Egor Bogatov [Wed, 6 May 2020 18:13:33 +0000 (21:13 +0300)]
Add basic OpenSSL support to Android (#35893)
Temp solution to enable `System.Security.Cryptography.Native` for Android while we figure out how to build/consume it properly.
#### Steps to enable:
1) Download and unzip `https://maven.google.com/com/android/ndk/thirdparty/openssl/1.1.1g-alpha-1/openssl-1.1.1g-alpha-1.aar`
2) Set these env variables:
```
export AndroidOpenSslHeaders="/Users/egorbo/prj/openssl-1.1.1g-alpha-1.aar/prefab/modules/ssl/include"
export AndroidOpenSslCryptoLib="/Users/egorbo/prj/openssl-1.1.1g-alpha-1.aar/prefab/modules/crypto/libs/android.x86_64/libcrypto.so"
export AndroidOpenSslLib="/Users/egorbo/prj/openssl-1.1.1g-alpha-1.aar/prefab/modules/ssl/libs/android.x86_64/libssl.so"
```
(make sure you use the right ABI in the paths, e.g. `android.x86_64` for `-arch x64` or `android.arm64-v8a` for `-arch arm64`)
3) build repo normally, e.g.
```
./build.sh -os Android -subset Mono+Libs
```
I am preparing a PR to update the docs and will include these temp hints.
Marie Píchová [Wed, 6 May 2020 18:06:19 +0000 (20:06 +0200)]
Moved NoWriteNoSeekStreamContent to WinHttpHandler. (#35898)
imhameed [Wed, 6 May 2020 17:49:13 +0000 (10:49 -0700)]
Enable more formerly-failing-on-mono tests. (#35849)
See:
- https://github.com/dotnet/runtime/issues/34387,
- https://github.com/dotnet/runtime/issues/34384, and
- https://github.com/dotnet/runtime/issues/34386.
David Wrighton [Wed, 6 May 2020 17:35:48 +0000 (10:35 -0700)]
Improve Crossgen2 performance via sorting improvements (#35870)
* Improve Crossgen2 performance
- Crossgen2 makes extensive use of sorting to achieve determinism
- Replace most signifcant sort operations with a parallel mergesort
- These sorts use rather expensive comparison functions and as such the standard sort algorithm which is tuned more to reduce memory usage than reduce comparisons is not ideal, and merge sort works better
- Primarily sorting is performed during the previously single threaded portions of the product. Adding parallelism yields easy wins
- In a large benchmark overall compilation performance improved from 106 seconds to 82 seconds
- Implementation makes extensive use of generics to share the implementation across the various apis that are used for sorting in crossgen2
- Add Detection on of duplicates in sort algorithm
- Perf result compiling a large set of assemblies
- Using standard sorts 121.6 seconds
- Using parallel merge sort 96.9 seconds
Carol Eidt [Wed, 6 May 2020 17:31:29 +0000 (10:31 -0700)]
More SuperPmi Tweaks (#35875)
* Add more queries for SuperPmi
* Support crossgen collection
Viktor Hofer [Wed, 6 May 2020 17:03:55 +0000 (19:03 +0200)]
Re-enable xunit.console code coverage (#35894)
Re-enable xunit console code coverage measurement until
https://github.com/coverlet-coverage/coverlet/issues/834 is fixed.
Alexis Christoforides [Wed, 6 May 2020 15:14:06 +0000 (11:14 -0400)]
[mono] Reintroduce an internal Marshal class method used by the runtime (#34900)
* [mono] Reintroduce an internal Marshal method used by the runtime
* Reenable test that is failing for a different reason now
See https://github.com/dotnet/runtime/issues/34374
Egor Bogatov [Wed, 6 May 2020 12:43:57 +0000 (15:43 +0300)]
[mono] Enable XHarness CLI for Android tests (#35786)
Stephen Toub [Wed, 6 May 2020 11:55:15 +0000 (07:55 -0400)]
Fix nullable warnings in HttpStress (#35862)
Stephen Toub [Wed, 6 May 2020 11:54:17 +0000 (07:54 -0400)]
Remove duplicate tests from System.Linq.Tests (#35840)
Stephen Toub [Wed, 6 May 2020 11:53:42 +0000 (07:53 -0400)]
Remove duplicate tests from System.IO.FileSystem.Tests (#35842)
Viktor Hofer [Wed, 6 May 2020 11:53:21 +0000 (13:53 +0200)]
Change netcoreapp5.0 to net5.0 (#35176)
* Change netcoreapp5.0 to net5.0
* Remove tfm downgrades
* Rename S.R.CS.Unsfae include folder
Santiago Fernandez Madero [Wed, 6 May 2020 10:30:53 +0000 (03:30 -0700)]
Update arcade from build
20200505.6 (#35871)
* Update arcade from build
20200505.6
* Fix package testing, settings.targets where not found due to updated tfm
* Stop invoking test target if SkipTests is false in installer
Co-authored-by: Viktor Hofer <viktor.hofer@microsoft.com>
monojenkins [Wed, 6 May 2020 09:53:12 +0000 (05:53 -0400)]
[interpreter] Fix incorrect arg storage for arm softfp interpreter (#35643)
In get_call_info, IS_HARD_FLOAT is checked to switch between using the FP registers and the GP registers. This was not done in arg_get_storage, which was always using the hard float registers. This would mean parameters to the native trampoline wouldn't get copied correctly, and return values would be incorrect as well.
I can't find a way to add this to tests, since no softfp test systems seem to be built. I'm working on testing the use cases I can think of on my softfp system just to be sure.
Any assignments to type RegTypeHFA were locked behind a IS_HARD_FLOAT check as well, so I think its safe to modify that case too.
Fixes mono/mono#14591
Peter Sollich [Wed, 6 May 2020 09:37:03 +0000 (02:37 -0700)]
Fixes to EtwCallback and EtwCallbackCommon: (#35793)
* Fixes to EtwCallback and EtwCallbackCommon:
- Windows only: only call GCHeapUtilities::RecordEventStateChange for EVENT_CONTROL_CODE_ENABLE_PROVIDER or EVENT_CONTROL_CODE_DISABLE_PROVIDER, *not* for EVENT_CONTROL_CODE_CAPTURE_STATE.
- Only call GCHeapUtilities::RecordEventStateChange for DotNETRuntime and DotNETRuntimePrivate providers
- Windows only: Consolidate level and keywords between EventPipe and ETW, so that disabling an event on one side does not disable it for the other side, if the other side is still interested in it.
Steve MacLean [Wed, 6 May 2020 05:47:14 +0000 (01:47 -0400)]
Revert "Restore Seperate test builds, but fix broken pipe (#35378)" (#35868)
* Revert "Restore Seperate test builds, but fix broken pipe (#35378)"
This reverts commit
6d64040a3a577e097e8d06ef72c10f700a17bf9c.
