5647ec6 | Nathan Daly | 11 November 2022, 23:49:51 UTC | Resolve method ambiguity logfield_contains(a::NamedTuple, b::Ignored) = true # method amibguity resolution | 13 November 2022, 19:16:15 UTC |
0a1393f | Nathan Daly | 10 November 2022, 20:58:10 UTC | Add ability to test_logs against kwargs. Currently, you cannot use `@test_logs` to test a log message's keyword arguments. This extends `@test_logs` to let you do this. Additionally, while I was here, I also changed `@test_logs` itself to let you specify _patterns_ using keyword args / NamedTuples, so that you can more easily specify parts of the LogRecord that come at the end (like the `kwargs`). Here is an example of a log message with kwargs: ```julia julia> @info "hi" x=2 ┌ Info: hi └ x = 2 ``` Which produces this LogRecord: ```julia LogRecord(Info, "hi", Main, Symbol("REPL[5]"), :Main_c95d07ae, "REPL[5]", 2, Base.Pairs(:x => 2)) ``` After this PR, you can test for this kind of log message by matching the `:kwargs` field to `@test_logs`: ```julia @test_logs (:info, "hi", Test.Ignored(), Test.Ignored(), Test.Ignored(), Test.Ignored(), Test.Ignored(), (;y=3)) @info("hi", x=2) ``` But obviously the above is way verbose, so as I described in the second paragraph, this PR also adds the following shorter syntax: ```julia @test_logs (level=:info, message="hi", kwargs=(;x=2)) @info("hi", x=2) ``` Also updated the docstring. | 13 November 2022, 19:16:15 UTC |
7650173 | Lilith Orion Hafner | 13 November 2022, 18:09:58 UTC | Update CONTRIBUTING.md (#47547) | 13 November 2022, 18:09:58 UTC |
f6f6e8f | Lilith Orion Hafner | 13 November 2022, 18:09:14 UTC | Switch Coveralls to Codecov in CONTRIBUTING.md (#47549) | 13 November 2022, 18:09:14 UTC |
3510eeb | Lilith Orion Hafner | 11 November 2022, 16:47:35 UTC | Don't show redundant typeinfo when printing LinRange (take 2) (#47521) | 11 November 2022, 16:47:35 UTC |
317211a | Andy Nowacki | 10 November 2022, 22:08:06 UTC | Dates: Error on construction/parsing with empty strings (#47117) When attempting to construct a `DateTime`, `Date` or `Time` from an `AbstractString`, throw an `ArgumentError` if the string is empty. Likewise, error when `parse`ing an empty string as one of these types, and return `nothing` from `tryparse`. This behavior differs from previously. Before, `Date` and `Time` would return default values of `Date(1)` and `Time(0)`, respectively, while `DateTime` would error without a `format` argument. With a `format` argument, it would return `DateTime(1)`. However, this appears not to have been explicitly intended, but rather a consequence of the way parsing was implemented; no tests for empty string parsing existed. This addresses #28090 and #43883; see discussion therein. Summary of changes: - Check for empty string in `Base.parse(::DateTime)` and throw if so. - Change documentation to mention this. - Add a compat notice to the docs contrasting the old behavior. | 10 November 2022, 22:08:06 UTC |
e0ba28a | Jakob Nybo Nissen | 10 November 2022, 18:37:38 UTC | Improve help for broadcasted && and ||, closes #47526. (#47527) Currently, the help text for `.&&` and `.||` says: `x .&& y` is akin to `broadcast(&&, x, y)`. However, that is invalid syntax. | 10 November 2022, 18:37:38 UTC |
c81456e | Dilum Aluthge | 10 November 2022, 18:29:26 UTC | Fix the whitespace check (#47533) | 10 November 2022, 18:29:26 UTC |
0af14f5 | Will Kimmerer | 10 November 2022, 17:58:55 UTC | Sparse NEWS.md Update (#47278) | 10 November 2022, 17:58:55 UTC |
3394bc5 | Fons van der Plas | 10 November 2022, 10:27:28 UTC | Fix escaping in `Base.show(::IO, ::DateFormat)`, fixes #45032. (#45259) | 10 November 2022, 10:27:28 UTC |
