platform_build_soong/ui/build/build.go

339 lines
11 KiB
Go
Raw Normal View History

Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
// Copyright 2017 Google Inc. All rights reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package build
import (
"io/ioutil"
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
"os"
"path/filepath"
"text/template"
"android/soong/ui/metrics"
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
)
// SetupOutDir ensures the out directory exists, and has the proper files to
// prevent kati from recursing into it.
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
func SetupOutDir(ctx Context, config Config) {
ensureEmptyFileExists(ctx, filepath.Join(config.OutDir(), "Android.mk"))
ensureEmptyFileExists(ctx, filepath.Join(config.OutDir(), "CleanSpec.mk"))
if !config.SkipKati() {
// Run soong_build with Kati for a hybrid build, e.g. running the
// AndroidMk singleton and postinstall commands. Communicate this to
// soong_build by writing an empty .soong.kati_enabled marker file in the
// soong_build output directory for the soong_build primary builder to
// know if the user wants to run Kati after.
//
// This does not preclude running Kati for *product configuration purposes*.
ensureEmptyFileExists(ctx, filepath.Join(config.SoongOutDir(), ".soong.kati_enabled"))
}
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
// The ninja_build file is used by our buildbots to understand that the output
// can be parsed as ninja output.
ensureEmptyFileExists(ctx, filepath.Join(config.OutDir(), "ninja_build"))
ensureEmptyFileExists(ctx, filepath.Join(config.OutDir(), ".out-dir"))
if buildDateTimeFile, ok := config.environ.Get("BUILD_DATETIME_FILE"); ok {
err := ioutil.WriteFile(buildDateTimeFile, []byte(config.buildDateTime), 0666) // a+rw
if err != nil {
ctx.Fatalln("Failed to write BUILD_DATETIME to file:", err)
}
} else {
ctx.Fatalln("Missing BUILD_DATETIME_FILE")
}
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
}
var combinedBuildNinjaTemplate = template.Must(template.New("combined").Parse(`
builddir = {{.OutDir}}
{{if .UseRemoteBuild }}pool local_pool
depth = {{.Parallel}}
{{end -}}
pool highmem_pool
depth = {{.HighmemParallel}}
Add a Kati-based packaging step The idea is that we'd move the installation and packaging tasks over to it, using data from Soong & the Kati reading Android.mk files. This would allow us to make more fundamental changes about how we package things without having to adjust makefiles throughout the tree. Possible use cases: * Moving some information from Soong's Android.mk output to a file read by the packaging step may allow us to read the Android.mk files less often, speeding up builds. * Refactoring our current two-stage ASAN builds to run the Kati build step twice, writing into different object directories, then have a single packaging step that reads both outputs. Soong already has the capability of writing out a single ninja file with all the asan combinations. * Running two build steps, one building the system-related modules using a "generic" device configuration, and one building the vendor modules using a specific device configuration. This could enforce a GSI/mainline system vs vendor split in a single build invocation. * If all installation is through this tool, it will be much easier to track what should no longer be installed on an incremental build, reducing the need for installclean. * Changing PRODUCT_PACKAGES should be a much faster operation, which means we could keep track of local additions to the images. Then `mma` would be more persistent, instead of installing something once, then never updating it again. Eventually we plan on switching from Kati to something Go-based, but this is a more incremental approach while we clean up everything else. Currently, this just moves the dist-for-goal handling over to the packaging step, so that we don't need to read Android.mk files when DIST_DIR changes, or we switch between dist vs not. Bug: 116968624 Bug: 117463001 Test: m nothing Change-Id: Idec5ac6f7c7475397ba0fb65bd3785128a7517df
2018-09-27 00:00:42 +02:00
{{if .HasKatiSuffix}}subninja {{.KatiBuildNinjaFile}}
subninja {{.KatiPackageNinjaFile}}
{{end -}}
Add a Kati-based packaging step The idea is that we'd move the installation and packaging tasks over to it, using data from Soong & the Kati reading Android.mk files. This would allow us to make more fundamental changes about how we package things without having to adjust makefiles throughout the tree. Possible use cases: * Moving some information from Soong's Android.mk output to a file read by the packaging step may allow us to read the Android.mk files less often, speeding up builds. * Refactoring our current two-stage ASAN builds to run the Kati build step twice, writing into different object directories, then have a single packaging step that reads both outputs. Soong already has the capability of writing out a single ninja file with all the asan combinations. * Running two build steps, one building the system-related modules using a "generic" device configuration, and one building the vendor modules using a specific device configuration. This could enforce a GSI/mainline system vs vendor split in a single build invocation. * If all installation is through this tool, it will be much easier to track what should no longer be installed on an incremental build, reducing the need for installclean. * Changing PRODUCT_PACKAGES should be a much faster operation, which means we could keep track of local additions to the images. Then `mma` would be more persistent, instead of installing something once, then never updating it again. Eventually we plan on switching from Kati to something Go-based, but this is a more incremental approach while we clean up everything else. Currently, this just moves the dist-for-goal handling over to the packaging step, so that we don't need to read Android.mk files when DIST_DIR changes, or we switch between dist vs not. Bug: 116968624 Bug: 117463001 Test: m nothing Change-Id: Idec5ac6f7c7475397ba0fb65bd3785128a7517df
2018-09-27 00:00:42 +02:00
subninja {{.SoongNinjaFile}}
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
`))
func createCombinedBuildNinjaFile(ctx Context, config Config) {
// If we're in SkipKati mode, skip creating this file if it already exists
if config.SkipKati() {
if _, err := os.Stat(config.CombinedNinjaFile()); err == nil || !os.IsNotExist(err) {
return
}
}
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
file, err := os.Create(config.CombinedNinjaFile())
if err != nil {
ctx.Fatalln("Failed to create combined ninja file:", err)
}
defer file.Close()
if err := combinedBuildNinjaTemplate.Execute(file, config); err != nil {
ctx.Fatalln("Failed to write combined ninja file:", err)
}
}
// These are bitmasks which can be used to check whether various flags are set e.g. whether to use Bazel.