* Add issues related to CI coverage outage
Tanner Gooding [Wed, 6 May 2020 04:01:42 +0000 (21:01 -0700)]
Adding a clarifying comment as to why HWIntrinsicInfo::lookupIval returns an inverted comparison op (#35867)
Vladimir Sadov [Wed, 6 May 2020 01:05:50 +0000 (18:05 -0700)]
Build an apphost with hostfxr and hostpolicy linked in (#35368)
* hostfxr: Build most of hostfxr as a static library
This is part of the work to create an apphost that bundles both hostfxr
and hostpolicy. The main distinction between the static and shared
versions of hostfxr is that the static version contains a hostpolicy
resolver that references the hostpolicy symbols directly rather than
loading them from a DLL.
* hostpolicy: Build as a static library
This change is part of the work to enable an apphost that bundles both
hostfxr and hostpolicy. There's no distinction between hostpolicy
that's built as a shared library and as a static library: the shared
library is built by linking an empty object file with the static
library.
* corehost: Allow linking of hostfxr and hostpolicy with apphost
Provide a hostfxr_iface class, that abstracts how the hostfxr functions called
by the early stage in the hosting layer is resolved.
* dotnet: Teach the muxer binary about hostfxr_iface
* apphost: Teach apphost about hostfxr_iface
This provides two implementations of hostfxr_iface: one for the static
apphost, which bundles hostfxr and hostpolicy, and another for the
conventional apphost, which loads them dynamically on startup.
* Add exports for hostfxr and policy
* Exports for unix
* EXPORTS_LINKER_OPTION
* use generateversionscript.awk from ENG
* Move fxr files out of static
* Fixes for Linux
* Fix for win-x86
* move HEADERS next to SOURCES similarly to other files.
* PR feedback (simplifying hostpolicy_resolver::try_get_dir for static host)
* Publish static_apphost to Microsoft.NETCore.App.Host
* bind to entry points without probing, when in a static host.
* Add a test case
* renamed hostfxr_iface --> hostfxr_resolver_t
* renamed shared --> standalone
* rename static_apphost --> singlefilehost
* Signing exclusions for singlefilehost
* switched StaticHost tst to a different asset (mostly a copy of StandaloneApp)
Co-authored-by: Leandro Pereira <leandro.pereira@microsoft.com>
Co-authored-by: Swaroop Sridhar <swaroop.sridhar@microsoft.com>
Nathan Ricci [Wed, 6 May 2020 00:57:26 +0000 (20:57 -0400)]
Create CI Lane for Mono in Interpreter Mode running the runtime tests (#35648)
Add CI lane for mono interpreter.
Kunal Pathak [Tue, 5 May 2020 23:44:43 +0000 (16:44 -0700)]
ARM64 intrinsic support for Vector64.Create() and Vector128.Create() (#35590)
* Make Vector64.Create() that takes multiple arguments use ARM64 intrinsic
* Make Vector128.Create() that takes multiple arguments use ARM64 intrinsic
* Intrinsify Vector64.Create() that takes single argument
* Intrinsify Vector64.Create() that takes single argument
* Fix edge case where int.MaxValue was failing if used as immediate
Santiago Fernandez Madero [Tue, 5 May 2020 20:36:55 +0000 (13:36 -0700)]
Disable NonDisposedSocket_SafeHandlesCollected test (#35847)
Stephen Toub [Tue, 5 May 2020 18:22:13 +0000 (14:22 -0400)]
Reduce unnecessary Regex match attempts for expressions beginning with atomic loops (#35824)
* Reduce unnecessary Regex match attempts for expressions beginning with atomic loops
* Fix corner-case overflows in ComputeMinLength
Aaron Robinson [Tue, 5 May 2020 18:03:02 +0000 (11:03 -0700)]
UnmanagedCallersOnlyAttribute in load_assembly_and_get_function_pointer (#35763)
* Add support for UnmanagedCallersOnlyAttribute in the load_assembly_and_get_function_pointer API.
* Handle x86 of UnmanagedCallersOnly in managed GetFunctionPointer() API.
Eirik Tsarpalis [Tue, 5 May 2020 17:55:22 +0000 (18:55 +0100)]
Migrate CBOR classes to System.Formats.Cbor namespace (#35808)
* Migrate CBOR classes to System.Formats.Cbor solution
* move cbor source files back S.Security.Cryptography.Encoding
* address feedback
* remove all diffs from product csproj
* move files to original locations
* make cbor production classes public
Stephen Toub [Tue, 5 May 2020 17:22:46 +0000 (13:22 -0400)]
Remove duplicate tests from System.Linq.Parallel.Tests (#35838)
Stephen Toub [Tue, 5 May 2020 16:58:29 +0000 (12:58 -0400)]
Disable preEnumerateTheories for running tests (#35837)
dotnet-maestro[bot] [Tue, 5 May 2020 15:50:54 +0000 (15:50 +0000)]
[master] Update dependencies from mono/linker dotnet/xharness (#35833)
* Update dependencies from https://github.com/mono/linker build
20200504.2
- Microsoft.NET.ILLink.Tasks: 5.0.0-preview.3.20230.3 -> 5.0.0-preview.3.20254.2
* Update dependencies from https://github.com/dotnet/xharness build
20200505.2
- Microsoft.DotNet.XHarness.Tests.Runners: 1.0.0-prerelease.20254.1 -> 1.0.0-prerelease.20255.2
Co-authored-by: dotnet-maestro[bot] <dotnet-maestro[bot]@users.noreply.github.com>
Kevin Jones [Tue, 5 May 2020 14:42:20 +0000 (10:42 -0400)]
Fix duplicate theories warning in S.S.C.Algorithms
Larry Ewing [Tue, 5 May 2020 14:01:53 +0000 (09:01 -0500)]
Merge pull request #35813 from monojenkins/sync-pr-19704-from-mono
[wasm][debugger] Fix type argument
Stefan Nikolei [Tue, 5 May 2020 13:51:22 +0000 (15:51 +0200)]
removed unused vars from System.DirectoryServices (#34161)
* removed unused vars from System.DirectoryServices
removed unused vars. added _ to unused return values
* use discard in GlobalCatalog instead of deleting
* removed unused var in ConfigSet
dotnet-maestro[bot] [Tue, 5 May 2020 13:37:21 +0000 (09:37 -0400)]
Update dependencies from https://github.com/dotnet/llvm-project build
20200504.1 (#35834)
- runtime.linux-x64.Microsoft.NETCore.Runtime.Mono.LLVM.Tools: 6.0.1-alpha.1.20224.3 -> 6.0.1-alpha.1.20254.1
- runtime.win-x64.Microsoft.NETCore.Runtime.Mono.LLVM.Tools: 6.0.1-alpha.1.20224.3 -> 6.0.1-alpha.1.20254.1
- runtime.win-x64.Microsoft.NETCore.Runtime.Mono.LLVM.Sdk: 6.0.1-alpha.1.20224.3 -> 6.0.1-alpha.1.20254.1
- runtime.osx.10.12-x64.Microsoft.NETCore.Runtime.Mono.LLVM.Tools: 6.0.1-alpha.1.20224.3 -> 6.0.1-alpha.1.20254.1
- runtime.osx.10.12-x64.Microsoft.NETCore.Runtime.Mono.LLVM.Sdk: 6.0.1-alpha.1.20224.3 -> 6.0.1-alpha.1.20254.1
- runtime.linux-x64.Microsoft.NETCore.Runtime.Mono.LLVM.Sdk: 6.0.1-alpha.1.20224.3 -> 6.0.1-alpha.1.20254.1
Co-authored-by: dotnet-maestro[bot] <dotnet-maestro[bot]@users.noreply.github.com>
Michal Strehovský [Tue, 5 May 2020 13:34:14 +0000 (15:34 +0200)]
Synchronize SR.cs (#35830)
Ports .NET Native change https://github.com/dotnet/corert/commit/
0ac83cb8d6f9ab10df616a608fcf6fdfa6eabe2b to the Runtime repo.