0a65af8 | Shuhei Kadowaki | 10 November 2022, 06:48:50 UTC | `AbstractInterpreter`: set up `ForwardableArgtypes` interface (#47505) This interface is necessary for us to implement IPO-able customized external lattice implementation. Additionally `InferenceResult(::MethodInstance, ::SimpleArgtypes)` constructor may be useful for debugging const-prop' inference or implement something like #29261. | 10 November 2022, 06:48:50 UTC |
1b2d37b | Dilum Aluthge | 10 November 2022, 04:07:12 UTC | Revert "Don't show redundant typeinfo when printing LinRange (#47509)" (#47516) | 10 November 2022, 04:07:12 UTC |
bedd14d | Shuhei Kadowaki | 10 November 2022, 02:27:18 UTC | follow up #47137, delete `setfield!` call to `must_be_codeinf` (#47508) We can come back to when exactly we need to turn this option on once we enable this option for Base. | 10 November 2022, 02:27:18 UTC |
dd62fac | Simon Byrne | 09 November 2022, 23:10:59 UTC | pass exitcode to atexit hooks (#47498) | 09 November 2022, 23:10:59 UTC |
83592cf | Oscar Smith | 09 November 2022, 16:39:13 UTC | faster `inv` for normal sized `ComplexF64` (#47255) * faster `inv` for normal sized `ComplexF64` | 09 November 2022, 16:39:13 UTC |
86bb1fc | Daniel Karrasch | 09 November 2022, 15:31:36 UTC | Fix erroneous generlization from #47107 (#47510) | 09 November 2022, 15:31:36 UTC |
afdc589 | Lilith Orion Hafner | 09 November 2022, 14:44:58 UTC | Don't show redundant typeinfo when printing LinRange (#47509) * Don't show redundant typeinfo when printing LinRange Co-authored-by: Lilith Hafner <Lilith.Hafner@gmail.com> | 09 November 2022, 14:44:58 UTC |
4e58d88 | woclass | 09 November 2022, 04:52:08 UTC | CONTRIBUTING.md: no submodules anymore (#47506) | 09 November 2022, 04:52:08 UTC |
473b698 | Shuhei Kadowaki | 09 November 2022, 02:48:54 UTC | optimizer: inline effect-free `:static_parameter` (#47490) JuliaLang/julia#45459 moved `:static_parameter` always to statement position as our optimizer assumes `:static_parameter` in value position effect-free. But it turns out that it can cause precision issue for semi-concrete interpretation as discovered at #47349, since the type of `:static_parameter` in statement position is widened when converted to compressed IR for cache. This commit follows up JuliaLang/julia#45459 so that we inline effect-free `:static_parameter` during IR conversion and get a more reasonable semi-concrete interpretation. | 09 November 2022, 02:48:54 UTC |
4ff526a | Shuhei Kadowaki | 08 November 2022, 22:51:33 UTC | optimizer: minor tweak on `insert_node!(::IncrementalCompact)` (#47491) So that it can take a new instruction without flag already computed. | 08 November 2022, 22:51:33 UTC |
bfaaad3 | Lilith Orion Hafner | 08 November 2022, 22:45:52 UTC | Warn folks away from metaprogramming (#47469) Co-authored-by: Martin Smit <57063134+jacobusmmsmit@users.noreply.github.com> Co-authored-by: Viral B. Shah <ViralBShah@users.noreply.github.com> | 08 November 2022, 22:45:52 UTC |
eb708d6 | apaz | 08 November 2022, 22:39:29 UTC | Probe and dlopen() the correct libstdc++ (#46976) * Probe if system libstdc++ is newer than ours If the system libstdc++ is detected to be newer, load it. Otherwise, load the one that we ship. This improves compatibility with external shared libraries that the user might have on their system. Fixes #34276 Co-authored-by: Jameson Nash <vtjnash@gmail.com> Co-authored-by: Elliot Saba <staticfloat@gmail.com> * Addressed review comments. * Change error handling in wrapper functions Co-authored-by: Jameson Nash <vtjnash@gmail.com> * Call write_wrapper three times instead of snprintf Co-authored-by: Jameson Nash <vtjnash@gmail.com> * Apply suggestions from code review Co-authored-by: Jameson Nash <vtjnash@gmail.com> * Update cli/loader_lib.c Co-authored-by: Jameson Nash <vtjnash@gmail.com> * Reordered reading and waiting to avoid a deadlock. * Fixed