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
const (
BuildNone = iota
BuildProductConfig = 1 << iota
BuildSoong = 1 << iota
BuildKati = 1 << iota
BuildNinja = 1 << iota
BuildBazel = 1 << iota
RunBuildTests = 1 << iota
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
BuildAll = BuildProductConfig | BuildSoong | BuildKati | BuildNinja
BuildAllWithBazel = BuildProductConfig | BuildSoong | BuildKati | BuildBazel
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
)
// checkProblematicFiles fails the build if existing Android.mk or CleanSpec.mk files are found at the root of the tree.
func checkProblematicFiles(ctx Context) {
files := []string{"Android.mk", "CleanSpec.mk"}
for _, file := range files {
if _, err := os.Stat(file); !os.IsNotExist(err) {
absolute := absPath(ctx, file)
ctx.Printf("Found %s in tree root. This file needs to be removed to build.\n", file)
ctx.Fatalf(" rm %s\n", absolute)
}
}
}
// checkCaseSensitivity issues a warning if a case-insensitive file system is being used.
func checkCaseSensitivity(ctx Context, config Config) {
outDir := config.OutDir()
lowerCase := filepath.Join(outDir, "casecheck.txt")
upperCase := filepath.Join(outDir, "CaseCheck.txt")
lowerData := "a"
upperData := "B"
if err := ioutil.WriteFile(lowerCase, []byte(lowerData), 0666); err != nil { // a+rw
ctx.Fatalln("Failed to check case sensitivity:", err)
}
if err := ioutil.WriteFile(upperCase, []byte(upperData), 0666); err != nil { // a+rw
ctx.Fatalln("Failed to check case sensitivity:", err)
}
res, err := ioutil.ReadFile(lowerCase)
if err != nil {
ctx.Fatalln("Failed to check case sensitivity:", err)
}
if string(res) != lowerData {
ctx.Println("************************************************************")
ctx.Println("You are building on a case-insensitive filesystem.")
ctx.Println("Please move your source tree to a case-sensitive filesystem.")
ctx.Println("************************************************************")
ctx.Fatalln("Case-insensitive filesystems not supported")
}
}
// help prints a help/usage message, via the build/make/help.sh script.
func help(ctx Context, config Config) {
cmd := Command(ctx, config, "help.sh", "build/make/help.sh")
cmd.Sandbox = dumpvarsSandbox
Add a unified status reporting UI This adds a new status package that merges the running of "actions" (ninja calls them edges) of multiple tools into one view of the current state, and gives that to a number of different outputs. For inputs: Kati's output parser has been rewritten (and moved) to map onto the StartAction/FinishAction API. A byproduct of this is that the build servers should be able to extract errors from Kati better, since they look like the errors that Ninja used to write. Ninja is no longer directly connected to the terminal, but its output is read via the protobuf frontend API, so it's just another tool whose output becomes merged together. multiproduct_kati loses its custom status routines, and uses the common one instead. For outputs: The primary output is the ui/terminal.Status type, which along with ui/terminal.Writer now controls everything about the terminal output. Today, this doesn't really change any behaviors, but having all terminal output going through here allows a more complicated (multi-line / full window) status display in the future. The tracer acts as an output of the status package, tracing all the action start / finish events. This replaces reading the .ninja_log file, so it now properly handles multiple output files from a single action. A new rotated log file (out/error.log, or out/dist/logs/error.log) just contains a description of all of the errors that happened during the current build. Another new compressed and rotated log file (out/verbose.log.gz, or out/dist/logs/verbose.log.gz) contains the full verbose (showcommands) log of every execution run by the build. Since this is now written on every build, the showcommands argument is now ignored -- if you want to get the commands run, look at the log file after the build. Test: m Test: <built-in tests> Test: NINJA_ARGS="-t list" m Test: check the build.trace.gz Test: check the new log files Change-Id: If1d8994890d43ef68f65aa10ddd8e6e06dc7013a
2018-05-18 01:37:09 +02:00
cmd.RunAndPrintOrFatal()
}
// checkRAM warns if there probably isn't enough RAM to complete a build.
