012af60 | Rafael Fourquet | 17 November 2019, 15:20:32 UTC | REPL: switch between modes via a menu The menu pops up via the `,` key (when at beginning of input). Modes selected via `,` are sticky. Help and shell modes remain non-sticky when entered via `;` and `?`. | 17 November 2019, 16:47:08 UTC |
4982b36 | Rafael Fourquet | 16 November 2019, 10:10:09 UTC | REPL: fix yank-twice activates region (#33814) Yanking twice was fooling `setmark` into believing that `^Space`, the keybinding for `setmark`, had been pressed twice, leading to activating the region. So let's make `setmark` behave as a "command" (and activate the region) only when called directly by the user, otherwise treat it as a low-level function. | 16 November 2019, 10:10:09 UTC |
9e20e03 | Rafael Fourquet | 16 November 2019, 09:20:56 UTC | REPL: allow tweaking the implicit IOContext of the REPL (#29249) This adds an `iocontext::Dict{Symbol,Any}` field to `LineEditREPL.options`, which can be initialized with `atreplinit`, and updated interactively, e.g. `Base.active_repl.options.iocontext[:compact] = true`. Fixes #20509. | 16 November 2019, 09:20:56 UTC |
d564afe | Rafael Fourquet | 16 November 2019, 09:00:05 UTC | docs: disambiguate what rand((2, 3)) does (fix #33309) (#33742) * docs: disambiguate what rand((2, 3)) does (fix #33309) * Update stdlib/Random/src/Random.jl Co-Authored-By: Matt Bauman <mbauman@juliacomputing.com> | 16 November 2019, 09:00:05 UTC |
6513811 | Rafael Fourquet | 16 November 2019, 08:52:21 UTC | REPL: implement "insert last word from previous history entry" (#33749) Bind this shell-like feature to the usual "meta-.". | 16 November 2019, 08:52:21 UTC |
6bf9f72 | Rafael Fourquet | 16 November 2019, 08:42:15 UTC | REPL: fix/add doc for \\M- and \\C- key specifiers (#33771) | 16 November 2019, 08:42:15 UTC |
b29d951 | Jameson Nash | 15 November 2019, 22:05:15 UTC | Makefile cleanups (#33826) fix #32852 | 15 November 2019, 22:05:15 UTC |
cf5b905 | Jeff Bezanson | 15 November 2019, 21:08:43 UTC | fix #33338, race condition in triggering method compilation (#33839) | 15 November 2019, 21:08:43 UTC |
44c923a | Jeff Bezanson | 15 November 2019, 21:01:47 UTC | remove spurious `&` in `get_next_task` (#33860) | 15 November 2019, 21:01:47 UTC |
ee3c88b | Rafael Fourquet | 15 November 2019, 20:58:48 UTC | rename test/bigint.jl to test/gmp.jl (#31992) For consistency with base/gmp.jl and test/mpfr.jl. | 15 November 2019, 20:58:48 UTC |
38be538 | Chris Foster | 15 November 2019, 20:22:37 UTC | Add line numbers in parser errors for disallowed space (#33855) | 15 November 2019, 20:22:37 UTC |
1b03919 | Jeff Bezanson | 15 November 2019, 19:28:45 UTC | avoid giving useless suggestions for method ambiguities (#33852) | 15 November 2019, 19:28:45 UTC |
7f63f00 | Rafael Fourquet | 15 November 2019, 16:54:10 UTC | REPL: make edit_title_case (M-c) call titlecase (#33851) When edit_title_case was implemented, `uppercasefirst` (née `ucfirst`) and `titlecase` were doing the same transformation on only one word. But `titlecase` now transforms non-leading letters to lowercase, which is more expected for the M-c keyboard combo, as it's the standard readline behavior (I believe). | 15 November 2019, 16:54:10 UTC |
bbccbac | Jameson Nash | 15 November 2019, 16:45:58 UTC | Merge pull request #33717 from JuliaLang/jn/allocated simplify allocated macro | 15 November 2019, 16:45:58 UTC |
4caf944 | Jameson Nash | 15 November 2019, 15:54:22 UTC | [REPL] exit when input stream closes (#33838) Instead of running a tight infinite loop. Seems somewhat preferable. | 15 November 2019, 15:54:22 UTC |
