Repository navigation
(See description!) Pull request for repo [ipc] branch [ipc-143_perf-dive-major] to base-branch [main]. - #165
Merged
Merged
Conversation
… deallocation-path overhaul; many new unit tests; assorted fixes; CI pipeline improvements (mirror of internal feature branch).
…-major (perf/stats/tests/fixes work described in their respective commits).
…/build trees before saving.
…ipc_transport_structured (build fixes).
…0 construct_at et al).
…h-only helper marked maybe_unused (transport_test).
…ort_structured, ipc_shm_arena_lend).
…ures inside jemalloc source, disable UBSAN for jemalloc builds. (Other sanitizer types cannot be handled this way by their nature.)
…ly gcc x debug-oriented) when building our code (not in the generated code); some new multi-SHM-provider tests, combined with existing such tests, can lead to out-of-memory in the build step.
…smaller translation units. Comment and/or doc changes.
…_shm, ipc_shm_arena_lend).
… sanitizer flags must be overridden per-package; the previous setting-based approach affected only the package ID, not the flags.
… the parallelism-reducing profile line landed in the wrong profile section (silently ignored); and in UBSAN config evict possibly-stale cached jemalloc, so it is rebuilt with the intended sanitizer-less flags.
…aries: on a non-exact cache restore, compare the profile flag sections against a snapshot recorded in the cache; evict on mismatch. (Conan package-id ignores flags arriving via conf/buildenv, so it cannot detect this itself.)
…) detection: its hard 64-held-mutexes capacity is exceeded by jemalloc, upon which its failure path hangs the process indefinitely. (Data-race detection is unaffected.)
…a sibling llvm-symbolizer version (17's crashes on our binaries, stalling tests); accordingly re-enable transport_test SHM-jemalloc sub-mode in that config. Also: temporary test-subset input, letting 2 parallel runs split one cell's test load.
…ersion) detector on generally -- disabling it only for the jemalloc-heavy unit-test batch, where it cannot function (its 64-held-mutexes cap + hang-on-internal-assertion-failure); bound the whole job at 120 min so any such hang fails loudly.
…Doxygen excludes and unit_test CMake updated for the removed arena_lend::Memory_manager.
…e (ipc-143_perf-dive-major), ipc_session (ipc-143_perf-dive-major), ipc_shm (ipc-143_perf-dive-major), ipc_shm_arena_lend (ipc-143_perf-dive-major).
…_arena_lend (ipc-143_perf-dive-major).
…_arena_lend (ipc-143_perf-dive-major).
…e (ipc-143_perf-dive-major), ipc_shm_arena_lend (ipc-143_perf-dive-major).
…pc-143_perf-dive-major), ipc_core (ipc-143_perf-dive-major), ipc_shm (ipc-143_perf-dive-major), ipc_shm_arena_lend (ipc-143_perf-dive-major).
…e (ipc-143_perf-dive-major).
…verview; multi-split universes chapter; full review (fixes incl. logic errors in example code; Unix domain sockets as the default transport). Comment and/or doc changes.
…pc-143_perf-dive-major), ipc_core (ipc-143_perf-dive-major), ipc_transport_structured (ipc-143_perf-dive-major), ipc_session (ipc-143_perf-dive-major), ipc_shm (ipc-143_perf-dive-major), ipc_shm_arena_lend (ipc-143_perf-dive-major).
…e runners => slow thread wake at times). Comment and/or doc changes.
…pc-143_perf-dive-major), ipc_core (ipc-143_perf-dive-major), ipc_transport_structured (ipc-143_perf-dive-major), ipc_shm (ipc-143_perf-dive-major), ipc_shm_arena_lend (ipc-143_perf-dive-major).
…e (ipc-143_perf-dive-major).
…nsport_structured (ipc-143_perf-dive-major).
…pc-143_perf-dive-major).
…_arena_lend (ipc-143_perf-dive-major).
…pc-143_perf-dive-major).
…ain), ipc_shm_arena_lend (ipc-143_perf-dive-major).
…e (ipc-143_perf-dive-major), ipc_transport_structured (ipc-143_perf-dive-major), ipc_session (ipc-143_perf-dive-major), ipc_shm (ipc-143_perf-dive-major), ipc_shm_arena_lend (ipc-143_perf-dive-major).
…nsport_structured (ipc-143_perf-dive-major), ipc_shm_arena_lend (ipc-143_perf-dive-major).
…e (main), ipc_transport_structured (main), ipc_session (main), ipc_shm (main), ipc_shm_arena_lend (main).
Contributor
Author
|
This also resolves #89 actually. |
Contributor
Author
|
Also it resolves #90 actually. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
fixes #143
fixes #164
fixes #121
fixes #145
fixes #83
fixes Flow-IPC/ipc_core#40
fixes Flow-IPC/ipc_core#41
fixes Flow-IPC/ipc_transport_structured#34
fixes Flow-IPC/ipc_session#38
fixes Flow-IPC/ipc_shm#34
related to Flow-IPC/flow#115
Final step of the Flow-IPC 3.0.0 line of work: Pulls in the 5 submodule PRs and Flow v3.0.0, and adds the 3.0.0 CHANGELOG. For the substance see
CHANGELOG.md(Highlights at the top) and the submodule PRs.flow: Release v3.0.0.ipc_core: (See description!) Pull request for repo [ipc_core] branch [ipc-143_perf-dive-major] to base-branch [main]. ipc_core#44ipc_transport_structured: (See description!) Pull request for repo [ipc_transport_structured] branch [ipc-143_perf-dive-major] to base-branch [main]. ipc_transport_structured#37ipc_session: (See description!) Pull request for repo [ipc_session] branch [ipc-143_perf-dive-major] to base-branch [main]. ipc_session#37ipc_shm: (See description!) Pull request for repo [ipc_shm] branch [ipc-143_perf-dive-major] to base-branch [main]. ipc_shm#33ipc_shm_arena_lend: (See description!) Pull request for repo [ipc_shm_arena_lend] branch [ipc-143_perf-dive-major] to base-branch [main]. ipc_shm_arena_lend#72conanfile.py).ipc_shm_arena_lend) is now scanned, so it appears in the generated Reference. A few files remain excluded (see explicit list inCMakeLists.txt).conanfile.py).conan/jemalloc: Vendored jemalloc 5.4.0 recipe, used by the "early-warning" pipeline jobs only.struc::Channelsend APIs takingMsg_out*.rpc_demo(capnp-RPC).transport_test: Expand scripted mode, including batch-receive and graceful-close scenarios; exercise-mode bug fix (tasks could run after theirApp_sessionwas destroyed); loosen certain timeouts.perf_demo: Bug fix (exit-time crash); correct serverAppexecutable path, as the new identity checks require.unit_test: Register the many new tests from the submodules; split heavy translation units, to reduce build RAM use; TSAN suppressions.rpc_demointegration-test steps.llvm-symbolizerfor clang-17 TSAN; no UBSAN for jemalloc itself.To code reviewer
This has been reviewed elsewhere, including with: 🤖 Claude Code