16529a8 | Jameson Nash | 08 January 2017, 03:21:39 UTC | drop ->tvars from TypeMapEntry | 08 January 2017, 14:40:38 UTC |
76a927c | Jameson Nash | 07 January 2017, 20:37:10 UTC | drop the inexact parameter and instead rely on subtyping computing correct answers | 08 January 2017, 14:40:37 UTC |
a034de4 | Jameson Nash | 07 January 2017, 00:57:46 UTC | improve codegen of unionall arrays | 07 January 2017, 05:27:33 UTC |
210d611 | Jameson Nash | 07 January 2017, 00:22:15 UTC | add a fastpath to type-equal for leaftype this is important because it is on the inner loop for `jl_inst_concrete_tupletype` (aka `jl_f_tuple`) | 07 January 2017, 05:27:33 UTC |
1872d59 | Jameson Nash | 03 January 2017, 05:52:55 UTC | fix another broken subtyping test | 07 January 2017, 05:27:33 UTC |
8dc1afe | Jameson Nash | 02 January 2017, 19:06:43 UTC | fix a broken subtyping test | 07 January 2017, 05:27:33 UTC |
4534b51 | Jameson Nash | 02 January 2017, 00:29:41 UTC | remove extraneous for loops for subtype algo fold the "more" flag into the first iteration of subtyping | 07 January 2017, 05:27:33 UTC |
dbe9c70 | Jeff Bezanson | 30 December 2016, 18:54:13 UTC | address review comments | 07 January 2017, 05:27:33 UTC |
fb5a0bb | Jeff Bezanson | 12 December 2016, 15:39:07 UTC | doc and NEWS updates for jb/subtype | 07 January 2017, 05:27:33 UTC |
7a60cf7 | Jeff Bezanson | 12 December 2016, 20:56:12 UTC | make ordered type cache more robust Allow multiple entries that are equal according to the ordered comparison predicate typekey_compare, but not equal according to jl_types_equal. This is mostly needed for types in replaced modules, which can be distinguished by pointer comparison but cannot be reliably ordered (since they differ only in memory address). | 07 January 2017, 05:27:32 UTC |
3c5163d | Jeff Bezanson | 22 November 2016, 22:25:44 UTC | more test fixes | 07 January 2017, 05:27:32 UTC |
1b89a92 | Jeff Bezanson | 15 November 2016, 05:17:32 UTC | get several test suites passing [ci skip] | 07 January 2017, 05:27:32 UTC |
63b48c8 | Jeff Bezanson | 15 November 2016, 21:19:32 UTC | clean up subtype tests [ci skip] | 07 January 2017, 05:27:32 UTC |
d4d07f7 | Jeff Bezanson | 04 November 2016, 22:23:18 UTC | make type declarations with triangular constraints work properly e.g. fixes #6721 fixes to type instantiation | 07 January 2017, 05:27:24 UTC |
75286b5 | Jeff Bezanson | 03 November 2016, 22:19:52 UTC | implement `where` syntax | 07 January 2017, 05:27:03 UTC |
fde4fdc | Jeff Bezanson | 03 November 2016, 19:50:05 UTC | remove old type intersection code [ci skip] | 07 January 2017, 05:27:03 UTC |
f6aa980 | Jeff Bezanson | 21 October 2016, 03:17:57 UTC | get all intersection tests passing [ci skip] | 07 January 2017, 05:27:03 UTC |
7fba081 | Jeff Bezanson | 25 October 2016, 18:54:33 UTC | fully move subtyping tests from test/core to test/subtype | 07 January 2017, 05:27:03 UTC |
99d733f | Jeff Bezanson | 19 October 2016, 22:58:14 UTC | add tests for type intersection [ci skip] | 07 January 2017, 05:27:03 UTC |
2b3e2e7 | Jeff Bezanson | 27 September 2016, 03:32:31 UTC | rough cut of type intersection [ci skip] | 07 January 2017, 05:27:03 UTC |
a64c591 | Jeff Bezanson | 13 September 2016, 18:59:46 UTC | subtype.c: fixes and more comments | 07 January 2017, 05:27:03 UTC |
8f36a8f | Jeff Bezanson | 13 September 2016, 04:51:44 UTC | add subtype test suite [ci skip] | 07 January 2017, 05:27:03 UTC |