Vladimir Sadov [Tue, 5 May 2020 12:50:50 +0000 (05:50 -0700)]
Enable test for MemoryBarrierProcessWide on arm64. Update comments. (#35817)
Linux kernels that we run in the lab should have reliable implementation of MemoryBarrierProcessWide now
Viktor Hofer [Tue, 5 May 2020 11:44:41 +0000 (13:44 +0200)]
Move test results into TestResults directory (#35829)
and avoid unique test results name as VSTest has issues with that.
Disabling capturing blame mode dumps at the moment until we have
cross-plat support for it.
Cleaning the results directory if the code coverage data collector is
used as there's currently not an easy way to distinguish between a new
and an old coverage results file. See tracking issue.
Zoltan Varga [Tue, 5 May 2020 11:33:23 +0000 (07:33 -0400)]
Preserve fields of AssemblyLoadContext since its accessed by runtime code through the MonoManagedAssemblyLoadContext structure. (#35792)
monojenkins [Tue, 5 May 2020 09:04:50 +0000 (05:04 -0400)]
Add soft debugger support to s390x (#35781)
* Add soft debugger support to s390x
- Implement the sdb trampoline
- Big-endian fixes for ppdb debugger:
- Fix utf16 strings
- Fix decoding of ImageDebugDirectory structures
- New implementation of OP_SEQ_POINT
- Enable successful unwind by defining markpoints
- Implement mono_arch_setup_resume_sighandler_ctx
- Implement unwind ops in epilog
* Fix delegate-invoke processing
- Handle up to 10 parameters
- Use sigparam information to correctly "slide" parameters
* Fix floating point negative zero for s390x
- Add a new instruction for use in OP_RxCONST -0
- Check for -0 in constant
* Minor indentation fixes
* Fix atomic add implementation
* Check for additional hardware facilities
Jan Vorlicek [Tue, 5 May 2020 08:31:41 +0000 (10:31 +0200)]
Merge pull request #35794 from janvorli/mark-event-test-unloadability-incompatible
Mark eventsourceerror test as unloadability incompatible
Swaroop Sridhar [Tue, 5 May 2020 06:33:34 +0000 (23:33 -0700)]
BundleProbe: Disable a test to keep rolling builds green. (#35828)
BundleProbe() neagtive tests are failing on Linux-musl-x64-release. #35755
Temporarily disable the tests to keep rolling builds green until the bug is fixed.
Simon Nattress [Tue, 5 May 2020 05:55:27 +0000 (22:55 -0700)]
Compile SERP (#35729)
Add R2RTest support for compiling specific larger applications.
Bruce Forstall [Tue, 5 May 2020 05:46:56 +0000 (22:46 -0700)]
Fix error with superpmi.py if `-output_mch_path` not used (#35822)
Sergey Andreenko [Tue, 5 May 2020 05:18:33 +0000 (22:18 -0700)]
Don't retype struct as primitive types in import. (#33225)
* Add two tests.
* Don't retype struct returns x64 windows.
* Turn on for Windows x64 by default for tests.
* fix a few typos/leftovers.
* Disable for Win64 by default.
There are 2 issues that prevent it from being enbaled by default. They are causing significant asm regressions.
Levi Broderick [Tue, 5 May 2020 04:59:46 +0000 (21:59 -0700)]
Address some new Roslyn nullability warnings (#35782)
Mike McLaughlin [Tue, 5 May 2020 04:00:10 +0000 (21:00 -0700)]
Fix a couple of Windows createdump issues (#35820)
* Fix ERROR_PARTIAL_COPY retry. MiniDumpWriteDump sets the last error to the HRESULT instead of win32 error code
* Fix createdump crash triggering on x86 runtime. The initialization code
was in the wrong place and not being called on the x86 coreclr.dll. Moved
it to early in EEStartupHelper.
Elinor Fung [Tue, 5 May 2020 01:57:40 +0000 (18:57 -0700)]
Separate registration for global ComWrappers instance for tracker support and marshalling (#35681)
Swaroop Sridhar [Tue, 5 May 2020 01:36:10 +0000 (18:36 -0700)]
BundleProbe: Disable a test to keep rolling builds green. (#35811)
BundleProbe("hostpolicy.dll") test is failing on Linux-musl-x64-release.
https://github.com/dotnet/runtime/issues/35755
Temporarily disable the test to keep rolling builds green until the bug is fixed.
Santiago Fernandez Madero [Tue, 5 May 2020 01:34:15 +0000 (18:34 -0700)]
Reduce usage of Win7 and Win8 queues, move to CI and add CI runs in Win arm/arm64 (#35690)
* Reduce usage of Win7 and Win8 queues, move to CI and add CI runs in Win arm/arm62
* Disable tests blocking clean rolling build
* Disable more failing tests
* Disable more failing tests
Matt Galbraith [Tue, 5 May 2020 00:27:31 +0000 (17:27 -0700)]
Move FreeBSD builds to DncEng Pool provider builds (#35537)
* Move FreeBSD builds to DncEng Pool provider builds
* PR feedback
* Update xplat-setup.yml
From reading @wfurt 's commit it looks like osGroup is actually FreeBSD here. FreeBSD_x64 seems to be ignored.