4d7701e | Pierre Haessig | 15 November 2019, 15:24:03 UTC | document return nothing in Functions man page (2nd PR) (#33028) * document return nothing in Functions man page + fix broken link to Markdown syntax in doc README | 15 November 2019, 15:24:03 UTC |
f5b961e | Daniel Karrasch | 15 November 2019, 14:36:51 UTC | add missing docstrings for BLAS routines (#33801) | 15 November 2019, 14:36:51 UTC |
4e76bcf | Alexander Seiler | 15 November 2019, 08:22:22 UTC | Fix left division for Hermitian sparse array and transpose/adjoint rhs (#33667) | 15 November 2019, 08:22:22 UTC |
360c102 | Ben Arthur | 15 November 2019, 00:42:08 UTC | mention Pkg's LRU cache in HISTORY.md (#33846) @KristofferC see https://github.com/JuliaLang/julia/pull/32651#issuecomment-546990023 | 15 November 2019, 00:42:08 UTC |
73a1caf | Jeff Bezanson | 15 November 2019, 00:26:08 UTC | remove an object allocation in the scheduler (#32835) - avoid allocating `gettask` closure - stop using a non-const global for eventloop() - remove some unused code | 15 November 2019, 00:26:08 UTC |
ecb1bff | Jeff Bezanson | 14 November 2019, 23:41:36 UTC | fix #33841, regression in let-bound function with kwargs (#33848) | 14 November 2019, 23:41:36 UTC |
dc9550f | Sebastian Pfitzner | 14 November 2019, 23:40:45 UTC | table column alignment in html output (#33849) | 14 November 2019, 23:40:45 UTC |
833b11a | Daniel Karrasch | 14 November 2019, 23:39:38 UTC | tighten signature of (complex mat) x (real matorvec) (#33843) | 14 November 2019, 23:39:38 UTC |
2e46386 | Jeff Bezanson | 14 November 2019, 23:38:48 UTC | use 1 object for `SpinLock` instead of 2 (#32836) | 14 November 2019, 23:38:48 UTC |
77a4d06 | Valentin Churavy | 14 November 2019, 15:21:22 UTC | Merge pull request #33234 from smallnamespace/smallnamespace-sqrt-fast Add sqrt_llvm_fast intrinsic (see #33220) | 14 November 2019, 15:21:22 UTC |
cb0cd91 | Elliot Saba | 14 November 2019, 07:15:47 UTC | Merge pull request #33819 from JuliaLang/sf/repl_realpath Fix `realpath()` assumptions in REPL test suite | 14 November 2019, 07:15:47 UTC |
f196d3f | Jeff Bezanson | 13 November 2019, 23:56:53 UTC | fix #33768, inference of `_apply` with strange argument list lengths (#33832) | 13 November 2019, 23:56:53 UTC |
08d1fe6 | Jeff Bezanson | 13 November 2019, 20:31:32 UTC | Merge pull request #33828 from JuliaLang/jn/gc-opt1 some small gc optimizations | 13 November 2019, 20:31:32 UTC |
baa6efd | Jameson Nash | 13 November 2019, 20:13:43 UTC | Merge pull request #33827 from JuliaLang/jn/sf/repl_realpath | 13 November 2019, 20:13:43 UTC |
1731d0a | Jameson Nash | 13 November 2019, 20:12:52 UTC | remove gc-token from WeakKeyDict (#33825) This was a hold-over from the old iteration protocol, which needed to maintain state between `done` and `next`. The `iteration` function of `Dict` has since been re-written to be safe for concurrent deletions. Replaces: #33756 Co-Authored-By: Hans-Peter Suter <hps@treetron.ch> | 13 November 2019, 20:12:52 UTC |
89a51fb | Takafumi Arakaki | 13 November 2019, 00:49:01 UTC | Fix compat docstring (#33813) | 13 November 2019, 00:49:01 UTC |
8333297 | Gerhard Aigner | 13 November 2019, 00:47:30 UTC | Fix findfirst with decreasing step range (#33809) Fixes #33808 | 13 November 2019, 00:47:30 UTC |
7358173 | Daniel Karrasch | 12 November 2019, 21:46:26 UTC | Automatically up-convert eltype in SPQR (#33601) | 12 November 2019, 21:46:26 UTC |
beee107 | Elliot Saba | 12 November 2019, 20:15:46 UTC | Add LIBEXECDIR relative folder constant (#33777) Add LIBEXECDIR relative folder constant | 12 November 2019, 20:15:46 UTC |
4ee78b4 | Jameson Nash | 12 November 2019, 19:15:49 UTC | fixup! Fix `realpath()` assumptions in REPL test suite delete no-op code | 12 November 2019, 19:45:52 UTC |