obvious issues. * Only load libstdc++ preemptively on linux. * Update cli/loader_lib.c Co-authored-by: Jameson Nash <vtjnash@gmail.com> * Update cli/loader_lib.c Co-authored-by: Jameson Nash <vtjnash@gmail.com> * Specified path to bundled libstdc++ on the command line. * Removed whitespace. * Update cli/Makefile Co-authored-by: Jameson Nash <vtjnash@gmail.com> * Handled make install stringreplace. * Correctly quoted stringreplace. * Added -Wl,--enable-new-dtags to prevent DT_RPATH for transitive dependencies * Updated news entry. * Added comment about environment variable. * patched rpath for libgfortran and libLLVM. * Added explaination to Make.inc * Removed trailing space * Removed patchelf for libgfortran, now that BB has been fixed. * Fixed typos and comments Co-authored-by: Max Horn <max@quendi.de> Co-authored-by: Mosè Giordano <mose@gnu.org> Co-authored-by: Jameson Nash <vtjnash@gmail.com> Co-authored-by: Elliot Saba <staticfloat@gmail.com> Co-authored-by: Max Horn <max@quendi.de> | 08 November 2022, 22:39:29 UTC |
f7d4edc | Lilith Orion Hafner | 08 November 2022, 19:54:51 UTC | Improve backwards compatibility in sorting (#47489) * Improve backwards compatibility in sorting Co-authored-by: Lilith Hafner <Lilith.Hafner@gmail.com> | 08 November 2022, 19:54:51 UTC |
cf9cac2 | Sudhansu | 08 November 2022, 17:09:30 UTC | Change the error messages in relpath function. (#47450) | 08 November 2022, 17:09:30 UTC |
5a323a6 | Éric Thiébaut | 08 November 2022, 11:25:57 UTC | Fix example for @atomicreplace (#47457) | 08 November 2022, 11:25:57 UTC |
d05fd69 | Daniel Karrasch | 08 November 2022, 09:46:53 UTC | Relax `StridedArray` signatures wherever possible (#47107) | 08 November 2022, 09:46:53 UTC |
26cb6d5 | Shuhei Kadowaki | 08 November 2022, 08:46:16 UTC | tweak test cases added in #47379 (#47487) - general inference test should go in `compiler/inference`, while `compiler/AbstractInterpreter` should keep cases that test `AbstractInterpreter` inference specifically - `only(Base.return_types(...))` is simpler than `code_typed(...)[1][1]` to implement a return type based test case. | 08 November 2022, 08:46:16 UTC |
dad08f2 | apaz | 08 November 2022, 07:25:50 UTC | Fast path constants in update_julia_type only if correct type (#47480) * Fast path constants in update_julia_type only if correct type Fixes #47247 Co-authored-by: Jameson Nash <vtjnash@gmail.com> | 08 November 2022, 07:25:50 UTC |
92cf557 | woclass | 08 November 2022, 07:23:34 UTC | devdoc: fix markdown title link (#47486) | 08 November 2022, 07:23:34 UTC |
bc39fd1 | apaz | 08 November 2022, 05:18:57 UTC | Removed attributes from arguments to gc_preserve_begin (#47482) LLVM adds the nonnull attribute on its own, which makes the verifier fail. Fixes #47245 | 08 November 2022, 05:18:57 UTC |
59fab46 | Jameson Nash | 08 November 2022, 05:18:31 UTC | signals: when exiting because of a signal, call `raise` instead of `exit` (#47459) * Be careful now to avoid `jl_exit` on foreign threads, which would try to adopt them now * Call `raise` (with enhancements as `jl_raise`) to exit due to handling a signal, which sets WTERMSIG instead of emulating it with WEXITSTATUS, and triggers the SIG_DFL behaviors (which may include a coredump, for SIGQUIT / `Ctrl-\`). * Use pthread_sigmask in preference to sigprocmask, since it is better specified, though probably usually identical. * Emulate all of that on Windows, to a small extent. | 08 November 2022, 05:18:31 UTC |
