP), DX - - MOVQ DX,CX - MOVQ 24(SI),DX - IMUL3Q $19,DX,AX - MOVQ AX,0(SP) - MULQ 16(CX) - MOVQ AX,R8 - MOVQ DX,R9 - MOVQ 32(SI),DX - IMUL3Q $19,DX,AX - MOVQ AX,8(SP) - MULQ 8(CX) - ADDQ AX,R8 - ADCQ DX,R9 - MOVQ 0(SI),AX - MULQ 0(CX) - ADDQ AX,R8 - ADCQ DX,R9 - MOVQ 0(SI),AX - MULQ 8(CX) - MOVQ AX,R10 - MOVQ DX,R11 - MOVQ 0(SI),AX - MULQ 16(CX) - MOVQ AX,R12 - MOVQ DX,R13 - MOVQ 0(SI),AX - MULQ 24(CX) - MOVQ AX,R14 - MOVQ DX,R15 - MOVQ 0(SI),AX - MULQ 32(CX) - MOVQ AX,BX - MOVQ DX,BP - MOVQ 8(SI),AX - MULQ 0(CX) - ADDQ AX,R10 - ADCQ DX,R11 - MOVQ 8(SI),AX - MULQ 8(CX) - ADDQ AX,R12 - ADCQ DX,R13 - MOVQ 8(SI),AX - MULQ 16(CX) - ADDQ AX,R14 - ADCQ DX,R15 - MOVQ 8(SI),AX - MULQ 24(CX) - ADDQ AX,BX - ADCQ DX,BP - MOVQ 8(SI),DX - IMUL3Q $19,DX,AX - MULQ 32(CX) - ADDQ AX,R8 - ADCQ DX,R9 - MOVQ 16(SI),AX - MULQ 0(CX) - ADDQ AX,R12 - ADCQ DX,R13 - MOVQ 16(SI),AX - MULQ 8(CX) - ADDQ AX,R14 - ADCQ DX,R15 - MOVQ 16(SI),AX - MULQ 16(CX) - ADDQ AX,BX - ADCQ DX,BP - MOVQ 16(SI),DX - IMUL3Q $19,DX,AX - MULQ 24(CX) - ADDQ AX,R8 - ADCQ DX,R9 - MOVQ 16(SI),DX - IMUL3Q $19,DX,AX - MULQ 32(CX) - ADDQ AX,R10 - ADCQ DX,R11 - MOVQ 24(SI),AX - MULQ 0(CX) - ADDQ AX,R14 - ADCQ DX,R15 - MOVQ 24(SI),AX - MULQ 8(CX) - ADDQ AX,BX - ADCQ DX,BP - MOVQ 0(SP),AX - MULQ 24(CX) - ADDQ AX,R10 - ADCQ DX,R11 - MOVQ 0(SP),AX - MULQ 32(CX) - ADDQ AX,R12 - ADCQ DX,R13 - MOVQ 32(SI),AX - MULQ 0(CX) - ADDQ AX,BX - ADCQ DX,BP - MOVQ 8(SP),AX - MULQ 16(CX) - ADDQ AX,R10 - ADCQ DX,R11 - MOVQ 8(SP),AX - MULQ 24(CX) - ADDQ AX,R12 - ADCQ DX,R13 - MOVQ 8(SP),AX - MULQ 32(CX) - ADDQ AX,R14 - ADCQ DX,R15 - MOVQ $REDMASK51,SI - SHLQ $13,R8,R9 - ANDQ SI,R8 - SHLQ $13,R10,R11 - ANDQ SI,R10 - ADDQ R9,R10 - SHLQ $13,R12,R13 - ANDQ SI,R12 - ADDQ R11,R12 - SHLQ $13,R14,R15 - ANDQ SI,R14 - ADDQ R13,R14 - SHLQ $13,BX,BP - ANDQ SI,BX - ADDQ R15,BX - IMUL3Q $19,BP,DX - ADDQ DX,R8 - MOVQ R8,DX - SHRQ $51,DX - ADDQ R10,DX - MOVQ DX,CX - SHRQ $51,DX - ANDQ SI,R8 - ADDQ R12,DX - MOVQ DX,R9 - SHRQ $51,DX - ANDQ SI,CX - ADDQ R14,DX - MOVQ DX,AX - SHRQ $51,DX - ANDQ SI,R9 - ADDQ BX,DX - MOVQ DX,R10 - SHRQ $51,DX - ANDQ SI,AX - IMUL3Q $19,DX,DX - ADDQ DX,R8 - ANDQ SI,R10 - MOVQ R8,0(DI) - MOVQ CX,8(DI) - MOVQ R9,16(DI) - MOVQ AX,24(DI) - MOVQ R10,32(DI) - RET diff --git a/vendor/golang.org/x/crypto/curve25519/square_amd64.s b/vendor/golang.org/x/crypto/curve25519/square_amd64.s deleted file mode 100644 index 07511a45af2..00000000000 --- a/vendor/golang.org/x/crypto/curve25519/square_amd64.s +++ /dev/null @@ -1,132 +0,0 @@ -// Copyright 2012 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// This code was translated into a form compatible with 6a from the public -// domain sources in SUPERCOP: https://bench.cr.yp.to/supercop.html - -// +build amd64,!gccgo,!appengine - -#include "const_amd64.h" - -// func square(out, in *[5]uint64) -TEXT ·square(SB),7,$0-16 - MOVQ out+0(FP), DI - MOVQ in+8(FP), SI - - MOVQ 0(SI),AX - MULQ 0(SI) - MOVQ AX,CX - MOVQ DX,R8 - MOVQ 0(SI),AX - SHLQ $1,AX - MULQ 8(SI) - MOVQ AX,R9 - MOVQ DX,R10 - MOVQ 0(SI),AX - SHLQ $1,AX - MULQ 16(SI) - MOVQ AX,R11 - MOVQ DX,R12 - MOVQ 0(SI),AX - SHLQ $1,AX - MULQ 24(SI) - MOVQ AX,R13 - MOVQ DX,R14 - MOVQ 0(SI),AX - SHLQ $1,AX - MULQ 32(SI) - MOVQ AX,R15 - MOVQ DX,BX - MOVQ 8(SI),AX - MULQ 8(SI) - ADDQ AX,R11 - ADCQ DX,R12 - MOVQ 8(SI),AX - SHLQ $1,AX - MULQ 16(SI) - ADDQ AX,R13 - ADCQ DX,R14 - MOVQ 8(SI),AX - SHLQ $1,AX - MULQ 24(SI) - ADDQ AX,R15 - ADCQ DX,BX - MOVQ 8(SI),DX - IMUL3Q $38,DX,AX - MULQ 32(SI) - ADDQ AX,CX - ADCQ DX,R8 - MOVQ 16(SI),AX - MULQ 16(SI) - ADDQ AX,R15 - ADCQ DX,BX - MOVQ 16(SI),DX - IMUL3Q $38,DX,AX - MULQ 24(SI) - ADDQ AX,CX - ADCQ DX,R8 - MOVQ 16(SI),DX - IMUL3Q $38,DX,AX - MULQ 32(SI) - ADDQ AX,R9 - ADCQ DX,R10 - MOVQ 24(SI),DX - IMUL3Q $19,DX,AX - MULQ 24(SI) - ADDQ AX,R9 - ADCQ DX,R10 - MOVQ 24(SI),DX - IMUL3Q $38,DX,AX - MULQ 32(SI) - ADDQ AX,R11 - ADCQ DX,R12 - MOVQ 32(SI),DX - IMUL3Q $19,DX,AX - MULQ 32(SI) - ADDQ AX,R13 - ADCQ DX,R14 - MOVQ $REDMASK51,SI - SHLQ $13,CX,R8 - ANDQ SI,CX - SHLQ $13,R9,R10 - ANDQ SI,R9 - ADDQ R8,R9 - SHLQ $13,R11,R12 - ANDQ SI,R11 - ADDQ R10,R11 - SHLQ $13,R13,R14 - ANDQ SI,R13 - ADDQ R12,R13 - SHLQ $13,R15,BX - ANDQ SI,R15 - ADDQ R14,R15 - IMUL3Q $19,BX,DX - ADDQ DX,CX - MOVQ CX,DX - SHRQ $51,DX - ADDQ R9,DX - ANDQ SI,CX - MOVQ DX,R8 - SHRQ $51,DX - ADDQ R11,DX - ANDQ SI,R8 - MOVQ DX,R9 - SHRQ $51,DX - ADDQ R13,DX - ANDQ SI,R9 - MOVQ DX,AX - SHRQ $51,DX - ADDQ R15,DX - ANDQ SI,AX - MOVQ DX,R10 - SHRQ $51,DX - IMUL3Q $19,DX,DX - ADDQ DX,CX - ANDQ SI,R10 - MOVQ CX,0(DI) - MOVQ R8,8(DI) - MOVQ R9,16(DI) - MOVQ AX,24(DI) - MOVQ R10,32(DI) - RET diff --git a/vendor/golang.org/x/crypto/internal/chacha20/chacha_arm64.go b/vendor/golang.org/x/crypto/internal/chacha20/chacha_arm64.go deleted file mode 100644 index ad74e23aef4..00000000000 --- a/vendor/golang.org/x/crypto/internal/chacha20/chacha_arm64.go +++ /dev/null @@ -1,31 +0,0 @@ -// Copyright 2018 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// +build go1.11 -// +build !gccgo - -package chacha20 - -const ( - haveAsm = true - bufSize = 256 -) - -//go:noescape -func xorKeyStreamVX(dst, src []byte, key *[8]uint32, nonce *[3]uint32, counter *uint32) - -func (c *Cipher) xorKeyStreamAsm(dst, src []byte) { - - if len(src) >= bufSize { - xorKeyStreamVX(dst, src, &c.key, &c.nonce, &c.counter) - } - - if len(src)%bufSize != 0 { - i := len(src) - len(src)%bufSize - c.buf = [bufSize]byte{} - copy(c.buf[:], src[i:]) - xorKeyStreamVX(c.buf[:], c.buf[:], &c.key, &c.nonce, &c.counter) - c.len = bufSize - copy(dst[i:], c.buf[:len(src)%bufSize]) - } -} diff --git a/vendor/golang.org/x/crypto/internal/chacha20/chacha_generic.go b/vendor/golang.org/x/crypto/internal/chacha20/chacha_generic.go deleted file mode 100644 index 6570847f5e0..00000000000 --- a/vendor/golang.org/x/crypto/internal/chacha20/chacha_generic.go +++ /dev/null @@ -1,264 +0,0 @@ -// Copyright 2016 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// Package ChaCha20 implements the core ChaCha20 function as specified -// in https://tools.ietf.org/html/rfc7539#section-2.3. -package chacha20 - -import ( - "crypto/cipher" - "encoding/binary" - - "golang.org/x/crypto/internal/subtle" -) - -// assert that *Cipher implements cipher.Stream -var _ cipher.Stream = (*Cipher)(nil) - -// Cipher is a stateful instance of ChaCha20 using a particular key -// and nonce. A *Cipher implements the cipher.Stream interface. -type Cipher struct { - key [8]uint32 - counter uint32 // incremented after each block - nonce [3]uint32 - buf [bufSize]byte // buffer for unused keystream bytes - len int // number of unused keystream bytes at end of buf -} - -// New creates a new ChaCha20 stream cipher with the given key and nonce. -// The initial counter value is set to 0. -func New(key [8]uint32, nonce [3]uint32) *Cipher { - return &Cipher{key: key, nonce: nonce} -} - -// ChaCha20 constants spelling "expand 32-byte k" -const ( - j0 uint32 = 0x61707865 - j1 uint32 = 0x3320646e - j2 uint32 = 0x79622d32 - j3 uint32 = 0x6b206574 -) - -func quarterRound(a, b, c, d uint32) (uint32, uint32, uint32, uint32) { - a += b - d ^= a - d = (d << 16) | (d >> 16) - c += d - b ^= c - b = (b << 12) | (b >> 20) - a += b - d ^= a - d = (d << 8) | (d >> 24) - c += d - b ^= c - b = (b << 7) | (b >> 25) - return a, b, c, d -} - -// XORKeyStream XORs each byte in the given slice with a byte from the -// cipher's key stream. Dst and src must overlap entirely or not at all. -// -// If len(dst) < len(src), XORKeyStream will panic. It is acceptable -// to pass a dst bigger than src, and in that case, XORKeyStream will -// only update dst[:len(src)] and will not touch the rest of dst. -// -// Multiple calls to XORKeyStream behave as if the concatenation of -// the src buffers was passed in a single run. That is, Cipher -// maintains state and does not reset at each XORKeyStream call. -func (s *Cipher) XORKeyStream(dst, src []byte) { - if len(dst) < len(src) { - panic("chacha20: output smaller than input") - } - if subtle.InexactOverlap(dst[:len(src)], src) { - panic("chacha20: invalid buffer overlap") - } - - // xor src with buffered keystream first - if s.len != 0 { - buf := s.buf[len(s.buf)-s.len:] - if len(src) < len(buf) { - buf = buf[:len(src)] - } - td, ts := dst[:len(buf)], src[:len(buf)] // BCE hint - for i, b := range buf { - td[i] = ts[i] ^ b - } - s.len -= len(buf) - if s.len != 0 { - return - } - s.buf = [len(s.buf)]byte{} // zero the empty buffer - src = src[len(buf):] - dst = dst[len(buf):] - } - - if len(src) == 0 { - return - } - if haveAsm { - if uint64(len(src))+uint64(s.counter)*64 > (1<<38)-64 { - panic("chacha20: counter overflow") - } - s.xorKeyStreamAsm(dst, src) - return - } - - // set up a 64-byte buffer to pad out the final block if needed - // (hoisted out of the main loop to avoid spills) - rem := len(src) % 64 // length of final block - fin := len(src) - rem // index of final block - if rem > 0 { - copy(s.buf[len(s.buf)-64:], src[fin:]) - } - - // pre-calculate most of the first round - s1, s5, s9, s13 := quarterRound(j1, s.key[1], s.key[5], s.nonce[0]) - s2, s6, s10, s14 := quarterRound(j2, s.key[2], s.key[6], s.nonce[1]) - s3, s7, s11, s15 := quarterRound(j3, s.key[3], s.key[7], s.nonce[2]) - - n := len(src) - src, dst = src[:n:n], dst[:n:n] // BCE hint - for i := 0; i < n; i += 64 { - // calculate the remainder of the first round - s0, s4, s8, s12 := quarterRound(j0, s.key[0], s.key[4], s.counter) - - // execute the second round - x0, x5, x10, x15 := quarterRound(s0, s5, s10, s15) - x1, x6, x11, x12 := quarterRound(s1, s6, s11, s12) - x2, x7, x8, x13 := quarterRound(s2, s7, s8, s13) - x3, x4, x9, x14 := quarterRound(s3, s4, s9, s14) - - // execute the remaining 18 rounds - for i := 0; i < 9; i++ { - x0, x4, x8, x12 = quarterRound(x0, x4, x8, x12) - x1, x5, x9, x13 = quarterRound(x1, x5, x9, x13) - x2, x6, x10, x14 = quarterRound(x2, x6, x10, x14) - x3, x7, x11, x15 = quarterRound(x3, x7, x11, x15) - - x0, x5, x10, x15 = quarterRound(x0, x5, x10, x15) - x1, x6, x11, x12 = quarterRound(x1, x6, x11, x12) - x2, x7, x8, x13 = quarterRound(x2, x7, x8, x13) - x3, x4, x9, x14 = quarterRound(x3, x4, x9, x14) - } - - x0 += j0 - x1 += j1 - x2 += j2 - x3 += j3 - - x4 += s.key[0] - x5 += s.key[1] - x6 += s.key[2] - x7 += s.key[3] - x8 += s.key[4] - x9 += s.key[5] - x10 += s.key[6] - x11 += s.key[7] - - x12 += s.counter - x13 += s.nonce[0] - x14 += s.nonce[1] - x15 += s.nonce[2] - - // increment the counter - s.counter += 1 - if s.counter == 0 { - panic("chacha20: counter overflow") - } - - // pad to 64 bytes if needed - in, out := src[i:], dst[i:] - if i == fin { - // src[fin:] has already been copied into s.buf before - // the main loop - in, out = s.buf[len(s.buf)-64:], s.buf[len(s.buf)-64:] - } - in, out = in[:64], out[:64] // BCE hint - - // XOR the key stream with the source and write out the result - xor(out[0:], in[0:], x0) - xor(out[4:], in[4:], x1) - xor(out[8:], in[8:], x2) - xor(out[12:], in[12:], x3) - xor(out[16:], in[16:], x4) - xor(out[20:], in[20:], x5) - xor(out[24:], in[24:], x6) - xor(out[28:], in[28:], x7) - xor(out[32:], in[32:], x8) - xor(out[36:], in[36:], x9) - xor(out[40:], in[40:], x10) - xor(out[44:], in[44:], x11) - xor(out[48:], in[48:], x12) - xor(out[52:], in[52:], x13) - xor(out[56:], in[56:], x14) - xor(out[60:], in[60:], x15) - } - // copy any trailing bytes out of the buffer and into dst - if rem != 0 { - s.len = 64 - rem - copy(dst[fin:], s.buf[len(s.buf)-64:]) - } -} - -// Advance discards bytes in the key stream until the next 64 byte block -// boundary is reached and updates the counter accordingly. If the key -// stream is already at a block boundary no bytes will be discarded and -// the counter will be unchanged. -func (s *Cipher) Advance() { - s.len -= s.len % 64 - if s.len == 0 { - s.buf = [len(s.buf)]byte{} - } -} - -// XORKeyStream crypts bytes from in to out using the given key and counters. -// In and out must overlap entirely or not at all. Counter contains the raw -// ChaCha20 counter bytes (i.e. block counter followed by nonce). -func XORKeyStream(out, in []byte, counter *[16]byte, key *[32]byte) { - s := Cipher{ - key: [8]uint32{ - binary.LittleEndian.Uint32(key[0:4]), - binary.LittleEndian.Uint32(key[4:8]), - binary.LittleEndian.Uint32(key[8:12]), - binary.LittleEndian.Uint32(key[12:16]), - binary.LittleEndian.Uint32(key[16:20]), - binary.LittleEndian.Uint32(key[20:24]), - binary.LittleEndian.Uint32(key[24:28]), - binary.LittleEndian.Uint32(key[28:32]), - }, - nonce: [3]uint32{ - binary.LittleEndian.Uint32(counter[4:8]), - binary.LittleEndian.Uint32(counter[8:12]), - binary.LittleEndian.Uint32(counter[12:16]), - }, - counter: binary.LittleEndian.Uint32(counter[0:4]), - } - s.XORKeyStream(out, in) -} - -// HChaCha20 uses the ChaCha20 core to generate a derived key from a key and a -// nonce. It should only be used as part of the XChaCha20 construction. -func HChaCha20(key *[8]uint32, nonce *[4]uint32) [8]uint32 { - x0, x1, x2, x3 := j0, j1, j2, j3 - x4, x5, x6, x7 := key[0], key[1], key[2], key[3] - x8, x9, x10, x11 := key[4], key[5], key[6], key[7] - x12, x13, x14, x15 := nonce[0], nonce[1], nonce[2], nonce[3] - - for i := 0; i < 10; i++ { - x0, x4, x8, x12 = quarterRound(x0, x4, x8, x12) - x1, x5, x9, x13 = quarterRound(x1, x5, x9, x13) - x2, x6, x10, x14 = quarterRound(x2, x6, x10, x14) - x3, x7, x11, x15 = quarterRound(x3, x7, x11, x15) - - x0, x5, x10, x15 = quarterRound(x0, x5, x10, x15) - x1, x6, x11, x12 = quarterRound(x1, x6, x11, x12) - x2, x7, x8, x13 = quarterRound(x2, x7, x8, x13) - x3, x4, x9, x14 = quarterRound(x3, x4, x9, x14) - } - - var out [8]uint32 - out[0], out[1], out[2], out[3] = x0, x1, x2, x3 - out[4], out[5], out[6], out[7] = x12, x13, x14, x15 - return out -} diff --git a/vendor/golang.org/x/crypto/internal/chacha20/chacha_ppc64le.go b/vendor/golang.org/x/crypto/internal/chacha20/chacha_ppc64le.go deleted file mode 100644 index d38a7d380a2..00000000000 --- a/vendor/golang.org/x/crypto/internal/chacha20/chacha_ppc64le.go +++ /dev/null @@ -1,53 +0,0 @@ -// Copyright 2019 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// +build ppc64le,!gccgo,!appengine - -package chacha20 - -import ( - "encoding/binary" -) - -var haveAsm = true - -const bufSize = 256 - -//go:noescape -func chaCha20_ctr32_vsx(out, inp *byte, len int, key *[8]uint32, counter *uint32) - -func (c *Cipher) xorKeyStreamAsm(dst, src []byte) { - // This implementation can handle buffers that aren't multiples of - // 256. - if len(src) >= bufSize { - chaCha20_ctr32_vsx(&dst[0], &src[0], len(src), &c.key, &c.counter) - } else if len(src)%bufSize != 0 { - chaCha20_ctr32_vsx(&c.buf[0], &c.buf[0], bufSize, &c.key, &c.counter) - start := len(src) - len(src)%bufSize - ts, td, tb := src[start:], dst[start:], c.buf[:] - // Unroll loop to XOR 32 bytes per iteration. - for i := 0; i < len(ts)-32; i += 32 { - td, tb = td[:len(ts)], tb[:len(ts)] // bounds check elimination - s0 := binary.LittleEndian.Uint64(ts[0:8]) - s1 := binary.LittleEndian.Uint64(ts[8:16]) - s2 := binary.LittleEndian.Uint64(ts[16:24]) - s3 := binary.LittleEndian.Uint64(ts[24:32]) - b0 := binary.LittleEndian.Uint64(tb[0:8]) - b1 := binary.LittleEndian.Uint64(tb[8:16]) - b2 := binary.LittleEndian.Uint64(tb[16:24]) - b3 := binary.LittleEndian.Uint64(tb[24:32]) - binary.LittleEndian.PutUint64(td[0:8], s0^b0) - binary.LittleEndian.PutUint64(td[8:16], s1^b1) - binary.LittleEndian.PutUint64(td[16:24], s2^b2) - binary.LittleEndian.PutUint64(td[24:32], s3^b3) - ts, td, tb = ts[32:], td[32:], tb[32:] - } - td, tb = td[:len(ts)], tb[:len(ts)] // bounds check elimination - for i, v := range ts { - td[i] = tb[i] ^ v - } - c.len = bufSize - (len(src) % bufSize) - } - -} diff --git a/vendor/golang.org/x/crypto/internal/chacha20/chacha_s390x.go b/vendor/golang.org/x/crypto/internal/chacha20/chacha_s390x.go deleted file mode 100644 index aad645b4476..00000000000 --- a/vendor/golang.org/x/crypto/internal/chacha20/chacha_s390x.go +++ /dev/null @@ -1,29 +0,0 @@ -// Copyright 2018 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// +build s390x,!gccgo,!appengine - -package chacha20 - -import ( - "golang.org/x/sys/cpu" -) - -var haveAsm = cpu.S390X.HasVX - -const bufSize = 256 - -// xorKeyStreamVX is an assembly implementation of XORKeyStream. It must only -// be called when the vector facility is available. -// Implementation in asm_s390x.s. -//go:noescape -func xorKeyStreamVX(dst, src []byte, key *[8]uint32, nonce *[3]uint32, counter *uint32, buf *[256]byte, len *int) - -func (c *Cipher) xorKeyStreamAsm(dst, src []byte) { - xorKeyStreamVX(dst, src, &c.key, &c.nonce, &c.counter, &c.buf, &c.len) -} - -// EXRL targets, DO NOT CALL! -func mvcSrcToBuf() -func mvcBufToDst() diff --git a/vendor/golang.org/x/crypto/openpgp/armor/armor.go b/vendor/golang.org/x/crypto/openpgp/armor/armor.go index 592d1864361..36a6804364c 100644 --- a/vendor/golang.org/x/crypto/openpgp/armor/armor.go +++ b/vendor/golang.org/x/crypto/openpgp/armor/armor.go @@ -62,10 +62,11 @@ var armorEndOfLine = []byte("-----") // lineReader wraps a line based reader. It watches for the end of an armor // block and records the expected CRC value. type lineReader struct { - in *bufio.Reader - buf []byte - eof bool - crc uint32 + in *bufio.Reader + buf []byte + eof bool + crc uint32 + crcSet bool } func (l *lineReader) Read(p []byte) (n int, err error) { @@ -87,6 +88,11 @@ func (l *lineReader) Read(p []byte) (n int, err error) { return 0, ArmorCorrupt } + if bytes.HasPrefix(line, armorEnd) { + l.eof = true + return 0, io.EOF + } + if len(line) == 5 && line[0] == '=' { // This is the checksum line var expectedBytes [3]byte @@ -108,6 +114,7 @@ func (l *lineReader) Read(p []byte) (n int, err error) { } l.eof = true + l.crcSet = true return 0, io.EOF } @@ -141,10 +148,8 @@ func (r *openpgpReader) Read(p []byte) (n int, err error) { n, err = r.b64Reader.Read(p) r.currentCRC = crc24(r.currentCRC, p[:n]) - if err == io.EOF { - if r.lReader.crc != uint32(r.currentCRC&crc24Mask) { - return 