- 12 Feb, 2022 3 commits
-
-
Yedidya Feldblum authored
Summary: When the Function signature is `noexcept`, rather than using `throw_exception`, in order to save a bit of generated code size. Reviewed By: luciang Differential Revision: D34190570 fbshipit-source-id: c0f9cbd6abffd9b1517d28126d8eb5c491089ea8
-
Yedidya Feldblum authored
Summary: TSAN detects and reports leaked threads in the child process, and exits nonzero. The parent process checks for zero exit code. Permit the TSAN exit code as well under TSAN. Differential Revision: D34191672 fbshipit-source-id: 359fe91e617d962af2705896c392001bdd458d7b
-
Yedidya Feldblum authored
Summary: By using a clever trick which "absorbs" attempts to force types for what should be subject to sfinae. Reviewed By: luciang Differential Revision: D34160080 fbshipit-source-id: a370c0240088cc78b8760e9f5736c3224842ae04
-
- 11 Feb, 2022 4 commits
-
-
Alex Hornby authored
Summary: Pull Request resolved: https://github.com/facebookexperimental/eden/pull/109 Bring back getdeps build of Mononoke so it can be built in opensource. Includes cargo generation and marking a couple of comments as text to stop cargo trying to build them as rust doc tests Successful GitHub Linux Build: https://github.com/facebookexperimental/eden/runs/5147129282?check_suite_focus=true Successful GitHub Mac Build: https://github.com/facebookexperimental/eden/runs/5147129271?check_suite_focus=true Reviewed By: markbt Differential Revision: D34114848 fbshipit-source-id: a18f4370048054b087d421deced29f0b7ba7ada1
-
Yedidya Feldblum authored
Reviewed By: Orvid Differential Revision: D34031459 fbshipit-source-id: 13e11ff462c5a975851970043d7e673d2905e237
-
Oguz Ulgen authored
Reviewed By: yfeldblum Differential Revision: D34154912 fbshipit-source-id: cb3528c4ab853c2b2dc82871f4ac1307d614f388
-
Adam Norton authored
Summary: Allow conversions to types containing a `dynamic`. Reviewed By: yfeldblum Differential Revision: D33823657 fbshipit-source-id: b09537639efad185114dfd5953c1707207ddfcdf
-
- 10 Feb, 2022 7 commits
-
-
Denis Dydychkin authored
Summary: ConcurrentHashMapTest.ConcurrentInsertClear is reported as flaky, see T111753832, T111745080, etc. Temporarily disabling the test while investigating the root cause. Preliminary observation is that folly::CPUThreadPoolExecutor/(anonymous namespace)::HazptrTPETag doesn't shutdown cleanly. Happens both to SIMD and regular version Reviewed By: magedm Differential Revision: D34147319 fbshipit-source-id: 8a9d9c71c4a42544df93fc48a324bd1299395719
-
Ajanthan Asogamoorthy authored
Summary: As title says, in this case implement just means call into openssl's implementation Reviewed By: mingtaoy Differential Revision: D33828545 fbshipit-source-id: 638adba4dd87dbccbc3886a512c52b5b8d44b5da
-
Ajanthan Asogamoorthy authored
Summary: To generically get the exported keying material from our transports. As the code comments say this is intended to be used to bind data to the stream/connection itself. Reviewed By: yfeldblum, knekritz Differential Revision: D33828649 fbshipit-source-id: c046eccfad5fc34188a310409ceb6c84d42e9f79
-
Chad Austin authored
Summary: Pull Request resolved: https://github.com/facebook/watchman/pull/998 A detailed description of this change is included in the new comment. Reviewed By: fanzeyi Differential Revision: D34124030 fbshipit-source-id: f37c458bae135357e16b256e09f0d6eb8f1f996d
-
Chad Austin authored
Summary: Pull Request resolved: https://github.com/facebook/folly/pull/1714 windows-2016 will be removed in March, and now that we've updated Boost, we can upgrade to windows-2019. Reviewed By: genevievehelsel Differential Revision: D34119171 fbshipit-source-id: caad432bebf6b8282844bf79f82b1fbc34f2df9e
-
Denis Dydychkin authored
