averello/swift-mirror
1
0
Fork 0
mirror of https://github.com/apple/swift.git synced 2025-12-21 12:14:44 +01:00
Code Issues Packages Projects Releases Wiki Activity
184,205 Commits 623 Branches 2,949 Tags
jgrynspan/windows-GUID-hashable-6.2.1
Commit Graph

3245 Commits

Author SHA1 Message Date
swift-ci
1aed6d25a0 Merge pull request #12724 from DougGregor/reenable-test-34462543 2017-11-02 15:08:21 -07:00
Joe Groff
a59af9855e Merge pull request #12725 from jckarter/key-path-optional-chain-out-of-bounds 
KeyPath: Fix out-of-bounds access when instantiating keypaths with optional chaining components.
 2017-11-02 22:34:33 +01:00
Joe Groff
993d795152 KeyPath: Fix out-of-bounds access when instantiating keypaths with optional chaining components. 
When we pre-scan the components of a key path pattern to determine its runtime type and instance size, we would short-circuit upon seeing an optional-chaining component, since that makes a key path definitely read-only, but the loop also accumulates the size of the instance we're supposed to allocate, so…bad stuff happened. Leave out the short-circuit, fixing SR-6096 | rdar://problem/34889333 .
 2017-11-02 13:32:35 -07:00
Doug Gregor
033bb33e40 Re-enable test failing due to the (now-removed) -F /S/L/PrivateFrameworks 
Fixes rdar://problem/34462543.
 2017-11-02 13:28:31 -07:00
Erik Eckstein
a13b8f1a1d Revert "Re-enable test failing due to the (now-removed) -F /S/L/PrivateFrameworks" 
This test still fails on a bot
rdar://problem/35284791

This reverts commit 9daf89cec5.
 2017-11-01 11:31:19 -07:00
Doug Gregor
9daf89cec5 Re-enable test failing due to the (now-removed) -F /S/L/PrivateFrameworks 
Fixes rdar://problem/34462543.
 2017-10-31 13:26:02 -07:00
Slava Pestov
3f872ecabc These tests now pass in resilient builds 2017-10-27 19:00:47 -07:00
gregomni
55edab880f Don't suggest adding 'as' in a pattern matching context where it will do no good. 2017-10-26 15:51:59 -07:00
Arnold Schwaighofer
3a0ac62776 Disable signaling NaN tests on i386 
i386 does not support signaling NaNs (SR-1515)

rdar://35085021
 2017-10-20 11:49:51 -07:00
Andrew Trick
b73352af7a Disable Unsafe[Raw]BufferPointer testing in optimized mode. 
The optimized-build behavior of UnsafeBufferPointer bounds/overflow
checking cannot be tested. The standard library always compiles with debug
checking enabled, so the behavior of the optimized test depends on whether
the inlining heuristics decide to inline these methods. To fix this, we need
a way to force @_inlineable UnsafeBufferPointer methods to be emitted inside
the client code, and thereby subject the stdlib implementation to the test
case's compile options.
 2017-10-19 21:12:09 -07:00
Philippe Hausler
b78b438897 Ensure that hashing data with zero bytes avoids empty allocations and fix bridged empty data hashes from de-referencing null values (#12509) 2017-10-19 17:53:19 -07:00
Greg Parker
f1fda3cd0d [test] Temporarily XFAIL two UnsafeBufferPointer tests pending rdar://35052802. 2017-10-18 10:02:54 -07:00
Greg Parker
72c65ffcaf Revert "[stdlib] Fix FloatingPoint.init(exactly:) (#11311)" 
This reverts commit c9f4df84f6.
It is causing test failures on 32-bit iOS simulator and on Linux.
 2017-10-18 01:17:08 -07:00
Nate Cook
c9f4df84f6 [stdlib] Fix FloatingPoint.init(exactly:) (#11311) 
* [stdlib] Fix FloatingPoint.init(exactly:)