6cb7669 | Jeff Bezanson | 13 September 2016, 03:23:33 UTC | fixes to show and serialize for new types improve static_show of UnionAll and Tuple | 07 January 2017, 05:27:02 UTC |
32a41e7 | Jeff Bezanson | 17 August 2016, 15:54:41 UTC | update type union and instantiation for new representation switch to new type system APIs in several places get things working well enough to load boot.jl subtype: add concrete matching rule and work on Vararg constraints improve diagonal subtype rule | 07 January 2017, 05:27:02 UTC |
886dba4 | Jeff Bezanson | 22 August 2016, 15:50:55 UTC | put subtyping algorithm skeleton in place | 07 January 2017, 05:27:02 UTC |
2bd0ccc | Jeff Bezanson | 07 September 2016, 18:06:54 UTC | update examples/juliatypes.jl - update for v0.5 - an algorithm improvement for vars that match unions | 07 January 2017, 05:27:02 UTC |
f552b94 | Jeff Bezanson | 06 January 2017, 16:51:03 UTC | fix #19897, regression in generated function reflection due to #19846 | 07 January 2017, 05:27:02 UTC |
77dfb35 | Jeff Bezanson | 06 January 2017, 21:42:50 UTC | Merge pull request #19875 from bjarthur/bja/printf add string precision to @printf | 06 January 2017, 21:42:50 UTC |
236a8af | Jeff Bezanson | 06 January 2017, 21:35:58 UTC | Merge pull request #19879 from pabloferz/pz/bc-typearg Some broadcast performance tweaks | 06 January 2017, 21:35:58 UTC |
cb5ad65 | Keno Fischer | 06 January 2017, 21:06:55 UTC | Merge pull request #19887 from JuliaLang/kf/kernelbug Without DirtyCOW, julia causes penguins to freeze | 06 January 2017, 21:06:55 UTC |
dd6ee63 | Katharine Hyatt | 06 January 2017, 17:41:36 UTC | Merge pull request #19842 from mpastell/markdown_cr Fix \r bugs in Markdown.parse | 06 January 2017, 17:41:36 UTC |
b36d53c | Jeff Bezanson | 06 January 2017, 17:02:56 UTC | Revert "add `<-` as an assignment-like operator" This reverts commit 1820cbb4df26a95ae42958fc2827eac14c4891c8. | 06 January 2017, 17:02:56 UTC |
fd2a882 | Stefan Karpinski | 06 January 2017, 16:53:43 UTC | Merge pull request #19750 from JuliaLang/kc/testset_bt_noprint do not print backtrace for TestSetExceptions and Pkg.test() exception. | 06 January 2017, 16:53:43 UTC |
5d90ab6 | Stefan Karpinski | 06 January 2017, 15:09:19 UTC | Deprecate `@test_approx_eq` (part of #4615) (#19880) | 06 January 2017, 15:09:19 UTC |
a861aaf | Katharine Hyatt | 06 January 2017, 13:19:23 UTC | Remove docs for deprecated cummin/cummax, more cross refs (#19882) | 06 January 2017, 13:19:23 UTC |
45d2fe0 | Simon Byrne | 06 January 2017, 13:05:23 UTC | Throw error on aborted rebase (#19651) * Throw error on aborted rebase `LibGit2.rebase!` will no longer silently abort if an error occurs. This changes the behaviour of one of the tests: for some reason LibGit2 is unable to automatically rebase this branch. * Correctly handle merge from unborn branch. Fixes earlier issue with rebase. * remove unnecessary comment * change equality * add some docs, and rename unborn -> orphan (consistent with CLI git) * correct branch name, close GitReference * close GitCommit | 06 January 2017, 13:05:23 UTC |
14a5dc8 | Mauro | 06 January 2017, 13:03:26 UTC | Added docs for ObjectIdDict and WeakKeyDict (#19515) Fixes #19498 | 06 January 2017, 13:03:26 UTC |
