platform/upstream/dotnet/runtime.git
4 years agoAdd support for ComWrappers-based RCWs to the special native WeakReference support...
Jeremy Koritzinsky [Fri, 8 May 2020 02:20:23 +0000 (19:20 -0700)]
Add support for ComWrappers-based RCWs to the special native WeakReference support. (#35819)

* Add support for ComWrappers-based RCWs to the special native WeakReference support.

Rename the WinRT weak reference handle to "native COM weak reference" handle.

* Fix line endings.

* Fix line endings (try 2).

* Revert change to prebuilt idl

* Try updating prebuilt again.

* Remove accidental duplicate of the test.

* PR feedback.

* PR Feedback.

* React to global ComWrappers changes.

* Add back WinRT enum member to prebuilt idl.

* Add back old enum member to idl for backcompat.

* Change definition of enum member to explicitly assign the same value.

* Code cleanup and go down a non-allocating route when possible.

* Switch to preemptive mode for the QI call.

* Fix contracts

* Add a check before calling GetComWeakReference so we only call it when the object has interop info attached.

* Apply early check to WeakReference<T> as well.

* Make sure we only make one call to PassiveGetSyncBlock instead of 2.

* PR Feedback.

* ComWrappersNative::GetIdentityForObject can trigger GC since we transition to and from pre-emptive when calling into the external QI.

4 years agoDelete task code that has been moved to the shared framework SDK in Arcade (#35405)
Chaojie [Fri, 8 May 2020 01:45:21 +0000 (09:45 +0800)]
Delete task code that has been moved to the shared framework SDK in Arcade (#35405)

* Delete task code that has been moved to the shared framework SDK in Arcade

* Remove RegenerateReadmeTable.cs

* Revert "Remove RegenerateReadmeTable.cs"

This reverts commit 037f425853e5587da95404f07130e36efd6cb2e3.

* BuildFPMToolPreReqs.cs is still needed for packaging Linux_x64

4 years agoRemove unnecessary recursive locking from Http2Stream.Complete (#36030)
Stephen Toub [Fri, 8 May 2020 00:58:18 +0000 (20:58 -0400)]
Remove unnecessary recursive locking from Http2Stream.Complete (#36030)

We just asserted we're holding the lock.  We don't need to take it again.

4 years agoAdd staticapphost template entry to signexclusion file (#36043)
Santiago Fernandez Madero [Fri, 8 May 2020 00:30:07 +0000 (17:30 -0700)]
Add staticapphost template entry to signexclusion file (#36043)

4 years agoMerge pull request #36032 from NikolaMilosavljevic/clean.publish
Nikola Milosavljevic [Fri, 8 May 2020 00:22:03 +0000 (17:22 -0700)]
Merge pull request #36032 from NikolaMilosavljevic/clean.publish

Do not package singlefilehost.exe in shared runtime

4 years agoCompile methods in canon form only (#36011)
David Wrighton [Fri, 8 May 2020 00:11:28 +0000 (17:11 -0700)]
Compile methods in canon form only  (#36011)

* Compile methods in canon form only
- The previous logic would compile methods that were not in canonical form
- These methods would never be used by the runtime, and contributed 70MB to a 170MB composite image

- Also fix incorrect intrinsics count value

4 years agoMake HttpResponseMessage.Content non-nullable (#35910)
Stephen Toub [Fri, 8 May 2020 00:02:37 +0000 (20:02 -0400)]
Make HttpResponseMessage.Content non-nullable (#35910)

* Make HttpResponseMessage.Content non-nullable

* Fix Microsoft.Extensions.Http test

4 years agoTranslate all ObjectDisposedExceptions to SocketExceptions in MultipleConnectAsync...
Stephen Toub [Fri, 8 May 2020 00:02:08 +0000 (20:02 -0400)]
Translate all ObjectDisposedExceptions to SocketExceptions in MultipleConnectAsync.AttemptConnection (#35965)

If ObjectDisposedExceptions occurred at unexpected times, we'd return that ObjectDisposedException rather than a SocketException for OperationAborted.  Higher-levels would then treat that arbitrary exception as a generic SocketError. So net net, this means that consuming code would get a SocketException for the more generic SocketError.SocketError rather than the more specific SocketError.OperationAborted.