This initializer wasn't actually checking the exact conversion. Corrects
the tests as well.
 2017-10-17 13:52:11 -05:00
Ben Cohen
a70e857d59 [stdlib] Fix issue with UTF16 index(_:offsetBy:limitedBy) (#12378) 
* Fix issue with empty string ranges

* Add tests for basic offsetBy operation on UTF16View.Index
 2017-10-16 13:44:51 -07:00
Kim Topley
1da4b04d99 Fixes overflow trap when creating DispatchTime objects with large uptimeNanoseconds values and re-enables the tests that failed. 
Adds some additional tests.
 2017-10-16 07:09:05 -07:00
Max Moiseev
93afb243ca Merge pull request #11917 from natecook1000/nc-uint-advance 
[stdlib] Fix strideable methods for large unsigned values
 2017-10-11 10:08:29 -07:00
Slava Pestov
d866ebf8f4 Merge pull request #11744 from gspiers/bugfix/SR-479-ownership-in-protocols 
[SE implementation] Ownership keyword removal in protocols
 2017-10-10 00:08:57 -07:00
Michael Ilseman
2c8393a78e Merge pull request #12347 from milseman/test_test_test 
[test] Use target-swift-frontend for x-compile tests. NFC
 2017-10-09 16:51:37 -07:00
Jacopo Andrea Giola
bcb9571079 Add fix in Casting.cpp for SR-6032 (#12298) 
Private classes doesn't add extra stuff when passed to String(describing:)
 2017-10-09 14:06:34 -07:00
Michael Ilseman
139293d6f8 [test] Use target-swift-frontend for x-compile tests. NFC 2017-10-09 12:57:51 -07:00
Arnold Schwaighofer
f691657f3e Use an autorelease pool on i386 instead 2017-10-06 14:41:56 -07:00
Arnold Schwaighofer
e015d7153e Change UNSUPPORTED: linux to REQUIRES: objc_interop 2017-10-06 14:12:53 -07:00
Arnold Schwaighofer
a4aad80156 Add a radar for the XFAIL 2017-10-06 14:09:39 -07:00
Arnold Schwaighofer
75a82d7a7c XFAIL i386 the autorelease return optimization seems to be defunct 2017-10-06 13:04:03 -07:00
Arnold Schwaighofer
513303fd63 Also fix and test uppercased and lowercased 2017-10-06 13:04:03 -07:00
Arnold Schwaighofer
ec5f40f12f runtime: Move String implementation stubs that want need the auto-released return value optimization to an ARC compiled file 
String's hashValue function is implemented in terms of Foundation's hash
function in a runtime function on darwin platforms. For non-ASCII strings we
will call str.decomposedStringWithCanonicalMapping inside this runtime function
which will allocate a new NSString and return the result in the current
autorelease pool. We implemented this function in a file compiled without ARC.
This meant that we would leak said NSString into the current active autorelease
pool.
This patch moves the implementation to a file compiled with ARC. ARC will insert
objc_retainAutoreleasedReturnValue call and on platforms that require it an
marker for the hand-off of the autoreleased return value optimization.