553b0b9 | Jacob Quinn | 08 November 2022, 04:55:01 UTC | Allow specifying waitfirst when waiting on a Condition (#47277) * Allow specifying waitfirst when waiting on a Condition I have a use-case where I use a `Condition` as a wait queue, where a singler waiter is woken up on `notify`. Once a waiter is woken up, it may need to re-wait, but we want it to get back in the wait queue _first_ instead of the FIFO behavior of regular `wait`. This allows the waiter to continue being notified until a logical condition is met and the next waiter in the queue will take the next notification. This PR proposes adding a keyword argument `waitfirst::Bool=false` to the `wait(c::Condition)` method that allows the caller to put itself _first_ in the `Condition` wait queue instead of last. I didn't consider whether other `wait` methods (Channel, Task, etc.) would also benefit from something like this to minimize the total changes to a specific use-case that I know is useful. * change waitfirst -> first as suggested by jameson | 08 November 2022, 04:55:01 UTC |
523c52f | Gabriel Baraldi | 08 November 2022, 00:14:37 UTC | Add some more details to the snapshot (#47359) | 08 November 2022, 00:14:37 UTC |
1fc4010 | Dilum Aluthge | 08 November 2022, 00:08:20 UTC | Fix the whitespace check (#47485) | 08 November 2022, 00:08:20 UTC |
9182326 | woclass | 07 November 2022, 23:07:08 UTC | Base: add new help function `isdebugbuild` (#47475) * Base: add new help function `isdebugbuild` * isdebugbuild: replace ccall with new fucntion * isdebugbuild: fix typo in docstring | 07 November 2022, 23:07:08 UTC |
4937986 | Kristoffer Carlsson | 07 November 2022, 20:14:27 UTC | update TOML from upstream library (#47448) * update TOML from upstream library | 07 November 2022, 20:14:27 UTC |
f9d15dc | Oscar Smith | 07 November 2022, 18:22:57 UTC | Faster rem_pio2 kernel for `Float32` (#47212) * faster rem_pio2 for Float32 | 07 November 2022, 18:22:57 UTC |
a41ae5b | Kristoffer Carlsson | 07 November 2022, 10:16:55 UTC | put back legacybindings for `libblas` and `liblapack` (#47477) | 07 November 2022, 10:16:55 UTC |
6354e2c | DilumAluthgeBot | 07 November 2022, 01:50:54 UTC | 🤖 Bump the SparseArrays stdlib from 3c2b65f to 311b4b4 (#47465) Co-authored-by: Dilum Aluthge <dilum@aluthge.com> | 07 November 2022, 01:50:54 UTC |
a611822 | Viral B. Shah | 06 November 2022, 22:31:44 UTC | Update gnome url (#47471) | 06 November 2022, 22:31:44 UTC |
1fc80e2 | spaette | 06 November 2022, 22:16:35 UTC | ui improvements (#47390) Co-authored-by: Viral B. Shah <ViralBShah@users.noreply.github.com> | 06 November 2022, 22:16:35 UTC |
01cc785 | DilumAluthgeBot | 06 November 2022, 21:51:44 UTC | 🤖 Bump the Pkg stdlib from b11ca0acd to ed6a5497e (#47470) | 06 November 2022, 21:51:44 UTC |
8af2e65 | Katharine Hyatt | 06 November 2022, 21:23:46 UTC | Remove unused internal n_waiters function (#30089) Co-authored-by: Jameson Nash <vtjnash@gmail.com> | 06 November 2022, 21:23:46 UTC |
4e2d13f | Erik Schnetter | 06 November 2022, 04:45:03 UTC | Allow long-form `--debug-info` option (#47257) | 06 November 2022, 04:45:03 UTC |
b0822b8 | Ian Butterworth | 06 November 2022, 04:24:56 UTC | Profile: save a heap snapshot after a profile peek via SIGUSR1/SIGINFO (via opt-in) (#47134) optionally save a heap snapshot after a profile peek via SIGUSR1/SIGINFO | 06 November 2022, 04:24:56 UTC |
ebc97ba | Dilum Aluthge | 06 November 2022, 01:54:53 UTC | Delete `.github/workflows/rerun_failed.yml` (#47460) | 06 November 2022, 01:54:53 UTC |
a768d0c | Lilith Orion Hafner | 06 November 2022, 00:11:44 UTC | rename buffer to scratch in sorting (#47172) Co-authored-by: Lilith Hafner <Lilith.Hafner@gmail.com> Co-authored-by: Kristoffer Carlsson <kcarlsson89@gmail.com> | 06 November 2022, 00:11:44 UTC |
27859f3 | Kristoffer Carlsson | 05 November 2022, 06:04:33 UTC | remove special case for homogeneous tuples in `all` and `any` (#47454) * remove special case for homogeneous tuples in `all` and `any` * Use loop-based `any/all` for homogenous tuples Unroll this loop at julia level wont gain any inference improvement, thus let LLVM unroll it if needed. Co-authored-by: N5N3 <2642243996@qq.com> | 05 November 2022, 06:04:33 UTC |
