aboutsummaryrefslogtreecommitdiff
path: root/stage1
Commit message (Collapse)AuthorAgeFilesLines
* add package to generated Build in build.scala and in-package discoveryChristopher Vogt2017-02-151-23/+31
| | | | | | | | | | | This should allow for build to add other builds to their dependencies and interact with them in a type-safe way. And ever regardless it seems like good practice to never have the same class existing in the same package or the top-level package even if they don’t end up on the same classpath. This might also help make stack traces easier to understand. Also improve error messages for mistakes with the build class, e.g. constructor, super classes, etc.
* now really cache classloaders for buildsChristopher Vogt2017-02-141-7/+3
| | | | should be safe now that we rebuild based on lastModified
* more debugging info for a bug observed with compiling dottyChristopher Vogt2017-02-141-1/+5
|
* ignore classes in randomly nested subdirectoriesChristopher Vogt2017-02-141-8/+13
| | | | useful for `dotty run <file>`
* support flatClassLoader and run at Dependency levelChristopher Vogt2017-02-141-0/+13
|
* implicitly pass classloader, might make code easierChristopher Vogt2017-02-145-59/+62
| | | | | and prepares for allowing `run` and `runFlat` at Dependency instead of Build level
* easier setting of projectDirectory in sub-buildsChristopher Vogt2017-02-132-3/+5
| | | | | by replacing context.projectDirectory by workingDirectory and using it as the default but allowing it to being overridden
* remove some unnecessary Context fields while we are modifying contextChristopher Vogt2017-02-133-5/+8
|
* Add support for nested builds and use it to replace cross compilationChristopher Vogt2017-02-131-1/+0
| | | | | | | | command which was previously hard-coded This will allow multi-project builds, too but we should first fix caching across instances and GitDependencies on sub-builds within other repositories.
* Merge pull request #337 from cvogt/minor-dotty-plugin-improvementsJan Christopher Vogt2017-02-121-19/+16
|\ | | | | Minor dotty plugin improvements
| * omit classpath when classpath is empty (= no dependencies, only jdk)Christopher Vogt2017-02-111-19/+16
| |
* | Urlencode ownerTrevor SIbanda2017-02-121-1/+1
|/
* upgrade zinc to 0.3.13Christopher Vogt2017-02-101-1/+1
|
* let Stage1 handle Stage2 exit codeChristopher Vogt2017-02-091-1/+1
|
* various smaller changesChristopher Vogt2017-02-093-5/+13
|
* idempotent change propagationChristopher Vogt2017-02-098-195/+183
| | | | | | using lastModified instead of a non-idempotent needsUpdate flag this fixes a bug where dependees would not be rebuilt if cbt exited or was killed after dependencies were already rebuilt.
* memoize task results across classes within a single runChristopher Vogt2017-02-098-90/+114
|
* make full dependencies available to compile instead of only classpathChristopher Vogt2017-02-013-4/+5
| | | | | this will make it possible to access lastModified times and cache them in the following commits
* replace flawed concurrent hashmap cache with consistent replacementChristopher Vogt2017-02-018-16/+9
| | | | | | | | | | | | | | | | | | | | | | | The concurrent hashmap approach to classloader caching was flawed. Assume you have two concurrently running builds A and B and projects P2 and P3 depending on project P1. And assume a time sequence where A compiles P1, then compiles P2, then P1’s sources change, then B compiles P1, then A compiles P3. At the end P2 and P3 will have different versions of P1 as their parent classloaders. This is inconsistent. The easiest way to work around this is making sure only one thread is changing the classloader cache during it’s entire run. This would mean either no concurrency or what we have done here, which is letting threads work on a copy of the cache and replace the original cache in the end using an atomic operation. This means the thread that finishes last wins, but for caching that’s fine. Worst case some things aren’t cached in a concurrent execution. This change also means that we don’t need concurrent hashmaps for the classloader cache anymore since no two theads will access the same hashmap. We still need a concurrent hashmap for the class caches inside of the classloaders as multiple threads can access the same classloaders.
* Merge separate hashmaps for persistent cache into oneChristopher Vogt2017-02-017-73/+57
| | | | | | | This isn’t type-safe, but re-using that same hashmap for both keys and classloaders allows to reduce the number of members in Context. Also we can re-use the same hashMap for other things as well in the coming commits, e.g. timestamps.