45d4277 | Jameson Nash | 12 November 2019, 19:12:21 UTC | fixup! Fix `realpath()` assumptions in REPL test suite | 12 November 2019, 19:45:52 UTC |
629d1a6 | Jameson Nash | 12 November 2019, 19:43:34 UTC | gc: add early-out for empty arrays | 12 November 2019, 19:43:34 UTC |
a77b2c1 | Jameson Nash | 12 November 2019, 19:43:12 UTC | codegen: remove some unnecessary write-barriers | 12 November 2019, 19:43:12 UTC |
ce646d5 | Rafael Fourquet | 12 November 2019, 17:41:25 UTC | fix "parallel computing" doc: remove extraneous "forwards" "Julia from `v0.5` forwards natively supports multi-threading" became "Julia forwards natively supports multi-threading", where the "forwards" doesn't make much sense anymore. | 12 November 2019, 17:41:28 UTC |
3d1c6e2 | Elliot Saba | 12 November 2019, 00:31:21 UTC | Fix `realpath()` assumptions in REPL test suite On non-case-sensitive platforms, it is possible for `cd` and `pwd()` to return values that are different from what `realpath()` of the same would give back. Therefore, we must adjust our tests to be more robust against this variation. This should hopefully fix some of the testing hangs on Windows | 12 November 2019, 00:31:21 UTC |
91effa2 | Rafael Fourquet | 11 November 2019, 17:28:37 UTC | rem docstring: fix missing space | 11 November 2019, 17:28:37 UTC |
1ed96b2 | Daniel Karrasch | 11 November 2019, 16:05:37 UTC | Dispatch even more to BLAS (#33743) * get MulAddMul out of the BLAS way, promote alpha/beta coeffs * fix ambiguity * simplify promotion, :crossedfingers: * remove promote_unless_bool, add symmetry check `syrk_wrapper!` (bugfix) * give BLAS another chance * extend coefficient promotion to sym(v/m)! and hem(v/m)! * fix typo | 11 November 2019, 16:05:37 UTC |
6eebbbe | Jeff Bezanson | 09 November 2019, 16:38:00 UTC | more latency improvements for #33615 (#33779) | 09 November 2019, 16:38:00 UTC |
7212602 | Sebastian Ament | 09 November 2019, 09:55:50 UTC | Casting low-rank CholeskyPivoted to Matrix (#33775) | 09 November 2019, 09:55:50 UTC |
8c34ad3 | Nathan Daly | 09 November 2019, 02:27:31 UTC | Add `lock(f, lock)` docstring to `lock`. (#33796) Currently the `lock(l) do ; ... end` method is not documented, and so is not discoverable. | 09 November 2019, 02:27:31 UTC |
07535dd | Stefan Karpinski | 08 November 2019, 17:01:09 UTC | temp cleanup: handle path collisions safely (#33786) | 08 November 2019, 17:01:09 UTC |
445a6fc | Mustafa | 07 November 2019, 15:28:03 UTC | test | 07 November 2019, 15:38:07 UTC |
025ac54 | Mustafa | 06 November 2019, 19:09:31 UTC | Add LIBEXECDIR relative folder constant | 07 November 2019, 15:38:07 UTC |
3b11bf3 | Rafael Fourquet | 06 November 2019, 20:42:03 UTC | REPL: fix bug when all the input buffer can't be printed (#33770) When the input buffer contained too many lines, it was still possible to edit it, but blindly for the n invisible lines: indeed, the first n lines where never shown even when the cursor was moved in their region. Fix that, by centering the cursor in the available space when all the lines can't be shown at the same time. Other strategies are possible, some nicer, but this is probably the simplest, by far, as it is "stateless": it doesn't need to remember what happened during the last refresh or to know what event triggered the refresh. | 06 November 2019, 20:42:03 UTC |
ba4f858 | Jeff Bezanson | 05 November 2019, 23:04:02 UTC | help load time regression in #33615 (#33757) | 05 November 2019, 23:04:02 UTC |
3608c84 | Jeff Bezanson | 05 November 2019, 22:38:01 UTC | use `===` for comparison to `nothing` and Symbols (#33764) | 05 November 2019, 22:38:01 UTC |