Tanner Gooding [Tue, 5 May 2020 00:01:36 +0000 (17:01 -0700)]
Adding basic support for recognizing and handling SIMD intrinsics as HW intrinsics (#35421)
* Adding basic support for recognizing and handling SIMD intrinsics as HW intrinsics
* Applying formatting patch
* Fixing a preprocessor concatenation for non windows
* Add a default case to workaround a compiler warning on FreeBSD
* Fixing a noway_assert to include GT_HWINTRINSIC
* Fixing some asserts that were being triggered
* Use getSIMDVectorRegisterByteLength
* Applying formatting patch
* Fixing ARM64 to use the actual type size
* Removing the [Intrinsic] attribute from some Vector2/3/4 methods which aren't intrinsic
* Updating SSE/SSE2 CompareGreaterThan and related functions to be table driven
* Fixing the SimdAsHWIntrinsic relational operations to match the GT_SIMD behavior
* Ensure that GT_HWINTRINSIC fixes the type for certain TYP_SIMD8
* Fixing the SimdAsHWIntrinsic Vector<int>.op_Multiply support to match the GT_SIMD behavior
* Fixing the SimdAsHWIntrinsic Vector2/3 Division to match the GT_SIMD behavior
* Porting Abs, Min, and Max to use the SimdAsHWIntrinsic support
* Minor fixups to the SSE2 codepath
* Applying formatting patch
* Fixing a check in lowering
* Mark SimdAsHWIntrinsic nodes so we can lookup the correct handle
* Adding the 3 operand overload for gtNewSimdAsHWIntrinsicNode
* Fixing BuildHWIntrinsic to properly take RMW into account
* Fixing the rationalize handling of GT_HWINTRINSIC to account for SIMD vs non-SIMD nodes
* Fixing the importer to not create SIMD nodes if featureSIMD is disabled
* Fixing the SSE4.2 implementation of CompareLessThan<long>
* Preserve the base type for subtraction/addition operations
* Applying formatting patch
* Responding to PR feedback
* Fixing a copy/paste error under reinterpret cast
* Fixing abs to expect 1 argument
* Adding method comment headers that were missing
* Removing unused table entries from SimdAsHWIntrinsic for Vector2/3/4
* Ensure we catch intrinsics from the Vector static class
* Fixing SSSE3_Abs and AVX2_Abs to get the base type from the first argument
* Ensure we adjust the class handle used for intrinsics from the Vector static class
* Ensure we populate the handle cache for clsHnd even if it isn't used
* Fix where we grab the base type from for the static Vector class
* Fixing ConditionalSelect and improving the messages used for impCloneExpr in SimdAsHWIntrinsic
* Ensure we clone the constVectorDup before using it
* Applying formatting patch
Maoni Stephens [Mon, 4 May 2020 23:05:20 +0000 (16:05 -0700)]
fix a regression introduced by Refactor handling of non-SOH generations (#1688)
this is supposed to get the user allocations, before #1688 it was getting gen0 and gen3. with the refactor it should get gen0 and all UOH allocations.
Miha Zupan [Mon, 4 May 2020 22:47:24 +0000 (00:47 +0200)]
Remove dead field in Uri (#33044)
* Remove dead _dnsSafeHost field on Uri
* Make Uri._iriParsing a property
* Simplify Uri OriginalString getter
* Simplify Uri.ToString()
* Remove duplicate IriParsing checks
* Remove UseOrigUncdStrOffset flag
* Cache IriParsing lookup to a local in CheckCanonical
* Update comments referring to m_ fields
* Remove more redundant checks
* Sort usings
* Remove unnecesarry AggressiveInlining attribute
Nathan Ricci [Mon, 4 May 2020 22:18:45 +0000 (18:18 -0400)]
Disable failing stack overflow test on mono. (#35812)
Santiago Fernandez Madero [Mon, 4 May 2020 22:17:14 +0000 (15:17 -0700)]
CultureInfo and CultureData use OS information to initialize culture routines on Windows (#35680)
* CultureInfo and CultureData use OS information to initialize culture routines on Windows
* PR Feedback
* Bring back removed assert
* Remove non meaninful asserts
* PR Feedback
Andrew Au [Mon, 4 May 2020 21:26:20 +0000 (14:26 -0700)]
Make corerun exit code more useful (#35684)
Egor Bogatov [Mon, 4 May 2020 21:06:17 +0000 (00:06 +0300)]
Optimize Vector4.Lerp (#35525)
* Optimize Vector4.Lerp
* Implement for Vector2 and Vector4
* Rollback Vector3
lewing [Mon, 4 May 2020 21:02:48 +0000 (21:02 +0000)]
[wasm][debugger] Fix type argument
fix numbers inspection in js
Matt Warren [Mon, 4 May 2020 20:24:51 +0000 (21:24 +0100)]
Fix-up styling to make the doc more readable (#35803)
Also remove unneeded '\' characters
Ganbarukamo41 [Mon, 4 May 2020 17:25:37 +0000 (02:25 +0900)]
Insert missing break (#35713)
Jan Vorlicek [Mon, 4 May 2020 15:14:06 +0000 (17:14 +0200)]
Mark eventsourceerror test as unloadability incompatible
This test was recently added and it has the same problem with
unloadability as the other tests marked as incompatible with
unloadability in https://github.com/dotnet/coreclr/pull/25954.
Anirudh Agnihotry [Mon, 4 May 2020 15:12:23 +0000 (08:12 -0700)]
Remove TargetsNet* from the runtime repo (#35589)
* feedbacl from previous time
* remove TargetsNet* from ItemGroup
* removing targetFramework from .targets
Ryan Lucia [Mon, 4 May 2020 14:56:03 +0000 (10:56 -0400)]
[runtime] Mono ALC embedding: use public types, correct hook signature (#35515)
dotnet-maestro[bot] [Mon, 4 May 2020 14:44:19 +0000 (14:44 +0000)]
Update dependencies from https://github.com/dotnet/xharness build
20200504.1 (#35787)
- Microsoft.DotNet.XHarness.Tests.Runners: 1.0.0-prerelease.20230.6 -> 1.0.0-prerelease.20254.1
Co-authored-by: dotnet-maestro[bot] <dotnet-maestro[bot]@users.noreply.github.com>
Jeremy Koritzinsky [Mon, 4 May 2020 14:13:50 +0000 (07:13 -0700)]
Remove deprecated build scripts. (#35115)
Jan Vorlicek [Mon, 4 May 2020 13:02:40 +0000 (15:02 +0200)]
Merge pull request #35667 from janvorli/fix-stackoverflow-report-at-native
Fix Unix stack overflow trace logging for native failure locations
Viktor Hofer [Mon, 4 May 2020 09:57:51 +0000 (11:57 +0200)]
Enable dotnet test in libraries (#35285)
* Enable dotnet test
* Update docs
* Use vstest in F5 scenarios
* Make dotnet test without framework switch work
* Code cleanup
* Only run code coverage conditionally
* Fix F5 condition for netcoreapp
* Downgrade sdk to check for helix submission failures
* Fix wrong conditions
* Add blame data collector
Viktor Hofer [Mon, 4 May 2020 09:45:15 +0000 (11:45 +0200)]
Update SDK and remove RID calculation in favor of RuntimeInformation (#35538)
* Remove GetTargetMachineInfo from runtime
* Update bootstrapped SDK
* Remove unused prop
* Rename RuntimeRID to HostRuntimeIdentifier
* Remove yml tmp copying
* Update SDK
* Move RuntimeOS into libraries
* Fix RuntimeOS move
Vlad Brezae [Mon, 4 May 2020 09:23:59 +0000 (12:23 +0300)]
[interp] Move label outside of main loop (#35778)
Wasm has control flow limitations which don't go well with gotos. Move label outside of main loop because it translates to very poor wasm code. Other gotos don't seem to negatively impact instruction dispatch code, so leave them alone, for now ..