* Remove classifier from pom files.mchav2016-11-251-12/+11
|
* Merge pull request #313 from cvogt/chris3Jan Christopher Vogt2016-11-131-2/+1
|\ | | | | minor cleanups regarding Context
| * minor cleanups regarding ContextChristopher Vogt2016-11-131-2/+1
| |
* | small refactoring simplifying Stage2Christopher Vogt2016-11-131-2/+7
|/
* Assert right build class. Fixes #174Christopher Vogt2016-11-131-0/+1
|
* comment regarding trigger filesChristopher Vogt2016-11-101-0/+3
|
* more sanity checks for cbt's classloadersChristopher Vogt2016-11-101-12/+27
|
* typesafe ===Christopher Vogt2016-11-101-0/+5
|
* fix java.lang.LinkageError due to cache invalidation bugChristopher Vogt2016-11-091-3/+10
| | | | | | | | | | | | cbtHasChanged does never change across one run. However classLoaderRecursion removes the corresponding classloader if needsUpdate is true. It is called multiple times meaning it removes and re-creates the classloader on every call leading to potentially multiple conflicting cbt classloaders. needsUpdate should never be returning true more than ones across a single run. Otherwise it can lead to this error: java.lang.LinkageError: loader constraint violation in interface itable initialization: when resolving method "Build.lib()Lcbt/Lib;" the class loader (instance of cbt/URLClassLoader) of the current class, Build, and the class loader (instance of cbt/URLClassLoader) for interface cbt/BaseBuild have different Class objects for the type cbt/Lib used in the signature
* Remove version from context, we can now use dynamic overridesChristopher Vogt2016-11-092-4/+0
|
* Upgrade zinc to 0.3.12Christopher Vogt2016-11-092-2/+2
|
* Add support for dynamic re-configuration.Christopher Vogt2016-11-081-0/+2
| | | | | The exact precedence rule of override code vs original code may still need to be tweaked as we go along.
* add context based task result cacheChristopher Vogt2016-11-082-0/+2
|
* filter out 'direct' meta task earlier from the argsChristopher Vogt2016-11-081-1/+1
|
* Fix bug in resource loading of MultiClassLoaderChristopher Vogt2016-11-071-1/+1
|
* allow blacklisting loggersChristopher Vogt2016-11-031-1/+10
|
* fix toString of cbt.URLClassLoaderChristopher Vogt2016-11-031-1/+1
|
* Add support for resources (still limited by CBT's classloader visibility)Christopher Vogt2016-11-031-0/+14
|
* web uitim-zh2016-10-281-0/+1
|
* swap out System.out and System.err in a way that affects JDK and ScalaChristopher Vogt2016-10-131-6/+29
| | | | | | | | Before it only affected jdk, because scala.Console captures our and err before the swap. This is needed when running main classes like Scaladoc or the compiler and wanting to redirect output to standard error
* do not interpret exceptions during redirectOutToErr as scalac crashesChristopher Vogt2016-10-131-22/+22
|
* Fix exit code trapping across classloadersChristopher Vogt2016-10-071-4/+4
| | | | | | | | | | A SecurityManager is once installed globally and stays the same across all classloaders. TrapSecurityManager was installed this way, but it looked up the `trapExitCode` in it's own classloader, while classes in other classloaders (when cbt was using another version of cbt) were writing it to their own classloader. This flag needs to be in a global place instead to fix this, so we'll put it straight into the TrapSecurityManager itself.
* make BuildInterfaces an abstract class so it can have default valuesChristopher Vogt2016-10-031-2/+2
| | | | | and it becomes slightly easier to maintain source compatibility between cbt versions
* fix pickOne error messagerockjam2016-10-031-1/+1
|
* clean up main class discovery mechanism and offer interactive choiceChristopher Vogt2016-10-021-36/+54
| | | | if multiple main classes are found
* Added run with package discoveryChavXO2016-09-291-0/+37
|
* Use correct main method for Dottydoc - the java interfaceChristopher Vogt2016-09-281-5/+11
| | | | and work around the fact that the main method is not static (huh?)
* use lower case instead of upper case for hashes to be consistent with SBTChristopher Vogt2016-09-191-3/+3
|
* fix checksum check for downloaded jars by ignoring caseChristopher Vogt2016-09-151-3/+3
|
* fix behavior of System.exit trappingChristopher Vogt2016-09-151-2/+4
| | | | | | | | | Installing one globally for the JVM live-time and make behavior dependent on a thread local variable seems safer than globally switching it out and having race conditions. Also now all other calls are forwarded to a potential Nailgun SecurityManager, which should fix some bugs.
generated by cgit v1.2.3 (git 2.39.1) at 2024-08-12 00:39:31 +0000