0, ArmorCorrupt - } + if err == io.EOF && r.lReader.crcSet && r.lReader.crc != uint32(r.currentCRC&crc24Mask) { + return 0, ArmorCorrupt } return diff --git a/vendor/golang.org/x/crypto/openpgp/elgamal/elgamal.go b/vendor/golang.org/x/crypto/openpgp/elgamal/elgamal.go index 73f4fe37859..72a6a739471 100644 --- a/vendor/golang.org/x/crypto/openpgp/elgamal/elgamal.go +++ b/vendor/golang.org/x/crypto/openpgp/elgamal/elgamal.go @@ -76,7 +76,9 @@ func Encrypt(random io.Reader, pub *PublicKey, msg []byte) (c1, c2 *big.Int, err // Bleichenbacher, Advances in Cryptology (Crypto '98), func Decrypt(priv *PrivateKey, c1, c2 *big.Int) (msg []byte, err error) { s := new(big.Int).Exp(c1, priv.X, priv.P) - s.ModInverse(s, priv.P) + if s.ModInverse(s, priv.P) == nil { + return nil, errors.New("elgamal: invalid private key") + } s.Mul(s, c2) s.Mod(s, priv.P) em := s.Bytes() diff --git a/vendor/golang.org/x/crypto/openpgp/packet/packet.go b/vendor/golang.org/x/crypto/openpgp/packet/packet.go index 5af64c5421b..9728d61d7aa 100644 --- a/vendor/golang.org/x/crypto/openpgp/packet/packet.go +++ b/vendor/golang.org/x/crypto/openpgp/packet/packet.go @@ -14,6 +14,7 @@ import ( "crypto/rsa" "io" "math/big" + "math/bits" "golang.org/x/crypto/cast5" "golang.org/x/crypto/openpgp/errors" @@ -100,33 +101,65 @@ func (r *partialLengthReader) Read(p []byte) (n int, err error) { type partialLengthWriter struct { w io.WriteCloser lengthByte [1]byte + sentFirst bool + buf []byte } +// RFC 4880 4.2.2.4: the first partial length MUST be at least 512 octets long. +const minFirstPartialWrite = 512 + func (w *partialLengthWriter) Write(p []byte) (n int, err error) { + off := 0 + if !w.sentFirst { + if len(w.buf) > 0 || len(p) < minFirstPartialWrite { + off = len(w.buf) + w.buf = append(w.buf, p...) + if len(w.buf) < minFirstPartialWrite { + return len(p), nil + } + p = w.buf + w.buf = nil + } + w.sentFirst = true + } + + power := uint8(30) for len(p) > 0 { - for power := uint(14); power < 32; power-- { - l := 1 << power - if len(p) >= l { - w.lengthByte[0] = 224 + uint8(power) - _, err = w.w.Write(w.lengthByte[:]) - if err != nil { - return - } - var m int - m, err = w.w.Write(p[:l]) - n += m - if err != nil { - return - } - p = p[l:] - break + l := 1 << power + if len(p) < l { + power = uint8(bits.Len32(uint32(len(p)))) - 1 + l = 1 << power + } + w.lengthByte[0] = 224 + power + _, err = w.w.Write(w.lengthByte[:]) + if err == nil { + var m int + m, err = w.w.Write(p[:l]) + n += m + } + if err != nil { + if n < off { + return 0, err } + return n - off, err } + p = p[l:] } - return + return n - off, nil } func (w *partialLengthWriter) Close() error { + if len(w.buf) > 0 { + // In this case we can't send a 512 byte packet. + // Just send what we have. + p := w.buf + w.sentFirst = true + w.buf = nil + if _, err := w.Write(p); err != nil { + return err + } + } + w.lengthByte[0] = 0 _, err := w.w.Write(w.lengthByte[:]) if err != nil { diff --git a/vendor/golang.org/x/crypto/poly1305/bits_compat.go b/vendor/golang.org/x/crypto/poly1305/bits_compat.go new file mode 100644 index 00000000000..157a69f61bd --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/bits_compat.go @@ -0,0 +1,39 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !go1.13 + +package poly1305 + +// Generic fallbacks for the math/bits intrinsics, copied from +// src/math/bits/bits.go. They were added in Go 1.12, but Add64 and Sum64 had +// variable time fallbacks until Go 1.13. + +func bitsAdd64(x, y, carry uint64) (sum, carryOut uint64) { + sum = x + y + carry + carryOut = ((x & y) | ((x | y) &^ sum)) >> 63 + return +} + +func bitsSub64(x, y, borrow uint64) (diff, borrowOut uint64) { + diff = x - y - borrow + borrowOut = ((^x & y) | (^(x ^ y) & diff)) >> 63 + return +} + +func bitsMul64(x, y uint64) (hi, lo uint64) { + const mask32 = 1<<32 - 1 + x0 := x & mask32 + x1 := x >> 32 + y0 := y & mask32 + y1 := y >> 32 + w0 := x0 * y0 + t := x1*y0 + w0>>32 + w1 := t & mask32 + w2 := t >> 32 + w1 += x0 * y1 + hi = x1*y1 + w2 + w1>>32 + lo = x * y + return +} diff --git a/vendor/golang.org/x/crypto/poly1305/bits_go1.13.go b/vendor/golang.org/x/crypto/poly1305/bits_go1.13.go new file mode 100644 index 00000000000..a0a185f0fc7 --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/bits_go1.13.go @@ -0,0 +1,21 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build go1.13 + +package poly1305 + +import "math/bits" + +func bitsAdd64(x, y, carry uint64) (sum, carryOut uint64) { + return bits.Add64(x, y, carry) +} + +func bitsSub64(x, y, borrow uint64) (diff, borrowOut uint64) { + return bits.Sub64(x, y, borrow) +} + +func bitsMul64(x, y uint64) (hi, lo uint64) { + return bits.Mul64(x, y) +} diff --git a/vendor/golang.org/x/crypto/poly1305/mac_noasm.go b/vendor/golang.org/x/crypto/poly1305/mac_noasm.go index a8dd589ae39..d118f30ed56 100644 --- a/vendor/golang.org/x/crypto/poly1305/mac_noasm.go +++ b/vendor/golang.org/x/crypto/poly1305/mac_noasm.go @@ -2,10 +2,8 @@ // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. -// +build !amd64,!ppc64le gccgo appengine +// +build !amd64,!ppc64le,!s390x gccgo purego package poly1305 type mac struct{ macGeneric } - -func newMAC(key *[32]byte) mac { return mac{newMACGeneric(key)} } diff --git a/vendor/golang.org/x/crypto/poly1305/poly1305.go b/vendor/golang.org/x/crypto/poly1305/poly1305.go index d076a562351..9d7a6af09fe 100644 --- a/vendor/golang.org/x/crypto/poly1305/poly1305.go +++ b/vendor/golang.org/x/crypto/poly1305/poly1305.go @@ -22,8 +22,16 @@ import "crypto/subtle" // TagSize is the size, in bytes, of a poly1305 authenticator. const TagSize = 16 -// Verify returns true if mac is a valid authenticator for m with the given -// key. +// Sum generates an authenticator for msg using a one-time key and puts the +// 16-byte result into out. Authenticating two different messages with the same +// key allows an attacker to forge messages at will. +func Sum(out *[16]byte, m []byte, key *[32]byte) { + h := New(key) + h.Write(m) + h.Sum(out[:0]) +} + +// Verify returns true if mac is a valid authenticator for m with the given key. func Verify(mac *[16]byte, m []byte, key *[32]byte) bool { var tmp [16]byte Sum(&tmp, m, key) @@ -40,10 +48,9 @@ func Verify(mac *[16]byte, m []byte, key *[32]byte) bool { // two different messages with the same key allows an attacker // to forge messages at will. func New(key *[32]byte) *MAC { - return &MAC{ - mac: newMAC(key), - finalized: false, - } + m := &MAC{} + initialize(key, &m.macState) + return m } // MAC is an io.Writer computing an authentication tag @@ -52,7 +59,7 @@ func New(key *[32]byte) *MAC { // MAC cannot be used like common hash.Hash implementations, // because using a poly1305 key twice breaks its security. // Therefore writing data to a running MAC after calling -// Sum causes it to panic. +// Sum or Verify causes it to panic. type MAC struct { mac // platform-dependent implementation @@ -65,10 +72,10 @@ func (h *MAC) Size() int { return TagSize } // Write adds more data to the running message authentication code. // It never returns an error. // -// It must not be called after the first call of Sum. +// It must not be called after the first call of Sum or Verify. func (h *MAC) Write(p []byte) (n int, err error) { if h.finalized { - panic("poly1305: write to MAC after Sum") + panic("poly1305: write to MAC after Sum or Verify") } return h.mac.Write(p) } @@ -81,3 +88,12 @@ func (h *MAC) Sum(b []byte) []byte { h.finalized = true return append(b, mac[:]...) } + +// Verify returns whether the authenticator of all data written to +// the message authentication code matches the expected value. +func (h *MAC) Verify(expected []byte) bool { + var mac [TagSize]byte + h.mac.Sum(&mac) + h.finalized = true + return subtle.ConstantTimeCompare(expected, mac[:]) == 1 +} diff --git a/vendor/golang.org/x/crypto/poly1305/sum_amd64.go b/vendor/golang.org/x/crypto/poly1305/sum_amd64.go index 2dbf42aa537..99e5a1d50ef 100644 --- a/vendor/golang.org/x/crypto/poly1305/sum_amd64.go +++ b/vendor/golang.org/x/crypto/poly1305/sum_amd64.go @@ -2,67 +2,46 @@ // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. -// +build amd64,!gccgo,!appengine +// +build !gccgo,!purego package poly1305 //go:noescape -func initialize(state *[7]uint64, key *[32]byte) +func update(state *macState, msg []byte) -//go:noescape -func update(state *[7]uint64, msg []byte) - -//go:noescape -func finalize(tag *[TagSize]byte, state *[7]uint64) - -// Sum generates an authenticator for m using a one-time key and puts the -// 16-byte result into out. Authenticating two different messages with the same -// key allows an attacker to forge messages at will. -func Sum(out *[16]byte, m []byte, key *[32]byte) { - h := newMAC(key) - h.Write(m) - h.Sum(out) -} - -func newMAC(key *[32]byte) (h mac) { - initialize(&h.state, key) - return -} - -type mac struct { - state [7]uint64 // := uint64{ h0, h1, h2, r0, r1, pad0, pad1 } - - buffer [TagSize]byte - offset int -} +// mac is a wrapper for macGeneric that redirects calls that would have gone to +// updateGeneric to update. +// +// Its Write and Sum methods are otherwise identical to the macGeneric ones, but +// using function pointers would carry a major performance cost. +type mac struct{ macGeneric } -func (h *mac) Write(p []byte) (n int, err error) { - n = len(p) +func (h *mac) Write(p []byte) (int, error) { + nn := len(p) if h.offset > 0 { - remaining := TagSize - h.offset - if n < remaining { - h.offset += copy(h.buffer[h.offset:], p) - return n, nil + n := copy(h.buffer[h.offset:], p) + if h.offset+n < TagSize { + h.offset += n + return nn, nil } - copy(h.buffer[h.offset:], p[:remaining]) - p = p[remaining:] + p = p[n:] h.offset = 0 - update(&h.state, h.buffer[:]) + update(&h.macState, h.buffer[:]) } - if nn := len(p) - (len(p) % TagSize); nn > 0 { - update(&h.state, p[:nn]) - p = p[nn:] + if n := len(p) - (len(p) % TagSize); n > 0 { + update(&h.macState, p[:n]) + p = p[n:] } if len(p) > 0 { h.offset += copy(h.buffer[h.offset:], p) } - return n, nil + return nn, nil } func (h *mac) Sum(out *[16]byte) { - state := h.state + state := h.macState if h.offset > 0 { update(&state, h.buffer[:h.offset]) } - finalize(out, &state) + finalize(out, &state.h, &state.s) } diff --git a/vendor/golang.org/x/crypto/poly1305/sum_amd64.s b/vendor/golang.org/x/crypto/poly1305/sum_amd64.s index 7d600f13cc8..8d394a212ee 100644 --- a/vendor/golang.org/x/crypto/poly1305/sum_amd64.s +++ b/vendor/golang.org/x/crypto/poly1305/sum_amd64.s @@ -2,7 +2,7 @@ // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. -// +build amd64,!gccgo,!appengine +// +build !gccgo,!purego #include "textflag.h" @@ -54,10 +54,6 @@ ADCQ t3, h1; \ ADCQ $0, h2 -DATA ·poly1305Mask<>+0x00(SB)/8, $0x0FFFFFFC0FFFFFFF -DATA ·poly1305Mask<>+0x08(SB)/8, $0x0FFFFFFC0FFFFFFC -GLOBL ·poly1305Mask<>(SB), RODATA, $16 - // func update(state *[7]uint64, msg []byte) TEXT ·update(SB), $0-32 MOVQ state+0(FP), DI @@ -110,39 +106,3 @@ done: MOVQ R9, 8(DI) MOVQ R10, 16(DI) RET - -// func initialize(state *[7]uint64, key *[32]byte) -TEXT ·initialize(SB), $0-16 - MOVQ state+0(FP), DI - MOVQ key+8(FP), SI - - // state[0...7] is initialized with zero - MOVOU 0(SI), X0 - MOVOU 16(SI), X1 - MOVOU ·poly1305Mask<>(SB), X2 - PAND X2, X0 - MOVOU X0, 24(DI) - MOVOU X1, 40(DI) - RET - -// func finalize(tag *[TagSize]byte, state *[7]uint64) -TEXT ·finalize(SB), $0-16 - MOVQ tag+0(FP), DI - MOVQ state+8(FP), SI - - MOVQ 0(SI), AX - MOVQ 8(SI), BX - MOVQ 16(SI), CX - MOVQ AX, R8 - MOVQ BX, R9 - SUBQ $0xFFFFFFFFFFFFFFFB, AX - SBBQ $0xFFFFFFFFFFFFFFFF, BX - SBBQ $3, CX - CMOVQCS R8, AX - CMOVQCS R9, BX - ADDQ 40(SI), AX - ADCQ 48(SI), BX - - MOVQ AX, 0(DI) - MOVQ BX, 8(DI) - RET diff --git a/vendor/golang.org/x/crypto/poly1305/sum_arm.go b/vendor/golang.org/x/crypto/poly1305/sum_arm.go deleted file mode 100644 index 5dc321c2f39..00000000000 --- a/vendor/golang.org/x/crypto/poly1305/sum_arm.go +++ /dev/null @@ -1,22 +0,0 @@ -// Copyright 2015 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// +build arm,!gccgo,!appengine,!nacl - -package poly1305 - -// This function is implemented in sum_arm.s -//go:noescape -func poly1305_auth_armv6(out *[16]byte, m *byte, mlen uint32, key *[32]byte) - -// Sum generates an authenticator for m using a one-time key and puts the -// 16-byte result into out. Authenticating two different messages with the same -// key allows an attacker to forge messages at will. -func Sum(out *[16]byte, m []byte, key *[32]byte) { - var mPtr *byte - if len(m) > 0 { - mPtr = &m[0] - } - poly1305_auth_armv6(out, mPtr, uint32(len(m)), key) -} diff --git a/vendor/golang.org/x/crypto/poly1305/sum_arm.s b/vendor/golang.org/x/crypto/poly1305/sum_arm.s deleted file mode 100644 index f70b4ac4845..00000000000 --- a/vendor/golang.org/x/crypto/poly1305/sum_arm.s +++ /dev/null @@ -1,427 +0,0 @@ -// Copyright 2015 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// +build arm,!gccgo,!appengine,!nacl - -#include "textflag.h" - -// This code was translated into a form compatible with 5a from the public -// domain source by Andrew Moon: github.com/floodyberry/poly1305-opt/blob/master/app/extensions/poly1305. - -DATA ·poly1305_init_constants_armv6<>+0x00(SB)/4, $0x3ffffff -DATA ·poly1305_init_constants_armv6<>+0x04(SB)/4, $0x3ffff03 -DATA ·poly1305_init_constants_armv6<>+0x08(SB)/4, $0x3ffc0ff -DATA ·poly1305_init_constants_armv6<>+0x0c(SB)/4, $0x3f03fff -DATA ·poly1305_init_constants_armv6<>+0x10(SB)/4, $0x00fffff -GLOBL ·poly1305_init_constants_armv6<>(SB), 8, $20 - -// Warning: the linker may use R11 to synthesize certain instructions. Please -// take care and verify that no synthetic instructions use it. - -TEXT poly1305_init_ext_armv6<>(SB), NOSPLIT, $0 - // Needs 16 bytes of stack and 64 bytes of space pointed to by R0. (It - // might look like it's only 60 bytes of space but the final four bytes - // will be written by another function.) We need to skip over four - // bytes of stack because that's saving the value of 'g'. - ADD $4, R13, R8 - MOVM.IB [R4-R7], (R8) - MOVM.IA.W (R1), [R2-R5] - MOVW $·poly1305_init_constants_armv6<>(SB), R7 - MOVW R2, R8 - MOVW R2>>26, R9 - MOVW R3>>20, g - MOVW R4>>14, R11 - MOVW R5>>8, R12 - ORR R3<<6, R9, R9 - ORR R4<<12, g, g - ORR R5<<18, R11, R11 - MOVM.IA (R7), [R2-R6] - AND R8, R2, R2 - AND R9, R3, R3 - AND g, R4, R4 - AND R11, R5, R5 - AND R12, R6, R6 - MOVM.IA.W [R2-R6], (R0) - EOR R2, R2, R2 - EOR R3, R3, R3 - EOR R4, R4, R4 - EOR R5, R5, R5 - EOR R6, R6, R6 - MOVM.IA.W [R2-R6], (R0) - MOVM.IA.W (R1), [R2-R5] - MOVM.IA [R2-R6], (R0) - ADD $20, R13, R0 - MOVM.DA (R0), [R4-R7] - RET - -#define MOVW_UNALIGNED(Rsrc, Rdst, Rtmp, offset) \ - MOVBU (offset+0)(Rsrc), Rtmp; \ - MOVBU Rtmp, (offset+0)(Rdst); \ - MOVBU (offset+1)(Rsrc), Rtmp; \ - MOVBU Rtmp, (offset+1)(Rdst); \ - MOVBU (offset+2)(Rsrc), Rtmp; \ - MOVBU Rtmp, (offset+2)(Rdst); \ - MOVBU (offset+3)(Rsrc), Rtmp; \ - MOVBU Rtmp, (offset+3)(Rdst) - -TEXT poly1305_blocks_armv6<>(SB), NOSPLIT, $0 - // Needs 24 bytes of stack for saved registers and then 88 bytes of - // scratch space after that. We assume that 24 bytes at (R13) have - // already been used: four bytes for the link register saved in the - // prelude of poly1305_auth_armv6, four bytes for saving the value of g - // in that function and 16 bytes of scratch space used around - // poly1305_finish_ext_armv6_skip1. - ADD $24, R13, R12 - MOVM.IB [R4-R8, R14], (R12) - MOVW R0, 88(R13) - MOVW R1, 92(R13) - MOVW R2, 96(R13) - MOVW R1, R14 - MOVW R2, R12 - MOVW 56(R0), R8 - WORD $0xe1180008 // TST R8, R8 not working see issue 5921 - EOR R6, R6, R6 - MOVW.EQ $(1<<24), R6 - MOVW R6, 84(R13) - ADD $116, R13, g - MOVM.IA (R0), [R0-R9] - MOVM.IA [R0-R4], (g) - CMP $16, R12 - BLO poly1305_blocks_armv6_done - -poly1305_blocks_armv6_mainloop: - WORD $0xe31e0003 // TST R14, #3 not working see issue 5921 - BEQ poly1305_blocks_armv6_mainloop_aligned - ADD $100, R13, g - MOVW_UNALIGNED(R14, g, R0, 0) - MOVW_UNALIGNED(R14, g, R0, 4) - MOVW_UNALIGNED(R14, g, R0, 8) - MOVW_UNALIGNED(R14, g, R0, 12) - MOVM.IA (g), [R0-R3] - ADD $16, R14 - B poly1305_blocks_armv6_mainloop_loaded - -poly1305_blocks_armv6_mainloop_aligned: - MOVM.IA.W (R14), [R0-R3] - -poly1305_blocks_armv6_mainloop_loaded: - MOVW R0>>26, g - MOVW R1>>20, R11 - MOVW R2>>14, R12 - MOVW R14, 92(R13) - MOVW R3>>8, R4 - ORR R1<<6, g, g - ORR R2<<12, R11, R11 - ORR R3<<18, R12, R12 - BIC $0xfc000000, R0, R0 - BIC $0xfc000000, g, g - MOVW 84(R13), R3 - BIC $0xfc000000, R11, R11 - BIC $0xfc000000, R12, R12 - ADD R0, R5, R5 - ADD g, R6, R6 - ORR R3, R4, R4 - ADD R11, R7, R7 - ADD $116, R13, R14 - ADD R12, R8, R8 - ADD R4, R9, R9 - MOVM.IA (R14), [R0-R4] - MULLU R4, R5, (R11, g) - MULLU R3, R5, (R14, R12) - MULALU R3, R6, (R11, g) - MULALU R2, R6, (R14, R12) - MULALU R2, R7, (R11, g) - MULALU R1, R7, (R14, R12) - ADD R4<<2, R4, R4 - ADD R3<<2, R3, R3 - MULALU R1, R8, (R11, g) - MULALU R0, R8, (R14, R12) - MULALU R0, R9, (R11, g) - MULALU R4, R9, (R14, R12) - MOVW g, 76(R13) - MOVW R11, 80(R13) - MOVW R12, 68(R13) - MOVW R14, 72(R13) - MULLU R2, R5, (R11, g) - MULLU R1, R5, (R14, R12) - MULALU R1, R6, (R11, g) - MULALU R0, R6, (R14, R12) - MULALU R0, R7, (R11, g) - MULALU R4, R7, (R14, R12) - ADD R2<<2, R2, R2 - ADD R1<<2, R1, R1 - MULALU R4, R8, (R11, g) - MULALU R3, R8, (R14, R12) - MULALU R3, R9, (R11, g) - MULALU R2, R9, (R14, R12) - MOVW g, 60(R13) - MOVW R11, 64(R13) - MOVW R12, 52(R13) - MOVW R14, 56(R13) - MULLU R0, R5, (R11, g) - MULALU R4, R6, (R11, g) - MULALU R3, R7, (R11, g) - MULALU R2, R8, (R11, g) - MULALU R1, R9, (R11, g) - ADD $52, R13, R0 - MOVM.IA (R0), [R0-R7] - MOVW g>>26, R12 - MOVW R4>>26, R14 - ORR R11<<6, R12, R12 - ORR R5<<6, R14, R14 - BIC $0xfc000000, g, g - BIC $0xfc000000, R4, R4 - ADD.S R12, R0, R0 - ADC $0, R1, R1 - ADD.S R14, R6, R6 - ADC $0, R7, R7 - MOVW R0>>26, R12 - MOVW R6>>26, R14 - ORR R1<<6, R12, R12 - ORR R7<<6, R14, R14 - BIC $0xfc000000, R0, R0 - BIC $0xfc000000, R6, R6 - ADD