SR-4889
rdar://32199117
 2017-10-06 11:12:22 -07:00
Greg Parker
8ca2225383 [test] XFAIL stdlib/Dispatch.swift on all ARM CPUs pending rdar://34751238. (#12283) 
Previous XFAIL attempt was only for 32-bit ARM, but 64-bit ARM is also affected.
 2017-10-04 19:12:40 -07:00
Greg Parker
b6b983a6d4 [test] XFAIL stdlib/Dispatch.swift on 32-bit ARM pending rdar://34751238. (#12259) 
Previous attempt to XFAIL (#12231) missed armv7s.
 2017-10-04 09:42:23 -07:00
swift-ci
efa58c7314 Merge pull request #9458 from linux-on-ibm-z/master-URBTest-fix 2017-10-03 11:32:55 -07:00
Greg Parker
cc188c9b55 [test] XFAIL stdlib/Dispatch.swift on 32-bit ARM pending rdar://34751238. (#12231) 2017-10-02 17:28:30 -07:00
Doug Gregor
772352e524 Add requirement StringProtocol.SubSequence : StringProtocol 2017-10-01 15:08:23 -07:00
Greg Spiers
c5d6e17d85 Do not allow ‘weak’ and ‘unowned’ keywords for property declarations in protocols. 2017-10-01 12:43:48 +01:00
Nate Cook
1acb38aa9e [stdlib] Add more tests for strideable behavior 2017-09-29 22:57:46 -05:00
swift-ci
caffda1b58 Merge pull request #12188 from moiseev/floating-modulo-message 2017-09-29 13:19:03 -07:00
Max Moiseev
6ea3644568 [stdlib] Clarify an availability message for % on floating point types 
<rdar://problem/27855641>
 2017-09-29 12:18:15 -07:00
swift-ci
920e42f51f Merge pull request #12151 from moiseev/ns-substring 2017-09-27 22:58:27 -07:00
Max Moiseev
7322b63c44 [overlay] Fix Foundation extensions to Substring 
When a substring gets bridged to NSString, it loses the initial offset,
therefore APIs that accept or return StringIndex ranges should handle
this case explicitly by adding/subtracting the substring start offset.

Fixes <rdar://problem/33873277>
 2017-09-27 20:09:01 -07:00
Karoy Lorentey
ed5b202c8f [stdlib] Implement -retainCount in _SwiftNativeNS*Base 
Fixes rdar://problem/28002554.
 2017-09-26 14:32:20 -07:00
Slava Pestov
a1ccab8a8d Merge pull request #12052 from slavapestov/remove-parent-field-from-type-metadata 
Remove parent field from type metadata
 2017-09-25 17:14:49 -07:00
Lance Parker
0787a4ec48 Merge pull request #12061 from lancep/CVarArgBoolConformance 
[stdlib] Bool should conform to the CVarArg protocol
 2017-09-25 16:06:37 -07:00
Slava Pestov
9bdb71b614 Runtime: Stop looking at parent type metadata in _buildDemanglingForNominalType() 
Wean the routine that builds a mangling tree from metadata from
looking at the parent metadata pointer, instead using the nested
depth to partition the arguments into a list of arguments for
each nesting depth.
 2017-09-25 15:45:16 -07:00
Lance Parker
07b9232235 Correctly handle big endian systems, updated the comment for va_arg usage 2017-09-25 12:38:18 -07:00
Max Moiseev
962895b652 Merge pull request #11965 from moiseev/arkit+metal 
[overlay] Add Metal, MetalKit overlays, and new APIs to ARKit
 2017-09-22 11:44:18 -07:00
Saleem Abdulrasool
572bd6f686 Merge pull request #12056 from compnerd/aarch64-test 
test: duplicate some conditions for AArch64
 2017-09-22 10:18:16 -07:00
Lance Parker
6056ebede1 Add tests for Bool’s conformance to CVarArg 2017-09-22 09:53:47 -07:00
Saleem Abdulrasool
99db384e2e test: duplicate some conditions for AArch64 
The Linux target has a target-cpu value of AArch64 rather than arm64.
Adjust the tests to account for the difference.
 2017-09-21 23:48:28 -07:00
ktopley-apple
0dc62017a1 Merge pull request #11927 from ktopley-apple/dispatch-time-overflows 
Fix overflow traps in DispatchTime/DispatchWallTime/DispatchTimeInterval
 2017-09-20 13:27:35 -07:00
Lance Parker
276f49af84 Merge pull request #11982 from lancep/comprehensiveArrayTests 
[stdlib] Comprehensive array tests
 2017-09-20 13:01:59 -07:00
Kim Topley
c5af2799c1 Fix overflow traps in DispatchTime/DispatchWallTime/DispatchTimeInterval. 
rdar://problem/32678302
 2017-09-20 12:16:05 -07:00