c093f92 | Jameson Nash | 05 November 2022, 04:45:23 UTC | signals: try harder to validate SEGV causes (#47234) Before turning a signal into a Julia exception, we would like to be fairly certain it is not: - caused by the GC (and going to hang the system waiting to finish) - occurring during GC (and going to corrupt the mark bits) - occurring on a "foreign" thread (and lacking an exception handler and/or ptls object) - actually an alignment fault - send by the user with `kill` - the page actually is non-existant The `msync` code calls can be directly checked instead by making sure the `code[0]` is not KERN_INVALID_ADDRESS, which answers the same question for us. Equivalent to how the check is done on other unix: `sig == SIGSEGV && info->si_code == SEGV_ACCERR` | 05 November 2022, 04:45:23 UTC |
b1c67ea | Steven G. Johnson | 04 November 2022, 15:11:09 UTC | more documentation for assignment vs mutation (#47434) Co-authored-by: Kristoffer Carlsson <kcarlsson89@gmail.com> Co-authored-by: Alex Arslan <ararslan@comcast.net> | 04 November 2022, 15:11:09 UTC |
4053f69 | Lilith Orion Hafner | 04 November 2022, 12:05:17 UTC | Make tiny function definition inline (style) (#47446) Follows up #46764 | 04 November 2022, 12:05:17 UTC |
6fc3d2c | Johnny Chen | 04 November 2022, 10:10:04 UTC | add `reinterpret` example for lazy array containers (#42006) Co-authored-by: Kristoffer <kcarlsson89@gmail.com> | 04 November 2022, 10:10:04 UTC |
dfa127f | Zachary P Christensen | 04 November 2022, 09:46:54 UTC | Effects for symbol comparisons (Fix #47424) (#47425) Co-authored-by: Shuhei Kadowaki <40514306+aviatesk@users.noreply.github.com> | 04 November 2022, 09:46:54 UTC |
bf92e83 | Nikitas Rontsis | 04 November 2022, 09:40:10 UTC | CircleCI caching: allow ftime to be ceil(ftime_req) in Base.stale_cachefile (#47433) * CircleCI caching: allow ftime to be ceil(ftime_req) in Base.stale_cachefile It appears that [caching functionalities](https://circleci.com/docs/caching/) provided by CircleCi, a leading CI/CD provider, can truncate timestamps to full seconds, resulting in re-compilations as below: ``` Rejecting stale cache file /root/.julia/compiled/v1.8/ComponentArrays/cYHSD_3rQji.ji (mtime 1.6673960929277816e9) because file /root/.julia/packages/ComponentArrays/YyD7i/src/ComponentArrays.jl ``` This PR relaxes the `is_stale` check to be robust against rounding-to-second timestamp mutations. I can provide a minimal CircleCI configuration file to reproduce if this is helpful. | 04 November 2022, 09:40:10 UTC |
6321e47 | Mosè Giordano | 04 November 2022, 09:37:51 UTC | [CompilerSupportLibraries_jll] Update to new build (#47444) This fixes the dependency of `libgcc_s.1.dylib` (which is a light compatibility shim) on `x86_64-apple-darwin-libgfortran5`, to make the library properly loadable. | 04 November 2022, 09:37:51 UTC |
3b5c058 | Lilith Orion Hafner | 04 November 2022, 02:51:55 UTC | remove unnecessary inline annotations in ryu (#47445) | 04 November 2022, 02:51:55 UTC |
242cdc5 | Nick Robinson | 04 November 2022, 02:00:14 UTC | Add `@allocations` macro for getting number of allocations (#47367) * Add `@allocs` macro for getting number of allocations * Rename to `@allocations` * Add to docs * Add news for `@allocations` | 04 November 2022, 02:00:14 UTC |
fadcbef | Shuhei Kadowaki | 04 November 2022, 01:16:54 UTC | use `Base.return_types` instead of `Core.Compiler.return_type` (#47439) | 04 November 2022, 01:16:54 UTC |
d0ba259 | Kristoffer Carlsson | 03 November 2022, 16:22:54 UTC | small update to the docs about ENV (#47440) | 03 November 2022, 16:22:54 UTC |