R14<<2, R14, R14 - ADD.S R12, R2, R2 - ADC $0, R3, R3 - ADD R14, g, g - MOVW R2>>26, R12 - MOVW g>>26, R14 - ORR R3<<6, R12, R12 - BIC $0xfc000000, g, R5 - BIC $0xfc000000, R2, R7 - ADD R12, R4, R4 - ADD R14, R0, R0 - MOVW R4>>26, R12 - BIC $0xfc000000, R4, R8 - ADD R12, R6, R9 - MOVW 96(R13), R12 - MOVW 92(R13), R14 - MOVW R0, R6 - CMP $32, R12 - SUB $16, R12, R12 - MOVW R12, 96(R13) - BHS poly1305_blocks_armv6_mainloop - -poly1305_blocks_armv6_done: - MOVW 88(R13), R12 - MOVW R5, 20(R12) - MOVW R6, 24(R12) - MOVW R7, 28(R12) - MOVW R8, 32(R12) - MOVW R9, 36(R12) - ADD $48, R13, R0 - MOVM.DA (R0), [R4-R8, R14] - RET - -#define MOVHUP_UNALIGNED(Rsrc, Rdst, Rtmp) \ - MOVBU.P 1(Rsrc), Rtmp; \ - MOVBU.P Rtmp, 1(Rdst); \ - MOVBU.P 1(Rsrc), Rtmp; \ - MOVBU.P Rtmp, 1(Rdst) - -#define MOVWP_UNALIGNED(Rsrc, Rdst, Rtmp) \ - MOVHUP_UNALIGNED(Rsrc, Rdst, Rtmp); \ - MOVHUP_UNALIGNED(Rsrc, Rdst, Rtmp) - -// func poly1305_auth_armv6(out *[16]byte, m *byte, mlen uint32, key *[32]key) -TEXT ·poly1305_auth_armv6(SB), $196-16 - // The value 196, just above, is the sum of 64 (the size of the context - // structure) and 132 (the amount of stack needed). - // - // At this point, the stack pointer (R13) has been moved down. It - // points to the saved link register and there's 196 bytes of free - // space above it. - // - // The stack for this function looks like: - // - // +--------------------- - // | - // | 64 bytes of context structure - // | - // +--------------------- - // | - // | 112 bytes for poly1305_blocks_armv6 - // | - // +--------------------- - // | 16 bytes of final block, constructed at - // | poly1305_finish_ext_armv6_skip8 - // +--------------------- - // | four bytes of saved 'g' - // +--------------------- - // | lr, saved by prelude <- R13 points here - // +--------------------- - MOVW g, 4(R13) - - MOVW out+0(FP), R4 - MOVW m+4(FP), R5 - MOVW mlen+8(FP), R6 - MOVW key+12(FP), R7 - - ADD $136, R13, R0 // 136 = 4 + 4 + 16 + 112 - MOVW R7, R1 - - // poly1305_init_ext_armv6 will write to the stack from R13+4, but - // that's ok because none of the other values have been written yet. - BL poly1305_init_ext_armv6<>(SB) - BIC.S $15, R6, R2 - BEQ poly1305_auth_armv6_noblocks - ADD $136, R13, R0 - MOVW R5, R1 - ADD R2, R5, R5 - SUB R2, R6, R6 - BL poly1305_blocks_armv6<>(SB) - -poly1305_auth_armv6_noblocks: - ADD $136, R13, R0 - MOVW R5, R1 - MOVW R6, R2 - MOVW R4, R3 - - MOVW R0, R5 - MOVW R1, R6 - MOVW R2, R7 - MOVW R3, R8 - AND.S R2, R2, R2 - BEQ poly1305_finish_ext_armv6_noremaining - EOR R0, R0 - ADD $8, R13, R9 // 8 = offset to 16 byte scratch space - MOVW R0, (R9) - MOVW R0, 4(R9) - MOVW R0, 8(R9) - MOVW R0, 12(R9) - WORD $0xe3110003 // TST R1, #3 not working see issue 5921 - BEQ poly1305_finish_ext_armv6_aligned - WORD $0xe3120008 // TST R2, #8 not working see issue 5921 - BEQ poly1305_finish_ext_armv6_skip8 - MOVWP_UNALIGNED(R1, R9, g) - MOVWP_UNALIGNED(R1, R9, g) - -poly1305_finish_ext_armv6_skip8: - WORD $0xe3120004 // TST $4, R2 not working see issue 5921 - BEQ poly1305_finish_ext_armv6_skip4 - MOVWP_UNALIGNED(R1, R9, g) - -poly1305_finish_ext_armv6_skip4: - WORD $0xe3120002 // TST $2, R2 not working see issue 5921 - BEQ poly1305_finish_ext_armv6_skip2 - MOVHUP_UNALIGNED(R1, R9, g) - B poly1305_finish_ext_armv6_skip2 - -poly1305_finish_ext_armv6_aligned: - WORD $0xe3120008 // TST R2, #8 not working see issue 5921 - BEQ poly1305_finish_ext_armv6_skip8_aligned - MOVM.IA.W (R1), [g-R11] - MOVM.IA.W [g-R11], (R9) - -poly1305_finish_ext_armv6_skip8_aligned: - WORD $0xe3120004 // TST $4, R2 not working see issue 5921 - BEQ poly1305_finish_ext_armv6_skip4_aligned - MOVW.P 4(R1), g - MOVW.P g, 4(R9) - -poly1305_finish_ext_armv6_skip4_aligned: - WORD $0xe3120002 // TST $2, R2 not working see issue 5921 - BEQ poly1305_finish_ext_armv6_skip2 - MOVHU.P 2(R1), g - MOVH.P g, 2(R9) - -poly1305_finish_ext_armv6_skip2: - WORD $0xe3120001 // TST $1, R2 not working see issue 5921 - BEQ poly1305_finish_ext_armv6_skip1 - MOVBU.P 1(R1), g - MOVBU.P g, 1(R9) - -poly1305_finish_ext_armv6_skip1: - MOVW $1, R11 - MOVBU R11, 0(R9) - MOVW R11, 56(R5) - MOVW R5, R0 - ADD $8, R13, R1 - MOVW $16, R2 - BL poly1305_blocks_armv6<>(SB) - -poly1305_finish_ext_armv6_noremaining: - MOVW 20(R5), R0 - MOVW 24(R5), R1 - MOVW 28(R5), R2 - MOVW 32(R5), R3 - MOVW 36(R5), R4 - MOVW R4>>26, R12 - BIC $0xfc000000, R4, R4 - ADD R12<<2, R12, R12 - ADD R12, R0, R0 - MOVW R0>>26, R12 - BIC $0xfc000000, R0, R0 - ADD R12, R1, R1 - MOVW R1>>26, R12 - BIC $0xfc000000, R1, R1 - ADD R12, R2, R2 - MOVW R2>>26, R12 - BIC $0xfc000000, R2, R2 - ADD R12, R3, R3 - MOVW R3>>26, R12 - BIC $0xfc000000, R3, R3 - ADD R12, R4, R4 - ADD $5, R0, R6 - MOVW R6>>26, R12 - BIC $0xfc000000, R6, R6 - ADD R12, R1, R7 - MOVW R7>>26, R12 - BIC $0xfc000000, R7, R7 - ADD R12, R2, g - MOVW g>>26, R12 - BIC $0xfc000000, g, g - ADD R12, R3, R11 - MOVW $-(1<<26), R12 - ADD R11>>26, R12, R12 - BIC $0xfc000000, R11, R11 - ADD R12, R4, R9 - MOVW R9>>31, R12 - SUB $1, R12 - AND R12, R6, R6 - AND R12, R7, R7 - AND R12, g, g - AND R12, R11, R11 - AND R12, R9, R9 - MVN R12, R12 - AND R12, R0, R0 - AND R12, R1, R1 - AND R12, R2, R2 - AND R12, R3, R3 - AND R12, R4, R4 - ORR R6, R0, R0 - ORR R7, R1, R1 - ORR g, R2, R2 - ORR R11, R3, R3 - ORR R9, R4, R4 - ORR R1<<26, R0, R0 - MOVW R1>>6, R1 - ORR R2<<20, R1, R1 - MOVW R2>>12, R2 - ORR R3<<14, R2, R2 - MOVW R3>>18, R3 - ORR R4<<8, R3, R3 - MOVW 40(R5), R6 - MOVW 44(R5), R7 - MOVW 48(R5), g - MOVW 52(R5), R11 - ADD.S R6, R0, R0 - ADC.S R7, R1, R1 - ADC.S g, R2, R2 - ADC.S R11, R3, R3 - MOVM.IA [R0-R3], (R8) - MOVW R5, R12 - EOR R0, R0, R0 - EOR R1, R1, R1 - EOR R2, R2, R2 - EOR R3, R3, R3 - EOR R4, R4, R4 - EOR R5, R5, R5 - EOR R6, R6, R6 - EOR R7, R7, R7 - MOVM.IA.W [R0-R7], (R12) - MOVM.IA [R0-R7], (R12) - MOVW 4(R13), g - RET diff --git a/vendor/golang.org/x/crypto/poly1305/sum_generic.go b/vendor/golang.org/x/crypto/poly1305/sum_generic.go index bab76ef0d83..c942a65904f 100644 --- a/vendor/golang.org/x/crypto/poly1305/sum_generic.go +++ b/vendor/golang.org/x/crypto/poly1305/sum_generic.go @@ -2,171 +2,309 @@ // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. +// This file provides the generic implementation of Sum and MAC. Other files +// might provide optimized assembly implementations of some of this code. + package poly1305 import "encoding/binary" -const ( - msgBlock = uint32(1 << 24) - finalBlock = uint32(0) -) +// Poly1305 [RFC 7539] is a relatively simple algorithm: the authentication tag +// for a 64 bytes message is approximately +// +// s + m[0:16] * r⁴ + m[16:32] * r³ + m[32:48] * r² + m[48:64] * r mod 2¹³⁰ - 5 +// +// for some secret r and s. It can be computed sequentially like +// +// for len(msg) > 0: +// h += read(msg, 16) +// h *= r +// h %= 2¹³⁰ - 5 +// return h + s +// +// All the complexity is about doing performant constant-time math on numbers +// larger than any available numeric type. -// sumGeneric generates an authenticator for msg using a one-time key and -// puts the 16-byte result into out. This is the generic implementation of -// Sum and should be called if no assembly implementation is available. func sumGeneric(out *[TagSize]byte, msg []byte, key *[32]byte) { h := newMACGeneric(key) h.Write(msg) h.Sum(out) } -func newMACGeneric(key *[32]byte) (h macGeneric) { - h.r[0] = binary.LittleEndian.Uint32(key[0:]) & 0x3ffffff - h.r[1] = (binary.LittleEndian.Uint32(key[3:]) >> 2) & 0x3ffff03 - h.r[2] = (binary.LittleEndian.Uint32(key[6:]) >> 4) & 0x3ffc0ff - h.r[3] = (binary.LittleEndian.Uint32(key[9:]) >> 6) & 0x3f03fff - h.r[4] = (binary.LittleEndian.Uint32(key[12:]) >> 8) & 0x00fffff - - h.s[0] = binary.LittleEndian.Uint32(key[16:]) - h.s[1] = binary.LittleEndian.Uint32(key[20:]) - h.s[2] = binary.LittleEndian.Uint32(key[24:]) - h.s[3] = binary.LittleEndian.Uint32(key[28:]) - return +func newMACGeneric(key *[32]byte) macGeneric { + m := macGeneric{} + initialize(key, &m.macState) + return m +} + +// macState holds numbers in saturated 64-bit little-endian limbs. That is, +// the value of [x0, x1, x2] is x[0] + x[1] * 2⁶⁴ + x[2] * 2¹²⁸. +type macState struct { + // h is the main accumulator. It is to be interpreted modulo 2¹³⁰ - 5, but + // can grow larger during and after rounds. It must, however, remain below + // 2 * (2¹³⁰ - 5). + h [3]uint64 + // r and s are the private key components. + r [2]uint64 + s [2]uint64 } type macGeneric struct { - h, r [5]uint32 - s [4]uint32 + macState buffer [TagSize]byte offset int } -func (h *macGeneric) Write(p []byte) (n int, err error) { - n = len(p) +// Write splits the incoming message into TagSize chunks, and passes them to +// update. It buffers incomplete chunks. +func (h *macGeneric) Write(p []byte) (int, error) { + nn := len(p) if h.offset > 0 { - remaining := TagSize - h.offset - if n < remaining { - h.offset += copy(h.buffer[h.offset:], p) - return n, nil + n := copy(h.buffer[h.offset:], p) + if h.offset+n < TagSize { + h.offset += n + return nn, nil } - copy(h.buffer[h.offset:], p[:remaining]) - p = p[remaining:] + p = p[n:] h.offset = 0 - updateGeneric(h.buffer[:], msgBlock, &(h.h), &(h.r)) + updateGeneric(&h.macState, h.buffer[:]) } - if nn := len(p) - (len(p) % TagSize); nn > 0 { - updateGeneric(p, msgBlock, &(h.h), &(h.r)) - p = p[nn:] + if n := len(p) - (len(p) % TagSize); n > 0 { + updateGeneric(&h.macState, p[:n]) + p = p[n:] } if len(p) > 0 { h.offset += copy(h.buffer[h.offset:], p) } - return n, nil + return nn, nil } -func (h *macGeneric) Sum(out *[16]byte) { - H, R := h.h, h.r +// Sum flushes the last incomplete chunk from the buffer, if any, and generates +// the MAC output. It does not modify its state, in order to allow for multiple +// calls to Sum, even if no Write is allowed after Sum. +func (h *macGeneric) Sum(out *[TagSize]byte) { + state := h.macState if h.offset > 0 { - var buffer [TagSize]byte - copy(buffer[:], h.buffer[:h.offset]) - buffer[h.offset] = 1 // invariant: h.offset < TagSize - updateGeneric(buffer[:], finalBlock, &H, &R) + updateGeneric(&state, h.buffer[:h.offset]) } - finalizeGeneric(out, &H, &(h.s)) + finalize(out, &state.h, &state.s) +} + +// [rMask0, rMask1] is the specified Poly1305 clamping mask in little-endian. It +// clears some bits of the secret coefficient to make it possible to implement +// multiplication more efficiently. +const ( + rMask0 = 0x0FFFFFFC0FFFFFFF + rMask1 = 0x0FFFFFFC0FFFFFFC +) + +// initialize loads the 256-bit key into the two 128-bit secret values r and s. +func initialize(key *[32]byte, m *macState) { + m.r[0] = binary.LittleEndian.Uint64(key[0:8]) & rMask0 + m.r[1] = binary.LittleEndian.Uint64(key[8:16]) & rMask1 + m.s[0] = binary.LittleEndian.Uint64(key[16:24]) + m.s[1] = binary.LittleEndian.Uint64(key[24:32]) +} + +// uint128 holds a 128-bit number as two 64-bit limbs, for use with the +// bits.Mul64 and bits.Add64 intrinsics. +type uint128 struct { + lo, hi uint64 +} + +func mul64(a, b uint64) uint128 { + hi, lo := bitsMul64(a, b) + return uint128{lo, hi} } -func updateGeneric(msg []byte, flag uint32, h, r *[5]uint32) { - h0, h1, h2, h3, h4 := h[0], h[1], h[2], h[3], h[4] - r0, r1, r2, r3, r4 := uint64(r[0]), uint64(r[1]), uint64(r[2]), uint64(r[3]), uint64(r[4]) - R1, R2, R3, R4 := r1*5, r2*5, r3*5, r4*5 - - for len(msg) >= TagSize { - // h += msg - h0 += binary.LittleEndian.Uint32(msg[0:]) & 0x3ffffff - h1 += (binary.LittleEndian.Uint32(msg[3:]) >> 2) & 0x3ffffff - h2 += (binary.LittleEndian.Uint32(msg[6:]) >> 4) & 0x3ffffff - h3 += (binary.LittleEndian.Uint32(msg[9:]) >> 6) & 0x3ffffff - h4 += (binary.LittleEndian.Uint32(msg[12:]) >> 8) | flag - - // h *= r - d0 := (uint64(h0) * r0) + (uint64(h1) * R4) + (uint64(h2) * R3) + (uint64(h3) * R2) + (uint64(h4) * R1) - d1 := (d0 >> 26) + (uint64(h0) * r1) + (uint64(h1) * r0) + (uint64(h2) * R4) + (uint64(h3) * R3) + (uint64(h4) * R2) - d2 := (d1 >> 26) + (uint64(h0) * r2) + (uint64(h1) * r1) + (uint64(h2) * r0) + (uint64(h3) * R4) + (uint64(h4) * R3) - d3 := (d2 >> 26) + (uint64(h0) * r3) + (uint64(h1) * r2) + (uint64(h2) * r1) + (uint64(h3) * r0) + (uint64(h4) * R4) - d4 := (d3 >> 26) + (uint64(h0) * r4) + (uint64(h1) * r3) + (uint64(h2) * r2) + (uint64(h3) * r1) + (uint64(h4) * r0) - - // h %= p - h0 = uint32(d0) & 0x3ffffff - h1 = uint32(d1) & 0x3ffffff - h2 = uint32(d2) & 0x3ffffff - h3 = uint32(d3) & 0x3ffffff - h4 = uint32(d4) & 0x3ffffff - - h0 += uint32(d4>>26) * 5 - h1 += h0 >> 26 - h0 = h0 & 0x3ffffff - - msg = msg[TagSize:] +func add128(a, b uint128) uint128 { + lo, c := bitsAdd64(a.lo, b.lo, 0) + hi, c := bitsAdd64(a.hi, b.hi, c) + if c != 0 { + panic("poly1305: unexpected overflow") } + return uint128{lo, hi} +} - h[0], h[1], h[2], h[3], h[4] = h0, h1, h2, h3, h4 +func shiftRightBy2(a uint128) uint128 { + a.lo = a.lo>>2 | (a.hi&3)<<62 + a.hi = a.hi >> 2 + return a } -func finalizeGeneric(out *[TagSize]byte, h *[5]uint32, s *[4]uint32) { - h0, h1, h2, h3, h4 := h[0], h[1], h[2], h[3], h[4] - - // h %= p reduction - h2 += h1 >> 26 - h1 &= 0x3ffffff - h3 += h2 >> 26 - h2 &= 0x3ffffff - h4 += h3 >> 26 - h3 &= 0x3ffffff - h0 += 5 * (h4 >> 26) - h4 &= 0x3ffffff - h1 += h0 >> 26 - h0 &= 0x3ffffff - - // h - p - t0 := h0 + 5 - t1 := h1 + (t0 >> 26) - t2 := h2 + (t1 >> 26) - t3 := h3 + (t2 >> 26) - t4 := h4 + (t3 >> 26) - (1 << 26) - t0 &= 0x3ffffff - t1 &= 0x3ffffff - t2 &= 0x3ffffff - t3 &= 0x3ffffff - - // select h if h < p else h - p - t_mask := (t4 >> 31) - 1 - h_mask := ^t_mask - h0 = (h0 & h_mask) | (t0 & t_mask) - h1 = (h1 & h_mask) | (t1 & t_mask) - h2 = (h2 & h_mask) | (t2 & t_mask) - h3 = (h3 & h_mask) | (t3 & t_mask) - h4 = (h4 & h_mask) | (t4 & t_mask) - - // h %= 2^128 - h0 |= h1 << 26 - h1 = ((h1 >> 6) | (h2 << 20)) - h2 = ((h2 >> 12) | (h3 << 14)) - h3 = ((h3 >> 18) | (h4 << 8)) - - // s: the s part of the key - // tag = (h + s) % (2^128) - t := uint64(h0) + uint64(s[0]) - h0 = uint32(t) - t = uint64(h1) + uint64(s[1]) + (t >> 32) - h1 = uint32(t) - t = uint64(h2) + uint64(s[2]) + (t >> 32) - h2 = uint32(t) - t = uint64(h3) + uint64(s[3]) + (t >> 32) - h3 = uint32(t) - - binary.LittleEndian.PutUint32(out[0:], h0) - binary.LittleEndian.PutUint32(out[4:], h1) - binary.LittleEndian.PutUint32(out[8:], h2) - binary.LittleEndian.PutUint32(out[12:], h3) +// updateGeneric absorbs msg into the state.h accumulator. For each chunk m of +// 128 bits of message, it computes +// +// h₊ = (h + m) * r mod 2¹³⁰ - 5 +// +// If the msg length is not a multiple of TagSize, it assumes the last +// incomplete chunk is the final one. +func updateGeneric(state *macState, msg []byte) { + h0, h1, h2 := state.h[0], state.h[1], state.h[2] + r0, r1 := state.r[0], state.r[1] + + for len(msg) > 0 { + var c uint64 + + // For the first step, h + m, we use a chain of bits.Add64 intrinsics. + // The resulting value of h might exceed 2¹³⁰ - 5, but will be partially + // reduced at the end of the multiplication below. + // + // The spec requires us to set a bit just above the message size, not to + // hide leading zeroes. For full chunks, that's 1 << 128, so we can just + // add 1 to the most significant (2¹²⁸) limb, h2. + if len(msg) >= TagSize { + h0, c = bitsAdd64(h0, binary.LittleEndian.Uint64(msg[0:8]), 0) + h1, c = bitsAdd64(h1, binary.LittleEndian.Uint64(msg[8:16]), c) + h2 += c + 1 + + msg = msg[TagSize:] + } else { + var buf [TagSize]byte + copy(buf[:], msg) + buf[len(msg)] = 1 + + h0, c = bitsAdd64(h0, binary.LittleEndian.Uint64(buf[0:8]), 0) + h1, c = bitsAdd64(h1, binary.LittleEndian.Uint64(buf[8:16]), c) + h2 += c + + msg = nil + } + + // Multiplication of big number limbs is similar to elementary school + // columnar multiplication. Instead of digits, there are 64-bit limbs. + // + // We are multiplying a 3 limbs number, h, by a 2 limbs number, r. + // + // h2 h1 h0 x + // r1 r0 = + // ---------------- + // h2r0 h1r0 h0r0 <-- individual 128-bit products + // + h2r1 h1r1 h0r1 + // ------------------------ + // m3 m2 m1 m0 <-- result in 128-bit overlapping limbs + // ------------------------ + // m3.hi m2.hi m1.hi m0.hi <-- carry propagation + // + m3.lo m2.lo m1.lo m0.lo + // ------------------------------- + // t4 t3 t2 t1 t0 <-- final result in 64-bit limbs + // + // The main difference from pen-and-paper multiplication is that we do + // carry propagation in a separate step, as if we wrote two digit sums + // at first (the 128-bit limbs), and then carried the tens all at once. + + h0r0 := mul64(h0, r0) + h1r0 := mul64(h1, r0) + h2r0 := mul64(h2, r0) + h0r1 := mul64(h0, r1) + h1r1 := mul64(h1, r1) + h2r1 := mul64(h2, r1) + + // Since h2 is known to be at most 7 (5 + 1 + 1), and r0 and r1 have their + // top 4 bits cleared by rMask{0,1}, we know that their product is not going + // to overflow 64 bits, so we can ignore the high part of the products. + // + // This also means that the product doesn't have a fifth limb (t4). + if h2r0.hi != 0 { + panic("poly1305: unexpected overflow") + } + if h2r1.hi != 0 { + panic("poly1305: unexpected overflow") + } + + m0 := h0r0 + m1 := add128(h1r0, h0r1) // These two additions don't overflow thanks again + m2 := add128(h2r0, h1r1) // to the 4 masked bits at the top of r0 and r1. + m3 := h2r1 + + t0 := m0.lo + t1, c := bitsAdd64(m1.lo, m0.hi, 0) + t2, c := bitsAdd64(m2.lo, m1.hi, c) + t3, _ := bitsAdd64(m3.lo, m2.hi, c) + + // Now we have the result as 4 64-bit limbs, and we need to reduce it + // modulo 2¹³⁰ - 5. The special shape of this Crandall prime lets us do + // a cheap partial reduction according to the reduction identity + // + // c * 2¹³⁰ + n = c * 5 + n mod 2¹³⁰ - 5 + // + // because 2¹³⁰ = 5 mod 2¹³⁰ - 5. Partial reduction since the result is + // likely to be larger than 2¹³⁰ - 5, but still small enough to fit the + // assumptions we make about h in the rest of the code. + // + // See also https://speakerdeck.com/gtank/engineering-prime-numbers?slide=23 + + // We split the final result at the 2¹³⁰ mark into h and cc, the carry. + // Note that the carry bits are effectively shifted left by 2, in other + // words, cc = c * 4 