Vladimir Sadov [Mon, 4 May 2020 00:10:09 +0000 (17:10 -0700)]
Incomplete read ordering in CastCache::Get (#35597)
* Incomplete read ordering in CastCache::Get
* Use LoadBarrier on the managed side as well.
* Rename to ReadMemoryBarrier
* PR feedback in the JIT
Stephen Toub [Sun, 3 May 2020 21:49:22 +0000 (17:49 -0400)]
Fix ManagedWebSocket.CloseAsync cancellation (#35715)
We allow at most one pending receive operation on a web socket, and CloseAsync needs to both send and receive, so if there's a pending receive already, it just reuses / waits for that existing one, which records if it sees a close frame. However, if the close operation is initiated with a cancellation token that's different from the cancellation token the receive operation was initiated with, the close won't respect the supplied cancellation token because it's just waiting on the existing receive. The fix is simply to register with this new cancellation token as well when there's an existing receive that we wait on.
Viktor Hofer [Sun, 3 May 2020 19:28:25 +0000 (21:28 +0200)]
Add error handling in build.sh (#35774)
Also pass help subset if it isn't set to invoke the help message.
Tarek Mahmoud Sayed [Sun, 3 May 2020 19:18:19 +0000 (12:18 -0700)]
Address the System.Diagnostics Feedback (#35582)
* Address the System.Diagnostics Feedback
* Address the feedback of the feedback :-)
* Use lock instead of InterlockedExchange+SpinWait
* Remove System.Linq usage
* More feedback addressing
* Fix compilation
Ben Adams [Sun, 3 May 2020 18:41:25 +0000 (19:41 +0100)]
SpanHelpers use nuint rather than byte* and IntPtr (#35765)
* SpanHelpers use nuint rather than byte* and IntPtr
* uint cast BitOperations
* Cast to uint before casting to nuint
Koundinya Veluri [Sun, 3 May 2020 18:13:19 +0000 (14:13 -0400)]
Parallelize epoll events on thread pool and process events in the same thread (#35330)
* Parallelize epoll events on thread pool and process events in the same thread
* Use interlocked write instead of volatile write
* Upon epoll notification for reads and writes to a socket, queue read work and process write work in same thread
* Readonly for Event struct and rename struct to SocketIOEvent
* Track and speculatively handle epoll events for synchronous operations on the epoll thread
* Prevent event scheduling threads from becoming long-running
* Non-speculatively schedule a work item to process epoll events upon first dequeue, delegating scheduling of more work items to other threads
monojenkins [Sun, 3 May 2020 17:14:33 +0000 (13:14 -0400)]
[Wasm] Additional signatures for SkiaSharp interop (#35595)
This change adds the set of signatures required for SkiaSharp to work untouched using the signatures already provided for the existing platform.
This change should not be required, as it is very specific to a library set. The m2n generator should allow for an external set of cookies, provided by the additional libraries linked by the packager. See https://github.com/mono/mono/issues/19670 for a more detailed issue on a possible solution.
/cc @mattleibow
---
Here's the list of enabled methods by this change:
```
VIIFFII :
System.Void SkiaSharp.SkiaApi::sk_canvas_draw_arc(System.IntPtr,SkiaSharp.SKRect*,System.Single,System.Single,System.Boolean,System.IntPtr)
VIFFFI :
System.Void SkiaSharp.SkiaApi::sk_canvas_draw_circle(System.IntPtr,System.Single,System.Single,System.Single,System.IntPtr)
VIFFFFI :
System.Void SkiaSharp.SkiaApi::sk_canvas_draw_line(System.IntPtr,System.Single,System.Single,System.Single,System.Single,System.IntPtr)
VIFFI :
System.Void SkiaSharp.SkiaApi::sk_canvas_draw_point(System.IntPtr,System.Single,System.Single,System.IntPtr)
VIFI :
System.Void SkiaSharp.SkiaApi::sk_codec_get_scaled_dimensions(System.IntPtr,System.Single,SkiaSharp.SKSizeI*)
IIIIIFII :
System.IntPtr SkiaSharp.SkiaApi::sk_font_break_text(System.IntPtr,System.Void*,System.IntPtr,SkiaSharp.SKTextEncoding,System.Single,System.Single*,System.IntPtr)
FII :
System.Single SkiaSharp.SkiaApi::sk_font_get_metrics(System.IntPtr,SkiaSharp.SKFontMetrics*)
VIIIIF :
System.Void SkiaSharp.SkiaApi::sk_font_get_xpos(System.IntPtr,System.UInt16*,System.Int32,System.Single*,System.Single)
FIIIIII :
System.Single SkiaSharp.SkiaApi::sk_font_measure_text(System.IntPtr,System.Void*,System.IntPtr,SkiaSharp.SKTextEncoding,SkiaSharp.SKRect*,System.IntPtr)
VIIIFFII :
System.Void SkiaSharp.SkiaApi::sk_text_utils_get_path(System.Void*,System.IntPtr,SkiaSharp.SKTextEncoding,System.Single,System.Single,System.IntPtr,System.IntPtr)
IFFFFIIII :
System.IntPtr SkiaSharp.SkiaApi::sk_imagefilter_new_arithmetic(System.Single,System.Single,System.Single,System.Single,System.Boolean,System.IntPtr,System.IntPtr,System.IntPtr)
IIIFFII :
System.IntPtr SkiaSharp.SkiaApi::sk_imagefilter_new_distant_lit_diffuse(SkiaSharp.SKPoint3*,System.UInt32,System.Single,System.Single,System.IntPtr,System.IntPtr)
IIIFFFII :
System.IntPtr SkiaSharp.SkiaApi::sk_imagefilter_new_distant_lit_specular(SkiaSharp.SKPoint3*,System.UInt32,System.Single,System.Single,System.Single,System.IntPtr,System.IntPtr)
IFFFFIIII :
System.IntPtr SkiaSharp.SkiaApi::sk_imagefilter_new_drop_shadow(System.Single,System.Single,System.Single,System.Single,System.UInt32,SkiaSharp.SKDropShadowImageFilterShadowMode,System.IntPtr,System.IntPtr)
IIIFFIIIII :
System.IntPtr SkiaSharp.SkiaApi::sk_imagefilter_new_matrix_convolution(SkiaSharp.SKSizeI*,System.Single*,System.Single,System.Single,SkiaSharp.SKPointI*,SkiaSharp.SKMatrixConvolutionTileMode,System.Boolean,System.IntPtr,System.IntPtr)