func checkRAM(ctx Context, config Config) {
if totalRAM := config.TotalRAM(); totalRAM != 0 {
ram := float32(totalRAM) / (1024 * 1024 * 1024)
ctx.Verbosef("Total RAM: %.3vGB", ram)
if ram <= 16 {
ctx.Println("************************************************************")
ctx.Printf("You are building on a machine with %.3vGB of RAM\n", ram)
ctx.Println("")
ctx.Println("The minimum required amount of free memory is around 16GB,")
ctx.Println("and even with that, some configurations may not work.")
ctx.Println("")
ctx.Println("If you run into segfaults or other errors, try reducing your")
ctx.Println("-j value.")
ctx.Println("************************************************************")
} else if ram <= float32(config.Parallel()) {
// Want at least 1GB of RAM per job.
ctx.Printf("Warning: high -j%d count compared to %.3vGB of RAM", config.Parallel(), ram)
ctx.Println("If you run into segfaults or other errors, try a lower -j value")
}
}
}
// Build the tree. The 'what' argument can be used to chose which components of
// the build to run, via checking various bitmasks.
func Build(ctx Context, config Config, what int) {
ctx.Verboseln("Starting build with args:", config.Arguments())
ctx.Verboseln("Environment:", config.Environment().Environ())
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
ctx.BeginTrace(metrics.Total, "total")
defer ctx.EndTrace()
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
if inList("help", config.Arguments()) {
help(ctx, config)
return
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
}
// Make sure that no other Soong process is running with the same output directory
buildLock := BecomeSingletonOrFail(ctx, config)
defer buildLock.Unlock()
if inList("clean", config.Arguments()) || inList("clobber", config.Arguments()) {
clean(ctx, config)
return
}
// checkProblematicFiles aborts the build if Android.mk or CleanSpec.mk are found at the root of the tree.
checkProblematicFiles(ctx)
checkRAM(ctx, config)
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
SetupOutDir(ctx, config)
// checkCaseSensitivity issues a warning if a case-insensitive file system is being used.
checkCaseSensitivity(ctx, config)
ensureEmptyDirectoriesExist(ctx, config.TempDir())
SetupPath(ctx, config)
if config.SkipConfig() {
ctx.Verboseln("Skipping Config as requested")
what = what &^ BuildProductConfig
}
if config.SkipKati() {
ctx.Verboseln("Skipping Kati as requested")
what = what &^ BuildKati
}
if config.StartGoma() {
// Ensure start Goma compiler_proxy
startGoma(ctx, config)
}
if config.StartRBE() {
// Ensure RBE proxy is started
startRBE(ctx, config)
}
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
if what&BuildProductConfig != 0 {
// Run make for product config
runMakeProductConfig(ctx, config)
}
// Everything below here depends on product config.
if inList("installclean", config.Arguments()) ||
inList("install-clean", config.Arguments()) {
installClean(ctx, config)
ctx.Println("Deleted images and staging directories.")
return
}
if inList("dataclean", config.Arguments()) ||
inList("data-clean", config.Arguments()) {
dataClean(ctx, config)
ctx.Println("Deleted data files.")
return
}
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
if what&BuildSoong != 0 {
// Run Soong
runSoong(ctx, config)
if config.Environment().IsEnvTrue("GENERATE_BAZEL_FILES") {
// Return early, if we're using Soong as the bp2build converter.