for the c in the reduction identity. + h0, h1, h2 = t0, t1, t2&maskLow2Bits + cc := uint128{t2 & maskNotLow2Bits, t3} + + // To add c * 5 to h, we first add cc = c * 4, and then add (cc >> 2) = c. + + h0, c = bitsAdd64(h0, cc.lo, 0) + h1, c = bitsAdd64(h1, cc.hi, c) + h2 += c + + cc = shiftRightBy2(cc) + + h0, c = bitsAdd64(h0, cc.lo, 0) + h1, c = bitsAdd64(h1, cc.hi, c) + h2 += c + + // h2 is at most 3 + 1 + 1 = 5, making the whole of h at most + // + // 5 * 2¹²⁸ + (2¹²⁸ - 1) = 6 * 2¹²⁸ - 1 + } + + state.h[0], state.h[1], state.h[2] = h0, h1, h2 +} + +const ( + maskLow2Bits uint64 = 0x0000000000000003 + maskNotLow2Bits uint64 = ^maskLow2Bits +) + +// select64 returns x if v == 1 and y if v == 0, in constant time. +func select64(v, x, y uint64) uint64 { return ^(v-1)&x | (v-1)&y } + +// [p0, p1, p2] is 2¹³⁰ - 5 in little endian order. +const ( + p0 = 0xFFFFFFFFFFFFFFFB + p1 = 0xFFFFFFFFFFFFFFFF + p2 = 0x0000000000000003 +) + +// finalize completes the modular reduction of h and computes +// +// out = h + s mod 2¹²⁸ +// +func finalize(out *[TagSize]byte, h *[3]uint64, s *[2]uint64) { + h0, h1, h2 := h[0], h[1], h[2] + + // After the partial reduction in updateGeneric, h might be more than + // 2¹³⁰ - 5, but will be less than 2 * (2¹³⁰ - 5). To complete the reduction + // in constant time, we compute t = h - (2¹³⁰ - 5), and select h as the + // result if the subtraction underflows, and t otherwise. + + hMinusP0, b := bitsSub64(h0, p0, 0) + hMinusP1, b := bitsSub64(h1, p1, b) + _, b = bitsSub64(h2, p2, b) + + // h = h if h < p else h - p + h0 = select64(b, h0, hMinusP0) + h1 = select64(b, h1, hMinusP1) + + // Finally, we compute the last Poly1305 step + // + // tag = h + s mod 2¹²⁸ + // + // by just doing a wide addition with the 128 low bits of h and discarding + // the overflow. + h0, c := bitsAdd64(h0, s[0], 0) + h1, _ = bitsAdd64(h1, s[1], c) + + binary.LittleEndian.PutUint64(out[0:8], h0) + binary.LittleEndian.PutUint64(out[8:16], h1) } diff --git a/vendor/golang.org/x/crypto/poly1305/sum_noasm.go b/vendor/golang.org/x/crypto/poly1305/sum_noasm.go deleted file mode 100644 index 8a9c2070b9f..00000000000 --- a/vendor/golang.org/x/crypto/poly1305/sum_noasm.go +++ /dev/null @@ -1,16 +0,0 @@ -// Copyright 2018 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// +build s390x,!go1.11 !arm,!amd64,!s390x,!ppc64le gccgo appengine nacl - -package poly1305 - -// Sum generates an authenticator for msg using a one-time key and puts the -// 16-byte result into out. Authenticating two different messages with the same -// key allows an attacker to forge messages at will. -func Sum(out *[TagSize]byte, msg []byte, key *[32]byte) { - h := newMAC(key) - h.Write(msg) - h.Sum(out) -} diff --git a/vendor/golang.org/x/crypto/poly1305/sum_ppc64le.go b/vendor/golang.org/x/crypto/poly1305/sum_ppc64le.go index 2402b6371bf..2e7a120b192 100644 --- a/vendor/golang.org/x/crypto/poly1305/sum_ppc64le.go +++ b/vendor/golang.org/x/crypto/poly1305/sum_ppc64le.go @@ -2,67 +2,46 @@ // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. -// +build ppc64le,!gccgo,!appengine +// +build !gccgo,!purego package poly1305 //go:noescape -func initialize(state *[7]uint64, key *[32]byte) +func update(state *macState, msg []byte) -//go:noescape -func update(state *[7]uint64, msg []byte) - -//go:noescape -func finalize(tag *[TagSize]byte, state *[7]uint64) - -// Sum generates an authenticator for m using a one-time key and puts the -// 16-byte result into out. Authenticating two different messages with the same -// key allows an attacker to forge messages at will. -func Sum(out *[16]byte, m []byte, key *[32]byte) { - h := newMAC(key) - h.Write(m) - h.Sum(out) -} - -func newMAC(key *[32]byte) (h mac) { - initialize(&h.state, key) - return -} - -type mac struct { - state [7]uint64 // := uint64{ h0, h1, h2, r0, r1, pad0, pad1 } - - buffer [TagSize]byte - offset int -} +// mac is a wrapper for macGeneric that redirects calls that would have gone to +// updateGeneric to update. +// +// Its Write and Sum methods are otherwise identical to the macGeneric ones, but +// using function pointers would carry a major performance cost. +type mac struct{ macGeneric } -func (h *mac) Write(p []byte) (n int, err error) { - n = len(p) +func (h *mac) Write(p []byte) (int, error) { + nn := len(p) if h.offset > 0 { - remaining := TagSize - h.offset - if n < remaining { - h.offset += copy(h.buffer[h.offset:], p) - return n, nil + n := copy(h.buffer[h.offset:], p) + if h.offset+n < TagSize { + h.offset += n + return nn, nil } - copy(h.buffer[h.offset:], p[:remaining]) - p = p[remaining:] + p = p[n:] h.offset = 0 - update(&h.state, h.buffer[:]) + update(&h.macState, h.buffer[:]) } - if nn := len(p) - (len(p) % TagSize); nn > 0 { - update(&h.state, p[:nn]) - p = p[nn:] + if n := len(p) - (len(p) % TagSize); n > 0 { + update(&h.macState, p[:n]) + p = p[n:] } if len(p) > 0 { h.offset += copy(h.buffer[h.offset:], p) } - return n, nil + return nn, nil } func (h *mac) Sum(out *[16]byte) { - state := h.state + state := h.macState if h.offset > 0 { update(&state, h.buffer[:h.offset]) } - finalize(out, &state) + finalize(out, &state.h, &state.s) } diff --git a/vendor/golang.org/x/crypto/poly1305/sum_ppc64le.s b/vendor/golang.org/x/crypto/poly1305/sum_ppc64le.s index 55c7167ec98..4e028138796 100644 --- a/vendor/golang.org/x/crypto/poly1305/sum_ppc64le.s +++ b/vendor/golang.org/x/crypto/poly1305/sum_ppc64le.s @@ -2,7 +2,7 @@ // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. -// +build ppc64le,!gccgo,!appengine +// +build !gccgo,!purego #include "textflag.h" @@ -58,7 +58,6 @@ DATA ·poly1305Mask<>+0x08(SB)/8, $0x0FFFFFFC0FFFFFFC GLOBL ·poly1305Mask<>(SB), RODATA, $16 // func update(state *[7]uint64, msg []byte) - TEXT ·update(SB), $0-32 MOVD state+0(FP), R3 MOVD msg_base+8(FP), R4 @@ -180,68 +179,3 @@ done: MOVD R9, 8(R3) MOVD R10, 16(R3) RET - -// func initialize(state *[7]uint64, key *[32]byte) -TEXT ·initialize(SB), $0-16 - MOVD state+0(FP), R3 - MOVD key+8(FP), R4 - - // state[0...7] is initialized with zero - // Load key - MOVD 0(R4), R5 - MOVD 8(R4), R6 - MOVD 16(R4), R7 - MOVD 24(R4), R8 - - // Address of key mask - MOVD $·poly1305Mask<>(SB), R9 - - // Save original key in state - MOVD R7, 40(R3) - MOVD R8, 48(R3) - - // Get mask - MOVD (R9), R7 - MOVD 8(R9), R8 - - // And with key - AND R5, R7, R5 - AND R6, R8, R6 - - // Save masked key in state - MOVD R5, 24(R3) - MOVD R6, 32(R3) - RET - -// func finalize(tag *[TagSize]byte, state *[7]uint64) -TEXT ·finalize(SB), $0-16 - MOVD tag+0(FP), R3 - MOVD state+8(FP), R4 - - // Get h0, h1, h2 from state - MOVD 0(R4), R5 - MOVD 8(R4), R6 - MOVD 16(R4), R7 - - // Save h0, h1 - MOVD R5, R8 - MOVD R6, R9 - MOVD $3, R20 - MOVD $-1, R21 - SUBC $-5, R5 - SUBE R21, R6 - SUBE R20, R7 - MOVD $0, R21 - SUBZE R21 - - // Check for carry - CMP $0, R21 - ISEL $2, R5, R8, R5 - ISEL $2, R6, R9, R6 - MOVD 40(R4), R8 - MOVD 48(R4), R9 - ADDC R8, R5 - ADDE R9, R6 - MOVD R5, 0(R3) - MOVD R6, 8(R3) - RET diff --git a/vendor/golang.org/x/crypto/poly1305/sum_s390x.go b/vendor/golang.org/x/crypto/poly1305/sum_s390x.go index ec99e07e9fb..958fedc0790 100644 --- a/vendor/golang.org/x/crypto/poly1305/sum_s390x.go +++ b/vendor/golang.org/x/crypto/poly1305/sum_s390x.go @@ -2,7 +2,7 @@ // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. -// +build s390x,go1.11,!gccgo,!appengine +// +build !gccgo,!purego package poly1305 @@ -10,33 +10,66 @@ import ( "golang.org/x/sys/cpu" ) -// poly1305vx is an assembly implementation of Poly1305 that uses vector +// updateVX is an assembly implementation of Poly1305 that uses vector // instructions. It must only be called if the vector facility (vx) is // available. //go:noescape -func poly1305vx(out *[16]byte, m *byte, mlen uint64, key *[32]byte) +func updateVX(state *macState, msg []byte) -// poly1305vmsl is an assembly implementation of Poly1305 that uses vector -// instructions, including VMSL. It must only be called if the vector facility (vx) is -// available and if VMSL is supported. -//go:noescape -func poly1305vmsl(out *[16]byte, m *byte, mlen uint64, key *[32]byte) - -// Sum generates an authenticator for m using a one-time key and puts the -// 16-byte result into out. Authenticating two different messages with the same -// key allows an attacker to forge messages at will. -func Sum(out *[16]byte, m []byte, key *[32]byte) { - if cpu.S390X.HasVX { - var mPtr *byte - if len(m) > 0 { - mPtr = &m[0] +// mac is a replacement for macGeneric that uses a larger buffer and redirects +// calls that would have gone to updateGeneric to updateVX if the vector +// facility is installed. +// +// A larger buffer is required for good performance because the vector +// implementation has a higher fixed cost per call than the generic +// implementation. +type mac struct { + macState + + buffer [16 * TagSize]byte // size must be a multiple of block size (16) + offset int +} + +func (h *mac) Write(p []byte) (int, error) { + nn := len(p) + if h.offset > 0 { + n := copy(h.buffer[h.offset:], p) + if h.offset+n < len(h.buffer) { + h.offset += n + return nn, nil + } + p = p[n:] + h.offset = 0 + if cpu.S390X.HasVX { + updateVX(&h.macState, h.buffer[:]) + } else { + updateGeneric(&h.macState, h.buffer[:]) } - if cpu.S390X.HasVXE && len(m) > 256 { - poly1305vmsl(out, mPtr, uint64(len(m)), key) + } + + tail := len(p) % len(h.buffer) // number of bytes to copy into buffer + body := len(p) - tail // number of bytes to process now + if body > 0 { + if cpu.S390X.HasVX { + updateVX(&h.macState, p[:body]) } else { - poly1305vx(out, mPtr, uint64(len(m)), key) + updateGeneric(&h.macState, p[:body]) } - } else { - sumGeneric(out, m, key) } + h.offset = copy(h.buffer[:], p[body:]) // copy tail bytes - can be 0 + return nn, nil +} + +func (h *mac) Sum(out *[TagSize]byte) { + state := h.macState + remainder := h.buffer[:h.offset] + + // Use the generic implementation if we have 2 or fewer blocks left + // to sum. The vector implementation has a higher startup time. + if cpu.S390X.HasVX && len(remainder) > 2*TagSize { + updateVX(&state, remainder) + } else if len(remainder) > 0 { + updateGeneric(&state, remainder) + } + finalize(out, &state.h, &state.s) } diff --git a/vendor/golang.org/x/crypto/poly1305/sum_s390x.s b/vendor/golang.org/x/crypto/poly1305/sum_s390x.s index ca5a309d867..0fa9ee6e0bf 100644 --- a/vendor/golang.org/x/crypto/poly1305/sum_s390x.s +++ b/vendor/golang.org/x/crypto/poly1305/sum_s390x.s @@ -2,115 +2,187 @@ // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. -// +build s390x,go1.11,!gccgo,!appengine +// +build !gccgo,!purego #include "textflag.h" -// Implementation of Poly1305 using the vector facility (vx). - -// constants -#define MOD26 V0 -#define EX0 V1 -#define EX1 V2 -#define EX2 V3 - -// temporaries -#define T_0 V4 -#define T_1 V5 -#define T_2 V6 -#define T_3 V7 -#define T_4 V8 - -// key (r) -#define R_0 V9 -#define R_1 V10 -#define R_2 V11 -#define R_3 V12 -#define R_4 V13 -#define R5_1 V14 -#define R5_2 V15 -#define R5_3 V16 -#define R5_4 V17 -#define RSAVE_0 R5 -#define RSAVE_1 R6 -#define RSAVE_2 R7 -#define RSAVE_3 R8 -#define RSAVE_4 R9 -#define R5SAVE_1 V28 -#define R5SAVE_2 V29 -#define R5SAVE_3 V30 -#define R5SAVE_4 V31 - -// message block -#define F_0 V18 -#define F_1 V19 -#define F_2 V20 -#define F_3 V21 -#define F_4 V22 - -// accumulator -#define H_0 V23 -#define H_1 V24 -#define H_2 V25 -#define H_3 V26 -#define H_4 V27 - -GLOBL ·keyMask<>(SB), RODATA, $16 -DATA ·keyMask<>+0(SB)/8, $0xffffff0ffcffff0f -DATA ·keyMask<>+8(SB)/8, $0xfcffff0ffcffff0f - -GLOBL ·bswapMask<>(SB), RODATA, $16 -DATA ·bswapMask<>+0(SB)/8, $0x0f0e0d0c0b0a0908 -DATA ·bswapMask<>+8(SB)/8, $0x0706050403020100 - -GLOBL ·constants<>(SB), RODATA, $64 -// MOD26 -DATA ·constants<>+0(SB)/8, $0x3ffffff -DATA ·constants<>+8(SB)/8, $0x3ffffff +// This implementation of Poly1305 uses the vector facility (vx) +// to process up to 2 blocks (32 bytes) per iteration using an +// algorithm based on the one described in: +// +// NEON crypto, Daniel J. Bernstein & Peter Schwabe +// https://cryptojedi.org/papers/neoncrypto-20120320.pdf +// +// This algorithm uses 5 26-bit limbs to represent a 130-bit +// value. These limbs are, for the most part, zero extended and +// placed into 64-bit vector register elements. Each vector +// register is 128-bits wide and so holds 2 of these elements. +// Using 26-bit limbs allows us plenty of headroom to accomodate +// accumulations before and after multiplication without +// overflowing either 32-bits (before multiplication) or 64-bits +// (after multiplication). +// +// In order to parallelise the operations required to calculate +// the sum we use two separate accumulators and then sum those +// in an extra final step. For compatibility with the generic +// implementation we perform this summation at the end of every +// updateVX call. +// +// To use two accumulators we must multiply the message blocks +// by r² rather than r. Only the final message block should be +// multiplied by r. +// +// Example: +// +// We want to calculate the sum (h) for a 64 byte message (m): +// +// h = m[0:16]r⁴ + m[16:32]r³ + m[32:48]r² + m[48:64]r +// +// To do this we split the calculation into the even indices +// and odd indices of the message. These form our SIMD 'lanes': +// +// h = m[ 0:16]r⁴ + m[32:48]r² + <- lane 0 +// m[16:32]r³ + m[48:64]r <- lane 1 +// +// To calculate this iteratively we refactor so that both lanes +// are written in terms of r² and r: +// +// h = (m[ 0:16]r² + m[32:48])r² + <- lane 0 +// (m[16:32]r² + m[48:64])r <- lane 1 +// ^ ^ +// | coefficients for second iteration +// coefficients for first iteration +// +// So in this case we would have two iterations. In the first +// both lanes are multiplied by r². In the second only the +// first lane is multiplied by r² and the second lane is +// instead multiplied by r. This gives use the odd and even +// powers of r that we need from the original equation. +// +// Notation: +// +// h - accumulator +// r - key +// m - message +// +// [a, b] - SIMD register holding two 64-bit values +// [a, b, c, d] - SIMD register holding four 32-bit values +// xᵢ[n] - limb n of variable x with bit width i +// +// Limbs are expressed in little endian order, so for 26-bit +// limbs x₂₆[4] will be the most significant limb and x₂₆[0] +// will be the least significant limb. + +// masking constants +#define MOD24 V0 // [0x0000000000ffffff, 0x0000000000ffffff] - mask low 24-bits +#define MOD26 V1 // [0x0000000003ffffff, 0x0000000003ffffff] - mask low 26-bits + +// expansion constants (see EXPAND macro) +#define EX0 V2 +#define EX1 V3 +#define EX2 V4 + +// key (r², r or 1 depending on context) +#define R_0 V5 +#define R_1 V6 +#define R_2 V7 +#define R_3 V8 +#define R_4 V9 + +// precalculated coefficients (5r², 5r or 0 depending on context) +#define R5_1 V10 +#define R5_2 V11 +#define R5_3 V12 +#define R5_4 V13 + +// message block (m) +#define M_0 V14 +#define M_1 V15 +#define M_2 V16 +#define M_3 V17 +#define M_4 V18 + +// accumulator (h) +#define H_0 V19 +#define H_1 V20 +#define H_2 V21 +#define H_3 V22 +#define H_4 V23 + +// temporary registers (for short-lived values) +#define T_0 V24 +#define T_1 V25 +#define T_2 V26 +#define T_3 V27 +#define T_4 V28 + +GLOBL ·constants<>(SB), RODATA, $0x30 // EX0 -DATA ·constants<>+16(SB)/8, $0x0006050403020100 -DATA ·constants<>+24(SB)/8, $0x1016151413121110 +DATA ·constants<>+0x00(SB)/8, $0x0006050403020100 +DATA ·constants<>+0x08(SB)/8, $0x1016151413121110 // EX1 -DATA ·constants<>+32(SB)/8, $0x060c0b0a09080706 -DATA ·constants<>+40(SB)/8, $0x161c1b1a19181716 +DATA ·constants<>+0x10(SB)/8, $0x060c0b0a09080706 +DATA ·constants<>+0x18(SB)/8, $0x161c1b1a19181716 // EX2 -DATA ·constants<>+48(SB)/8, $0x0d0d0d0d0d0f0e0d -DATA ·constants<>+56(SB)/8, $0x1d1d1d1d1d1f1e1d - -// h = (f*g) % (2**130-5) [partial reduction] +DATA ·constants<>+0x20(SB)/8, $0x0d0d0d0d0d0f0e0d +DATA ·constants<>+0x28(SB)/8, $0x1d1d1d1d1d1f1e1d + +// MULTIPLY multiplies each lane of f and g, partially reduced +// modulo 2¹³⁰ - 5. The result, h, consists of partial products +// in each lane that need to be reduced further to produce the +// final result. +// +// h₁₃₀ = (f₁₃₀g₁₃₀) % 2¹³⁰ + (5f₁₃₀g₁₃₀) / 2¹³⁰ +// +// Note that the multiplication by 5 of the high bits is +// achieved by precalculating the multiplication of four of the +// g coefficients by 5. These are g51-g54. #define MULTIPLY(f0, f1, f2, f3, f4, g0, g1, g2, g3, g4, g51, g52, g53, g54, h0, h1, h2, h3, h4) \ VMLOF f0, g0, h0 \ - VMLOF f0, g1, h1 \ - VMLOF f0, g2, h2 \ VMLOF f0, g3, h3 \ + VMLOF f0, g1, h1 \ VMLOF f0, g4, h4 \ + VMLOF f0, g2, h2 \ VMLOF f1, g54, T_0 \ - VMLOF f1, g0, T_1 \ - VMLOF f1, g1, T_2 \ VMLOF f1, g2, T_3 \ + VMLOF f1, g0, T_1 \ VMLOF f1, g3, T_4 \ + VMLOF f1, g1, T_2 \ VMALOF f2, g53, h0, h0 \ - VMALOF f2, g54, h1, h1 \ - VMALOF f2, g0, h2, h2 \ VMALOF f2, g1, h3, h3 \ + VMALOF f2, g54, h1, h1 \ VMALOF f2, g2, h4, h4 \ + VMALOF f2, g0, h2, h2 \ VMALOF f3, g52, T_0, T_0 \ - VMALOF f3, g53, T_1, T_1 \ - VMALOF f3, g54, T_2, T_2 \ VMALOF f3, g0, T_3, T_3 \ + VMALOF f3, g53, T_1, T_1 \ VMALOF f3, g1, T_4, T_4 \ + VMALOF f3, g54, T_2, T_2 \ VMALOF f4, g51, h0, h0 \ - VMALOF f4, g52, h1, h1 \ - VMALOF f4, g53, h2, h2 \ VMALOF f4, g54, h3, h3 \ + VMALOF f4, g52, h1, h1 \ VMALOF f4, g0, h4, h4 \ + VMALOF f4, g53, h2, h2 \ VAG T_0, h0, h0 \ - VAG T_1, h1, h1 \ - VAG T_2, h2, h2 \ VAG T_3, h3, h3 \ - VAG T_4, h4, h4 - -// carry h0->h1 h3->h4, h1->h2 h4->h0, h0->h1 h2->h3, h3->h4 + VAG T_1, h1, h1 \ + VAG T_4, h4, h4 \ + VAG T_2, h2, h2 + +// REDUCE performs the following carry operations in four +// stages, as specified in Bernstein & Schwabe: +// +// 1: h₂₆[0]->h₂₆[1] h₂₆[3]->h₂₆[4] +// 2: h₂₆[1]->h₂₆[2] h₂₆[4]->h₂₆[0] +// 3: h₂₆[0]->h₂₆[1] h₂₆[2]->h₂₆[3] +// 4: h₂₆[3]->h₂₆[4] +// +// The result is that all of the limbs are limited to 26-bits +// except for h₂₆[1] and h₂₆[4] which are limited to 27-bits. +// +// Note that