b382456 | Jacob Quinn | 03 November 2022, 14:20:51 UTC | Fix typo/boolean usage in stream eof comments (#46504) Co-authored-by: Jameson Nash <vtjnash@gmail.com> | 03 November 2022, 14:20:51 UTC |
d9bcc5f | Shuhei Kadowaki | 03 November 2022, 13:41:46 UTC | fix up `Conditional`-forwarding for vararg methods (#47438) | 03 November 2022, 13:41:46 UTC |
9f572b6 | Michael Abbott | 03 November 2022, 13:14:22 UTC | Ignore ANSI colour codes in printing Diagonal, etc (#47430) | 03 November 2022, 13:14:22 UTC |
65c6bda | Kristoffer Carlsson | 03 November 2022, 12:46:50 UTC | add an example of using `sizeof` on a struct with padding (#47342) | 03 November 2022, 12:46:50 UTC |
b0c2ad9 | Udoh Jeremiah | 03 November 2022, 11:03:20 UTC | in operator with missing on Dicts keys and values (#46695) Co-authored-by: Kristoffer Carlsson <kcarlsson89@gmail.com> | 03 November 2022, 11:03:20 UTC |
4a1301a | Udoh Jeremiah | 03 November 2022, 10:57:49 UTC | Order of operation in a composed function (#46696) Co-authored-by: Kristoffer Carlsson <kcarlsson89@gmail.com> | 03 November 2022, 10:57:49 UTC |
b8a77da | Sasank | 03 November 2022, 10:47:53 UTC | Fix isdone for empty product iterators, fixes #43921 (#43947) * Fix the issue #43921 * add a test Co-authored-by: Kristoffer <kcarlsson89@gmail.com> | 03 November 2022, 10:47:53 UTC |
bfb0098 | woclass | 03 November 2022, 10:47:31 UTC | [docs/base] add a type tree for `Number` (#45975) * [docs/base] add a type tree https://github.com/JuliaLang/julia/issues/24741 Generated using the following code with manual modifications ```jl using AbstractTrees AbstractTrees.children(x::Type) = subtypes(x) print_tree(Number) ``` * [docs/base] mark "Abstract Type" * [docs/base] add a blurb for the tree * Update numbers.md Co-authored-by: Kristoffer Carlsson <kcarlsson89@gmail.com> | 03 November 2022, 10:47:31 UTC |
33731d9 | apaz | 03 November 2022, 10:46:06 UTC | Handle incorrect call type in abstract_invoke (#47379) Abstract interpreter was returning Any where calling the method would result in a MethodError, implying that the result should have been that the code was unreachable. Fixes #46839 | 03 November 2022, 10:46:06 UTC |
10fcec9 | Mosè Giordano | 03 November 2022, 09:36:39 UTC | [CompilerSupportLibraries_jll] Upgrade to libraries from GCC 12 (#45582) * [CompilerSupportLibraries_jll] Upgrade to libraries from GCC 12 * [CompilerSupportLibraries_jll] Fix soname of libgcc_s on x86_64-darwin * [CompilerSupportLibraries_jll] Update to v0.6.1 to get latest libgcc_s on x86_64-darwin * Add libstdc++ to list of preloaded libraries * Try to build on Windows with Buildkite * Debug printing * Revert "Try to build on Windows with Buildkite" This reverts commit f973004ec8566a2c0a864bb572440b419b82e394. * More debugging * Revert "More debugging" This reverts commit 8d05d3289d709a7ac0dc5d363cb5258c61efb58d. * Revert "Debug printing" This reverts commit 4c14b581a88eac131d2d4d23ca40b9c7a0d1fb9a. * Revert "Add libstdc++ to list of preloaded libraries" This reverts commit c2c182db043e97bd19247de4eda0501f35c7853b. * [CompilerSupportLibraries_jll] Use new build with fix for Windows libstdc++ | 03 November 2022, 09:36:39 UTC |
0ff2373 | Jishnu Bhattacharya | 03 November 2022, 00:36:25 UTC | Avoid overflow in first/last (continue #45843) (#47186) * Avoid overflow in first/last * remove inbounds annotation * use checked_length | 03 November 2022, 00:36:25 UTC |
0ee0888 | Lilith Orion Hafner | 02 November 2022, 20:46:18 UTC | Improve documentation of unary `:` again (#47009) * Improve documentation of unary : again Co-authored-by: Hendrik Ranocha <ranocha@users.noreply.github.com> Co-authored-by: Kristoffer Carlsson <kcarlsson89@gmail.com> | 02 November 2022, 20:46:18 UTC |