Summary: ConcurrentHashMap's clear() implementation had a race condition where bucket_count_ / chunk_count_ might be changed after they are read, resulting in smaller than required replacement buffer being created. This change addresses race condition and adds test to reproduce the issue. Reviewed By: magedm Differential Revision: D34022761 fbshipit-source-id: 1c8e8c382adf3839cbb4031e85314a02d963a762
-
Yedidya Feldblum authored
Reviewed By: ot, luciang Differential Revision: D32974082 fbshipit-source-id: b5bc35a56ef7600790938a73af586154b1631903
-
- 09 Feb, 2022 10 commits
-
-
Alex Hornby authored
Summary: Pull Request resolved: https://github.com/facebookexperimental/rust-shed/pull/26 Pull Request resolved: https://github.com/facebookexperimental/eden/pull/104 No need for the checked in generated code anymore Successful external CI Linux run on Git Hub PR: https://github.com/facebookexperimental/eden/runs/5130405984?check_suite_focus=true Mac PR run fails with a SSL cert issue, will look separately Reviewed By: chadaustin, mitrandir77 Differential Revision: D33840545 fbshipit-source-id: eafc2a0e2191d438fd828adeebc2e318d319025f
-
Joseph Hwang authored
Summary: This change should not make a functional difference to any existing callers. The underlying data structure stores a TimePoint. Reviewed By: yfeldblum Differential Revision: D34060591 fbshipit-source-id: b2b4d7f0d7fed8220cffb6b1dba5082d21073686
-
Dead Code Bot authored
Reviewed By: pranavtbhat Differential Revision: D34103812 fbshipit-source-id: e02b728d468bf091e8b4521f8f8746a986a2130d
-
Kurt Tomlinson authored
Summary: Define at least one virtual method in a cpp file. Differential Revision: D33891879 fbshipit-source-id: 3d2e3177b112c5fc7b2b84ed8bc41c5d0e15127e
-
Philip Pronin authored
Summary: `EDFThreadPoolExecutor` is currently the only implementation of `SoftRealTimeExecutor`, and it implements multi-add functionality (which allows us to wake up idle threads just once on batch-add, as well as reduces planning overhead), let's expose this at an interface level. Reviewed By: ot, luciang Differential Revision: D34095644 fbshipit-source-id: ba59300a4c51599d7c6cbfaaabe3d6013be7c37f
-
Chad Austin authored
Summary: Watchman advertises support for FreeBSD but it's regressed a couple ways. Ensure it builds on FreeBSD again with a Vagrantfile. Reviewed By: ahornby Differential Revision: D33989289 fbshipit-source-id: ff906bc219974bafbc349740ce6bddfe5bd2908f
-
Chad Austin authored
Summary: Add some basic support for FreeBSD to getdeps. Reviewed By: ahornby Differential Revision: D33989129 fbshipit-source-id: 42ff5f160b7e19c12196bb2e52a726f7815487bd
-
Chad Austin authored
Summary: `unsigned long` and `size_t` are not the same type on Clang Windows. Reviewed By: yfeldblum Differential Revision: D34086667 fbshipit-source-id: adba5d960cb4c8b8e13b62f419d5b4179b1525e1
-
Xavier Deguillard authored
Summary: This is an old copy of Mercurial, saved to keep the Mercurial servers running until we completely pull the plug on them. Neither EdenFS nor EdenSCM should depend on it. Reviewed By: mitrandir77 Differential Revision: D34081890 fbshipit-source-id: 78e2c07fae75b3f25aa492ece03308f736d42769
-
Yedidya Feldblum authored
Summary: In a way which also happens to improve build speed. It turns out that the bulk of the improvement comes from removing the use of `std::is_nothrow_move_constructible` since it may internally have a `std::declval` in there. But that's not the only source of improvements. ```name=trunk $ foundation/scripts/run-perf-compile --syntax folly/futures/test/FutureTest.cpp 8,367,373,932 instructions:uP ``` ```name=branch $ foundation/scripts/run-perf-compile --syntax folly/futures/test/FutureTest.cpp 8,101,850,897 instructions:uP ``` Reviewed By: luciang Differential Revision: D32973629 fbshipit-source-id: 37c27763a4f622750719db11fee7cb23c0b0291e
-
- 08 Feb, 2022 10 commits