4 years agoUpdate dependencies from https://github.com/dotnet/xharness build 20200506.2 (#35963)
dotnet-maestro[bot] [Thu, 7 May 2020 23:39:04 +0000 (23:39 +0000)]
Update dependencies from https://github.com/dotnet/xharness build 20200506.2 (#35963)

- Microsoft.DotNet.XHarness.Tests.Runners: 1.0.0-prerelease.20255.4 -> 1.0.0-prerelease.20256.2

Co-authored-by: dotnet-maestro[bot] <dotnet-maestro[bot]@users.noreply.github.com>
4 years agoDo not package singlefilehost.exe in shared runtime
Nikola Milosavljevic (CLR) [Thu, 7 May 2020 21:36:44 +0000 (14:36 -0700)]
Do not package singlefilehost.exe in shared runtime

4 years agoadd requirement doc for FreeBSD (#35704)
Tomas Weinfurt [Thu, 7 May 2020 20:47:41 +0000 (13:47 -0700)]
add requirement doc for FreeBSD (#35704)

* add freebsd requirements

* corrections

* format

* feedback from review

* more cleanup

Co-authored-by: Tomas Weinfurt <furt@Shining.local>
4 years agoimprove handling of handshake failure (#35549)
Tomas Weinfurt [Thu, 7 May 2020 20:46:43 +0000 (13:46 -0700)]
improve handling of handshake failure (#35549)

* improve handling of handshake failure

* more cleanup

* remove console reference

* add quotes around parameter

* enable ServerAsyncAuthenticate_MismatchProtocols_Fails

* cleanup ssl2 test

* fix http2

* feedback from review

* add back two missing empty lines

Co-authored-by: Tomas Weinfurt <furt@Shining.local>
4 years agoRemove TargetFramework reads in eng/testing (#35954)
Viktor Hofer [Thu, 7 May 2020 20:43:49 +0000 (22:43 +0200)]
Remove TargetFramework reads in eng/testing (#35954)

4 years agoMove tfmsuffix file to fix DesignTimeBuild (#35969)
Viktor Hofer [Thu, 7 May 2020 20:40:53 +0000 (22:40 +0200)]
Move tfmsuffix file to fix DesignTimeBuild (#35969)

4 years agoAdding Linux support to System.DirectoryServices.Protocols (#35380)
Jose Perez Rodriguez [Thu, 7 May 2020 20:37:38 +0000 (13:37 -0700)]
Adding Linux support to System.DirectoryServices.Protocols (#35380)

4 years agoUse type system entities instead of strings for node identity (#35918)
Simon Nattress [Thu, 7 May 2020 20:26:49 +0000 (13:26 -0700)]
Use type system entities instead of strings for node identity (#35918)

The calling method for interface dispatches is used to provide imports specific to each call site (actually currently a shared one per calling method). Replace the use of mangled calling method name with its `MethodDesc` object as the key for the interface dispatch cell.

This saves ~4% working set measured on a large application.

Also, removed mangled name as the key for method-specific readonly data blobs the JIT allocates. This was much more modest since most methods don't need a readonly blob.

4 years agoDelete FEATURE_CORRUPTING_EXCEPTIONS (#35938)
Jan Kotas [Thu, 7 May 2020 18:58:16 +0000 (11:58 -0700)]
Delete FEATURE_CORRUPTING_EXCEPTIONS (#35938)

* Delete HCORMODULE

* Delete FEATURE_CORRUPTING_EXCEPTIONS

Ability to catch corrupting exceptions was disabled in .NET Core since early days for security reasons. This is still disabled after this change, but achieved in much simpler way without the full corrupting exceptions machinery.