although each limb is aligned at 26-bit intervals +// they may contain values that exceed 2²⁶ - 1, hence the need +// to carry the excess bits in each limb. #define REDUCE(h0, h1, h2, h3, h4) \ VESRLG $26, h0, T_0 \ VESRLG $26, h3, T_1 \ @@ -136,144 +208,155 @@ DATA ·constants<>+56(SB)/8, $0x1d1d1d1d1d1f1e1d VN MOD26, h3, h3 \ VAG T_2, h4, h4 -// expand in0 into d[0] and in1 into d[1] +// EXPAND splits the 128-bit little-endian values in0 and in1 +// into 26-bit big-endian limbs and places the results into +// the first and second lane of d₂₆[0:4] respectively. +// +// The EX0, EX1 and EX2 constants are arrays of byte indices +// for permutation. The permutation both reverses the bytes +// in the input and ensures the bytes are copied into the +// destination limb ready to be shifted into their final +// position. #define EXPAND(in0, in1, d0, d1, d2, d3, d4) \ - VGBM $0x0707, d1 \ // d1=tmp - VPERM in0, in1, EX2, d4 \ VPERM in0, in1, EX0, d0 \ VPERM in0, in1, EX1, d2 \ - VN d1, d4, d4 \ + VPERM in0, in1, EX2, d4 \ VESRLG $26, d0, d1 \ VESRLG $30, d2, d3 \ VESRLG $4, d2, d2 \ - VN MOD26, d0, d0 \ - VN MOD26, d1, d1 \ - VN MOD26, d2, d2 \ - VN MOD26, d3, d3 - -// pack h4:h0 into h1:h0 (no carry) -#define PACK(h0, h1, h2, h3, h4) \ - VESLG $26, h1, h1 \ - VESLG $26, h3, h3 \ - VO h0, h1, h0 \ - VO h2, h3, h2 \ - VESLG $4, h2, h2 \ - VLEIB $7, $48, h1 \ - VSLB h1, h2, h2 \ - VO h0, h2, h0 \ - VLEIB $7, $104, h1 \ - VSLB h1, h4, h3 \ - VO h3, h0, h0 \ - VLEIB $7, $24, h1 \ - VSRLB h1, h4, h1 - -// if h > 2**130-5 then h -= 2**130-5 -#define MOD(h0, h1, t0, t1, t2) \ - VZERO t0 \ - VLEIG $1, $5, t0 \ - VACCQ h0, t0, t1 \ - VAQ h0, t0, t0 \ - VONE t2 \ - VLEIG $1, $-4, t2 \ - VAQ t2, t1, t1 \ - VACCQ h1, t1, t1 \ - VONE t2 \ - VAQ t2, t1, t1 \ - VN h0, t1, t2 \ - VNC t0, t1, t1 \ - VO t1, t2, h0 - -// func poly1305vx(out *[16]byte, m *byte, mlen uint64, key *[32]key) -TEXT ·poly1305vx(SB), $0-32 - // This code processes up to 2 blocks (32 bytes) per iteration - // using the algorithm described in: - // NEON crypto, Daniel J. Bernstein & Peter Schwabe - // https://cryptojedi.org/papers/neoncrypto-20120320.pdf - LMG out+0(FP), R1, R4 // R1=out, R2=m, R3=mlen, R4=key - - // load MOD26, EX0, EX1 and EX2 + VN MOD26, d0, d0 \ // [in0₂₆[0], in1₂₆[0]] + VN MOD26, d3, d3 \ // [in0₂₆[3], in1₂₆[3]] + VN MOD26, d1, d1 \ // [in0₂₆[1], in1₂₆[1]] + VN MOD24, d4, d4 \ // [in0₂₆[4], in1₂₆[4]] + VN MOD26, d2, d2 // [in0₂₆[2], in1₂₆[2]] + +// func updateVX(state *macState, msg []byte) +TEXT ·updateVX(SB), NOSPLIT, $0 + MOVD state+0(FP), R1 + LMG msg+8(FP), R2, R3 // R2=msg_base, R3=msg_len + + // load EX0, EX1 and EX2 MOVD $·constants<>(SB), R5 - VLM (R5), MOD26, EX2 - - // setup r - VL (R4), T_0 - MOVD $·keyMask<>(SB), R6 - VL (R6), T_1 - VN T_0, T_1, T_0 - EXPAND(T_0, T_0, R_0, R_1, R_2, R_3, R_4) - - // setup r*5 - VLEIG $0, $5, T_0 - VLEIG $1, $5, T_0 - - // store r (for final block) - VMLOF T_0, R_1, R5SAVE_1 - VMLOF T_0, R_2, R5SAVE_2 - VMLOF T_0, R_3, R5SAVE_3 - VMLOF T_0, R_4, R5SAVE_4 - VLGVG $0, R_0, RSAVE_0 - VLGVG $0, R_1, RSAVE_1 - VLGVG $0, R_2, RSAVE_2 - VLGVG $0, R_3, RSAVE_3 - VLGVG $0, R_4, RSAVE_4 - - // skip r**2 calculation + VLM (R5), EX0, EX2 + + // generate masks + VGMG $(64-24), $63, MOD24 // [0x00ffffff, 0x00ffffff] + VGMG $(64-26), $63, MOD26 // [0x03ffffff, 0x03ffffff] + + // load h (accumulator) and r (key) from state + VZERO T_1 // [0, 0] + VL 0(R1), T_0 // [h₆₄[0], h₆₄[1]] + VLEG $0, 16(R1), T_1 // [h₆₄[2], 0] + VL 24(R1), T_2 // [r₆₄[0], r₆₄[1]] + VPDI $0, T_0, T_2, T_3 // [h₆₄[0], r₆₄[0]] + VPDI $5, T_0, T_2, T_4 // [h₆₄[1], r₆₄[1]] + + // unpack h and r into 26-bit limbs + // note: h₆₄[2] may have the low 3 bits set, so h₂₆[4] is a 27-bit value + VN MOD26, T_3, H_0 // [h₂₆[0], r₂₆[0]] + VZERO H_1 // [0, 0] + VZERO H_3 // [0, 0] + VGMG $(64-12-14), $(63-12), T_0 // [0x03fff000, 0x03fff000] - 26-bit mask with low 12 bits masked out + VESLG $24, T_1, T_1 // [h₆₄[2]<<24, 0] + VERIMG $-26&63, T_3, MOD26, H_1 // [h₂₆[1], r₂₆[1]] + VESRLG $+52&63, T_3, H_2 // [h₂₆[2], r₂₆[2]] - low 12 bits only + VERIMG $-14&63, T_4, MOD26, H_3 // [h₂₆[1], r₂₆[1]] + VESRLG $40, T_4, H_4 // [h₂₆[4], r₂₆[4]] - low 24 bits only + VERIMG $+12&63, T_4, T_0, H_2 // [h₂₆[2], r₂₆[2]] - complete + VO T_1, H_4, H_4 // [h₂₆[4], r₂₆[4]] - complete + + // replicate r across all 4 vector elements + VREPF $3, H_0, R_0 // [r₂₆[0], r₂₆[0], r₂₆[0], r₂₆[0]] + VREPF $3, H_1, R_1 // [r₂₆[1], r₂₆[1], r₂₆[1], r₂₆[1]] + VREPF $3, H_2, R_2 // [r₂₆[2], r₂₆[2], r₂₆[2], r₂₆[2]] + VREPF $3, H_3, R_3 // [r₂₆[3], r₂₆[3], r₂₆[3], r₂₆[3]] + VREPF $3, H_4, R_4 // [r₂₆[4], r₂₆[4], r₂₆[4], r₂₆[4]] + + // zero out lane 1 of h + VLEIG $1, $0, H_0 // [h₂₆[0], 0] + VLEIG $1, $0, H_1 // [h₂₆[1], 0] + VLEIG $1, $0, H_2 // [h₂₆[2], 0] + VLEIG $1, $0, H_3 // [h₂₆[3], 0] + VLEIG $1, $0, H_4 // [h₂₆[4], 0] + + // calculate 5r (ignore least significant limb) + VREPIF $5, T_0 + VMLF T_0, R_1, R5_1 // [5r₂₆[1], 5r₂₆[1], 5r₂₆[1], 5r₂₆[1]] + VMLF T_0, R_2, R5_2 // [5r₂₆[2], 5r₂₆[2], 5r₂₆[2], 5r₂₆[2]] + VMLF T_0, R_3, R5_3 // [5r₂₆[3], 5r₂₆[3], 5r₂₆[3], 5r₂₆[3]] + VMLF T_0, R_4, R5_4 // [5r₂₆[4], 5r₂₆[4], 5r₂₆[4], 5r₂₆[4]] + + // skip r² calculation if we are only calculating one block CMPBLE R3, $16, skip - // calculate r**2 - MULTIPLY(R_0, R_1, R_2, R_3, R_4, R_0, R_1, R_2, R_3, R_4, R5SAVE_1, R5SAVE_2, R5SAVE_3, R5SAVE_4, H_0, H_1, H_2, H_3, H_4) - REDUCE(H_0, H_1, H_2, H_3, H_4) - VLEIG $0, $5, T_0 - VLEIG $1, $5, T_0 - VMLOF T_0, H_1, R5_1 - VMLOF T_0, H_2, R5_2 - VMLOF T_0, H_3, R5_3 - VMLOF T_0, H_4, R5_4 - VLR H_0, R_0 - VLR H_1, R_1 - VLR H_2, R_2 - VLR H_3, R_3 - VLR H_4, R_4 - - // initialize h - VZERO H_0 - VZERO H_1 - VZERO H_2 - VZERO H_3 - VZERO H_4 + // calculate r² + MULTIPLY(R_0, R_1, R_2, R_3, R_4, R_0, R_1, R_2, R_3, R_4, R5_1, R5_2, R5_3, R5_4, M_0, M_1, M_2, M_3, M_4) + REDUCE(M_0, M_1, M_2, M_3, M_4) + VGBM $0x0f0f, T_0 + VERIMG $0, M_0, T_0, R_0 // [r₂₆[0], r²₂₆[0], r₂₆[0], r²₂₆[0]] + VERIMG $0, M_1, T_0, R_1 // [r₂₆[1], r²₂₆[1], r₂₆[1], r²₂₆[1]] + VERIMG $0, M_2, T_0, R_2 // [r₂₆[2], r²₂₆[2], r₂₆[2], r²₂₆[2]] + VERIMG $0, M_3, T_0, R_3 // [r₂₆[3], r²₂₆[3], r₂₆[3], r²₂₆[3]] + VERIMG $0, M_4, T_0, R_4 // [r₂₆[4], r²₂₆[4], r₂₆[4], r²₂₆[4]] + + // calculate 5r² (ignore least significant limb) + VREPIF $5, T_0 + VMLF T_0, R_1, R5_1 // [5r₂₆[1], 5r²₂₆[1], 5r₂₆[1], 5r²₂₆[1]] + VMLF T_0, R_2, R5_2 // [5r₂₆[2], 5r²₂₆[2], 5r₂₆[2], 5r²₂₆[2]] + VMLF T_0, R_3, R5_3 // [5r₂₆[3], 5r²₂₆[3], 5r₂₆[3], 5r²₂₆[3]] + VMLF T_0, R_4, R5_4 // [5r₂₆[4], 5r²₂₆[4], 5r₂₆[4], 5r²₂₆[4]] loop: - CMPBLE R3, $32, b2 - VLM (R2), T_0, T_1 - SUB $32, R3 - MOVD $32(R2), R2 - EXPAND(T_0, T_1, F_0, F_1, F_2, F_3, F_4) - VLEIB $4, $1, F_4 - VLEIB $12, $1, F_4 + CMPBLE R3, $32, b2 // 2 or fewer blocks remaining, need to change key coefficients + + // load next 2 blocks from message + VLM (R2), T_0, T_1 + + // update message slice + SUB $32, R3 + MOVD $32(R2), R2 + + // unpack message blocks into 26-bit big-endian limbs + EXPAND(T_0, T_1, M_0, M_1, M_2, M_3, M_4) + + // add 2¹²⁸ to each message block value + VLEIB $4, $1, M_4 + VLEIB $12, $1, M_4 multiply: - VAG H_0, F_0, F_0 - VAG H_1, F_1, F_1 - VAG H_2, F_2, F_2 - VAG H_3, F_3, F_3 - VAG H_4, F_4, F_4 - MULTIPLY(F_0, F_1, F_2, F_3, F_4, R_0, R_1, R_2, R_3, R_4, R5_1, R5_2, R5_3, R5_4, H_0, H_1, H_2, H_3, H_4) + // accumulate the incoming message + VAG H_0, M_0, M_0 + VAG H_3, M_3, M_3 + VAG H_1, M_1, M_1 + VAG H_4, M_4, M_4 + VAG H_2, M_2, M_2 + + // multiply the accumulator by the key coefficient + MULTIPLY(M_0, M_1, M_2, M_3, M_4, R_0, R_1, R_2, R_3, R_4, R5_1, R5_2, R5_3, R5_4, H_0, H_1, H_2, H_3, H_4) + + // carry and partially reduce the partial products REDUCE(H_0, H_1, H_2, H_3, H_4) + CMPBNE R3, $0, loop finish: - // sum vectors + // sum lane 0 and lane 1 and put the result in lane 1 VZERO T_0 VSUMQG H_0, T_0, H_0 - VSUMQG H_1, T_0, H_1 - VSUMQG H_2, T_0, H_2 VSUMQG H_3, T_0, H_3 + VSUMQG H_1, T_0, H_1 VSUMQG H_4, T_0, H_4 + VSUMQG H_2, T_0, H_2 - // h may be >= 2*(2**130-5) so we need to reduce it again + // reduce again after summation + // TODO(mundaym): there might be a more efficient way to do this + // now that we only have 1 active lane. For example, we could + // simultaneously pack the values as we reduce them. REDUCE(H_0, H_1, H_2, H_3, H_4) - // carry h1->h4 + // carry h[1] through to h[4] so that only h[4] can exceed 2²⁶ - 1 + // TODO(mundaym): in testing this final carry was unnecessary. + // Needs a proof before it can be removed though. VESRLG $26, H_1, T_1 VN MOD26, H_1, H_1 VAQ T_1, H_2, H_2 @@ -284,95 +367,137 @@ finish: VN MOD26, H_3, H_3 VAQ T_3, H_4, H_4 - // h is now < 2*(2**130-5) - // pack h into h1 (hi) and h0 (lo) - PACK(H_0, H_1, H_2, H_3, H_4) - - // if h > 2**130-5 then h -= 2**130-5 - MOD(H_0, H_1, T_0, T_1, T_2) - - // h += s - MOVD $·bswapMask<>(SB), R5 - VL (R5), T_1 - VL 16(R4), T_0 - VPERM T_0, T_0, T_1, T_0 // reverse bytes (to big) - VAQ T_0, H_0, H_0 - VPERM H_0, H_0, T_1, H_0 // reverse bytes (to little) - VST H_0, (R1) - + // h is now < 2(2¹³⁰ - 5) + // Pack each lane in h₂₆[0:4] into h₁₂₈[0:1]. + VESLG $26, H_1, H_1 + VESLG $26, H_3, H_3 + VO H_0, H_1, H_0 + VO H_2, H_3, H_2 + VESLG $4, H_2, H_2 + VLEIB $7, $48, H_1 + VSLB H_1, H_2, H_2 + VO H_0, H_2, H_0 + VLEIB $7, $104, H_1 + VSLB H_1, H_4, H_3 + VO H_3, H_0, H_0 + VLEIB $7, $24, H_1 + VSRLB H_1, H_4, H_1 + + // update state + VSTEG $1, H_0, 0(R1) + VSTEG $0, H_0, 8(R1) + VSTEG $1, H_1, 16(R1) RET -b2: +b2: // 2 or fewer blocks remaining CMPBLE R3, $16, b1 - // 2 blocks remaining - SUB $17, R3 - VL (R2), T_0 - VLL R3, 16(R2), T_1 - ADD $1, R3 + // Load the 2 remaining blocks (17-32 bytes remaining). + MOVD $-17(R3), R0 // index of final byte to load modulo 16 + VL (R2), T_0 // load full 16 byte block + VLL R0, 16(R2), T_1 // load final (possibly partial) block and pad with zeros to 16 bytes + + // The Poly1305 algorithm requires that a 1 bit be appended to + // each message block. If the final block is less than 16 bytes + // long then it is easiest to insert the 1 before the message + // block is split into 26-bit limbs. If, on the other hand, the + // final message block is 16 bytes long then we append the 1 bit + // after expansion as normal. MOVBZ $1, R0 - CMPBEQ R3, $16, 2(PC) - VLVGB R3, R0, T_1 - EXPAND(T_0, T_1, F_0, F_1, F_2, F_3, F_4) + MOVD $-16(R3), R3 // index of byte in last block to insert 1 at (could be 16) + CMPBEQ R3, $16, 2(PC) // skip the insertion if the final block is 16 bytes long + VLVGB R3, R0, T_1 // insert 1 into the byte at index R3 + + // Split both blocks into 26-bit limbs in the appropriate lanes. + EXPAND(T_0, T_1, M_0, M_1, M_2, M_3, M_4) + + // Append a 1 byte to the end of the second to last block. + VLEIB $4, $1, M_4 + + // Append a 1 byte to the end of the last block only if it is a + // full 16 byte block. CMPBNE R3, $16, 2(PC) - VLEIB $12, $1, F_4 - VLEIB $4, $1, F_4 - - // setup [r²,r] - VLVGG $1, RSAVE_0, R_0 - VLVGG $1, RSAVE_1, R_1 - VLVGG $1, RSAVE_2, R_2 - VLVGG $1, RSAVE_3, R_3 - VLVGG $1, RSAVE_4, R_4 - VPDI $0, R5_1, R5SAVE_1, R5_1 - VPDI $0, R5_2, R5SAVE_2, R5_2 - VPDI $0, R5_3, R5SAVE_3, R5_3 - VPDI $0, R5_4, R5SAVE_4, R5_4 + VLEIB $12, $1, M_4 + + // Finally, set up the coefficients for the final multiplication. + // We have previously saved r and 5r in the 32-bit even indexes + // of the R_[0-4] and R5_[1-4] coefficient registers. + // + // We want lane 0 to be multiplied by r² so that can be kept the + // same. We want lane 1 to be multiplied by r so we need to move + // the saved r value into the 32-bit odd index in lane 1 by + // rotating the 64-bit lane by 32. + VGBM $0x00ff, T_0 // [0, 0xffffffffffffffff] - mask lane 1 only + VERIMG $32, R_0, T_0, R_0 // [_, r²₂₆[0], _, r₂₆[0]] + VERIMG $32, R_1, T_0, R_1 // [_, r²₂₆[1], _, r₂₆[1]] + VERIMG $32, R_2, T_0, R_2 // [_, r²₂₆[2], _, r₂₆[2]] + VERIMG $32, R_3, T_0, R_3 // [_, r²₂₆[3], _, r₂₆[3]] + VERIMG $32, R_4, T_0, R_4 // [_, r²₂₆[4], _, r₂₆[4]] + VERIMG $32, R5_1, T_0, R5_1 // [_, 5r²₂₆[1], _, 5r₂₆[1]] + VERIMG $32, R5_2, T_0, R5_2 // [_, 5r²₂₆[2], _, 5r₂₆[2]] + VERIMG $32, R5_3, T_0, R5_3 // [_, 5r²₂₆[3], _, 5r₂₆[3]] + VERIMG $32, R5_4, T_0, R5_4 // [_, 5r²₂₆[4], _, 5r₂₆[4]] MOVD $0, R3 BR multiply skip: - VZERO H_0 - VZERO H_1 - VZERO H_2 - VZERO H_3 - VZERO H_4 - CMPBEQ R3, $0, finish -b1: - // 1 block remaining - SUB $1, R3 - VLL R3, (R2), T_0 - ADD $1, R3 +b1: // 1 block remaining + + // Load the final block (1-16 bytes). This will be placed into + // lane 0. + MOVD $-1(R3), R0 + VLL R0, (R2), T_0 // pad to 16 bytes with zeros + + // The Poly1305 algorithm requires that a 1 bit be appended to + // each message block. If the final block is less than 16 bytes + // long then it is easiest to insert the 1 before the message + // block is split into 26-bit limbs. If, on the other hand, the + // final message block is 16 bytes long then we append the 1 bit + // after expansion as normal. MOVBZ $1, R0 CMPBEQ R3, $16, 2(PC) VLVGB R3, R0, T_0 - VZERO T_1 - EXPAND(T_0, T_1, F_0, F_1, F_2, F_3, F_4) + + // Set the message block in lane 1 to the value 0 so that it + // can be accumulated without affecting the final result. + VZERO T_1 + + // Split the final message block into 26-bit limbs in lane 0. + // Lane 1 will be contain 0. + EXPAND(T_0, T_1, M_0, M_1, M_2, M_3, M_4) + + // Append a 1 byte to the end of the last block only if it is a + // full 16 byte block. CMPBNE R3, $16, 2(PC) - VLEIB $4, $1, F_4 - VLEIG $1, $1, R_0 - VZERO R_1 - VZERO R_2 - VZERO R_3 - VZERO R_4 - VZERO R5_1 - VZERO R5_2 - VZERO R5_3 - VZERO R5_4 - - // setup [r, 1] - VLVGG $0, RSAVE_0, R_0 - VLVGG $0, RSAVE_1, R_1 - VLVGG $0, RSAVE_2, R_2 - VLVGG $0, RSAVE_3, R_3 - VLVGG $0, RSAVE_4, R_4 - VPDI $0, R5SAVE_1, R5_1, R5_1 - VPDI $0, R5SAVE_2, R5_2, R5_2 - VPDI $0, R5SAVE_3, R5_3, R5_3 - VPDI $0, R5SAVE_4, R5_4, R5_4 + VLEIB $4, $1, M_4 + + // We have previously saved r and 5r in the 32-bit even indexes + // of the R_[0-4] and R5_[1-4] coefficient registers. + // + // We want lane 0 to be multiplied by r so we need to move the + // saved r value into the 32-bit odd index in lane 0. We want + // lane 1 to be set to the value 1. This makes multiplication + // a no-op. We do this by setting lane 1 in every register to 0 + // and then just setting the 32-bit index 3 in R_0 to 1. + VZERO T_0 + MOVD $0, R0 + MOVD $0x10111213, R12 + VLVGP R12, R0, T_1 // [_, 0x10111213, _, 0x00000000] + VPERM T_0, R_0, T_1, R_0 // [_, r₂₆[0], _, 0] + VPERM T_0, R_1, T_1, R_1 // [_, r₂₆[1], _, 0] + VPERM T_0, R_2, T_1, R_2 // [_, r₂₆[2], _, 0] + VPERM T_0, R_3, T_1, R_3 // [_, r₂₆[3], _, 0] + VPERM T_0, R_4, T_1, R_4 // [_, r₂₆[4], _, 0] + VPERM T_0, R5_1, T_1, R5_1 // [_, 5r₂₆[1], _, 0] + VPERM T_0, R5_2, T_1, R5_2 // [_, 5r₂₆[2], _, 0] + VPERM T_0, R5_3, T_1, R5_3 // [_, 5r₂₆[3], _, 0] + VPERM T_0, R5_4, T_1, R5_4 // [_, 5r₂₆[4], _, 0] + + // Set the value of lane 1 to be 1. + VLEIF $3, $1, R_0 // [_, r₂₆[0], _, 1] MOVD $0, R3 BR multiply diff --git a/vendor/golang.org/x/crypto/poly1305/sum_vmsl_s390x.s b/vendor/golang.org/x/crypto/poly1305/sum_vmsl_s390x.s deleted file mode 100644 index e60bbc1d7f8..00000000000 --- a/vendor/golang.org/x/crypto/poly1305/sum_vmsl_s390x.s +++ /dev/null @@ -1,909 +0,0 @@ -// Copyright 2018 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -// +build s390x,go1.11,!gccgo,!appengine - -#include "textflag.h" - -// Implementation of Poly1305 using the vector facility (vx) and the VMSL instruction. - -// constants -#define EX0 V1 -#define EX1 V2 -#define EX2 V3 - -// temporaries -#define T_0 V4 -#define T_1 V5 -#define T_2 V6 -#define T_3 V7 -#define T_4 V8 -#define T_5 V9 -#define T_6 V10 -#define T_7 V11 -#define T_8 V12 -#define T_9 V13 -#define T_10 V14 - -// r**2 & r**4 -#define R_0 V15 -#define R_1 V16 -#define R_2 V17 -#define R5_1 V18 -#define R5_2 V19 -// key (r) -#define RSAVE_0 R7 -#define RSAVE_1 R8 -#define RSAVE_2 R9 -#define R5SAVE_1 R10 -#define R5SAVE_2 R11 - -// message block -#define M0 V20 -#define M1 V21 -#define M2 V22 -#define M3 V23 -#define M4 V24 -#define M5 V25 - -// accumulator -#define H0_0 V26 -#define H1_0 V27 -#define H2_0 V28 -#define H0_1 V29 -#define H1_1 V30 -#define H2_1 V31 - -GLOBL ·keyMask<>(SB), RODATA, $16 -DATA ·keyMask<>+0(SB)/8, $0xffffff0ffcffff0f -DATA ·keyMask<>+8(SB)/8, $0xfcffff0ffcffff0f - -GLOBL ·bswapMask<>(SB), RODATA, $16 -DATA ·bswapMask<>+0(SB)/8, $0x0f0e0d0c0b0a0908 -DATA ·bswapMask<>+8(SB)/8, $0x0706050403020100 - -GLOBL ·constants<>(SB), RODATA, $48 -// EX0 -DATA ·constants<>+0(SB)/8, $0x18191a1b1c1d1e1f -DATA ·constants<>+8(SB)/8, $0x0000050403020100 -// EX1 -DATA ·constants<>+16(SB)/8, $0x18191a1b1c1d1e1f -DATA ·constants<>+24(SB)/8, $0x00000a0908070605 -// EX2 -DATA ·constants<>+32(SB)/8, $0x18191a1b1c1d1e1f -DATA ·constants<>+40(SB)/8, $0x0000000f0e0d0c0b - -GLOBL ·c<>(SB), RODATA, $48 -// EX0 -DATA ·c<>+0(SB)/8, $0x0000050403020100 -DATA ·c<>+8(SB)/8, $0x0000151413121110 -// EX1 -DATA ·c<>+16(SB)/8, $0x00000a0908070605 -DATA ·c<>+24(SB)/8, $0x00001a1918171615 -// EX2 -DATA ·c<>+32(SB)/8, $0x0000000f0e0d0c0b -DATA ·c<>+40(SB)/8, $0x0000001f1e1d1c1b - -GLOBL ·reduce<>(SB), RODATA, $32 -// 44 bit -DATA ·reduce<>+0(SB)/8, $0x0 -DATA ·reduce<>+8(SB)/8, $0xfffffffffff -// 42 bit -DATA ·reduce<>+16(SB)/8, $0x0 -DATA ·reduce<>+24(SB)/8, $0x3ffffffffff - -// h = (f*g) % (2**130-5) [partial reduction] -// uses T_0...T_9 temporary registers -// input: m02_0, m02_1, m02_2, m13_0, m13_1, m13_2, r_0, r_1, r_2, r5_1, r5_2, m4_0, m4_1, m4_2, m5_0, m5_1, m5_2 -// temp: t0, t1, t2, t3, t4, t5, t6, t7, t8, t9 -// output: m02_0, m02_1, m02_2, m13_0, m13_1, m13_2 -#define MULTIPLY(m02_0, m02_1, m02_2, m13_0, m13_1, m13_2, r_0, r_1, r_2, r5_1, r5_2, m4_0, m4_1, m4_2, m5_0, m5_1, m5_2, t0, t1, t2, t3, t4, t5, t6, t7, t8, t9) \ - \ // Eliminate the dependency for the last 2 VMSLs - VMSLG m02_0, r_2, m4_2, m4_2 \ - VMSLG m13_0, r_2, m5_2, m5_2 \ // 8 VMSLs pipelined - VMSLG m02_0, r_0, m4_0, m4_0 \ - VMSLG m02_1, r5_2, V0, T_0 \ - VMSLG