-
-
Chad Austin authored
Summary: libevent 2.1.8 is not aware of the arc4random deprecation on FreeBSD, so upgrade to the latest. Reviewed By: xavierd Differential Revision: D33984680 fbshipit-source-id: b7d2b869bc47d49d8f29df1b318d8f2067204aae
-
Dan Melnic authored
Summary: Adding AsyncSSLSocket::performReadv Reviewed By: afrind Differential Revision: D33717297 fbshipit-source-id: 413f50fcaf1d59cec98765a095a5a5964f634e23
-
Chad Austin authored
Summary: The new boost has an MSVC detection algorithm that 1) I don't understand and 2) does not even seem deterministic. Since our internal builds are pinned to vc142, force that in the manifest. Reviewed By: fanzeyi Differential Revision: D34085081 fbshipit-source-id: d8ce4d8e28aad14f6fa60f9227dafb546ce321e5
-
Alex Hornby authored
Summary: Pull Request resolved: https://github.com/facebookexperimental/eden/pull/101 public_autocargo already existed for edenfs, but needed to move up a level so to make edenfs thrift definition visible via autocargo. Reviewed By: markbt Differential Revision: D32693522 fbshipit-source-id: 4c070cfa2d6294a28c17697338f0309dcab1576f
-
Alex Hornby authored
Summary: Pull Request resolved: https://github.com/facebookexperimental/eden/pull/103 Automate maintenance of the edenscm_* github actions yamls Add job file and name options and support for the Rust install section Reviewed By: fanzeyi Differential Revision: D34044422 fbshipit-source-id: 7d5f07d37bab1eff5de30a88e710dbf7479ca192
-
Alex Hornby authored
Summary: Fix the getdeps.py generation sub command and regenerate github actions before making changes to support actions generation for eden_scm Reviewed By: fanzeyi Differential Revision: D34044243 fbshipit-source-id: 0039d04f25af4c842145dc142dae6b9996fc8046
-
Yedidya Feldblum authored
Summary: For global variables with nontrivial destructors, the compiler inserts calls to `atexit` to destroy the variables at shutdown time. For global variables with trivial destructors, the compiler elides such calls. Revise `Indestructible` to have a trivial destructor. Some compilers optimize further and remove such calls even for some nontrivial destructors, but not all do so. Also remove move-construction and move-assignment. And remove actual `constexpr`-constructibility since `::new` is not a constant expression. Reviewed By: chadaustin Differential Revision: D34026842 fbshipit-source-id: 9ab2f957cbe9a67809a6c99e6b9370c1f4315777
-
Yedidya Feldblum authored
Reviewed By: ot, luciang Differential Revision: D33864764 fbshipit-source-id: 9839f578bd0fe5e8f69d5299287a59cb471f8cf3
-
Yedidya Feldblum authored
Summary: Whether a value is passed in registers also depends on its size, not just whether it is trivially copyable. (It also depends on its position in the arg-list but that is another step in complexity.) Reviewed By: luciang Differential Revision: D33079529 fbshipit-source-id: ae7383d61491001e54892cf077e45556801bdcf4
-
Yedidya Feldblum authored
Differential Revision: D34056336 fbshipit-source-id: adf4c0b96d6fd3f5e15c7bfefe7e908e85baf5b1
-
- 07 Feb, 2022 2 commits
-
-
Yedidya Feldblum authored
Summary: It turns out that instantiating this function declaration template so widely measurably affects build speed. Prefer alternative techniques, such as `FOLLY_DECLVAL`. ```name=trunk $ foundation/scripts/run-perf-compile --syntax folly/futures/test/FutureTest.cpp 8,462,001,355 instructions:uP ``` ```name=branch $ foundation/scripts/run-perf-compile --syntax folly/futures/test/FutureTest.cpp 8,362,569,934 instructions:uP ``` Reviewed By: Gownta, luciang Differential Revision: D32973365 fbshipit-source-id: 6930f75e04e0bb9ac4216b08645e67b98ca2c2f3
-
Yedidya Feldblum authored
Summary: Coalesce the inline-asm. Use invokers. Reviewed By: Gownta Differential Revision: D32302511 fbshipit-source-id: 6edf75c6ef820ed80e751d965174e5acfaac957e