* Delete unused bSecurityCheck argument of GetComIPFromObjectRef

4 years agoRe-enable default interface methods tests on ARM64 (#35889)
Michal Strehovský [Thu, 7 May 2020 18:35:26 +0000 (20:35 +0200)]
Re-enable default interface methods tests on ARM64 (#35889)

These are all disabled on an unrelated bug. I couldn't dig out how this happened. If they are failing, I would want to see a high priority bug to look at them.

For some reason tiering-related tests were also disabled on the mystery bug.

4 years agoChange epoll thread count (#35800)
Adam Sitnik [Thu, 7 May 2020 18:29:52 +0000 (20:29 +0200)]
Change epoll thread count (#35800)

* use struct wrapper for better perf

* check the most common case first

* don't access static variable in a loop

* use Span instead of raw pointers

* change the heuristic, single epoll thread is not always enough

* simplify the heuristic and add a comment

* apply the naming suggestions

4 years ago[Arm64] ASIMD Rename AddReturningHighNarrow to AddHighNarrow (#35919)
Egor Chesakov [Thu, 7 May 2020 17:30:27 +0000 (10:30 -0700)]
[Arm64] ASIMD Rename AddReturningHighNarrow to AddHighNarrow (#35919)

* Rename AddReturning{Rounded}HighNarrow to Add{Rounded}HighNarrow in AdvSimd.PlatformNotSupported.cs AdvSimd.cs

* Rename SubtractReturning{Rounded}HighNarrow to Subtract{Rounded}HighNarrow in AdvSimd.PlatformNotSupported.cs AdvSimd.cs

* Rename Add/SubtractReturning{Rounded}HighNarrow to Add/Subtract{Rounded}HighNarrow in hwintrinsiclistarm64.h

* Update System.Runtime.Intrinsics.Experimental.cs

* Rename Add/SubtractReturning{Rounded}HighNarrow to Add/Subtract{Rounded}HighNarrow in GenerateTests.csx

* Update Helpers.cs Helpers.tt

* Update AdvSimd/

4 years agoFix the stackoverflow test output checks (#35914)
Jan Vorlicek [Thu, 7 May 2020 17:26:42 +0000 (19:26 +0200)]
Fix the stackoverflow test output checks (#35914)

* Fix the stackoverflow test output checks

The System.Threading.ThreadHelper.ThreadStart can tail call the
System.Threading.ExecutionContext.RunInternal (and some other methods in
the runtime as well) and thus it would not be visible on the stack
trace. So the fix is to not to look at the System.Threading.ThreadHelper.ThreadStart
in the stack trace and use a method in the test itself instead. Since
the test is compiled with optimizations disabled, JIT should not do any
"interesting" things.

* Add NoInlining attribute and do a little unification

4 years agoUse C++11 alias templates instead of macro-defined types to implement specialized...
Jeremy Koritzinsky [Thu, 7 May 2020 17:24:20 +0000 (10:24 -0700)]
Use C++11 alias templates instead of macro-defined types to implement specialized holder wrappers. (#35863)

4 years agoFix crashes when tracing enabled and Socket disposed during connect (#35964)
Stephen Toub [Thu, 7 May 2020 16:38:33 +0000 (12:38 -0400)]
Fix crashes when tracing enabled and Socket disposed during connect (#35964)

Given the right (wrong) sequence of events, if tracing is enabled and a Socket is disposed of during a connect operation, the tracing code might try to access properties on the Socket that throw ObjectDisposedException, and these exceptions may propagate and crash the process.

4 years agoFix broken links on Botr doc (#35973)
Matt Warren [Thu, 7 May 2020 16:38:13 +0000 (17:38 +0100)]
Fix broken links on Botr doc (#35973)

Sorry, I messed these up in my previous PR

4 years agoRemove superflous CheckReleased call which will slow down Event Log Reading up to...
Alois-xx [Thu, 7 May 2020 16:31:56 +0000 (18:31 +0200)]
Remove superflous CheckReleased call which will slow down Event Log Reading up to 20x times. (#35911)