IIIFFII :
System.IntPtr SkiaSharp.SkiaApi::sk_imagefilter_new_point_lit_diffuse(SkiaSharp.SKPoint3*,System.UInt32,System.Single,System.Single,System.IntPtr,System.IntPtr)
IIIFFFII :
System.IntPtr SkiaSharp.SkiaApi::sk_imagefilter_new_point_lit_specular(SkiaSharp.SKPoint3*,System.UInt32,System.Single,System.Single,System.Single,System.IntPtr,System.IntPtr)
IIIFFIFFII :
System.IntPtr SkiaSharp.SkiaApi::sk_imagefilter_new_spot_lit_diffuse(SkiaSharp.SKPoint3*,SkiaSharp.SKPoint3*,System.Single,System.Single,System.UInt32,System.Single,System.Single,System.IntPtr,System.IntPtr)
IIIFFIFFFII :
System.IntPtr SkiaSharp.SkiaApi::sk_imagefilter_new_spot_lit_specular(SkiaSharp.SKPoint3*,SkiaSharp.SKPoint3*,System.Single,System.Single,System.UInt32,System.Single,System.Single,System.Single,System.IntPtr,System.IntPtr)
FIFFF :
System.Single SkiaSharp.SkiaApi::sk_3dview_dot_with_normal(System.IntPtr,System.Single,System.Single,System.Single)
VIFFF :
System.Void SkiaSharp.SkiaApi::sk_3dview_translate(System.IntPtr,System.Single,System.Single,System.Single)
VIFFI :
System.Void SkiaSharp.SkiaApi::sk_matrix_map_vector(SkiaSharp.SKMatrix*,System.Single,System.Single,SkiaSharp.SKPoint*)
VIFFI :
System.Void SkiaSharp.SkiaApi::sk_matrix_map_xy(SkiaSharp.SKMatrix*,System.Single,System.Single,SkiaSharp.SKPoint*)
FIII :
System.Single SkiaSharp.SkiaApi::sk_matrix44_get(System.IntPtr,System.Int32,System.Int32)
VIFFF :
System.Void SkiaSharp.SkiaApi::sk_matrix44_post_scale(System.IntPtr,System.Single,System.Single,System.Single)
VIFFF :
System.Void SkiaSharp.SkiaApi::sk_matrix44_post_translate(System.IntPtr,System.Single,System.Single,System.Single)
VIFFF :
System.Void SkiaSharp.SkiaApi::sk_matrix44_pre_scale(System.IntPtr,System.Single,System.Single,System.Single)
VIFFF :
System.Void SkiaSharp.SkiaApi::sk_matrix44_pre_translate(System.IntPtr,System.Single,System.Single,System.Single)
VIIIF :
System.Void SkiaSharp.SkiaApi::sk_matrix44_set(System.IntPtr,System.Int32,System.Int32,System.Single)
VIFFF :
System.Void SkiaSharp.SkiaApi::sk_matrix44_set_scale(System.IntPtr,System.Single,System.Single,System.Single)
VIFFF :
System.Void SkiaSharp.SkiaApi::sk_matrix44_set_translate(System.IntPtr,System.Single,System.Single,System.Single)
IIIIIF :
System.Boolean SkiaSharp.SkiaApi::sk_paint_get_fill_path(System.IntPtr,System.IntPtr,System.IntPtr,SkiaSharp.SKRect*,System.Single)
VIIFF :
System.Void SkiaSharp.SkiaApi::sk_path_add_arc(System.IntPtr,SkiaSharp.SKRect*,System.Single,System.Single)
VIFFFI :
System.Void SkiaSharp.SkiaApi::sk_path_add_circle(System.IntPtr,System.Single,System.Single,System.Single,SkiaSharp.SKPathDirection)
VIFFFIIFF :
System.Void SkiaSharp.SkiaApi::sk_path_arc_to(System.IntPtr,System.Single,System.Single,System.Single,SkiaSharp.SKPathArcSize,SkiaSharp.SKPathDirection,System.Single,System.Single)
VIFFFIIFF :
System.Void SkiaSharp.SkiaApi::sk_path_rarc_to(System.IntPtr,System.Single,System.Single,System.Single,SkiaSharp.SKPathArcSize,SkiaSharp.SKPathDirection,System.Single,System.Single)
IFI :
System.IntPtr SkiaSharp.SkiaApi::sk_path_effect_create_2d_line(System.Single,SkiaSharp.SKMatrix*)
IFFI :
System.IntPtr SkiaSharp.SkiaApi::sk_path_effect_create_discrete(System.Single,System.Single,System.UInt32)
IFFI :
System.IntPtr SkiaSharp.SkiaApi::sk_path_effect_create_trim(System.Single,System.Single,SkiaSharp.SKTrimPathEffectMode)
VIIFFFF :
System.Void SkiaSharp.SkiaApi::sk_rrect_set_nine_patch(System.IntPtr,SkiaSharp.SKRect*,System.Single,System.Single,System.Single,System.Single)
VIIFF :
System.Void SkiaSharp.SkiaApi::sk_rrect_set_rect_xy(System.IntPtr,SkiaSharp.SKRect*,System.Single,System.Single)
IFIII :
System.IntPtr SkiaSharp.SkiaApi::sk_shader_new_lerp(System.Single,System.IntPtr,System.IntPtr,SkiaSharp.SKMatrix*)
IFFIFI :
System.IntPtr SkiaSharp.SkiaApi::sk_shader_new_perlin_noise_fractal_noise(System.Single,System.Single,System.Int32,System.Single,SkiaSharp.SKSizeI*)
IFFIF :
System.IntPtr SkiaSharp.SkiaApi::sk_shader_new_perlin_noise_improved_noise(System.Single,System.Single,System.Int32,System.Single)
IFFIFI :
System.IntPtr SkiaSharp.SkiaApi::sk_shader_new_perlin_noise_turbulence(System.Single,System.Single,System.Int32,System.Single,SkiaSharp.SKSizeI*)
IIFIIIII :
System.IntPtr SkiaSharp.SkiaApi::sk_shader_new_radial_gradient(SkiaSharp.SKPoint*,System.Single,System.UInt32*,System.Single*,System.Int32,SkiaSharp.SKShaderTileMode,SkiaSharp.SKMatrix*)
IIFIIIIII :
System.IntPtr SkiaSharp.SkiaApi::sk_shader_new_radial_gradient_color4f(SkiaSharp.SKPoint*,System.Single,SkiaSharp.SKColorF*,System.IntPtr,System.Single*,System.Int32,SkiaSharp.SKShaderTileMode,SkiaSharp.SKMatrix*)
IIIIIIFFI :
System.IntPtr SkiaSharp.SkiaApi::sk_shader_new_sweep_gradient(SkiaSharp.SKPoint*,System.UInt32*,System.Single*,System.Int32,SkiaSharp.SKShaderTileMode,System.Single,System.Single,SkiaSharp.SKMatrix*)
IIIIIIIFFI :
System.IntPtr SkiaSharp.SkiaApi::sk_shader_new_sweep_gradient_color4f(SkiaSharp.SKPoint*,SkiaSharp.SKColorF*,System.IntPtr,System.Single*,System.Int32,SkiaSharp.SKShaderTileMode,System.Single,System.Single,SkiaSharp.SKMatrix*)
IIFIFIIIII :
System.IntPtr SkiaSharp.SkiaApi::sk_shader_new_two_point_conical_gradient(SkiaSharp.SKPoint*,System.Single,SkiaSharp.SKPoint*,System.Single,System.UInt32*,System.Single*,System.Int32,SkiaSharp.SKShaderTileMode,SkiaSharp.SKMatrix*)
IIFIFIIIIII :
System.IntPtr SkiaSharp.SkiaApi::sk_shader_new_two_point_conical_gradient_color4f(SkiaSharp.SKPoint*,System.Single,SkiaSharp.SKPoint*,System.Single,SkiaSharp.SKColorF*,System.IntPtr,System.Single*,System.Int32,SkiaSharp.SKShaderTileMode,SkiaSharp.SKMatrix*)