1ccfd76 | Mauro | 06 January 2017, 13:03:11 UTC | WeakKeyDict constructors to mirror Dict's (#19566) - constructors for Pairs and iterables - updated tests - unrelated: changed wrapped dict to a tighter type | 06 January 2017, 13:03:11 UTC |
62c423b | Steven G. Johnson | 06 January 2017, 13:02:27 UTC | WIP: custom Unicode normalization for Julia identifiers (#19464) * implement custom Julia Unicode normalization for confusable characters in identifiers * whoops * separated julia_charmap into its own file to make it easier to update * normalize fullwidth -> halfwidth in identifiers, ala NFKC * make \varepsilon complete to ε (u+03b5), fixes #14751 * docs for canonicalization * normalize fullwidth characters during parsing (fixes #5903) * typo * tests * be more cautious about normalizing chars when parsing, so as not to normalize string literals * test fullwidth numeric literals and parens * typo/clarification * update to utf8proc-2.1 * checksum for utf8proc 2.1 * moved symbol-normalization test from test/core to test/parse * Revert "be more cautious about normalizing chars when parsing, so as not to normalize string literals" This reverts commit 81033fa0a0f11f9e42907e3348c64b95f070cac4. * Revert "normalize fullwidth characters during parsing (fixes #5903)" This reverts commit cf61972a68bc76b4f2fdbebc5a9fb3e5c9619338. * remove more references to fullwidth normalization * rm fullwidth identifier normalization | 06 January 2017, 13:02:27 UTC |
5407a0d | Stefan Karpinski | 06 January 2017, 12:52:35 UTC | test/runtests.jl: fix code for the on tests case (#19881) This would fail if there were no testset-style tests run. | 06 January 2017, 12:52:35 UTC |
8192aff | Steven G. Johnson | 06 January 2017, 12:22:04 UTC | don't warn about redefining methods in Main (closes #18725) (#19888) | 06 January 2017, 12:22:04 UTC |
5c421ff | Ben Arthur | 05 January 2017, 14:38:17 UTC | add string precision to @printf | 06 January 2017, 12:14:18 UTC |
32fd14c | Steven G. Johnson | 06 January 2017, 12:12:31 UTC | show current-world number of methods for display(::Function) (#19889) | 06 January 2017, 12:12:31 UTC |
4a6bbb8 | Keno Fischer | 06 January 2017, 10:56:30 UTC | Fix reflection test failure on 32bit (#19885) The bug here is that one of the pointers gets zero extended (isymb->first), while the other (Fptr) gets sign extended, so comparing the two breaks down. The simplest way to fix, is just to use the appropriate integer size for the platform we're on. | 06 January 2017, 10:56:30 UTC |
5b6ff4a | Jeff Bezanson | 06 January 2017, 04:03:05 UTC | Merge pull request #19800 from ararslan/aa/farewell-collections RFC: Move PriorityQueue and heap functions out of Base | 06 January 2017, 04:03:05 UTC |
84d9eac | Steven G. Johnson | 06 January 2017, 02:23:54 UTC | REPL help for LaTeX tab completion (#19858) * make it possible to get REPL help directed to a caller-specified io::IO * add REPL help for latex tab completion | 06 January 2017, 02:23:54 UTC |
c8312d3 | Keno Fischer | 06 January 2017, 01:06:12 UTC | Disable /proc/self/mem memmgr on Linux < 4.10 The DirtyCOW exploit mitigation patch in Linux (torvalds/linux@19be0eaff), introduced a kernel bug, that would cause the kernel to hang if an application tries to use /proc/<pid>/mem (or ptrace) to bypass page protections on pages that are backed by transparent huge pages (though only if the write causes a COW resolution). We make use of this feature in our cg memory manager on Linux, but since we can't predict whether or