Co-authored-by: Alois Kraus <alois.kraus@siemens-healthineers.com>
4 years agoFix a potential memory leak due to EventPipe buffer allocation (#35924)
Sung Yoon Whang [Thu, 7 May 2020 16:02:02 +0000 (09:02 -0700)]
Fix a potential memory leak due to EventPipe buffer allocation (#35924)

* simplest fix

* make the buffer size page-aligned

* CR feedback

* Add comment about memset

4 years agoRemove FromAsync usage from UDPClient (#35719)
Stephen Toub [Thu, 7 May 2020 15:26:51 +0000 (11:26 -0400)]
Remove FromAsync usage from UDPClient (#35719)

4 years ago[interp] Remove unnecessary code (#35952)
Vlad Brezae [Thu, 7 May 2020 15:22:33 +0000 (18:22 +0300)]
[interp] Remove unnecessary code (#35952)

param_count is always 1 and it doesn't need to be computed at runtime.

4 years agoFix buffer overflow in assert (#35902)
Carol Eidt [Thu, 7 May 2020 15:12:26 +0000 (08:12 -0700)]
Fix buffer overflow in assert (#35902)

4 years agoSkip processing async operations from epoll thread (#35940)
Koundinya Veluri [Thu, 7 May 2020 14:27:08 +0000 (10:27 -0400)]
Skip processing async operations from epoll thread (#35940)

Fixes https://github.com/dotnet/runtime/pull/35330#discussion_r420749150 by skipping state transitions when an async operation needs to be processed.

4 years agoFix build issue w/ libs.tests subset on mono (#35905)
Steve Pfister [Thu, 7 May 2020 14:11:42 +0000 (10:11 -0400)]
Fix build issue w/ libs.tests subset on mono (#35905)

Resolves local issue building mono+lib+libs.tests where we didn't copy hostpolicy/libhostpolicy into testhost/shared.

4 years ago[master] Update dependencies from 3 repositories (#35890)
dotnet-maestro[bot] [Thu, 7 May 2020 14:00:09 +0000 (14:00 +0000)]
[master] Update dependencies from 3 repositories (#35890)

* Update dependencies from https://github.com/mono/linker build 20200506.1

- Microsoft.NET.ILLink.Tasks: 5.0.0-preview.3.20254.2 -> 5.0.0-preview.3.20256.1

* Update dependencies from https://github.com/dotnet/runtime-assets build 20200506.1

- System.ComponentModel.TypeConverter.TestData: 5.0.0-beta.20251.1 -> 5.0.0-beta.20256.1
- System.Drawing.Common.TestData: 5.0.0-beta.20251.1 -> 5.0.0-beta.20256.1
- System.IO.Compression.TestData: 5.0.0-beta.20251.1 -> 5.0.0-beta.20256.1
- System.IO.Packaging.TestData: 5.0.0-beta.20251.1 -> 5.0.0-beta.20256.1
- System.Net.TestData: 5.0.0-beta.20251.1 -> 5.0.0-beta.20256.1
- System.Private.Runtime.UnicodeData: 5.0.0-beta.20251.1 -> 5.0.0-beta.20256.1
- System.Security.Cryptography.X509Certificates.TestData: 5.0.0-beta.20251.1 -> 5.0.0-beta.20256.1
- System.Windows.Extensions.TestData: 5.0.0-beta.20251.1 -> 5.0.0-beta.20256.1

* Update dependencies from https://github.com/dotnet/xharness build 20200505.4

- Microsoft.DotNet.XHarness.Tests.Runners: 1.0.0-prerelease.20255.2 -> 1.0.0-prerelease.20255.4

* Update dependencies from https://github.com/mono/linker build 20200506.5

- Microsoft.NET.ILLink.Tasks: 5.0.0-preview.3.20254.2 -> 5.0.0-preview.3.20256.5

Co-authored-by: dotnet-maestro[bot] <dotnet-maestro[bot]@users.noreply.github.com>
4 years ago[mono] Fix 32bit Android build (#35907)
Egor Bogatov [Thu, 7 May 2020 13:36:10 +0000 (16:36 +0300)]
[mono] Fix 32bit Android build (#35907)