2fcb1b1 | Fredrik Bagge Carlson | 05 November 2019, 12:22:53 UTC | Update docstring of nullspace (#33767) * Update docstring of nullspace * Update stdlib/LinearAlgebra/src/dense.jl Co-Authored-By: Andreas Noack <andreas@noack.dk> | 05 November 2019, 12:22:53 UTC |
2c187bd | Jeff Bezanson | 04 November 2019, 21:38:31 UTC | fix an error in `dump(NamedTuple)` (#33723) | 04 November 2019, 21:38:31 UTC |
d5d4c69 | Jameson Nash | 04 November 2019, 16:00:55 UTC | Distributed: fix warning generated by test (#33746) | 04 November 2019, 16:00:55 UTC |
f45b6be | Jeff Fessler | 04 November 2019, 13:12:03 UTC | Document customizable binary operators (#33479) * Document customizable binary operators I learned indirectly through comments on a PR that \otimes and \oplus and \odot can be extended as binary operators yet I cannot find this fact anywhere in the documentation. Search the docs for `oplus` yields nothing. Somewhere there must be a list of which characters can be extended this way and it should be documented in the manual. This PR is an attempt to get that ball rolling! * Add url for list and fix prefix typo * Expand docs of infix binary operators. | 04 November 2019, 13:12:03 UTC |
5294219 | Keno Fischer | 03 November 2019, 18:28:07 UTC | Fix non-uniformity in nothrow queries with PartialTypeVar (#33751) We were missing a case in the lattice operation for ⊑(::PartialTypeVar, ::Type{TypeVar}) causing non-uniformity for various nothrow queries. Additionally `apply_type_nothrow` was using `===` rather than `⊑`, causing additional problems there even after the fix. | 03 November 2019, 18:28:07 UTC |
c7e4b99 | klaff | 02 November 2019, 09:58:01 UTC | Update arrays.md (#33748) Adding parenthetical to state that `using LinearAlgebra` is needed to access `I` to create identity matrix. | 02 November 2019, 09:58:01 UTC |
c9a2553 | Elliot Saba | 02 November 2019, 05:45:11 UTC | Force `patchelf` to use 64KB page size on aarch64/powerpc64le (#33745) Force `patchelf` to use 64KB page size on aarch64/powerpc64le | 02 November 2019, 05:45:11 UTC |
ffb3ed0 | Elliot Saba | 02 November 2019, 01:02:09 UTC | Force `PATCHELF` override on `fixup-libgfortran.sh` as well | 02 November 2019, 01:02:09 UTC |
3f0e7d6 | Matt Bauman | 01 November 2019, 22:30:28 UTC | Merge pull request #33533 from JuliaLang/mb/fast-partitions Make Iterators.partition split arrays into views for faster and easier parallelism | 01 November 2019, 22:30:28 UTC |
2f9eebd | Jameson Nash | 01 November 2019, 21:26:41 UTC | Merge pull request #33624 from JuliaLang/jn/runtests-error runtests: show errors and continue | 01 November 2019, 21:26:41 UTC |
5ceed92 | Elliot Saba | 01 November 2019, 20:01:14 UTC | Force `patchelf` to use 64KB page size on aarch64/powerpc64le This fixes problems with binaries on these two platforms that often use these large page sizes. | 01 November 2019, 20:11:13 UTC |
7b9b7e1 | Matt Bauman | 01 November 2019, 17:09:19 UTC | Document the SubArray type (#32931) * Document the SubArray type but not its methods as you should typically use `view` to construct it. Small part of #28697. * Update base/subarray.jl Co-Authored-By: Tim Holy <tim.holy@gmail.com> * Update base/subarray.jl Co-Authored-By: Tim Holy <tim.holy@gmail.com> | 01 November 2019, 17:09:19 UTC |
96be27c | Jameson Nash | 21 October 2019, 13:34:54 UTC | Test: apply no-specialize to all of the code This code (esp. the constructors), may see lots of different types (expressions and results), but it is not performance-sensative code. | 01 November 2019, 16:44:37 UTC |
3c01c60 | Jameson Nash | 21 October 2019, 13:32:25 UTC | runtests: recycle workers after they run into internal errors Previously, if a work crashed, we would rapidly schedule all future jobs to run on that worker. That did not work very well. Now we will tear down the bad worker and start a new one to run the remaining tests. | 01 November 2019, 16:44:37 UTC |