not our mappings will be backed by transparent huge pages, the only safe thing to do is to fall back to dual maps on kernels that potentially have this issue. Since the problematic commit is a fix for a high-profile exploit, it is quite likely that it is included in most stable kernels by now. I have a patch pending at http://marc.info/?l=linux-mm&m=148359462417378&w=2, which I expect will fix this in the kernel for 4.10. However, we'll have to disable the use of /proc/self/mem for all prior kernel versions to avoid locking up the kernel. | 06 January 2017, 01:06:12 UTC |
f550489 | Yichao Yu | 06 January 2017, 00:37:31 UTC | Fix c++-compat warnings | 06 January 2017, 00:37:31 UTC |
6eaec18 | Keno Fischer | 05 January 2017, 22:13:05 UTC | Merge pull request #19823 from JuliaLang/kf/fixcpuidthings Fixes for cpuid specific binaries patch | 05 January 2017, 22:13:05 UTC |
d7625dd | Alex Arslan | 05 January 2017, 19:55:02 UTC | Add ORIGIN to rpath for libgit2, libssh2, and mbedtls on FreeBSD (#19870) | 05 January 2017, 19:55:02 UTC |
49859b9 | Pablo Zubieta | 05 January 2017, 08:12:40 UTC | broadcast performance tweaks | 05 January 2017, 17:42:09 UTC |
fd2928b | Tony Kelman | 05 January 2017, 17:20:22 UTC | Merge branch 'master' into aa/farewell-collections | 05 January 2017, 17:20:22 UTC |
d4707a0 | Jeff Bezanson | 05 January 2017, 16:45:10 UTC | Merge pull request #19765 from JuliaLang/jb/leftarrow RFC: add `<-` as an assignment-like operator | 05 January 2017, 16:45:10 UTC |
1f59fbf | Jeff Bezanson | 05 January 2017, 16:30:12 UTC | Merge pull request #19869 from JuliaLang/yyc/typeinf/fast_cmp Fix inference of fastmath floating point comparisons | 05 January 2017, 16:30:12 UTC |
f25d5a7 | Steven G. Johnson | 05 January 2017, 16:22:24 UTC | more verbose multi-line display(c) for Char (#19847) * more verbose multi-line display(c) for Char | 05 January 2017, 16:22:24 UTC |
b53cd8b | Martin Holters | 05 January 2017, 16:19:40 UTC | Make broadcast return BitArray even if it cannot be inferred (#19854) * Make broadcast return BitArray even if it cannot be inferred | 05 January 2017, 16:19:40 UTC |
7a80fc9 | Jeff Bezanson | 05 January 2017, 16:16:30 UTC | Merge pull request #19865 from JuliaLang/jb/fix8925 fix #8925, `global const (a,b) = (1,2)` should work | 05 January 2017, 16:16:30 UTC |
f3fd7ac | Jeff Bezanson | 05 January 2017, 16:15:59 UTC | Merge pull request #19868 from JuliaLang/jb/fix7669 fix #7669, always parse assignments the same inside macro calls | 05 January 2017, 16:15:59 UTC |
fa9dfb8 | Martin Holters | 05 January 2017, 16:12:47 UTC | Support speye for abstract element type (#19871) | 05 January 2017, 16:12:47 UTC |
20b704a | Matt Bauman | 05 January 2017, 14:47:31 UTC | Faster view creation (#19259) * Ensure SubArray constructor inlines * Don't bother calculating the size of the SubArray during creation All it needs is the dimensionality of the indexing result. * Deprecate three-argument `SubArray` constructor * Wrap long lines | 05 January 2017, 14:47:31 UTC |
e22198b | Simon Byrne | 05 January 2017, 14:05:20 UTC | Add BigInt methods for op_with_overflow (#19859) Fixes issue mentioned in https://github.com/JuliaLang/julia/pull/17152#issuecomment-269951160 | 05 January 2017, 14:05:20 UTC |
8d369bf | Ben Arthur | 05 January 2017, 14:03:18 UTC | added logging() for redirection of info/warn/error messages (#16213) | 05 January 2017, 14:03:18 UTC |