return
}
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
}
if what&BuildKati != 0 {
// Run ckati
genKatiSuffix(ctx, config)
runKatiCleanSpec(ctx, config)
runKatiBuild(ctx, config)
Add a Kati-based packaging step The idea is that we'd move the installation and packaging tasks over to it, using data from Soong & the Kati reading Android.mk files. This would allow us to make more fundamental changes about how we package things without having to adjust makefiles throughout the tree. Possible use cases: * Moving some information from Soong's Android.mk output to a file read by the packaging step may allow us to read the Android.mk files less often, speeding up builds. * Refactoring our current two-stage ASAN builds to run the Kati build step twice, writing into different object directories, then have a single packaging step that reads both outputs. Soong already has the capability of writing out a single ninja file with all the asan combinations. * Running two build steps, one building the system-related modules using a "generic" device configuration, and one building the vendor modules using a specific device configuration. This could enforce a GSI/mainline system vs vendor split in a single build invocation. * If all installation is through this tool, it will be much easier to track what should no longer be installed on an incremental build, reducing the need for installclean. * Changing PRODUCT_PACKAGES should be a much faster operation, which means we could keep track of local additions to the images. Then `mma` would be more persistent, instead of installing something once, then never updating it again. Eventually we plan on switching from Kati to something Go-based, but this is a more incremental approach while we clean up everything else. Currently, this just moves the dist-for-goal handling over to the packaging step, so that we don't need to read Android.mk files when DIST_DIR changes, or we switch between dist vs not. Bug: 116968624 Bug: 117463001 Test: m nothing Change-Id: Idec5ac6f7c7475397ba0fb65bd3785128a7517df
2018-09-27 00:00:42 +02:00
runKatiPackage(ctx, config)
ioutil.WriteFile(config.LastKatiSuffixFile(), []byte(config.KatiSuffix()), 0666) // a+rw
} else {
// Load last Kati Suffix if it exists
if katiSuffix, err := ioutil.ReadFile(config.LastKatiSuffixFile()); err == nil {
ctx.Verboseln("Loaded previous kati config:", string(katiSuffix))
config.SetKatiSuffix(string(katiSuffix))
}
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
}
// Write combined ninja file
createCombinedBuildNinjaFile(ctx, config)
distGzipFile(ctx, config, config.CombinedNinjaFile())
if what&RunBuildTests != 0 {
testForDanglingRules(ctx, config)
}
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
if what&BuildNinja != 0 {
if what&BuildKati != 0 {
installCleanIfNecessary(ctx, config)
}
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
// Run ninja
runNinja(ctx, config)
}
// Currently, using Bazel requires Kati and Soong to run first, so check whether to run Bazel last.
if what&BuildBazel != 0 {
runBazel(ctx, config)
}
Add a Go replacement for our top-level Make wrapper Right now this mostly just copies what Make is doing in build/core/ninja.mk and build/core/soong.mk. The only major feature it adds is a rotating log file with some verbose logging. There is one major functional difference -- you cannot override random Make variables during the Make phase anymore. The environment variable is set, and if Make uses ?= or the equivalent, it can still use those variables. We already made this change for Kati, which also loads all of the same code and actually does the build, so it has been half-removed for a while. The only "UI" this implements is what I'll call "Make Emulation" mode -- it's expected that current command lines will continue working, and we'll explore alternate user interfaces later. We're still using Make as a wrapper, but all it does is call into this single Go program, it won't even load the product configuration. Once this is default, we can start moving individual users over to using this directly (still in Make emulation mode), skipping the Make wrapper. Ideas for the future: * Generating trace files showing time spent in Make/Kati/Soong/Ninja (also importing ninja traces into the same stream). I had this working in a previous version of this patch, but removed it to keep the size down and focus on the current features. * More intelligent SIGALRM handling, once we fully remove the Make wrapper (which hides the SIGALRM) * Reading the experimental binary output stream from Ninja, so that we can always save the verbose log even if we're not printing it out to the console Test: USE_SOONG_UI=true m -j blueprint_tools Change-Id: I884327b9a8ae24499eb6c56f6e1ad26df1cfa4e4
2016-08-22 00:17:17 +02:00
}
// distGzipFile writes a compressed copy of src to the distDir if dist is enabled. Failures
// are printed but non-fatal.
func distGzipFile(ctx Context, config Config, src string, subDirs ...string) {
if !config.Dist() {
return
}
subDir := filepath.Join(subDirs...)
destDir := filepath.Join(config.RealDistDir(), "soong_ui", subDir)
if err := os.MkdirAll(destDir, 0777); err != nil { // a+rwx
ctx.Printf("failed to mkdir %s: %s", destDir, err.Error())
}
if err := gzipFileToDir(src, destDir); err != nil {
ctx.Printf("failed to dist %s: %s", filepath.Base(src), err.Error())
}
}
// distFile writes a copy of src to the distDir if dist is enabled. Failures are printed but
// non-fatal.
func distFile(ctx Context, config Config, src string, subDirs ...string) {
if !config.Dist() {
return
}
subDir := filepath.Join(subDirs...)
destDir := filepath.Join(config.RealDistDir(), "soong_ui", subDir)
if err := os.MkdirAll(destDir, 0777); err != nil { // a+rwx
ctx.Printf("failed to mkdir %s: %s", destDir, err.Error())
}
if _, err := copyFile(src, filepath.Join(destDir, filepath.Base(src))); err != nil {
ctx.Printf("failed to dist %s: %s", filepath.Base(src), err.Error())
}
}