-
- 06 Feb, 2022 4 commits
-
-
Zino Benaissa authored
Summary: Define `heap_vector_map` and `heap_vector_set`. These feature a fast lookup (find operation) which does not rely on `std::lower_bound` or similar to find keys. Map elements are not laid out in sorted order. Instead, they are laid out in heap order, which is also known as Eytzinger order. A heap order layout is optimal to maximize memory and cache locality of lookup operations. A heap vector set containing the elements 0-7 lays out the elements as: elements in sorted order heap_container[0] = 4 <-- middle heap_container[1] = 2 heap_container[2] = 6 heap_container[3] = 1 heap_container[4] = 3 heap_container[5] = 5 heap_container[6] = 7 <-- max heap_container[7] = 0 <-- min `heap_vector_map` (referred below as HM) is almost a drop-in replacements for `sorted_vector_map` (referred as SM). All SM APIs are supported, and are semantically equal, with very few exceptions. Although HM works correctly for any key value SM supports, the speedup of find operation (random key search) is typically limited to maps with small key types with comparison operations subject to inlining. Measurements suggest up to 2x speedup of find operation for HM compared to SM. The key tradeoff is that HM has slower insertion and removal operations and slower iteration. Prefer HM if updates and traversals operations are rare, and lookups are the dominant operation. Of course, SM itself has slow insertion and removal operations so, if these operations are sufficiently common, another map type entirely would be preferable. Key similarities: 1) They have the same in-situ sizes: `sizeof(HM) == sizeof(SM)`. 2) Insertions and removals. As expected, these operations are slower for HM. Both SM and HM need to move elements while preserving sorted or heap order but the SM operation can be much simpler. 3) A fast construction of HM from a sorted vector or from a SM. 4) An iterator that follows sorted order compatible with SM. This is complex, though. When sorted iteration order is not required, it is faster to use the enclosed container's iteration order. Key differences: 1) iterate() is a new API that provides the iterator range of the container vector. iterate() enables the fastest traversal of the heap container elements. For example, for (auto& element : HS.iterator()) { // traversed as laid out in memory std::cout << e << ", "; } for above examples, the loop prints: 4, 2, 6, 1, 3, 5, 7, 0, for (auto& element : HS) { // key-sorted std::cout << e << ", "; } for above examples, the loop prints: 0, 1, 2, 3, 4, 5, 6, 7, 2) data() is purposely not provided because it does not point to the first elements. If the start address is needed, use instead HM.iterate().data() Reviewed By: Gownta Differential Revision: D32128733 fbshipit-source-id: 1df7372720b969ee7a84004ded101db132e0c224
-
Yedidya Feldblum authored
Summary: It turns out that `std::decay` becomes expensive when as widely instantiated as it is. ```name=trunk $ foundation/scripts/run-perf-compile --syntax folly/futures/test/FutureTest.cpp 6,336,421,499 instructions:uP ``` ```name=branch $ foundation/scripts/run-perf-compile --syntax folly/futures/test/FutureTest.cpp 6,287,269,480 instructions:uP ``` Reviewed By: Gownta, luciang Differential Revision: D32973204 fbshipit-source-id: 75d52fd5ced58f50518913d3c00709bcee27c87b
-
Yedidya Feldblum authored
Reviewed By: ahornby Differential Revision: D34030118 fbshipit-source-id: 2229527eebb147a9c2eea90be78ab72e63e2fe3f
-
Alex Hornby authored
Summary: Folly has an optional libsodium dependency. Declare that in folly getdeps manifest so that the relevant code is built and tested in CI. Reviewed By: yfeldblum Differential Revision: D34028478 fbshipit-source-id: 5e7c2f86602b94a3a975907adf8bd04f49ec33c0
-