4 years agoImprove HTTP/2 scaling performance (#35694)
Stephen Toub [Thu, 7 May 2020 10:08:20 +0000 (06:08 -0400)]
Improve HTTP/2 scaling performance (#35694)

* Rearrange Http2Connection.SendHeadersAsync to increase scale

Many concurrent streams are currently bottlenecking in SendHeadersAsync.  They:
- take the headers serialization lock
- request a stream credit
- serialize all their headers to a buffer stored on the connection
- take the write lock
- write out the headers
- release the write lock
- release the headers serialization lock

Instead of using a header buffer pooled on the connection, we can instead temporarily grab ArrayPool buffers into which we serialize the headers, which allows us to eliminate the headers serialization lock.  With that, we get:
- request a stream credit
- serialize the headers to a pooled buffer
- take the write lock
- write out the headers
- release the write lock

This has a significant impact on the ability for many concurrent streams to scale, as all of the header serialization work happens outside of the lock.

* Use ContinueWith instead of async/await in Ignore/LogExceptions

The current implementation will always invoke the continuation, even in the majority case where there's no failure (and if there is a failure, it's cheaper not to throw it again).  We can avoid that and decrease both overhead and allocation in the common case.

* Remove CopyToAsync wrapping task from SendRequestBodyAsync

* Reduce contention on semaphore locks

We're seeing a lot of contention trying to acquire the monitors inside of semaphore slims, even when the contention is to release the semaphore, which should entail minimal contention (and delay of a release will just cause more contention).  We can use a small interlocked gate to add a fast path.

* Stop clearing ArrayBuffer's pooled byte arrays

We already weren't clearing as the array grew, nor are we clearing in most other places in the library.  It's not clear why we were clearing in this one spot, but the zero'ing is showing up meaningfully in profiles.

* Add several gRPC known headers

* Avoid dictionary lookup in HttpMethod.Normalize

* Avoid allocating strings for known Content-Types

* Combine AcquireWriteLockAsync into StartWriteAsync

There's not a good reason to keep them separate, and StartWriteAsync is AcquireWriteLockAsync's only caller.

* Remove GetCancelableWaiterTask async method

We can achieve the same thing without the extra async method by putting the cleanup logic into the awaiter's GetResult.

* Address PR feedback

4 years agoSimplify build on macOS (#35570)
Cédric Luthi [Thu, 7 May 2020 09:12:29 +0000 (11:12 +0200)]
Simplify build on macOS (#35570)

* Check for OpenSSL (and pkg-config) in the check_prereqs() function
* Simplify OpenSSL installation instructions
* Automatically pick the OpenSSL version installed by Homebrew by exporting the proper PKG_CONFIG_PATH environment variable
* Improve the error message if cmake can't find OpenSSL

The previous instructions were asking the user to add symbolic links inside  /usr/local/lib/ and /usr/local/lib/pkgconfig/ for OpenSSL related files. There is no need for a complicated setup with symbolic links when the PKG_CONFIG_PATH environment variable points to the right pkgconfig, which is now done in build-commons.sh.

Co-Authored-By: Jan Kotas <jkotas@microsoft.com>
Co-Authored-By: Jeremy Barton <jbarton@microsoft.com>
4 years agoReenable unmanagedcallersonly related tests (#35926)
Aaron Robinson [Thu, 7 May 2020 05:10:52 +0000 (22:10 -0700)]
Reenable unmanagedcallersonly related tests (#35926)

* Reenable UnmanagedCallersOnlyAttribute tests

* Reenable Vector2_3_4 tests.

* Reenable MarshalSeqStruct/ReversePInvoke tests.