2220eb5 | Udoh Jeremiah | 02 November 2022, 18:27:28 UTC | add info on colon operator (#46892) Co-authored-by: Michael Abbott <32575566+mcabbott@users.noreply.github.com> Co-authored-by: Kristoffer Carlsson <kcarlsson89@gmail.com> | 02 November 2022, 18:27:28 UTC |
e6baa6b | Udoh Jeremiah | 02 November 2022, 18:26:05 UTC | Remove complexity and ambiguity from function example (#47110) * Remove complexity and ambiguity from function example Co-authored-by: Kristoffer Carlsson <kcarlsson89@gmail.com> | 02 November 2022, 18:26:05 UTC |
b9da8a9 | Valentin Churavy | 02 November 2022, 13:21:59 UTC | Merge pull request #47419 from JuliaLang/jn/cgmemmgr-init-mac cgmemmgr,macos: add missing shared_map_lock initialization | 02 November 2022, 13:21:59 UTC |
5d6d830 | Lilith Orion Hafner | 02 November 2022, 13:17:51 UTC | Stop incorrectly documenting the default sorting algorithms (#47303) * Stop incorrectly documenting the default sorting algorithms Co-authored-by: Petr Vana <petvana@centrum.cz> | 02 November 2022, 13:17:51 UTC |
67fedc8 | Max Horn | 02 November 2022, 11:13:27 UTC | Fix some typos found by codespell (#47423) Co-authored-by: Kristoffer Carlsson <kcarlsson89@gmail.com> | 02 November 2022, 11:13:27 UTC |
e304ad8 | Mosè Giordano | 02 November 2022, 11:10:27 UTC | [deps] Do not build CLI of libgit2 (#47422) * [deps] Do not build CLI of libgit2 The build system of this component is broken upstream, we don't need it, let's just don't build it. * [LibGit2_jll] Update build which doesn't include the CLI | 02 November 2022, 11:10:27 UTC |
e9ca677 | Steven G. Johnson | 02 November 2022, 10:40:37 UTC | many types should define broadcastable(o) = Ref(o) (#44093) Co-authored-by: Fredrik Ekre <ekrefredrik@gmail.com> | 02 November 2022, 10:40:37 UTC |
df06375 | Daniel Karrasch | 02 November 2022, 09:12:42 UTC | Make `norm` handle `missing`s (#40790) | 02 November 2022, 09:12:42 UTC |
22fe28b | Daniel Karrasch | 02 November 2022, 09:04:50 UTC | Fix LAPACK test: don't call `sygvd!` on `undef`ed matrix (#47397) | 02 November 2022, 09:04:50 UTC |
02fe132 | Lilith Orion Hafner | 02 November 2022, 08:44:44 UTC | Fix error and formatting in subnormal docstring (#47035) * Fix error and formatting in subnormal docstring Fixup for #46899 | 02 November 2022, 08:44:44 UTC |
8a9589d | Elliot Saba | 02 November 2022, 08:42:49 UTC | Initialize memory to satisfy `valgrind` (#47132) When debugging Julia with `valgrind`, the uninitialized value of `ch` used within an `if` statement causes many warnings; let's just initialize it. | 02 November 2022, 08:42:49 UTC |
9e1dac0 | caipengxiang | 02 November 2022, 08:41:26 UTC | move src and dst to next buffer position (#47229) Co-authored-by: Gabriel Baraldi <baraldigabriel@gmail.com> | 02 November 2022, 08:41:26 UTC |
de5559f | Adam B | 02 November 2022, 08:36:35 UTC | Simplify definition of ntuple (#34978) Since macro @ntuple already exists in Base.Cartesian, using it directly simplifies the definition of ntuple instead of an ad-hoc recreation here. Co-authored-by: Kristoffer Carlsson <kcarlsson89@gmail.com> | 02 November 2022, 08:36:35 UTC |
d553fed | Lilith Orion Hafner | 02 November 2022, 08:33:15 UTC | Add QuickSort's stability to news (#47330) * Add QuickSort's stability to news * Mention stability by default (thanks @petvana) | 02 November 2022, 08:33:15 UTC |
c9362a5 | apaz | 02 November 2022, 08:31:19 UTC | Improve error message for non-distinct static param names (#47368) For code generated by macros, it may be otherwise difficult to tell which parameters are not distinct. Fixes #46622 | 02 November 2022, 08:31:19 UTC |
40e2ab6 | jariji | 02 November 2022, 08:30:46 UTC | Document the behavior of zip() (#47370) `zip(xss...)` terminates when its shortest argument does -- the initial case is infinite. https://github.com/JuliaLang/julia/issues/43821#issuecomment-1013816939 | 02 November 2022, 08:30:46 UTC |