376d848 | Jameson Nash | 21 October 2019, 13:30:38 UTC | runtests: handle Exceptions better Previously, we would only record and not print the Error itself. (Still only using [] for the stacktrace though.) | 01 November 2019, 16:44:37 UTC |
776b72a | Jameson Nash | 21 October 2019, 13:27:11 UTC | runtests: use the IO's lock (if available) to reduce the occurance of intermingled output | 01 November 2019, 16:44:37 UTC |
ebed874 | Jameson Nash | 21 October 2019, 13:25:57 UTC | runtest: use textwidth instead of length While these strings all currently happen to be ascii, which have text-width==length, this is better practice. | 01 November 2019, 16:44:37 UTC |
e0321e4 | Lyndon White | 01 November 2019, 15:32:43 UTC | minor code cleanup in abstractarraymath.jl (#33527) * minor code cleanup in abstractarraymath.jl * Update base/abstractarraymath.jl * Update base/abstractarraymath.jl * Apply suggestions from code review | 01 November 2019, 15:32:42 UTC |
09a0822 | Matt Bauman | 01 November 2019, 14:41:20 UTC | Merge pull request #33620 from JuliaLang/mb/sortinteger fix #33611, sorting of arrays with non-Int axes | 01 November 2019, 14:41:20 UTC |
7793ca9 | Takafumi Arakaki | 01 November 2019, 06:48:54 UTC | Mention that handle_message has to handle non-Symbol id (#33732) | 01 November 2019, 06:48:54 UTC |
47e49ab | Matt Bauman | 31 October 2019, 20:10:42 UTC | Merge branch 'master' into mb/fast-partitions | 31 October 2019, 20:10:42 UTC |
e472b14 | Matt Bauman | 31 October 2019, 20:05:10 UTC | NFC: style changes per review shorter lines, more spaces, and comments (because I had already forgotten how this worked myself) | 31 October 2019, 20:05:10 UTC |
4f649c2 | Daniel Karrasch | 31 October 2019, 15:48:29 UTC | fix diagonal Cholesky factorization and return Diagonal Cholesky factor (#33600) * return diagonal from diagonal cholesky * fix diagonal cholesky, adjust tests | 31 October 2019, 15:48:29 UTC |
93ebf3e | Venkatesh Dayananda | 31 October 2019, 13:03:25 UTC | fix kill docstring: since SIGTERM is not exported (#33726) | 31 October 2019, 13:03:25 UTC |
b36a8c6 | Jameson Nash | 29 October 2019, 15:26:59 UTC | add compile heuristic to all perf macros | 30 October 2019, 23:23:04 UTC |
b9546c8 | Jameson Nash | 29 October 2019, 15:24:41 UTC | make allocated macro more reliable Changes testset to avoid compiler heuristics (copyast) that disables inference. And changes the allocated macro to rely less on inference to elid allocations for the machinary itself. | 30 October 2019, 23:23:03 UTC |
350548c | Jameson Nash | 29 October 2019, 14:56:07 UTC | GC diff: avoid generating lots of signedness checks | 30 October 2019, 23:22:14 UTC |
f0772c5 | Jameson Nash | 30 October 2019, 23:21:40 UTC | Merge pull request #33532 from JuliaLang/jn/error Be careful with internal error handling. | 30 October 2019, 23:21:40 UTC |
42e9490 | Eric Davies | 30 October 2019, 21:43:03 UTC | Document the pass-through specialization heuristic for Type, Vararg, and Function (#32817) | 30 October 2019, 21:43:03 UTC |
f593456 | Jameson Nash | 11 October 2019, 18:30:44 UTC | Libdl: return correct error code Previously, we were returning the error code from `stat` instead (success) in some cases. Also make it unicode text, while we're here. Fix #15117 | 30 October 2019, 19:35:25 UTC |
254be62 | Jameson Nash | 11 October 2019, 06:13:38 UTC | error: add versions of systemerror that accept an error code This is helpful when doing proper cleanup, which might otherwise clear the error code. | 30 October 2019, 19:34:58 UTC |
