Age | Commit message (Collapse) | Author | |
---|---|---|---|
2022-08-19 | main: honor configuration for target_file name | Klaus Aehlig | |
... also when determining default module or target. | |||
2022-08-09 | RemoteExecutionClient: print debug string | Alberto Sartori | |
2022-08-05 | dump-targets: don't dump source trees | Klaus Aehlig | |
We have to include explicit tree references into our target-result map, as these targets can provide a tree definition, if the root is not a git root or compatible mode is used. Nevertheless, from a user's persepective, trees are just source references, like files. Therefore, do not include them, when dumping the map of analyzed targets. In this way, we also avoid the ambiguity in the dump between an explicit tree reference and a defined target with the same name. | |||
2022-08-05 | Artifact reporting: repeat the failed artifacts | Klaus Aehlig | |
In targets that produce a large list of artifacts, e.g., a simple test suite realized by "install", it is sometimes hard to find the failed artifacts after being made aware of them by the warning. Therefore, at info level, repeat a list of the failed artifacts close to the end of the output. | |||
2022-08-05 | Move analysis out of main.cpp | Klaus Aehlig | |
Analysing a target is a pretty self-contained task. So move it into a library of its own, to keep main.cpp more managable. | |||
2022-08-05 | Make ANSI escape sequences optional | Klaus Aehlig | |
2022-08-05 | Version: Bump up to 1.0.0~beta1 | Oliver Reiche | |
2022-08-05 | Describe: Move describe code to separate library | Oliver Reiche | |
2022-08-05 | InstallCas: Print warning for empty size in object-id | Oliver Reiche | |
2022-08-05 | InstallCas: Moved install-cas code to separate library | Oliver Reiche | |
2022-08-05 | BazelNetwork: Use bytestream for reading unknown size blobs | Oliver Reiche | |
... otherwise actual blob size might exceed the maximum transfer size of the CAS client. Therefore, we always have to use the bytestream client if the size is unknown. | |||
2022-08-05 | Parse git tree objects instead of protobuf messages as action result in ↵ | Sascha Roloff | |
native mode | |||
2022-08-05 | Add more tracing messages to remote execution client | Sascha Roloff | |
2022-08-05 | Git tree based root digest creation | Sascha Roloff | |
2022-08-05 | BazelResponse: Support collecting output dirs in native mode | Oliver Reiche | |
2022-08-05 | LocalExecution: Use Git tree format | Oliver Reiche | |
- LocalStorage Add tree CAS and support reading Git trees - LocalAction: Create Git tree for output directory - LocalApi: Support availability and upload of Git trees - LocalStorage: Support dumping tree to stream in native mode | |||
2022-08-05 | LocalCAS: Support storing trees | Oliver Reiche | |
2022-08-05 | CLI: Add flag for dumping raw tree objects | Oliver Reiche | |
2022-08-05 | BazelMsgFactory: Support dumping Git tree to string | Oliver Reiche | |
2022-08-05 | BazelMsgFactory: Implement reading object infos from Git tree | Oliver Reiche | |
2022-08-05 | BazelMsgFactory: Implement creating Git tree from fs path | Oliver Reiche | |
2022-08-05 | ObjectInfo: Support stringify with unknown size | Oliver Reiche | |
2022-08-05 | Upload known git trees if not available in CAS | Sascha Roloff | |
2022-08-05 | Allow to read git trees from global CAS as well as local workspace root | Sascha Roloff | |
2022-08-05 | Git tree can return its raw string data | Sascha Roloff | |
2022-08-05 | Introduced batch availability check for remote execution api | Sascha Roloff | |
2022-08-05 | Target map makes use of known trees and avoids directory traversal | Sascha Roloff | |
2022-08-05 | Directory entries can detect themselves as known trees and return according ↵ | Sascha Roloff | |
description | |||
2022-08-05 | Git tree can return size | Sascha Roloff | |
2022-08-05 | Modified artifact digest to provide wire digest on demand | Sascha Roloff | |
2022-08-05 | Introduced class with helper functions to support native remote execution ↵ | Sascha Roloff | |
protocol | |||
2022-08-05 | GitCAS: Support reading/creating trees without filesystem IO | Oliver Reiche | |
2022-08-05 | GitCAS: Add create tree via libgit2's treebuilder | Oliver Reiche | |
2022-08-05 | Version: Bump up to 0.1.1+devel | Oliver Reiche | |
2022-08-05 | Version: Bump up to 0.1.1 | Oliver Reiche | |
2022-08-05 | GitCAS: Implement reading git tree via libgit2 | Oliver Reiche | |
2022-08-05 | ExecutionResponse: Fix artifacts reserve size | Oliver Reiche | |
2022-08-02 | Utils: Drop ill-formed audit checks for truncating JSON | Oliver Reiche | |
2022-07-27 | describe: Support serializing to JSON | Oliver Reiche | |
2022-07-27 | describe: Support describing by rule name | Oliver Reiche | |
2022-07-27 | install-cas now accepts --log-limit and --log-file | Alberto Sartori | |
2022-07-08 | In install-cas be more liberal in parsing artifact identifiers | Klaus Aehlig | |
2022-07-07 | Traverser: Bring task system down gracefully on error | Oliver Reiche | |
2022-07-07 | TaskSystem: Implement shutdown | Oliver Reiche | |
2022-07-06 | TaskSystem: Support wait for finish | Oliver Reiche | |
2022-07-06 | TaskSystem: Fix early shutdown | Oliver Reiche | |
... conceptually, it was possible that a previous task decrements the `num_threads_running_` counter before it is incremented by the next task. Therefore, we have to unify the queue and thread status in a single counter (`total_workload_`) and ensure that woken threads increment it before decrementing it for popping a queue. | |||
2022-07-06 | When requesting action input, also provide action information | Klaus Aehlig | |
Extend the --request-action-input option to also provide command line, environment, etc, of the requested action. As this option is mainly for debugging, having that information available via the analyse subcommand is useful; for build/install that change does not make a difference anyway. | |||
2022-07-05 | Make remote endpoint specification part of analysis | Klaus Aehlig | |
The remote execution endpoint shards the target-level cache; still, for analysis, we want to specify a particular target-level cache, e.g., to analyse the inputs of a particular action that failed remotely. Note that the action identifier depends on the target-level cache in question, due to the extensional projection implicit in target-level caching. | |||
2022-07-05 | Remove option --persistent | Klaus Aehlig | |
... as the --request-action-input option provides a clean way of restoring the inputs to a specific action, without the need of interacting with the internal directory structure of the tool. In particular, the new option also works for remote builds. | |||
2022-07-05 | Implement staging of action inputs | Klaus Aehlig | |
Add an option to change the requested result to be (as artifacts) the input stage of a specified action. In this way, the inputs to individual actions can conveniently be inspected, e.g., for local debugging of why an action failed. |