5378a31 | Steven G. Johnson | 02 November 2022, 08:28:13 UTC | support Unicode 15 via utf8proc 2.8 (#47392) * support Unicode 15 via utf8proc 2.8 * link PR in NEWS | 02 November 2022, 08:28:13 UTC |
c694139 | Lilith Orion Hafner | 02 November 2022, 08:27:49 UTC | don't warn about skipping tests that were not skipped (#47395) Co-authored-by: Lilith Hafner <Lilith.Hafner@gmail.com> | 02 November 2022, 08:27:49 UTC |
92e6d9c | Páll Haraldsson | 02 November 2022, 08:21:40 UTC | Bugfix; do not allocate more than needed (#47411) | 02 November 2022, 08:21:40 UTC |
59b52df | Dilum Aluthge | 02 November 2022, 03:29:24 UTC | Delete the `.github/workflows/statuses.yml` (#47426) | 02 November 2022, 03:29:24 UTC |
7de3033 | DilumAluthgeBot | 02 November 2022, 03:27:17 UTC | 🤖 Bump the NetworkOptions stdlib from 8ce1e10 to 791fa05 (#47412) Co-authored-by: Dilum Aluthge <dilum@aluthge.com> | 02 November 2022, 03:27:17 UTC |
55d4e45 | apaz | 02 November 2022, 00:35:29 UTC | Detect bad arguments to code_native (#47264) The types for `code_native()` must be given as a tuple of types. Added checking to `signature_type()` with `to_tuple_type()` to validate this. Fixes #45414 | 02 November 2022, 00:35:29 UTC |
e31aec0 | Udoh Jeremiah | 01 November 2022, 20:48:50 UTC | add hvncat to the cat see also section (#47398) | 01 November 2022, 20:48:50 UTC |
3661a08 | Jameson Nash | 01 November 2022, 18:31:11 UTC | cgmemmgr,macos: add missing shared_map_lock initialization | 01 November 2022, 18:31:11 UTC |
de7b6d5 | Kiran | 01 November 2022, 16:45:22 UTC | Merge pull request #47402 from JuliaLang/kp/task-stacktraces Unrevert 46845 (Add `jl_print_task_backtraces()`) | 01 November 2022, 16:45:22 UTC |
abce8af | Jameson Nash | 01 November 2022, 14:49:51 UTC | hide kwcall methods from the backtrace (#47373) They are not strictly internal, but also not strictly external, so just assume they should be hidden. Perhaps, in the future, we should figure out a way to signal directly in the line number value that this line can be hidden, so only exactly the inner call gets hidden, but not errors that happen in the kwsorter itself (for example, for computing default values). Then all of these could print comparable backtraces (rather than just hiding the kwcall, as this PR does): ``` julia> f(; a, b) = error() f (generic function with 1 method) julia> f() ERROR: UndefKeywordError: keyword argument `a` not assigned Stacktrace: [1] f() @ Main ./REPL[1]:1 [2] top-level scope @ REPL[2]:1 julia> f(b=1) ERROR: UndefKeywordError: keyword argument `a` not assigned Stacktrace: [1] kwcall(::NamedTuple{(:b,), Tuple{Int64}}, ::typeof(f)) @ Main ./REPL[1]:1 [2] top-level scope @ REPL[3]:1 julia> f(a=1) ERROR: UndefKeywordError: keyword argument `b` not assigned Stacktrace: [1] kwcall(::NamedTuple{(:a,), Tuple{Int64}}, ::typeof(f)) @ Main ./REPL[1]:1 [2] top-level scope @ REPL[4]:1 julia> f(a=1, b=2) ERROR: Stacktrace: [1] error() @ Base ./error.jl:44 [2] f(; a::Int64, b::Int64) @ Main ./REPL[1]:1 [3] kwcall(::NamedTuple{(:a, :b), Tuple{Int64, Int64}}, ::typeof(f)) @ Main ./REPL[1]:1 [4] top-level scope @ REPL[5]:1 ``` Fix #47319 | 01 November 2022, 14:49:51 UTC |
5707b25 | K Pamnany | 01 November 2022, 14:29:09 UTC | Use an atomic to load `jl_all_tls_states` | 01 November 2022, 14:29:09 UTC |
70ecc4b | apaz | 01 November 2022, 12:43:00 UTC | loader_fputs convert to utf16 only for win console (#47406) We use loader_fputs to write error messages, and when we write them to files we want those files to be utf8. This should fix a broken test introduced in #47343. | 01 November 2022, 12:43:00 UTC |
1c963b6 | K Pamnany | 01 November 2022, 12:09:45 UTC | Use an atomic to load `jl_n_threads` | 01 November 2022, 12:09:45 UTC |