3defa72 | Jameson Nash | 11 October 2019, 06:14:39 UTC | error: protect internal state from changing errno externally fix #32180 | 30 October 2019, 19:34:58 UTC |
4c775e7 | Jameson Nash | 18 October 2019, 16:06:17 UTC | [GCChecker] update to run on Darwin and simplify some of the code and makefile rules | 30 October 2019, 19:34:58 UTC |
4e31741 | Alexander Seiler | 30 October 2019, 16:24:18 UTC | Prevent `OverflowError` when multiplying empty sparse matrix and its adjoint (fixes #33365) (#33619) | 30 October 2019, 16:24:18 UTC |
a92f566 | Ian Butterworth | 30 October 2019, 16:20:00 UTC | Add pkgdir(m::Module) (#33128) | 30 October 2019, 16:20:00 UTC |
1ae9902 | Jameson Nash | 30 October 2019, 16:17:39 UTC | codegen: avoid bad tbaa paths (#33706) This avoids cases where getMostGenericTBAA could return broken answers. fix #33590 | 30 October 2019, 16:17:39 UTC |
c819cde | Lyndon White | 30 October 2019, 15:54:55 UTC | Tweak log-level descriptions (#33716) Co-Authored-By: Fredrik Ekre <ekrefredrik@gmail.com> | 30 October 2019, 15:54:55 UTC |
306182f | Rafael Fourquet | 30 October 2019, 15:45:25 UTC | improve REPL printing of sets (#33300) | 30 October 2019, 15:45:25 UTC |
b92a35d | Liozou | 30 October 2019, 14:44:46 UTC | Reorder self-referential types checking to prevent inline allocation (#33718) | 30 October 2019, 14:44:46 UTC |
d35134c | Kristoffer Carlsson | 30 October 2019, 13:44:01 UTC | add an example to div for mixed types (#33714) | 30 October 2019, 13:44:01 UTC |
ffdee15 | Keno Fischer | 29 October 2019, 21:41:24 UTC | Add test-revise-* Makefile targets (#33682) These targets use Revise to load any modifications to base into the process before running tests. This is very handy when editing Base and wanting to run the appropriate tests as a sanity check but without waiting 3 minutes for a whole system image rebuild. Of course it was previously possible to do this through the REPL by doing the 3-4 steps manually, but I was generally too lazy too do that, so here's the automatic option. | 29 October 2019, 21:41:24 UTC |
592748a | Daniel Karrasch | 29 October 2019, 11:33:11 UTC | Promote eltype in BlasFloat matrix multiplication (#32587) Co-Authored-By: Fredrik Ekre <ekrefredrik@gmail.com> | 29 October 2019, 11:33:11 UTC |
962634d | Takafumi Arakaki | 29 October 2019, 08:25:33 UTC | Suggestion: Use different precompilation cache path for different system image (#29914) * Include path to system image in package_slug * Include path to julia binary in package_slug * Mix `image_file` and `julia_bin` into project slug | 29 October 2019, 08:25:33 UTC |
49fb792 | Nathan Daly | 29 October 2019, 06:39:43 UTC | Fix `div` docstring to describe division, after #33040 (#33703) It looks like in #33040 the `div` docstring was accidentally incorrectly copied from the `rem` docstring, so it currently describes the `rem` operation, not `div`. This commit changes that docstring to correctly describe integer division with custom rounding. | 29 October 2019, 06:39:43 UTC |
1b6101a | Jeff Bezanson | 29 October 2019, 06:35:26 UTC | fix #33675, unpredictable `summarysize` on isbitsunion arrays (#33713) | 29 October 2019, 06:35:26 UTC |
97f142f | Michael Abbott | 29 October 2019, 02:43:12 UTC | Tuple constructors for one-element containers (#33674) | 29 October 2019, 02:43:12 UTC |
aeb934e | simeonschaub | 28 October 2019, 21:57:56 UTC | better syntax error for := (#33694) | 28 October 2019, 21:57:56 UTC |
5e23b84 | Rafael Fourquet | 28 October 2019, 19:20:23 UTC | speedup BitSet iteration (#33690) | 28 October 2019, 19:20:23 UTC |
309ffc3 | Valentin Churavy | 28 October 2019, 17:40:53 UTC | Merge pull request #33678 from JuliaLang/vc/nosink remove SinkingPass from pass pipeline | 28 October 2019, 17:40:53 UTC |