VIIIFFII :
System.Void SkiaSharp.SkiaApi::sk_textblob_builder_alloc_run(System.IntPtr,System.IntPtr,System.Int32,System.Single,System.Single,SkiaSharp.SKRect*,SkiaSharp.SKRunBufferInternal*)
VIIIFII :
System.Void SkiaSharp.SkiaApi::sk_textblob_builder_alloc_run_pos_h(System.IntPtr,System.IntPtr,System.Int32,System.Single,SkiaSharp.SKRect*,SkiaSharp.SKRunBufferInternal*)
VIIIFFIII :
System.Void SkiaSharp.SkiaApi::sk_textblob_builder_alloc_run_text(System.IntPtr,System.IntPtr,System.Int32,System.Single,System.Single,System.Int32,SkiaSharp.SKRect*,SkiaSharp.SKRunBufferInternal*)
VIIIFIII :
System.Void SkiaSharp.SkiaApi::sk_textblob_builder_alloc_run_text_pos_h(System.IntPtr,System.IntPtr,System.Int32,System.Single,System.Int32,SkiaSharp.SKRect*,SkiaSharp.SKRunBufferInternal*)
```
Co-authored-by: jeromelaban <jeromelaban@users.noreply.github.com>
dotnet-maestro[bot] [Sun, 3 May 2020 17:11:05 +0000 (13:11 -0400)]
Update dependencies from https://github.com/dotnet/runtime-assets build
20200501.1 (#35772)
- System.ComponentModel.TypeConverter.TestData: 5.0.0-beta.20228.1 -> 5.0.0-beta.20251.1
- System.IO.Compression.TestData: 5.0.0-beta.20228.1 -> 5.0.0-beta.20251.1
- System.IO.Packaging.TestData: 5.0.0-beta.20228.1 -> 5.0.0-beta.20251.1
- System.Net.TestData: 5.0.0-beta.20228.1 -> 5.0.0-beta.20251.1
- System.Private.Runtime.UnicodeData: 5.0.0-beta.20228.1 -> 5.0.0-beta.20251.1
- System.Security.Cryptography.X509Certificates.TestData: 5.0.0-beta.20228.1 -> 5.0.0-beta.20251.1
- System.Windows.Extensions.TestData: 5.0.0-beta.20228.1 -> 5.0.0-beta.20251.1
Co-authored-by: dotnet-maestro[bot] <dotnet-maestro[bot]@users.noreply.github.com>
Hugh Bellamy [Sun, 3 May 2020 14:37:37 +0000 (14:37 +0000)]
Fix some incorrect warnings in LiveBuilds.targets (#35770)
Stephen Toub [Sun, 3 May 2020 13:07:30 +0000 (09:07 -0400)]
Remove FromAsync usage from TcpListener (#35714)
Cijo Thomas [Sun, 3 May 2020 10:33:59 +0000 (03:33 -0700)]
Fix minor typo in comment. (#35768)
Tomáš Rylek [Sun, 3 May 2020 07:59:27 +0000 (09:59 +0200)]
Baseline SimpleSIMDProgram failing with CG2; fix typo in build-test.cmd (#35757)
Jan Kotas [Sun, 3 May 2020 05:50:41 +0000 (22:50 -0700)]
Revert "Add Reverse Diagnostics Server (#33307)" (#35767)
This reverts commit
629dba55058b58a126d08fc37db30b0cfa844dde.
Aaron Robinson [Sun, 3 May 2020 00:39:18 +0000 (17:39 -0700)]
Ensure the function pointer being returned is from an active method. (#35750)
Bruce Forstall [Sat, 2 May 2020 22:58:51 +0000 (15:58 -0700)]
Enable src/JIT/Methodical/explicit/coverage tests on non-Windows (#34250)
Eirik Tsarpalis [Sat, 2 May 2020 21:34:00 +0000 (22:34 +0100)]
[CBOR] Implement Conformance Levels (#35246)
* implement writing sorted keys in maps
* rename NoConformance to NonStrict
* add tests for strict mode map writes
* extract common conformance level logic to helper class
* implement reader map conformance levels
* add missing struct labels
* NonStrict -> Lax
* implement CborWriter indefinite length conformance levels
* implement CborWriter tag conformance levels
* remove preserve float representation check
* implement CborReader integer encoding conformance
* implement CborReader indefinte-length conformance
* implement CborReader tag conformance level validation
* factor frame struct out of checkpoint struct
* use HashSet for reader key uniqueness
* Update src/libraries/System.Security.Cryptography.Encoding/tests/Cbor/CborReader.Integer.cs
Co-Authored-By: Jeremy Barton <jbarton@microsoft.com>
* address tuple label naming consistency
* address PR feedback
* address tuple label style
* implement appropriate rollback semantics for key conformance validation
* rename CborWriter.ToArray() => CborWriter.GetEncoding(); add a span encoding method
* implement patching for indefinite-length items
* minor CborReader string refactoring
* add support for indefinite byte string literals in reader helpers
* Implement CborReader.SkipToParent();
* add a validateConformance parameter to Skip* methods
* implement CborWriter encodeIndefiniteLengths parameter
* make ReadStartMap() and ReadStartArray() return int
* strip IndefiniteLength suffix from write methods
* replace SortedList use with HashSet
* pool key encoding range HashSets
* implement writing and reading multiple root-level values
* I am disappoint
* Update src/libraries/System.Security.Cryptography.Encoding/tests/Cbor/CborConformanceLevel.cs
Co-authored-by: Jeremy Barton <jbarton@microsoft.com>
* Update src/libraries/System.Security.Cryptography.Encoding/tests/Cbor/CborReader.Array.cs
Co-authored-by: Jeremy Barton <jbarton@microsoft.com>
* Update src/libraries/System.Security.Cryptography.Encoding/tests/Cbor/CborReader.Tag.cs
Co-authored-by: Jeremy Barton <jbarton@microsoft.com>
* Update src/libraries/System.Security.Cryptography.Encoding/tests/Cbor/CborReader.cs
Co-authored-by: Jeremy Barton <jbarton@microsoft.com>
* Update src/libraries/System.Security.Cryptography.Encoding/tests/Cbor/CborWriter.Map.cs
Co-authored-by: Jeremy Barton <jbarton@microsoft.com>
* address PR feedback
* clean up code and add comments; remove unneeded synchronization
* rework CborWriter map sorting implementation
* store current major type in a field
Co-authored-by: Jeremy Barton <jbarton@microsoft.com>
Stephen Toub [Sat, 2 May 2020 20:42:39 +0000 (16:42 -0400)]
Improve ContinueWith perf with NotOn* options (#35575)
* Delete dead code from Task.InternalCancel
The bCancelNonExecutingOnly argument was always false.