cbc6670 | Tony Kelman | 05 January 2017, 11:46:50 UTC | Merge pull request #19855 from JuliaLang/sb/libgit2/25 Upgrade libgit2 to v0.25.0 | 05 January 2017, 11:46:50 UTC |
2356ebb | Tony Kelman | 05 January 2017, 06:38:15 UTC | Merge pull request #19863 from JuliaLang/jr/perfdocs update test/perf README (fix #19853) | 05 January 2017, 06:38:15 UTC |
7aa6d41 | Jeff Bezanson | 04 January 2017, 23:28:56 UTC | fix #7669, always parse assignments the same inside macro calls | 05 January 2017, 06:02:00 UTC |
cf68974 | Jeff Bezanson | 05 January 2017, 05:58:32 UTC | Merge pull request #19867 from JuliaLang/jb/repstring remove `RepString` type | 05 January 2017, 05:58:32 UTC |
26af9c0 | Jeff Bezanson | 04 January 2017, 19:32:37 UTC | fix #8925, `global const (a,b) = (1,2)` should work | 05 January 2017, 04:16:20 UTC |
c56d59b | Yichao Yu | 05 January 2017, 01:43:43 UTC | Fix inference of fastmath floating point comparisons | 05 January 2017, 01:43:43 UTC |
cfa2312 | Eric Davies | 04 January 2017, 22:03:39 UTC | Fix #19655 (TypeError on empty stacktrace) (#19656) * Fix #19655 * Test for #19655 | 04 January 2017, 22:03:39 UTC |
55aa1d4 | Jeff Bezanson | 04 January 2017, 21:43:54 UTC | remove `RepString` type | 04 January 2017, 21:51:38 UTC |
7648ad6 | Augustin Yiptong | 04 January 2017, 21:39:59 UTC | splitdrive under windows (#19695) * splitdrive under windows * splitdrive under windows tests * more test * move tests outside loop over S * changed test cases to S("...") | 04 January 2017, 21:39:59 UTC |
1820cbb | Jeff Bezanson | 29 December 2016, 21:43:41 UTC | add `<-` as an assignment-like operator | 04 January 2017, 21:37:02 UTC |
c5c417c | David P. Sanders | 04 January 2017, 19:40:06 UTC | BigFloat constructors with given precision and/or rounding mode (#17217) | 04 January 2017, 19:40:06 UTC |
28bcb4b | Jarrett Revels | 04 January 2017, 18:44:06 UTC | update test/perf README (fix #19853) | 04 January 2017, 19:23:45 UTC |
6ce0102 | Jane Herriman | 04 January 2017, 19:09:43 UTC | Move docs inline from helpdb/Base.jl (#19674) | 04 January 2017, 19:09:43 UTC |
794c0ac | Tony Kelman | 04 January 2017, 18:37:03 UTC | Merge pull request #19810 from JuliaLang/tk/llvm-patches Add 3 LLVM patches to fix issue #19792 | 04 January 2017, 18:37:03 UTC |
8eb55ab | Kristoffer Carlsson | 04 January 2017, 18:34:16 UTC | RFC: jump to numbered stackframe with hotkey (#19680) * jump to numbered stackframe with hotkey CTRL+Q * fix grammar * add pr ref to NEWS | 04 January 2017, 18:34:16 UTC |
7eadb55 | Spencer Russell | 04 January 2017, 16:05:24 UTC | updates enum capitalization to be more consistent with Julia style (#19506) | 04 January 2017, 16:05:24 UTC |
d8c9b0c | Raj Krishnan | 04 January 2017, 15:47:54 UTC | Docs: Readme reflects correct package name | 04 January 2017, 15:47:54 UTC |
44502c1 | Raj Krishnan | 04 January 2017, 15:35:21 UTC | Docs: Fixes #19853 Removed dead links and replaced abandoned benchmark package. | 04 January 2017, 15:35:21 UTC |
12111e4 | Tony Kelman | 04 January 2017, 15:26:03 UTC | Merge pull request #19728 from kumarkrishna/testset_bitarray Add testset to bitarray | 04 January 2017, 15:26:03 UTC |
99b5d34 | Tony Kelman | 04 January 2017, 15:05:17 UTC | Upgrade libgit2 to v0.25.0 rebase patches fixes #19793 | 04 January 2017, 15:05:41 UTC |