m02_0, r_1, m4_1, m4_1 \ - VMSLG m02_1, r_0, V0, T_1 \ - VMSLG m02_1, r_1, V0, T_2 \ - VMSLG m02_2, r5_1, V0, T_3 \ - VMSLG m02_2, r5_2, V0, T_4 \ - VMSLG m13_0, r_0, m5_0, m5_0 \ - VMSLG m13_1, r5_2, V0, T_5 \ - VMSLG m13_0, r_1, m5_1, m5_1 \ - VMSLG m13_1, r_0, V0, T_6 \ - VMSLG m13_1, r_1, V0, T_7 \ - VMSLG m13_2, r5_1, V0, T_8 \ - VMSLG m13_2, r5_2, V0, T_9 \ - VMSLG m02_2, r_0, m4_2, m4_2 \ - VMSLG m13_2, r_0, m5_2, m5_2 \ - VAQ m4_0, T_0, m02_0 \ - VAQ m4_1, T_1, m02_1 \ - VAQ m5_0, T_5, m13_0 \ - VAQ m5_1, T_6, m13_1 \ - VAQ m02_0, T_3, m02_0 \ - VAQ m02_1, T_4, m02_1 \ - VAQ m13_0, T_8, m13_0 \ - VAQ m13_1, T_9, m13_1 \ - VAQ m4_2, T_2, m02_2 \ - VAQ m5_2, T_7, m13_2 \ - -// SQUARE uses three limbs of r and r_2*5 to output square of r -// uses T_1, T_5 and T_7 temporary registers -// input: r_0, r_1, r_2, r5_2 -// temp: TEMP0, TEMP1, TEMP2 -// output: p0, p1, p2 -#define SQUARE(r_0, r_1, r_2, r5_2, p0, p1, p2, TEMP0, TEMP1, TEMP2) \ - VMSLG r_0, r_0, p0, p0 \ - VMSLG r_1, r5_2, V0, TEMP0 \ - VMSLG r_2, r5_2, p1, p1 \ - VMSLG r_0, r_1, V0, TEMP1 \ - VMSLG r_1, r_1, p2, p2 \ - VMSLG r_0, r_2, V0, TEMP2 \ - VAQ TEMP0, p0, p0 \ - VAQ TEMP1, p1, p1 \ - VAQ TEMP2, p2, p2 \ - VAQ TEMP0, p0, p0 \ - VAQ TEMP1, p1, p1 \ - VAQ TEMP2, p2, p2 \ - -// carry h0->h1->h2->h0 || h3->h4->h5->h3 -// uses T_2, T_4, T_5, T_7, T_8, T_9 -// t6, t7, t8, t9, t10, t11 -// input: h0, h1, h2, h3, h4, h5 -// temp: t0, t1, t2, t3, t4, t5, t6, t7, t8, t9, t10, t11 -// output: h0, h1, h2, h3, h4, h5 -#define REDUCE(h0, h1, h2, h3, h4, h5, t0, t1, t2, t3, t4, t5, t6, t7, t8, t9, t10, t11) \ - VLM (R12), t6, t7 \ // 44 and 42 bit clear mask - VLEIB $7, $0x28, t10 \ // 5 byte shift mask - VREPIB $4, t8 \ // 4 bit shift mask - VREPIB $2, t11 \ // 2 bit shift mask - VSRLB t10, h0, t0 \ // h0 byte shift - VSRLB t10, h1, t1 \ // h1 byte shift - VSRLB t10, h2, t2 \ // h2 byte shift - VSRLB t10, h3, t3 \ // h3 byte shift - VSRLB t10, h4, t4 \ // h4 byte shift - VSRLB t10, h5, t5 \ // h5 byte shift - VSRL t8, t0, t0 \ // h0 bit shift - VSRL t8, t1, t1 \ // h2 bit shift - VSRL t11, t2, t2 \ // h2 bit shift - VSRL t8, t3, t3 \ // h3 bit shift - VSRL t8, t4, t4 \ // h4 bit shift - VESLG $2, t2, t9 \ // h2 carry x5 - VSRL t11, t5, t5 \ // h5 bit shift - VN t6, h0, h0 \ // h0 clear carry - VAQ t2, t9, t2 \ // h2 carry x5 - VESLG $2, t5, t9 \ // h5 carry x5 - VN t6, h1, h1 \ // h1 clear carry - VN t7, h2, h2 \ // h2 clear carry - VAQ t5, t9, t5 \ // h5 carry x5 - VN t6, h3, h3 \ // h3 clear carry - VN t6, h4, h4 \ // h4 clear carry - VN t7, h5, h5 \ // h5 clear carry - VAQ t0, h1, h1 \ // h0->h1 - VAQ t3, h4, h4 \ // h3->h4 - VAQ t1, h2, h2 \ // h1->h2 - VAQ t4, h5, h5 \ // h4->h5 - VAQ t2, h0, h0 \ // h2->h0 - VAQ t5, h3, h3 \ // h5->h3 - VREPG $1, t6, t6 \ // 44 and 42 bit masks across both halves - VREPG $1, t7, t7 \ - VSLDB $8, h0, h0, h0 \ // set up [h0/1/2, h3/4/5] - VSLDB $8, h1, h1, h1 \ - VSLDB $8, h2, h2, h2 \ - VO h0, h3, h3 \ - VO h1, h4, h4 \ - VO h2, h5, h5 \ - VESRLG $44, h3, t0 \ // 44 bit shift right - VESRLG $44, h4, t1 \ - VESRLG $42, h5, t2 \ - VN t6, h3, h3 \ // clear carry bits - VN t6, h4, h4 \ - VN t7, h5, h5 \ - VESLG $2, t2, t9 \ // multiply carry by 5 - VAQ t9, t2, t2 \ - VAQ t0, h4, h4 \ - VAQ t1, h5, h5 \ - VAQ t2, h3, h3 \ - -// carry h0->h1->h2->h0 -// input: h0, h1, h2 -// temp: t0, t1, t2, t3, t4, t5, t6, t7, t8 -// output: h0, h1, h2 -#define REDUCE2(h0, h1, h2, t0, t1, t2, t3, t4, t5, t6, t7, t8) \ - VLEIB $7, $0x28, t3 \ // 5 byte shift mask - VREPIB $4, t4 \ // 4 bit shift mask - VREPIB $2, t7 \ // 2 bit shift mask - VGBM $0x003F, t5 \ // mask to clear carry bits - VSRLB t3, h0, t0 \ - VSRLB t3, h1, t1 \ - VSRLB t3, h2, t2 \ - VESRLG $4, t5, t5 \ // 44 bit clear mask - VSRL t4, t0, t0 \ - VSRL t4, t1, t1 \ - VSRL t7, t2, t2 \ - VESRLG $2, t5, t6 \ // 42 bit clear mask - VESLG $2, t2, t8 \ - VAQ t8, t2, t2 \ - VN t5, h0, h0 \ - VN t5, h1, h1 \ - VN t6, h2, h2 \ - VAQ t0, h1, h1 \ - VAQ t1, h2, h2 \ - VAQ t2, h0, h0 \ - VSRLB t3, h0, t0 \ - VSRLB t3, h1, t1 \ - VSRLB t3, h2, t2 \ - VSRL t4, t0, t0 \ - VSRL t4, t1, t1 \ - VSRL t7, t2, t2 \ - VN t5, h0, h0 \ - VN t5, h1, h1 \ - VESLG $2, t2, t8 \ - VN t6, h2, h2 \ - VAQ t0, h1, h1 \ - VAQ t8, t2, t2 \ - VAQ t1, h2, h2 \ - VAQ t2, h0, h0 \ - -// expands two message blocks into the lower halfs of the d registers -// moves the contents of the d registers into upper halfs -// input: in1, in2, d0, d1, d2, d3, d4, d5 -// temp: TEMP0, TEMP1, TEMP2, TEMP3 -// output: d0, d1, d2, d3, d4, d5 -#define EXPACC(in1, in2, d0, d1, d2, d3, d4, d5, TEMP0, TEMP1, TEMP2, TEMP3) \ - VGBM $0xff3f, TEMP0 \ - VGBM $0xff1f, TEMP1 \ - VESLG $4, d1, TEMP2 \ - VESLG $4, d4, TEMP3 \ - VESRLG $4, TEMP0, TEMP0 \ - VPERM in1, d0, EX0, d0 \ - VPERM in2, d3, EX0, d3 \ - VPERM in1, d2, EX2, d2 \ - VPERM in2, d5, EX2, d5 \ - VPERM in1, TEMP2, EX1, d1 \ - VPERM in2, TEMP3, EX1, d4 \ - VN TEMP0, d0, d0 \ - VN TEMP0, d3, d3 \ - VESRLG $4, d1, d1 \ - VESRLG $4, d4, d4 \ - VN TEMP1, d2, d2 \ - VN TEMP1, d5, d5 \ - VN TEMP0, d1, d1 \ - VN TEMP0, d4, d4 \ - -// expands one message block into the lower halfs of the d registers -// moves the contents of the d registers into upper halfs -// input: in, d0, d1, d2 -// temp: TEMP0, TEMP1, TEMP2 -// output: d0, d1, d2 -#define EXPACC2(in, d0, d1, d2, TEMP0, TEMP1, TEMP2) \ - VGBM $0xff3f, TEMP0 \ - VESLG $4, d1, TEMP2 \ - VGBM $0xff1f, TEMP1 \ - VPERM in, d0, EX0, d0 \ - VESRLG $4, TEMP0, TEMP0 \ - VPERM in, d2, EX2, d2 \ - VPERM in, TEMP2, EX1, d1 \ - VN TEMP0, d0, d0 \ - VN TEMP1, d2, d2 \ - VESRLG $4, d1, d1 \ - VN TEMP0, d1, d1 \ - -// pack h2:h0 into h1:h0 (no carry) -// input: h0, h1, h2 -// output: h0, h1, h2 -#define PACK(h0, h1, h2) \ - VMRLG h1, h2, h2 \ // copy h1 to upper half h2 - VESLG $44, h1, h1 \ // shift limb 1 44 bits, leaving 20 - VO h0, h1, h0 \ // combine h0 with 20 bits from limb 1 - VESRLG $20, h2, h1 \ // put top 24 bits of limb 1 into h1 - VLEIG $1, $0, h1 \ // clear h2 stuff from lower half of h1 - VO h0, h1, h0 \ // h0 now has 88 bits (limb 0 and 1) - VLEIG $0, $0, h2 \ // clear upper half of h2 - VESRLG $40, h2, h1 \ // h1 now has upper two bits of result - VLEIB $7, $88, h1 \ // for byte shift (11 bytes) - VSLB h1, h2, h2 \ // shift h2 11 bytes to the left - VO h0, h2, h0 \ // combine h0 with 20 bits from limb 1 - VLEIG $0, $0, h1 \ // clear upper half of h1 - -// if h > 2**130-5 then h -= 2**130-5 -// input: h0, h1 -// temp: t0, t1, t2 -// output: h0 -#define MOD(h0, h1, t0, t1, t2) \ - VZERO t0 \ - VLEIG $1, $5, t0 \ - VACCQ h0, t0, t1 \ - VAQ h0, t0, t0 \ - VONE t2 \ - VLEIG $1, $-4, t2 \ - VAQ t2, t1, t1 \ - VACCQ h1, t1, t1 \ - VONE t2 \ - VAQ t2, t1, t1 \ - VN h0, t1, t2 \ - VNC t0, t1, t1 \ - VO t1, t2, h0 \ - -// func poly1305vmsl(out *[16]byte, m *byte, mlen uint64, key *[32]key) -TEXT ·poly1305vmsl(SB), $0-32 - // This code processes 6 + up to 4 blocks (32 bytes) per iteration - // using the algorithm described in: - // NEON crypto, Daniel J. Bernstein & Peter Schwabe - // https://cryptojedi.org/papers/neoncrypto-20120320.pdf - // And as moddified for VMSL as described in - // Accelerating Poly1305 Cryptographic Message Authentication on the z14 - // O'Farrell et al, CASCON 2017, p48-55 - // https://ibm.ent.box.com/s/jf9gedj0e9d2vjctfyh186shaztavnht - - LMG out+0(FP), R1, R4 // R1=out, R2=m, R3=mlen, R4=key - VZERO V0 // c - - // load EX0, EX1 and EX2 - MOVD $·constants<>(SB), R5 - VLM (R5), EX0, EX2 // c - - // setup r - VL (R4), T_0 - MOVD $·keyMask<>(SB), R6 - VL (R6), T_1 - VN T_0, T_1, T_0 - VZERO T_2 // limbs for r - VZERO T_3 - VZERO T_4 - EXPACC2(T_0, T_2, T_3, T_4, T_1, T_5, T_7) - - // T_2, T_3, T_4: [0, r] - - // setup r*20 - VLEIG $0, $0, T_0 - VLEIG $1, $20, T_0 // T_0: [0, 20] - VZERO T_5 - VZERO T_6 - VMSLG T_0, T_3, T_5, T_5 - VMSLG T_0, T_4, T_6, T_6 - - // store r for final block in GR - VLGVG $1, T_2, RSAVE_0 // c - VLGVG $1, T_3, RSAVE_1 // c - VLGVG $1, T_4, RSAVE_2 // c - VLGVG $1, T_5, R5SAVE_1 // c - VLGVG $1, T_6, R5SAVE_2 // c - - // initialize h - VZERO H0_0 - VZERO H1_0 - VZERO H2_0 - VZERO H0_1 - VZERO H1_1 - VZERO H2_1 - - // initialize pointer for reduce constants - MOVD $·reduce<>(SB), R12 - - // calculate r**2 and 20*(r**2) - VZERO R_0 - VZERO R_1 - VZERO R_2 - SQUARE(T_2, T_3, T_4, T_6, R_0, R_1, R_2, T_1, T_5, T_7) - REDUCE2(R_0, R_1, R_2, M0, M1, M2, M3, M4, R5_1, R5_2, M5, T_1) - VZERO R5_1 - VZERO R5_2 - VMSLG T_0, R_1, R5_1, R5_1 - VMSLG T_0, R_2, R5_2, R5_2 - - // skip r**4 calculation if 3 blocks or less - CMPBLE R3, $48, b4 - - // calculate r**4 and 20*(r**4) - VZERO T_8 - VZERO T_9 - VZERO T_10 - SQUARE(R_0, R_1, R_2, R5_2, T_8, T_9, T_10, T_1, T_5, T_7) - REDUCE2(T_8, T_9, T_10, M0, M1, M2, M3, M4, T_2, T_3, M5, T_1) - VZERO T_2 - VZERO T_3 - VMSLG T_0, T_9, T_2, T_2 - VMSLG T_0, T_10, T_3, T_3 - - // put r**2 to the right and r**4 to the left of R_0, R_1, R_2 - VSLDB $8, T_8, T_8, T_8 - VSLDB $8, T_9, T_9, T_9 - VSLDB $8, T_10, T_10, T_10 - VSLDB $8, T_2, T_2, T_2 - VSLDB $8, T_3, T_3, T_3 - - VO T_8, R_0, R_0 - VO T_9, R_1, R_1 - VO T_10, R_2, R_2 - VO T_2, R5_1, R5_1 - VO T_3, R5_2, R5_2 - - CMPBLE R3, $80, load // less than or equal to 5 blocks in message - - // 6(or 5+1) blocks - SUB $81, R3 - VLM (R2), M0, M4 - VLL R3, 80(R2), M5 - ADD $1, R3 - MOVBZ $1, R0 - CMPBGE R3, $16, 2(PC) - VLVGB R3, R0, M5 - MOVD $96(R2), R2 - EXPACC(M0, M1, H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_0, T_1, T_2, T_3) - EXPACC(M2, M3, H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_0, T_1, T_2, T_3) - VLEIB $2, $1, H2_0 - VLEIB $2, $1, H2_1 - VLEIB $10, $1, H2_0 - VLEIB $10, $1, H2_1 - - VZERO M0 - VZERO M1 - VZERO M2 - VZERO M3 - VZERO T_4 - VZERO T_10 - EXPACC(M4, M5, M0, M1, M2, M3, T_4, T_10, T_0, T_1, T_2, T_3) - VLR T_4, M4 - VLEIB $10, $1, M2 - CMPBLT R3, $16, 2(PC) - VLEIB $10, $1, T_10 - MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, T_10, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) - REDUCE(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_10, M0, M1, M2, M3, M4, T_4, T_5, T_2, T_7, T_8, T_9) - VMRHG V0, H0_1, H0_0 - VMRHG V0, H1_1, H1_0 - VMRHG V0, H2_1, H2_0 - VMRLG V0, H0_1, H0_1 - VMRLG V0, H1_1, H1_1 - VMRLG V0, H2_1, H2_1 - - SUB $16, R3 - CMPBLE R3, $0, square - -load: - // load EX0, EX1 and EX2 - MOVD $·c<>(SB), R5 - VLM (R5), EX0, EX2 - -loop: - CMPBLE R3, $64, add // b4 // last 4 or less blocks left - - // next 4 full blocks - VLM (R2), M2, M5 - SUB $64, R3 - MOVD $64(R2), R2 - REDUCE(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_10, M0, M1, T_0, T_1, T_3, T_4, T_5, T_2, T_7, T_8, T_9) - - // expacc in-lined to create [m2, m3] limbs - VGBM $0x3f3f, T_0 // 44 bit clear mask - VGBM $0x1f1f, T_1 // 40 bit clear mask - VPERM M2, M3, EX0, T_3 - VESRLG $4, T_0, T_0 // 44 bit clear mask ready - VPERM M2, M3, EX1, T_4 - VPERM M2, M3, EX2, T_5 - VN T_0, T_3, T_3 - VESRLG $4, T_4, T_4 - VN T_1, T_5, T_5 - VN T_0, T_4, T_4 - VMRHG H0_1, T_3, H0_0 - VMRHG H1_1, T_4, H1_0 - VMRHG H2_1, T_5, H2_0 - VMRLG H0_1, T_3, H0_1 - VMRLG H1_1, T_4, H1_1 - VMRLG H2_1, T_5, H2_1 - VLEIB $10, $1, H2_0 - VLEIB $10, $1, H2_1 - VPERM M4, M5, EX0, T_3 - VPERM M4, M5, EX1, T_4 - VPERM M4, M5, EX2, T_5 - VN T_0, T_3, T_3 - VESRLG $4, T_4, T_4 - VN T_1, T_5, T_5 - VN T_0, T_4, T_4 - VMRHG V0, T_3, M0 - VMRHG V0, T_4, M1 - VMRHG V0, T_5, M2 - VMRLG V0, T_3, M3 - VMRLG V0, T_4, M4 - VMRLG V0, T_5, M5 - VLEIB $10, $1, M2 - VLEIB $10, $1, M5 - - MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, M5, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) - CMPBNE R3, $0, loop - REDUCE(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_10, M0, M1, M3, M4, M5, T_4, T_5, T_2, T_7, T_8, T_9) - VMRHG V0, H0_1, H0_0 - VMRHG V0, H1_1, H1_0 - VMRHG V0, H2_1, H2_0 - VMRLG V0, H0_1, H0_1 - VMRLG V0, H1_1, H1_1 - VMRLG V0, H2_1, H2_1 - - // load EX0, EX1, EX2 - MOVD $·constants<>(SB), R5 - VLM (R5), EX0, EX2 - - // sum vectors - VAQ H0_0, H0_1, H0_0 - VAQ H1_0, H1_1, H1_0 - VAQ H2_0, H2_1, H2_0 - - // h may be >= 2*(2**130-5) so we need to reduce it again - // M0...M4 are used as temps here - REDUCE2(H0_0, H1_0, H2_0, M0, M1, M2, M3, M4, T_9, T_10, H0_1, M5) - -next: // carry h1->h2 - VLEIB $7, $0x28, T_1 - VREPIB $4, T_2 - VGBM $0x003F, T_3 - VESRLG $4, T_3 - - // byte shift - VSRLB T_1, H1_0, T_4 - - // bit shift - VSRL T_2, T_4, T_4 - - // clear h1 carry bits - VN T_3, H1_0, H1_0 - - // add carry - VAQ T_4, H2_0, H2_0 - - // h is now < 2*(2**130-5) - // pack h into h1 (hi) and h0 (lo) - PACK(H0_0, H1_0, H2_0) - - // if h > 2**130-5 then h -= 2**130-5 - MOD(H0_0, H1_0, T_0, T_1, T_2) - - // h += s - MOVD $·bswapMask<>(SB), R5 - VL (R5), T_1 - VL 16(R4), T_0 - VPERM T_0, T_0, T_1, T_0 // reverse bytes (to big) - VAQ T_0, H0_0, H0_0 - VPERM H0_0, H0_0, T_1, H0_0 // reverse bytes (to little) - VST H0_0, (R1) - RET - -add: - // load EX0, EX1, EX2 - MOVD $·constants<>(SB), R5 - VLM (R5), EX0, EX2 - - REDUCE(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_10, M0, M1, M3, M4, M5, T_4, T_5, T_2, T_7, T_8, T_9) - VMRHG V0, H0_1, H0_0 - VMRHG V0, H1_1, H1_0 - VMRHG V0, H2_1, H2_0 - VMRLG V0, H0_1, H0_1 - VMRLG V0, H1_1, H1_1 - VMRLG V0, H2_1, H2_1 - CMPBLE R3, $64, b4 - -b4: - CMPBLE R3, $48, b3 // 3 blocks or less - - // 4(3+1) blocks remaining - SUB $49, R3 - VLM (R2), M0, M2 - VLL R3, 48(R2), M3 - ADD $1, R3 - MOVBZ $1, R0 - CMPBEQ R3, $16, 2(PC) - VLVGB R3, R0, M3 - MOVD $64(R2), R2 - EXPACC(M0, M1, H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_0, T_1, T_2, T_3) - VLEIB $10, $1, H2_0 - VLEIB $10, $1, H2_1 - VZERO M0 - VZERO M1 - VZERO M4 - VZERO M5 - VZERO T_4 - VZERO T_10 - EXPACC(M2, M3, M0, M1, M4, M5, T_4, T_10, T_0, T_1, T_2, T_3) - VLR T_4, M2 - VLEIB $10, $1, M4 - CMPBNE R3, $16, 2(PC) - VLEIB $10, $1, T_10 - MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M4, M5, M2, T_10, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) - REDUCE(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_10, M0, M1, M3, M4, M5, T_4, T_5, T_2, T_7, T_8, T_9) - VMRHG V0, H0_1, H0_0 - VMRHG V0, H1_1, H1_0 - VMRHG V0, H2_1, H2_0 - VMRLG V0, H0_1, H0_1 - VMRLG V0, H1_1, H1_1 - VMRLG V0, H2_1, H2_1 - SUB $16, R3 - CMPBLE R3, $0, square // this condition must always hold true! - -b3: - CMPBLE R3, $32, b2 - - // 3 blocks remaining - - // setup [r²,r] - VSLDB $8, R_0, R_0, R_0 - VSLDB $8, R_1, R_1, R_1 - VSLDB $8, R_2, R_2, R_2 - VSLDB $8, R5_1, R5_1, R5_1 - VSLDB $8, R5_2, R5_2, R5_2 - - VLVGG $1, RSAVE_0, R_0 - VLVGG $1, RSAVE_1, R_1 - VLVGG $1, RSAVE_2, R_2 - VLVGG $1, R5SAVE_1, R5_1 - VLVGG $1, R5SAVE_2, R5_2 - - // setup [h0, h1] - VSLDB $8, H0_0, H0_0, H0_0 - VSLDB $8, H1_0, H1_0, H1_0 - VSLDB $8, H2_0, H2_0, H2_0 - VO H0_1, H0_0, H0_0 - VO H1_1, H1_0, H1_0 - VO H2_1, H2_0, H2_0 - VZERO H0_1 - VZERO H1_1 - VZERO H2_1 - - VZERO M0 - VZERO M1 - VZERO M2 - VZERO M3 - VZERO M4 - VZERO M5 - - // H*[r**2, r] - MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, M5, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) - REDUCE2(H0_0, H1_0, H2_0, M0, M1, M2, M3, M4, H0_1, H1_1, T_10, M5) - - SUB $33, R3 - VLM (R2), M0, M1 - VLL R3, 32(R2), M2 - ADD $1, R3 - MOVBZ $1, R0 - CMPBEQ R3, $16, 2(PC) - VLVGB R3, R0, M2 - - // H += m0 - VZERO T_1 - VZERO T_2 - VZERO T_3 - EXPACC2(M0, T_1, T_2, T_3, T_4, T_5, T_6) - VLEIB $10, $1, T_3 - VAG H0_0, T_1, H0_0 - VAG H1_0, T_2, H1_0 - VAG H2_0, T_3, H2_0 - - VZERO M0 - VZERO M3 - VZERO M4 - VZERO M5 - VZERO T_10 - - // (H+m0)*r - MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M3, M4, M5, V0, T_10, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) - REDUCE2(H0_0, H1_0, H2_0, M0, M3, M4, M5, T_10, H0_1, H1_1, H2_1, T_9) - - // H += m1 - VZERO V0 - VZERO T_1 - VZERO T_2 - VZERO T_3 - EXPACC2(M1, T_1, T_2, T_3, T_4, T_5, T_6) - VLEIB $10, $1, T_3 - VAQ H0_0, T_1, H0_0 - VAQ H1_0, T_2, H1_0 - VAQ H2_0, T_3, H2_0 - REDUCE2(H0_0, H1_0, H2_0, M0, M3, M4, M5, T_9, H0_1, H1_1, H2_1, T_10) - - // [H, m2] * [r**2, r] - EXPACC2(M2, H0_0, H1_0, H2_0, T_1, T_2, T_3) - CMPBNE R3, $16, 2(PC) - VLEIB $10, $1, H2_0 - VZERO M0 - VZERO M1 - VZERO M2 - VZERO M3 - VZERO M4 - VZERO M5 - MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, M5, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) - REDUCE2(H0_0, H1_0, H2_0, M0, M1, M2, M3, M4, H0_1, H1_1, M5, T_10) - SUB $16, R3 - CMPBLE R3, $0, next // this condition must always hold true! - -b2: - CMPBLE R3, $16, b1 - - // 2 blocks remaining - - // setup [r²,r] - VSLDB $8, R_0, R_0, R_0 - VSLDB $8, R_1, R_1, R_1 - VSLDB $8, R_2, R_2, R_2 - VSLDB $8, R5_1, R5_1, R5_1 - VSLDB $8, R5_2, R5_2, R5_2 - - VLVGG $1, RSAVE_0, R_0 - VLVGG $1, RSAVE_1, R_1 - VLVGG $1, RSAVE_2, R_2 - VLVGG $1, R5SAVE_1, R5_1 - VLVGG $1, R5SAVE_2, R5_2 - - // setup [h0, h1] - VSLDB $8, H0_0, H0_0, H0_0 - VSLDB $8, H1_0, H1_0, H1_0 - VSLDB $8, H2_0, H2_0, H2_0 - VO H0_1, H0_0, H0_0 - VO H1_1, H1_0, H1_0 - VO H2_1, H2_0, H2_0 - VZERO H0_1 - VZERO H1_1 - VZERO H2_1 - - VZERO M0 - VZERO M1 - VZERO M2 - VZERO M3 - VZERO M4 - VZERO M5 - - // H*[r**2, r] - MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, M5, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) - REDUCE(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_10, M0, M1, M2, M3, M4, T_4, T_5, T_2, T_7, T_8, T_9) - VMRHG V0, H0_1, H0_0 - VMRHG V0, H1_1, H1_0 - VMRHG V0, H2_1, H2_0 - VMRLG V0, H0_1, H0_1 - VMRLG V0, H1_1, H1_1 - VMRLG V0, H2_1, H2_1 - - // move h to the left and 0s at the right - VSLDB $8, H0_0, H0_0, H0_0 - VSLDB $8, H1_0, H1_0, H1_0 - VSLDB $8, H2_0, H2_0, H2_0 - - // get message blocks and append 1 to start - SUB $17, R3 - VL (R2), M0 - VLL R3, 16(R2), M1 - ADD $1, R3 - MOVBZ $1, R0 - CMPBEQ R3, $16, 2(PC) - VLVGB R3, R0, M1 - VZERO T_6 - VZERO T_7 - VZERO T_8 - EXPACC2(M0, T_6, T_7, T_8, T_1, T_2, T_3) - EXPACC2(M1, T_6, T_7, T_8, T_1, T_2, T_3) - VLEIB $2, $1, T_8 - CMPBNE R3, $16, 2(PC) - VLEIB $10, $1, T_8 - - // add [m0, m1] to h - VAG H0_0, T_6, H0_0 - VAG H1_0, T_7, H1_0 - VAG H2_0, T_8, H2_0 - - VZERO M2 - VZERO M3 - VZERO M4 - VZERO M5 - VZERO T_10 - VZERO