* Rename StandardTaskContinuation to ContinueWithTaskContinuation
* Remove unused return value from InternalCancel
* Combine two overloads
There's no point in these being separate, with one being small and only and always called from the other.
* Streamline ContinueWith completions for NotOn*
We can avoid unnecessarily instantiating the contingent properties bag, as well as a few interlocked operations.
Santiago Fernandez Madero [Sat, 2 May 2020 19:58:22 +0000 (12:58 -0700)]
Use GDI+1.1 API set in Windows 8 and later (#35169)
* Use GDI+1.1 API set in Windows 8 and later
* PR Feedback, add meaningful test
* Update System.Drawing test package from runtime-assets
* PR Feedback and skip test on full framework
Aaron Robinson [Sat, 2 May 2020 19:40:49 +0000 (12:40 -0700)]
Fix spurious error message in libnethost.lib scenarios (#35739)
1) Empty paths don't exist, so no attribute will be found.
2) Remove string copy.
Stephen Toub [Sat, 2 May 2020 19:24:59 +0000 (15:24 -0400)]
Remove FromAsync usage from System.Net.NetworkInformation (#35710)
Eirik Tsarpalis [Sat, 2 May 2020 18:59:41 +0000 (19:59 +0100)]
store current major type in a field
Eirik Tsarpalis [Sat, 2 May 2020 18:10:47 +0000 (19:10 +0100)]
rework CborWriter map sorting implementation
Tomas Weinfurt [Sat, 2 May 2020 16:09:57 +0000 (09:09 -0700)]
build HTTP on FreeBSD once again (#35742)
Stephen Toub [Sat, 2 May 2020 16:02:25 +0000 (12:02 -0400)]
Remove FromAsync usage from SendPacketsAsync on Unix (#35513)
This call is in the Unix implementation, and the Unix implementation of DisconnectAsync just calls the synchronous Disconnect... so we can just call it directly and skip all the unnecessary layering.
Carol Eidt [Sat, 2 May 2020 15:58:57 +0000 (08:58 -0700)]
Fix ZeroInit of finally vars (#35723)
Finally vars need zero-init'ing if we are enregistering them, but they will not necessarily be live-in to the first block, so we have to take care not to zero-init its last associated register in that case.
Tomas Weinfurt [Sat, 2 May 2020 09:11:44 +0000 (02:11 -0700)]
fix handling of COMPlus_ZapDisable (#35741)
Jan Kotas [Sat, 2 May 2020 07:52:34 +0000 (00:52 -0700)]
Fix CoreCLR incremental builds (#35732)
Layomi Akinrinade [Sat, 2 May 2020 06:18:53 +0000 (02:18 -0400)]
Merge pull request #35700 from layomia/clean_up
Clean up misc areas in JsonSerializer
Anton Lapounov [Sat, 2 May 2020 02:55:58 +0000 (19:55 -0700)]
Support ARM64 HVAs in Crossgen2 (#35576)
• Change FieldLayoutAlgorithm.ComputeValueTypeShapeCharacteristics method to compute the homogeneous aggregate element type and cache it in the existing field. That allows to remove all ComputeHomogeneousFloatAggregateElementType methods.
• Change MetadataFieldLayoutAlgorithm.ComputeHomogeneousAggregateCharacteristic to compute HVAs in addition to HFAs.
• Change CorInfoImpl.getHFAType JIT callback to handle HVAs. Note that returning ELEMENT_TYPE_VALUETYPE indicates the TYP_SIMD16 type (see Compiler::GetHfaType).
• Change TypeFixupSignature.EncodeTypeLayout to handle HVAs.
• Support HVAs in the ArgIterator class.
• Fix TransitionBlock.OffsetFromGCRefMapPos for ARM64. R2RDump used to dump incorrect offsets.
• Use TransitionBlock.OffsetFromGCRefMapPos in GCRefMapBuilder.GetCallRefMap to simplify logic.
• Remove ARM64 .NET Native-specific code from TransitionBlock.cs and ArgIterator.cs files.
Minor:
• Remove a redundant GetVectorSize check in MethodTable::GetHFAType.
• Improve an assertion in Compiler::raUpdateRegStateForArg.
• Fix comments in JIT code.
Layomi Akinrinade [Fri, 1 May 2020 17:21:21 +0000 (10:21 -0700)]
Move collection tests to dedicated class + more clean up
Stephen Toub [Sat, 2 May 2020 00:27:00 +0000 (20:27 -0400)]
Fix test after TcpClient change (#35721)
My change to TcpClient.ConnectAsync resulted in a sporadic test failure. ConnectAsync had been implemented in terms of Begin/EndConnect. It would call BeginConnect, and then the callback invoked when the asynchronous operation completed would call EndConnect. The very first thing EndConnect would do is check whether the TcpClient was disposed, and throw an ObjectDisposedException if it is, ignoring the actual result or failure in the callback. By changing the ConnectAsync implementation to bypass Begin/EndConnect, it's also skipping that check, and so if you dispose of the TcpClient while the operation is in flight, you'll end up with a SocketError.OperationAborted SocketException (exactly what you get from Socket) rather than an ObjectDisposedException.
Since SocketException was already possible from TcpClient.ConnectAsync (and is documented as such), since we want an actual error to propagate rather than hiding it with a different one, since this is about behavior after dispose, and since this is effectively an implementation detail that was leaking out, I've opted to fix the test rather than change the implementation.
Egor Chesakov [Fri, 1 May 2020 23:30:22 +0000 (16:30 -0700)]
[Arm64] Fix JIT/HardwareIntrinsics/Arm/AdvSimd tests (#35726)
* Set LargestVectorSize to 16 where it must be in GenerateTests.csx
* Update AdvSimd/
Eugene Rozenfeld [Fri, 1 May 2020 19:50:55 +0000 (12:50 -0700)]
Remove an unnecessary check from fgMakeOutgoingStructArgCopy. (#35701)
The old-style helper tail calls required the jit to copy implicit by-ref
args. After #341 we are using old-style helper tail calls only for x86,
which doesn't have implicit by-ref parameters. So the check is no longer
necessary.
This is a no-diffs cleanup change.
Stephen Toub [Fri, 1 May 2020 18:56:52 +0000 (14:56 -0400)]
Fix comment on CancellationTokenRegistration.Token (#35708)