2e0b001 | Simon Byrne | 04 January 2017, 13:04:24 UTC | deprecate vectorised frexp cf #19817 | 04 January 2017, 13:04:24 UTC |
5f265f7 | Tony Kelman | 04 January 2017, 11:38:11 UTC | Update appveyor llvm binaries for latest patches | 04 January 2017, 12:24:16 UTC |
10ca4b2 | Tony Kelman | 04 January 2017, 09:56:36 UTC | Add a patch for detection of AVX512 on Skylake | 04 January 2017, 11:48:16 UTC |
3092476 | Tony Kelman | 02 January 2017, 14:16:32 UTC | Adjust context lines in llvm-D27389.patch | 04 January 2017, 10:24:45 UTC |
8563fc5 | Yuri | 04 January 2017, 09:23:09 UTC | Re-quine quine.jl (#19715) In May 2015, a sweeping [license header change](https://github.com/JuliaLang/julia/commit/fde191ecc9199aa48feb8fed08f9dde7bb699aed) added a standard license comment to the top of each of Julia's source files. This had the side effect of making quine.jl no longer a true [quine](http://www.madore.org/~david/computers/quine.html), since its printed output was no longer identical to the original source file. This pull request restores quine.jl's quine-ness. | 04 January 2017, 09:23:09 UTC |
205a8cd | Tony Kelman | 04 January 2017, 09:14:27 UTC | Merge pull request #19808 from JuliaLang/ksh/docex New Year's Day of Doctests | 04 January 2017, 09:14:27 UTC |
888f6b5 | Michael Hatherly | 04 January 2017, 09:08:08 UTC | Upgrade doc deps (#19834) Fixes #19826. | 04 January 2017, 09:08:08 UTC |
362b271 | Matti Pastell | 04 January 2017, 09:07:43 UTC | Fix Markdown.latex for paragraphs and code (#19832) | 04 January 2017, 09:07:43 UTC |
da20cee | Katharine Hyatt | 04 January 2017, 09:01:25 UTC | Convert BLAS tests to use testsets (#19820) | 04 January 2017, 09:01:25 UTC |
2e08102 | Martin Holters | 04 January 2017, 08:12:07 UTC | Remove broadcast_elwise_op and deprecate promote_eltype_op (#19814) * Remove broadcast_elwise_op and deprecate promote_eltype_op * Fix typo in comment | 04 January 2017, 08:12:07 UTC |
5d2e538 | Jeff Bezanson | 04 January 2017, 04:00:58 UTC | Merge pull request #19846 from JuliaLang/jb/fix18838 fix #18838, populate `source` field of generated functions | 04 January 2017, 04:00:58 UTC |
887815c | Stefan Karpinski | 04 January 2017, 01:47:35 UTC | Merge pull request #17155 from JuliaLang/sk/funcnegcomp function composition (f∘g) and negation (!f) | 04 January 2017, 01:47:35 UTC |
7bb7f13 | Jeff Bezanson | 04 January 2017, 01:18:14 UTC | fix #19848, doc build failing on `CompoundPeriod` | 04 January 2017, 01:18:14 UTC |
39e62e7 | Keno Fischer | 02 January 2017, 20:00:42 UTC | Fixes for cpuid specific binaries patch Fixes #19822 | 04 January 2017, 01:15:16 UTC |
fce015b | Kristoffer Carlsson | 29 December 2016, 12:27:10 UTC | no longer print backtraces to internal functions that run tests | 03 January 2017, 22:57:27 UTC |
6a662e4 | Wonseok Shin | 03 January 2017, 22:01:15 UTC | Generalize isapprox() to support nested arrays (#19637) * Generalize isapprox() to support nested arrays - Implement `promote_leaf_eltypes()` to find the type promoted from all leaf types of a given nested array. | 03 January 2017, 22:01:15 UTC |
b9106a5 | Jeff Bezanson | 03 January 2017, 21:57:53 UTC | Merge pull request #19828 from stevengj/getfield_nodep drop special support in the parser for deprecating f.(1) getfield syntax | 03 January 2017, 21:57:53 UTC |
8b708b4 | Jeff Bezanson | 03 January 2017, 21:21:16 UTC | fix #18838, populate `source` field of generated functions | 03 January 2017, 21:25:44 UTC |