M0 - - // at this point R_0 .. R5_2 look like [r**2, r] - MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M2, M3, M4, M5, T_10, M0, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) - REDUCE2(H0_0, H1_0, H2_0, M2, M3, M4, M5, T_9, H0_1, H1_1, H2_1, T_10) - SUB $16, R3, R3 - CMPBLE R3, $0, next - -b1: - CMPBLE R3, $0, next - - // 1 block remaining - - // setup [r²,r] - VSLDB $8, R_0, R_0, R_0 - VSLDB $8, R_1, R_1, R_1 - VSLDB $8, R_2, R_2, R_2 - VSLDB $8, R5_1, R5_1, R5_1 - VSLDB $8, R5_2, R5_2, R5_2 - - VLVGG $1, RSAVE_0, R_0 - VLVGG $1, RSAVE_1, R_1 - VLVGG $1, RSAVE_2, R_2 - VLVGG $1, R5SAVE_1, R5_1 - VLVGG $1, R5SAVE_2, R5_2 - - // setup [h0, h1] - VSLDB $8, H0_0, H0_0, H0_0 - VSLDB $8, H1_0, H1_0, H1_0 - VSLDB $8, H2_0, H2_0, H2_0 - VO H0_1, H0_0, H0_0 - VO H1_1, H1_0, H1_0 - VO H2_1, H2_0, H2_0 - VZERO H0_1 - VZERO H1_1 - VZERO H2_1 - - VZERO M0 - VZERO M1 - VZERO M2 - VZERO M3 - VZERO M4 - VZERO M5 - - // H*[r**2, r] - MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, M5, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) - REDUCE2(H0_0, H1_0, H2_0, M0, M1, M2, M3, M4, T_9, T_10, H0_1, M5) - - // set up [0, m0] limbs - SUB $1, R3 - VLL R3, (R2), M0 - ADD $1, R3 - MOVBZ $1, R0 - CMPBEQ R3, $16, 2(PC) - VLVGB R3, R0, M0 - VZERO T_1 - VZERO T_2 - VZERO T_3 - EXPACC2(M0, T_1, T_2, T_3, T_4, T_5, T_6)// limbs: [0, m] - CMPBNE R3, $16, 2(PC) - VLEIB $10, $1, T_3 - - // h+m0 - VAQ H0_0, T_1, H0_0 - VAQ H1_0, T_2, H1_0 - VAQ H2_0, T_3, H2_0 - - VZERO M0 - VZERO M1 - VZERO M2 - VZERO M3 - VZERO M4 - VZERO M5 - MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, M5, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) - REDUCE2(H0_0, H1_0, H2_0, M0, M1, M2, M3, M4, T_9, T_10, H0_1, M5) - - BR next - -square: - // setup [r²,r] - VSLDB $8, R_0, R_0, R_0 - VSLDB $8, R_1, R_1, R_1 - VSLDB $8, R_2, R_2, R_2 - VSLDB $8, R5_1, R5_1, R5_1 - VSLDB $8, R5_2, R5_2, R5_2 - - VLVGG $1, RSAVE_0, R_0 - VLVGG $1, RSAVE_1, R_1 - VLVGG $1, RSAVE_2, R_2 - VLVGG $1, R5SAVE_1, R5_1 - VLVGG $1, R5SAVE_2, R5_2 - - // setup [h0, h1] - VSLDB $8, H0_0, H0_0, H0_0 - VSLDB $8, H1_0, H1_0, H1_0 - VSLDB $8, H2_0, H2_0, H2_0 - VO H0_1, H0_0, H0_0 - VO H1_1, H1_0, H1_0 - VO H2_1, H2_0, H2_0 - VZERO H0_1 - VZERO H1_1 - VZERO H2_1 - - VZERO M0 - VZERO M1 - VZERO M2 - VZERO M3 - VZERO M4 - VZERO M5 - - // (h0*r**2) + (h1*r) - MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, M5, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) - REDUCE2(H0_0, H1_0, H2_0, M0, M1, M2, M3, M4, T_9, T_10, H0_1, M5) - BR next diff --git a/vendor/golang.org/x/crypto/ssh/certs.go b/vendor/golang.org/x/crypto/ssh/certs.go index 00ed9923e78..916c840b698 100644 --- a/vendor/golang.org/x/crypto/ssh/certs.go +++ b/vendor/golang.org/x/crypto/ssh/certs.go @@ -17,12 +17,14 @@ import ( // These constants from [PROTOCOL.certkeys] represent the algorithm names // for certificate types supported by this package. const ( - CertAlgoRSAv01 = "ssh-rsa-cert-v01@openssh.com" - CertAlgoDSAv01 = "ssh-dss-cert-v01@openssh.com" - CertAlgoECDSA256v01 = "ecdsa-sha2-nistp256-cert-v01@openssh.com" - CertAlgoECDSA384v01 = "ecdsa-sha2-nistp384-cert-v01@openssh.com" - CertAlgoECDSA521v01 = "ecdsa-sha2-nistp521-cert-v01@openssh.com" - CertAlgoED25519v01 = "ssh-ed25519-cert-v01@openssh.com" + CertAlgoRSAv01 = "ssh-rsa-cert-v01@openssh.com" + CertAlgoDSAv01 = "ssh-dss-cert-v01@openssh.com" + CertAlgoECDSA256v01 = "ecdsa-sha2-nistp256-cert-v01@openssh.com" + CertAlgoECDSA384v01 = "ecdsa-sha2-nistp384-cert-v01@openssh.com" + CertAlgoECDSA521v01 = "ecdsa-sha2-nistp521-cert-v01@openssh.com" + CertAlgoSKECDSA256v01 = "sk-ecdsa-sha2-nistp256-cert-v01@openssh.com" + CertAlgoED25519v01 = "ssh-ed25519-cert-v01@openssh.com" + CertAlgoSKED25519v01 = "sk-ssh-ed25519-cert-v01@openssh.com" ) // Certificate types distinguish between host and user @@ -37,6 +39,7 @@ const ( type Signature struct { Format string Blob []byte + Rest []byte `ssh:"rest"` } // CertTimeInfinity can be used for OpenSSHCertV01.ValidBefore to indicate that @@ -411,8 +414,8 @@ func (c *CertChecker) CheckCert(principal string, cert *Certificate) error { return nil } -// SignCert sets c.SignatureKey to the authority's public key and stores a -// Signature, by authority, in the certificate. +// SignCert signs the certificate with an authority, setting the Nonce, +// SignatureKey, and Signature fields. func (c *Certificate) SignCert(rand io.Reader, authority Signer) error { c.Nonce = make([]byte, 32) if _, err := io.ReadFull(rand, c.Nonce); err != nil { @@ -429,12 +432,14 @@ func (c *Certificate) SignCert(rand io.Reader, authority Signer) error { } var certAlgoNames = map[string]string{ - KeyAlgoRSA: CertAlgoRSAv01, - KeyAlgoDSA: CertAlgoDSAv01, - KeyAlgoECDSA256: CertAlgoECDSA256v01, - KeyAlgoECDSA384: CertAlgoECDSA384v01, - KeyAlgoECDSA521: CertAlgoECDSA521v01, - KeyAlgoED25519: CertAlgoED25519v01, + KeyAlgoRSA: CertAlgoRSAv01, + KeyAlgoDSA: CertAlgoDSAv01, + KeyAlgoECDSA256: CertAlgoECDSA256v01, + KeyAlgoECDSA384: CertAlgoECDSA384v01, + KeyAlgoECDSA521: CertAlgoECDSA521v01, + KeyAlgoSKECDSA256: CertAlgoSKECDSA256v01, + KeyAlgoED25519: CertAlgoED25519v01, + KeyAlgoSKED25519: CertAlgoSKED25519v01, } // certToPrivAlgo returns the underlying algorithm for a certificate algorithm. @@ -518,6 +523,12 @@ func parseSignatureBody(in []byte) (out *Signature, rest []byte, ok bool) { return } + switch out.Format { + case KeyAlgoSKECDSA256, CertAlgoSKECDSA256v01, KeyAlgoSKED25519, CertAlgoSKED25519v01: + out.Rest = in + return out, nil, ok + } + return out, in, ok } diff --git a/vendor/golang.org/x/crypto/ssh/cipher.go b/vendor/golang.org/x/crypto/ssh/cipher.go index a65a923be3d..8bd6b3daff5 100644 --- a/vendor/golang.org/x/crypto/ssh/cipher.go +++ b/vendor/golang.org/x/crypto/ssh/cipher.go @@ -16,9 +16,8 @@ import ( "hash" "io" "io/ioutil" - "math/bits" - "golang.org/x/crypto/internal/chacha20" + "golang.org/x/crypto/chacha20" "golang.org/x/crypto/poly1305" ) @@ -120,7 +119,7 @@ var cipherModes = map[string]*cipherMode{ chacha20Poly1305ID: {64, 0, newChaCha20Cipher}, // CBC mode is insecure and so is not included in the default config. - // (See http://www.isg.rhul.ac.uk/~kp/SandPfinal.pdf). If absolutely + // (See https://www.ieee-security.org/TC/SP2013/papers/4977a526.pdf). If absolutely // needed, it's possible to specify a custom Config to enable it. // You should expect that an active attacker can recover plaintext if // you do. @@ -642,8 +641,8 @@ const chacha20Poly1305ID = "chacha20-poly1305@openssh.com" // the methods here also implement padding, which RFC4253 Section 6 // also requires of stream ciphers. type chacha20Poly1305Cipher struct { - lengthKey [8]uint32 - contentKey [8]uint32 + lengthKey [32]byte + contentKey [32]byte buf []byte } @@ -656,21 +655,21 @@ func newChaCha20Cipher(key, unusedIV, unusedMACKey []byte, unusedAlgs directionA buf: make([]byte, 256), } - for i := range c.contentKey { - c.contentKey[i] = binary.LittleEndian.Uint32(key[i*4 : (i+1)*4]) - } - for i := range c.lengthKey { - c.lengthKey[i] = binary.LittleEndian.Uint32(key[(i+8)*4 : (i+9)*4]) - } + copy(c.contentKey[:], key[:32]) + copy(c.lengthKey[:], key[32:]) return c, nil } func (c *chacha20Poly1305Cipher) readCipherPacket(seqNum uint32, r io.Reader) ([]byte, error) { - nonce := [3]uint32{0, 0, bits.ReverseBytes32(seqNum)} - s := chacha20.New(c.contentKey, nonce) - var polyKey [32]byte + nonce := make([]byte, 12) + binary.BigEndian.PutUint32(nonce[8:], seqNum) + s, err := chacha20.NewUnauthenticatedCipher(c.contentKey[:], nonce) + if err != nil { + return nil, err + } + var polyKey, discardBuf [32]byte s.XORKeyStream(polyKey[:], polyKey[:]) - s.Advance() // skip next 32 bytes + s.XORKeyStream(discardBuf[:], discardBuf[:]) // skip the next 32 bytes encryptedLength := c.buf[:4] if _, err := io.ReadFull(r, encryptedLength); err != nil { @@ -678,7 +677,11 @@ func (c *chacha20Poly1305Cipher) readCipherPacket(seqNum uint32, r io.Reader) ([ } var lenBytes [4]byte - chacha20.New(c.lengthKey, nonce).XORKeyStream(lenBytes[:], encryptedLength) + ls, err := chacha20.NewUnauthenticatedCipher(c.lengthKey[:], nonce) + if err != nil { + return nil, err + } + ls.XORKeyStream(lenBytes[:], encryptedLength) length := binary.BigEndian.Uint32(lenBytes[:]) if length > maxPacket { @@ -724,11 +727,15 @@ func (c *chacha20Poly1305Cipher) readCipherPacket(seqNum uint32, r io.Reader) ([ } func (c *chacha20Poly1305Cipher) writeCipherPacket(seqNum uint32, w io.Writer, rand io.Reader, payload []byte) error { - nonce := [3]uint32{0, 0, bits.ReverseBytes32(seqNum)} - s := chacha20.New(c.contentKey, nonce) - var polyKey [32]byte + nonce := make([]byte, 12) + binary.BigEndian.PutUint32(nonce[8:], seqNum) + s, err := chacha20.NewUnauthenticatedCipher(c.contentKey[:], nonce) + if err != nil { + return err + } + var polyKey, discardBuf [32]byte s.XORKeyStream(polyKey[:], polyKey[:]) - s.Advance() // skip next 32 bytes + s.XORKeyStream(discardBuf[:], discardBuf[:]) // skip the next 32 bytes // There is no blocksize, so fall back to multiple of 8 byte // padding, as described in RFC 4253, Sec 6. @@ -748,7 +755,11 @@ func (c *chacha20Poly1305Cipher) writeCipherPacket(seqNum uint32, w io.Writer, r } binary.BigEndian.PutUint32(c.buf, uint32(1+len(payload)+padding)) - chacha20.New(c.lengthKey, nonce).XORKeyStream(c.buf, c.buf[:4]) + ls, err := chacha20.NewUnauthenticatedCipher(c.lengthKey[:], nonce) + if err != nil { + return err + } + ls.XORKeyStream(c.buf, c.buf[:4]) c.buf[4] = byte(padding) copy(c.buf[5:], payload) packetEnd := 5 + len(payload) + padding diff --git a/vendor/golang.org/x/crypto/ssh/internal/bcrypt_pbkdf/bcrypt_pbkdf.go b/vendor/golang.org/x/crypto/ssh/internal/bcrypt_pbkdf/bcrypt_pbkdf.go new file mode 100644 index 00000000000..af81d266546 --- /dev/null +++ b/vendor/golang.org/x/crypto/ssh/internal/bcrypt_pbkdf/bcrypt_pbkdf.go @@ -0,0 +1,93 @@ +// Copyright 2014 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// Package bcrypt_pbkdf implements bcrypt_pbkdf(3) from OpenBSD. +// +// See https://flak.tedunangst.com/post/bcrypt-pbkdf and +// https://cvsweb.openbsd.org/cgi-bin/cvsweb/src/lib/libutil/bcrypt_pbkdf.c. +package bcrypt_pbkdf + +import ( + "crypto/sha512" + "errors" + "golang.org/x/crypto/blowfish" +) + +const blockSize = 32 + +// Key derives a key from the password, salt and rounds count, returning a +// []byte of length keyLen that can be used as cryptographic key. +func Key(password, salt []byte, rounds, keyLen int) ([]byte, error) { + if rounds < 1 { + return nil, errors.New("bcrypt_pbkdf: number of rounds is too small") + } + if len(password) == 0 { + return nil, errors.New("bcrypt_pbkdf: empty password") + } + if len(salt) == 0 || len(salt) > 1<<20 { + return nil, errors.New("bcrypt_pbkdf: bad salt length") + } + if keyLen > 1024 { + return nil, errors.New("bcrypt_pbkdf: keyLen is too large") + } + + numBlocks := (keyLen + blockSize - 1) / blockSize + key := make([]byte, numBlocks*blockSize) + + h := sha512.New() + h.Write(password) + shapass := h.Sum(nil) + + shasalt := make([]byte, 0, sha512.Size) + cnt, tmp := make([]byte, 4), make([]byte, blockSize) + for block := 1; block <= numBlocks; block++ { + h.Reset() + h.Write(salt) + cnt[0] = byte(block >> 24) + cnt[1] = byte(block >> 16) + cnt[2] = byte(block >> 8) + cnt[3] = byte(block) + h.Write(cnt) + bcryptHash(tmp, shapass, h.Sum(shasalt)) + + out := make([]byte, blockSize) + copy(out, tmp) + for i := 2; i <= rounds; i++ { + h.Reset() + h.Write(tmp) + bcryptHash(tmp, shapass, h.Sum(shasalt)) + for j := 0; j < len(out); j++ { + out[j] ^= tmp[j] + } + } + + for i, v := range out { + key[i*numBlocks+(block-1)] = v + } + } + return key[:keyLen], nil +} + +var magic = []byte("OxychromaticBlowfishSwatDynamite") + +func bcryptHash(out, shapass, shasalt []byte) { + c, err := blowfish.NewSaltedCipher(shapass, shasalt) + if err != nil { + panic(err) + } + for i := 0; i < 64; i++ { + blowfish.ExpandKey(shasalt, c) + blowfish.ExpandKey(shapass, c) + } + copy(out, magic) + for i := 0; i < 32; i += 8 { + for j := 0; j < 64; j++ { + c.Encrypt(out[i:i+8], out[i:i+8]) + } + } + // Swap bytes due to different endianness. + for i := 0; i < 32; i += 4 { + out[i+3], out[i+2], out[i+1], out[i] = out[i], out[i+1], out[i+2], out[i+3] + } +} diff --git a/vendor/golang.org/x/crypto/ssh/kex.go b/vendor/golang.org/x/crypto/ssh/kex.go index 16072004b17..7eedb209fa7 100644 --- a/vendor/golang.org/x/crypto/ssh/kex.go +++ b/vendor/golang.org/x/crypto/ssh/kex.go @@ -212,7 +212,7 @@ func (group *dhGroup) Server(c packetConn, randSource io.Reader, magics *handsha HostKey: hostKeyBytes, Signature: sig, Hash: crypto.SHA1, - }, nil + }, err } // ecdh performs Elliptic Curve Diffie-Hellman key exchange as @@ -572,7 +572,7 @@ func (gex *dhGEXSHA) diffieHellman(theirPublic, myPrivate *big.Int) (*big.Int, e return new(big.Int).Exp(theirPublic, myPrivate, gex.p), nil } -func (gex *dhGEXSHA) Client(c packetConn, randSource io.Reader, magics *handshakeMagics) (*kexResult, error) { +func (gex dhGEXSHA) Client(c packetConn, randSource io.Reader, magics *handshakeMagics) (*kexResult, error) { // Send GexRequest kexDHGexRequest := kexDHGexRequestMsg{ MinBits: dhGroupExchangeMinimumBits, @@ -677,7 +677,7 @@ func (gex *dhGEXSHA) Client(c packetConn, randSource io.Reader, magics *handshak // Server half implementation of the Diffie Hellman Key Exchange with SHA1 and SHA256. // // This is a minimal implementation to satisfy the automated tests. -func (gex *dhGEXSHA) Server(c packetConn, randSource io.Reader, magics *handshakeMagics, priv Signer) (result *kexResult, err error) { +func (gex dhGEXSHA) Server(c packetConn, randSource io.Reader, magics *handshakeMagics, priv Signer) (result *kexResult, err error) { // Receive GexRequest packet, err := c.readPacket() if err != nil { diff --git a/vendor/golang.org/x/crypto/ssh/keys.go b/vendor/golang.org/x/crypto/ssh/keys.go index 969804794fb..31f26349a05 100644 --- a/vendor/golang.org/x/crypto/ssh/keys.go +++ b/vendor/golang.org/x/crypto/ssh/keys.go @@ -7,6 +7,8 @@ package ssh import ( "bytes" "crypto" + "crypto/aes" + "crypto/cipher" "crypto/dsa" "crypto/ecdsa" "crypto/elliptic" @@ -25,17 +27,20 @@ import ( "strings" "golang.org/x/crypto/ed25519" + "golang.org/x/crypto/ssh/internal/bcrypt_pbkdf" ) // These constants represent the algorithm names for key types supported by this // package. const ( - KeyAlgoRSA = "ssh-rsa" - KeyAlgoDSA = "ssh-dss" - KeyAlgoECDSA256 = "ecdsa-sha2-nistp256" - KeyAlgoECDSA384 = "ecdsa-sha2-nistp384" - KeyAlgoECDSA521 = "ecdsa-sha2-nistp521" - KeyAlgoED25519 = "ssh-ed25519" + KeyAlgoRSA = "ssh-rsa" + KeyAlgoDSA = "ssh-dss" + KeyAlgoECDSA256 = "ecdsa-sha2-nistp256" + KeyAlgoSKECDSA256 = "sk-ecdsa-sha2-nistp256@openssh.com" + KeyAlgoECDSA384 = "ecdsa-sha2-nistp384" + KeyAlgoECDSA521 = "ecdsa-sha2-nistp521" + KeyAlgoED25519 = "ssh-ed25519" + KeyAlgoSKED25519 = "sk-ssh-ed25519@openssh.com" ) // These constants represent non-default signature algorithms that are supported @@ -58,9 +63,13 @@ func parsePubKey(in []byte, algo string) (pubKey PublicKey, rest []byte, err err return parseDSA(in) case KeyAlgoECDSA256, KeyAlgoECDSA384, KeyAlgoECDSA521: return parseECDSA(in) + case KeyAlgoSKECDSA256: + return parseSKECDSA(in) case KeyAlgoED25519: return parseED25519(in) - case CertAlgoRSAv01, CertAlgoDSAv01, CertAlgoECDSA256v01, CertAlgoECDSA384v01, CertAlgoECDSA521v01, CertAlgoED25519v01: + case KeyAlgoSKED25519: + return parseSKEd25519(in) + case CertAlgoRSAv01, CertAlgoDSAv01, CertAlgoECDSA256v01, CertAlgoECDSA384v01, CertAlgoECDSA521v01, CertAlgoSKECDSA256v01, CertAlgoED25519v01, CertAlgoSKED25519v01: cert, err := parseCert(in, certToPrivAlgo(algo)) if err != nil { return nil, nil, err @@ -553,9 +562,11 @@ func parseED25519(in []byte) (out PublicKey, rest []byte, err error) { return nil, nil, err } - key := ed25519.PublicKey(w.KeyBytes) + if l := len(w.KeyBytes); l != ed25519.PublicKeySize { + return nil, nil, fmt.Errorf("invalid size %d for Ed25519 public key", l) + } - return (ed25519PublicKey)(key), w.Rest, nil + return ed25519PublicKey(w.KeyBytes), w.Rest, nil } func (k ed25519PublicKey) Marshal() []byte { @@ -573,9 +584,11 @@ func (k ed25519PublicKey) Verify(b []byte, sig *Signature) error { if sig.Format != k.Type() { return fmt.Errorf("ssh: signature type %s for key type %s", sig.Format, k.Type()) } + if l := len(k); l != ed25519.PublicKeySize { + return fmt.Errorf("ssh: invalid size %d for Ed25519 public key", l) + } - edKey := (ed25519.PublicKey)(k) - if ok := ed25519.Verify(edKey, b, sig.Blob); !ok { + if ok := ed25519.Verify(ed25519.PublicKey(k), b, sig.Blob); !ok { return errors.New("ssh: signature did not verify") } @@ -685,6 +698,224 @@ func (k *ecdsaPublicKey) CryptoPublicKey() crypto.PublicKey { return (*ecdsa.PublicKey)(k) } +// skFields holds the additional fields present in U2F/FIDO2 signatures. +// See openssh/PROTOCOL.u2f 'SSH U2F Signatures' for details. +type skFields struct { + // Flags contains U2F/FIDO2 flags such as 'user present' + Flags byte + // Counter is a monotonic signature counter which can be + // used to detect concurrent use of a private key, should + // it be extracted from hardware. + Counter uint32 +} + +type skECDSAPublicKey struct { + // application is a URL-like string, typically "ssh:" for SSH. + // see openssh/PROTOCOL.u2f for details. + application string + ecdsa.PublicKey +} + +func (k *skECDSAPublicKey) Type() string { + return KeyAlgoSKECDSA256 +} + +func (k *skECDSAPublicKey) nistID() string { + return "nistp256" +} + +func parseSKECDSA(in []byte) (out PublicKey, rest []byte, err error) { + var w struct { + Curve string + KeyBytes []byte + Application string + Rest []byte `ssh:"rest"` + } + + if err := Unmarshal(in, &w); err != nil { + return nil, nil, err + } + + key := new(skECDSAPublicKey) + key.application = w.Application + + if w.Curve != "nistp256" { + return nil, nil, errors.New("ssh: unsupported curve") + } + key.Curve = elliptic.P256() + + key.X, key.Y = elliptic.Unmarshal(key.Curve, w.KeyBytes) + if key.X == nil || key.Y == nil { + return nil, nil, errors.New("ssh: invalid curve point") + } + + return key, w.Rest, nil +} + +func (k *skECDSAPublicKey) Marshal() []byte { + // See RFC 5656, section 3.1. + keyBytes := elliptic.Marshal(k.Curve, k.X, k.Y) + w := struct { + Name string + ID string + Key []byte + Application string + }{ + k.Type(), + k.nistID(), + keyBytes, + k.application, + } + + return Marshal(&w) +} + +func (k *skECDSAPublicKey) Verify(data []byte, sig *Signature) error { + if sig.Format != k.Type() { + return fmt.Errorf("ssh: signature type %s for key type %s", sig.Format, k.Type()) + } + + h := ecHash(k.Curve).New() + h.Write([]byte(k.application)) + appDigest := h.Sum(nil) + + h.Reset() + h.Write(data) + dataDigest := h.Sum(nil) + + var ecSig struct { + R *big.Int + S *big.Int + } + if err := Unmarshal(sig.Blob, &ecSig); err != nil { + return err + } + + var skf skFields + if err := Unmarshal(sig.Rest, &skf); err != nil { + return err + } + + blob := struct { + ApplicationDigest []byte `ssh:"rest"` + Flags byte + Counter uint32 + MessageDigest []byte `ssh:"rest"` + }{ + appDigest, + skf.Flags, + skf.Counter, + dataDigest, + } + + original := Marshal(blob) + + h.Reset() + h.Write(original) + digest := h.Sum(nil) + + if ecdsa.Verify((*ecdsa.PublicKey)(&k.PublicKey), digest, ecSig.R, ecSig.S) { + return nil + } + return errors.New("ssh: signature did not verify") +} + +type skEd25519PublicKey struct { + // application is a URL-like string, typically "ssh:" for SSH. + // see openssh/PROTOCOL.u2f for details. + application string + ed25519.PublicKey +} + +func (k *skEd25519PublicKey) Type() string { + return KeyAlgoSKED25519 +} + +func parseSKEd25519(in []byte) (out PublicKey, rest []byte, err error) { + var w struct { + KeyBytes []byte + Application string + Rest []byte `ssh:"rest"` + } + + if err := Unmarshal(in, &w); err != nil { + return nil, nil, err + } + + if l := len(w.KeyBytes); l != ed25519.PublicKeySize { + return nil, nil, fmt.Errorf("invalid size %d for Ed25519 public key", l) + } + + key := new(skEd25519PublicKey) + key.application = w.Application + key.PublicKey = ed25519.PublicKey(w.KeyBytes) + + return key, w.Rest, nil +} + +func (k *skEd25519PublicKey) Marshal() []byte { + w := struct { + Name string + KeyBytes []byte + Application string + }{ + KeyAlgoSKED25519, + []byte(k.PublicKey), + k.application, + } + return Marshal(&w) +} + +func (k *skEd25519PublicKey) Verify(data []byte, sig *Signature) error { + if sig.Format != k.Type() { + return fmt.Errorf("ssh: signature type %s for key type %s", sig.Format, k.Type()) + } + if l := len(k.PublicKey); l != ed25519.PublicKeySize { + return fmt.Errorf("invalid size %d for Ed25519 public key", l) + } + + h := sha256.New() + h.Write([]byte(k.application)) + appDigest := h.Sum(nil) + + h.Reset() + h.Write(data) + dataDigest := h.Sum(nil) + + var edSig struct { + Signature []byte `ssh:"rest"` + } + + if err := Unmarshal(sig.Blob, &edSig); err != nil { + return err + } + + var skf skFields + if err := Unmarshal(sig.Rest, &skf); err != nil { + return err + } + + blob := struct { + ApplicationDigest []byte `ssh:"rest"` + Flags byte + Counter uint32 + MessageDigest []byte `ssh:"rest"` + }{ + appDigest, + skf.Flags, + skf.Counter, + dataDigest, + } + + original := Marshal(blob) + + if ok := ed25519.Verify(k.PublicKey, original, edSig.Signature); !ok { + return errors.New("ssh: signature did not verify") + } + + return nil +} + // NewSignerFromKey takes an *rsa.PrivateKey, *dsa.PrivateKey, // *ecdsa.PrivateKey or any other crypto.Signer and returns a // corresponding Signer instance. ECDSA keys must use P-256, P-384 or @@ -830,14 +1061,18 @@ func NewPublicKey(key interface{}) (PublicKey, error) { case *dsa.PublicKey: return (*dsaPublicKey)(key), nil case ed25519.PublicKey: - return (ed25519PublicKey)(key), nil + if l := len(key); l != ed25519.PublicKeySize { + return nil, fmt.Errorf("ssh: invalid size %d for Ed25519 public key", l) + } + return ed25519PublicKey(key), nil default: return nil, fmt.Errorf("ssh: unsupported key type %T", key) } } // ParsePrivateKey returns a Signer from a PEM encoded private key. It supports -// the same keys as ParseRawPrivateKey. +// the same keys as ParseRawPrivateKey. If the private key is encrypted, it +// will return a PassphraseMissingError. func ParsePrivateKey(pemBytes []byte) (Signer, error) { key, err := ParseRawPrivateKey(pemBytes) if err != nil { @@ -850,8 +1085,8 @@ func ParsePrivateKey(pemBytes []byte) (Signer, error) { // ParsePrivateKeyWithPassphrase returns a Signer from a PEM encoded private // key and passphrase. It supports the same keys as // ParseRawPrivateKeyWithPassphrase. -func ParsePrivateKeyWithPassphrase(pemBytes, passPhrase []byte) (Signer, error) { - key, err := ParseRawPrivateKeyWithPassphrase(pemBytes, passPhrase) +func ParsePrivateKeyWithPassphrase(pemBytes, passphrase []byte) (Signer, error) { + key, err := ParseRawPrivateKeyWithPassphrase(pemBytes, passphrase) if err != nil { return nil, err } @@ -867,8 +1102,21 @@ func encryptedBlock(block *pem.Block) bool { return strings.Contains(block.Headers["Proc-Type"], "ENCRYPTED") } +// A PassphraseMissingError indicates that parsing this private key requires a +// passphrase. Use ParsePrivateKeyWithPassphrase. +type PassphraseMissingError struct { + // PublicKey will be set if the private key format includes an unencrypted + // public key along with the encrypted private key. + PublicKey PublicKey +} + +func (*PassphraseMissingError) Error() string { + return "ssh: this private key is passphrase protected" +} + // ParseRawPrivateKey returns a private key from a PEM encoded private key. It -// supports RSA (PKCS#1), PKCS#8, DSA (OpenSSL), and ECDSA private keys. +// supports RSA (PKCS#1), PKCS#8, DSA (OpenSSL), and ECDSA private keys. If the +// private key is encrypted, it will return a PassphraseMissingError. func ParseRawPrivateKey(pemBytes []byte) (interface{}, error) { block, _ := pem.Decode(pemBytes) if block == nil { @@ -876,7 +1124,7 @@ func ParseRawPrivateKey(pemBytes []byte) (interface{}, error) { } if encryptedBlock(block) { - return nil, errors.New("ssh: cannot decode encrypted private keys") + return nil, &PassphraseMissingError{} } switch block.Type { @@ -890,33 +1138,35 @@ func ParseRawPrivateKey(pemBytes []byte) (interface{}, error) { case "DSA PRIVATE KEY": return ParseDSAPrivateKey(block.Bytes) case "OPENSSH PRIVATE KEY": - return parseOpenSSHPrivateKey(block.Bytes) + return parseOpenSSHPrivateKey(block.Bytes, unencryptedOpenSSHKey) default: return nil, fmt.Errorf("ssh: unsupported key type %q", block.Type) } } // ParseRawPrivateKeyWithPassphrase returns a private key decrypted with -// passphrase from a PEM encoded private key. If wrong passphrase, return -// x509.IncorrectPasswordError. -func ParseRawPrivateKeyWithPassphrase(pemBytes, passPhrase []byte) (interface{}, error) { +// passphrase from a PEM encoded private key. If the passphrase is wrong, it +// will return x509.IncorrectPasswordError. +func ParseRawPrivateKeyWithPassphrase(pemBytes, passphrase []byte) (interface{}, error) { block, _ := pem.Decode(pemBytes) if block == nil { return nil, errors.New("ssh: no key found") } - buf := block.Bytes - if encryptedBlock(block) { - if x509.IsEncryptedPEMBlock(block) { - var err error - buf, err = x509.DecryptPEMBlock(block, passPhrase) - if err != nil { - if err == x509.IncorrectPasswordError { - return nil, err - } - return nil, fmt.Errorf("ssh: cannot decode encrypted private keys: %v", err) - } + if block.Type == "OPENSSH PRIVATE KEY" { + return parseOpenSSHPrivateKey(block.Bytes, passphraseProtectedOpenSSHKey(passphrase)) + } + + if !encryptedBlock(block) || !x509.IsEncryptedPEMBlock(block) { + return nil, errors.New("ssh: not an encrypted key") + } + + buf, err := x509.DecryptPEMBlock(block, passphrase) + if err != nil { + if err == x509.IncorrectPasswordError { + return nil, err } + return nil, fmt.Errorf("ssh: cannot decode encrypted private keys: %v", err) } switch block.Type { @@ -926,8 +1176,6 @@ func ParseRawPrivateKeyWithPassphrase(pemBytes, passPhrase []byte) (interface{}, return x509.ParseECPrivateKey(buf) case "DSA PRIVATE KEY": return ParseDSAPrivateKey(buf) - case "OPENSSH PRIVATE KEY": - return parseOpenSSHPrivateKey(buf) default: return nil, fmt.Errorf("ssh: unsupported key type %q", block.Type) } @@ -965,9 +1213,68 @@ func ParseDSAPrivateKey(der []byte) (*dsa.PrivateKey, error) { }, nil } -// Implemented based on the documentation at -// https://github.com/openssh/openssh-portable/blob/master/PROTOCOL.key -func parseOpenSSHPrivateKey(key []byte) (crypto.PrivateKey, error) { +func unencryptedOpenSSHKey(cipherName, kdfName, kdfOpts string, privKeyBlock []byte) ([]byte, error) { + if kdfName != "none" || cipherName != "none" { + return nil, &PassphraseMissingError{} + } + if kdfOpts != "" { + return nil, errors.New("ssh: invalid openssh private key") + } + return privKeyBlock, nil +} + +func passphraseProtectedOpenSSHKey(passphrase []byte) openSSHDecryptFunc { + return func(cipherName, kdfName, kdfOpts string, privKeyBlock []byte) ([]byte, error) { + if kdfName == "none" || cipherName == "none" { + return nil, errors.New("ssh: key is not password protected") + } + if kdfName != "bcrypt" { + return nil, fmt.Errorf("ssh: unknown KDF %q, only supports %q", kdfName, "bcrypt") + } + + var opts struct { + Salt string + Rounds uint32 + } + if err := Unmarshal([]byte(kdfOpts), &opts); err != nil { + return nil, err + } + + k, err := bcrypt_pbkdf.Key(passphrase, []byte(opts.Salt), int(opts.Rounds), 32+16) + if err != nil { + return nil, err + } + key, iv := k[:32], k[32:] + + c, err := aes.NewCipher(key) + if err != nil { + return nil, err + } + switch cipherName { + case "aes256-ctr": + ctr := cipher.NewCTR(c, iv) + ctr.XORKeyStream(privKeyBlock, privKeyBlock) + case "aes256-cbc": + if len(privKeyBlock)%c.BlockSize() != 0 { + return nil, fmt.Errorf("ssh: invalid encrypted private key length, not a multiple of the block size") + } + cbc := cipher.NewCBCDecrypter(c, iv) + cbc.CryptBlocks(privKeyBlock, privKeyBlock) + default: + return nil, fmt.Errorf("ssh: unknown cipher %q, only supports %q or %q", cipherName, "aes256-ctr", "aes256-cbc") + } + + return privKeyBlock, nil + } +} + +type openSSHDecryptFunc func(CipherName, KdfName, KdfOpts string, PrivKeyBlock []byte) ([]byte, error) + +// parseOpenSSHPrivateKey parses an OpenSSH private key, using the decrypt +// function to unwrap the encrypted portion. unencryptedOpenSSHKey can be used +// as the decrypt function to parse an unencrypted private key. See +// https://github.com/openssh/openssh-portable/blob/master/PROTOCOL.key. +func parseOpenSSHPrivateKey(key []byte, decrypt openSSHDecryptFunc) (crypto.PrivateKey, error) { const magic = "openssh-key-v1\x00" if len(key) < len(magic) || string(key[:len(magic)]) != magic { return nil, errors.New("ssh: invalid openssh private key format") @@ -986,9 +1293,22 @@ func parseOpenSSHPrivateKey(key []byte) (crypto.PrivateKey, error) { if err := Unmarshal(remaining, &w); err != nil { return nil, err } + if w.NumKeys != 1 { + // We only support single key files, and so does OpenSSH. + // https://github.com/openssh/openssh-portable/blob/4103a3ec7/sshkey.c#L4171 + return nil, errors.New("ssh: multi-key files are not supported") + } - if w.KdfName != "none" || w.CipherName != "none" { - return nil, errors.New("ssh: cannot decode encrypted private keys") + privKeyBlock, err := decrypt(w.CipherName, w.KdfName, w.KdfOpts, w.PrivKeyBlock) + if err != nil { + if err, ok := err.(*PassphraseMissingError); ok { + pub, errPub := ParsePublicKey(w.PubKey) + if errPub != nil { + return nil, fmt.Errorf("ssh: failed to parse embedded public key: %v", errPub) + } + err.PublicKey = pub + } + return nil, err } pk1 := struct { @@ -998,15 +1318,13 @@ func parseOpenSSHPrivateKey(key []byte) (crypto.PrivateKey, error) { Rest []byte `ssh:"rest"` }{} - if err := Unmarshal(w.PrivKeyBlock, &pk1); err != nil { - return nil, err - } - - if pk1.Check1 != pk1.Check2 { - return nil, errors.New("ssh: checkint mismatch") + if err := Unmarshal(privKeyBlock, &pk1); err != nil || pk1.Check1 != pk1.Check2 { + if w.CipherName != "none" { + return nil, x509.IncorrectPasswordError + } + return nil, errors.New("ssh: malformed OpenSSH key") } - // we only handle ed25519 and rsa keys currently switch pk1.Keytype { case KeyAlgoRSA: // https://github.com/openssh/openssh-portable/blob/master/sshkey.c#L2760-L2773 @@ -1025,10 +1343,8 @@ func parseOpenSSHPrivateKey(key []byte) (crypto.PrivateKey, error) { return nil, err } - for i, b := range key.Pad { - if int(b) != i+1 { - return nil, errors.New("ssh: padding not as expected") - } + if err := checkOpenSSHKeyPadding(key.Pad); err != nil { + return nil, err } pk := &rsa.PrivateKey{ @@ -1063,20 +1379,78 @@ func parseOpenSSHPrivateKey(key []byte) (crypto.PrivateKey, error) { return nil, errors.New("ssh: private key unexpected length") } - for i, b := range key.Pad { - if int(b) != i+1 { - return nil, errors.New("ssh: padding not as expected") - } + if err := checkOpenSSHKeyPadding(key.Pad); err != nil { + return nil, err } pk := ed25519.PrivateKey(make([]byte, ed25519.PrivateKeySize)) copy(pk, key.Priv) return &pk, nil + case KeyAlgoECDSA256, KeyAlgoECDSA384, KeyAlgoECDSA521: + key := struct { + Curve string + Pub []byte + D *big.Int + Comment string + Pad []byte `ssh:"rest"` + }{} + + if err := Unmarshal(pk1.Rest, &key); err != nil { + return nil, err + } + + if err := checkOpenSSHKeyPadding(key.Pad); err != nil { + return nil, err + } + + var curve elliptic.Curve + switch key.Curve { + case "nistp256": + curve = elliptic.P256() + case "nistp384": + curve = elliptic.P384() + case "nistp521": + curve = elliptic.P521() + default: + return nil, errors.New("ssh: unhandled elliptic curve: " + key.Curve) + } + + X, Y := elliptic.Unmarshal(curve, key.Pub) + if X == nil || Y == nil { + return nil, errors.New("ssh: failed to unmarshal public key") + } + + if key.D.Cmp(curve.Params().N) >= 0 { + return nil, errors.New("ssh: scalar is out of range") + } + + x, y := curve.ScalarBaseMult(key.D.Bytes()) + if x.Cmp(X) != 0 || y.Cmp(Y) != 0 { + return nil, errors.New("ssh: public key does not match private key") + } + + return &ecdsa.PrivateKey{ + PublicKey: ecdsa.PublicKey{ + Curve: curve, + X: X, + Y: Y, + }, + D: key.D, + }, nil default: return nil, errors.New("ssh: unhandled key type") } } +func checkOpenSSHKeyPadding(pad []byte) error { + for i, b := range pad { + if int(b) != i+1 { + return errors.New("ssh: padding not as expected") + } + } + return nil +} + // FingerprintLegacyMD5 returns the user presentation of the key's // fingerprint as described by RFC 4716 section 4. func FingerprintLegacyMD5(pubKey PublicKey) string { diff --git a/vendor/golang.org/x/crypto/ssh/server.go b/vendor/golang.org/x/crypto/ssh/server.go index 7a5a1d7ad39..7d42a8c88d2 100644 --- a/vendor/golang.org/x/crypto/ssh/server.go +++ b/vendor/golang.org/x/crypto/ssh/server.go @@ -284,8 +284,8 @@ func (s *connection) serverHandshake(config *ServerConfig) (*Permissions, error) func isAcceptableAlgo(algo string) bool { switch algo { - case KeyAlgoRSA, KeyAlgoDSA, KeyAlgoECDSA256, KeyAlgoECDSA384, KeyAlgoECDSA521, KeyAlgoED25519, - CertAlgoRSAv01, CertAlgoDSAv01, CertAlgoECDSA256v01, CertAlgoECDSA384v01, CertAlgoECDSA521v01, CertAlgoED25519v01: + case KeyAlgoRSA, KeyAlgoDSA, KeyAlgoECDSA256, KeyAlgoECDSA384, KeyAlgoECDSA521, KeyAlgoSKECDSA256, KeyAlgoED25519, KeyAlgoSKED25519, + CertAlgoRSAv01, CertAlgoDSAv01, CertAlgoECDSA256v01, CertAlgoECDSA384v01, CertAlgoECDSA521v01, CertAlgoSKECDSA256v01, CertAlgoED25519v01, CertAlgoSKED25519v01: return true } return false diff --git a/vendor/modules.txt b/vendor/modules.txt index dd2668ca47f..8a4e50612dd 100644 --- a/vendor/modules.txt +++ b/vendor/modules.txt @@ -487,14 +487,14 @@ github.com/hashicorp/aws-sdk-go-base github.com/hashicorp/errwrap # github.com/hashicorp/go-cleanhttp v0.5.1 github.com/hashicorp/go-cleanhttp -# github.com/hashicorp/go-getter v1.4.0 +# github.com/hashicorp/go-getter v1.4.2-0.20200106182914-9813cbd4eb02 github.com/hashicorp/go-getter github.com/hashicorp/go-getter/helper/url # github.com/hashicorp/go-hclog v0.10.0 github.com/hashicorp/go-hclog # github.com/hashicorp/go-multierror v1.1.0 github.com/hashicorp/go-multierror -# github.com/hashicorp/go-plugin v1.2.0 +# github.com/hashicorp/go-plugin v1.3.0 github.com/hashicorp/go-plugin github.com/hashicorp/go-plugin/internal/plugin # github.com/hashicorp/go-safetemp v1.0.0 @@ -530,7 +530,7 @@ github.com/hashicorp/hcl/v2/hclwrite github.com/hashicorp/hcl/v2/json # github.com/hashicorp/logutils v1.0.0 github.com/hashicorp/logutils -# github.com/hashicorp/terraform-config-inspect v0.0.0-20191115094559-17f92b0546e8 +# github.com/hashicorp/terraform-config-inspect v0.0.0-20191212124732-c6ae6269b9d7 github.com/hashicorp/terraform-config-inspect/tfconfig # github.com/hashicorp/terraform-json v0.5.0 github.com/hashicorp/terraform-json @@ -773,14 +773,14 @@ go.opencensus.io/trace go.opencensus.io/trace/internal go.opencensus.io/trace/propagation go.opencensus.io/trace/tracestate -# golang.org/x/crypto v0.0.0-20191011191535-87dc89f01550 +# golang.org/x/crypto v0.0.0-20200510223506-06a226fb4e37 golang.org/x/crypto/bcrypt golang.org/x/crypto/blowfish golang.org/x/crypto/cast5 +golang.org/x/crypto/chacha20 golang.org/x/crypto/curve25519 golang.org/x/crypto/ed25519 golang.org/x/crypto/ed25519/internal/edwards25519 -golang.org/x/crypto/internal/chacha20 golang.org/x/crypto/internal/subtle golang.org/x/crypto/openpgp golang.org/x/crypto/openpgp/armor @@ -790,6 +790,7 @@ golang.org/x/crypto/openpgp/packet golang.org/x/crypto/openpgp/s2k golang.org/x/crypto/poly1305 golang.org/x/crypto/ssh +golang.org/x/crypto/ssh/internal/bcrypt_pbkdf # golang.org/x/mod v0.2.0 golang.org/x/mod/internal/lazyregexp golang.org/x/mod/modfile