4 years agousing the new sdk hook (#35799)
Anirudh Agnihotry [Thu, 7 May 2020 05:10:09 +0000 (22:10 -0700)]
using the new sdk hook (#35799)

* using the new sdk home

* moving file to eng

4 years agoAdd app-local support for ICU (#35383)
Santiago Fernandez Madero [Thu, 7 May 2020 01:09:52 +0000 (18:09 -0700)]
Add app-local support for ICU (#35383)

* Add app-local support for ICU

* Fix android build

* Fix Unix distros with an old ICU version

* PR Feedback

* PR Feedback

* Add ICU data validation when initializing

* Fix linux build

4 years agoStop spamming System.Net.Http.Tests output with HTTP/2 skipped tests (#35913)
Stephen Toub [Thu, 7 May 2020 01:02:49 +0000 (21:02 -0400)]
Stop spamming System.Net.Http.Tests output with HTTP/2 skipped tests (#35913)

There are a bunch of tests that don't make sense on HTTP/2, but end up executing because of how our tests are structured.  As a result, the test itself checks whether HTTP/2 is applicable and bails if it isn't.  Currently some of these are bailing by throwing a SkipTestException, which results in a message being output on the console.  Such messages are meant for tests that can be run in some circumstances, e.g. a stress test that needs an environment variable set.  They aren't needed when a test is never going to run.

I also marked the stress tests as outerloop so they don't spam the console when just running innerloop.

4 years agoHave SIMDIntrinsicUpperSave and SIMDIntrinsicUpperRestore set a baseType if one can...
Tanner Gooding [Thu, 7 May 2020 00:54:39 +0000 (17:54 -0700)]
Have SIMDIntrinsicUpperSave and SIMDIntrinsicUpperRestore set a baseType if one can't otherwise be determined (#35899)

4 years agoEnsure we account for intrinsics which have gtType=TYP_SIMD16 but gtSIMDSize=32 ...
Tanner Gooding [Thu, 7 May 2020 00:54:27 +0000 (17:54 -0700)]
Ensure we account for intrinsics which have gtType=TYP_SIMD16 but gtSIMDSize=32 (#35904)

4 years agoDisable JIT.HardwareIntrinsics NotSupported runtime tests on Mono (#35901)
Santiago Fernandez Madero [Wed, 6 May 2020 22:00:17 +0000 (15:00 -0700)]
Disable JIT.HardwareIntrinsics NotSupported runtime tests on Mono (#35901)

* Disable JIT.HardwareIntrinsics NotSupported runtime tests

* Only disable on Mono

4 years agoFix ETWFireEvent macro to emit events correctly (#35737)
Sung Yoon Whang [Wed, 6 May 2020 20:45:39 +0000 (13:45 -0700)]
Fix ETWFireEvent macro to emit events correctly (#35737)

4 years agoReduce allocated object cost of debug info (#35873)
David Wrighton [Wed, 6 May 2020 20:26:53 +0000 (13:26 -0700)]
Reduce allocated object cost of debug info (#35873)

Reduce allocated object cost of debug info, by converting it into its final form as its extracted from the JIT and not as its emitted into the file. Also has a small benefit of increased parallelization as its now created in a parallel portion of the build instead of a sequential portion of the build. Overall size win of 0.5% observed.

4 years agoAdd methods to the compilation graph without a full Entrypoint (#35879)
David Wrighton [Wed, 6 May 2020 20:26:14 +0000 (13:26 -0700)]
Add methods to the compilation graph without a full Entrypoint (#35879)

* Add methods to the compilation graph without a full Entrypoint
- There is quite a lot of content needed to provide an actual entrypoint for a method
- These cases don't actually require the entrypoint, they simple express a desire for the method to be compiled if possible
- This change both reduces the final file size substantially as well as reduces the memory footprint of the compiler

4 years agoRyuJIT: x*2 -> x+x; x*1 -> x (fp) (#33024)
Egor Bogatov [Wed, 6 May 2020 19:00:31 +0000 (22:00 +0300)]
RyuJIT:  x*2 -> x+x; x*1 -> x (fp) (#33024)

* Optimize "x * 2" to "x + x" for floats

* Add tests

* Fix typo

* Address feedback

* Move to morph.cpp, also add "x*1" optimization

* clean up

* clean up

* update op2

* Fix incorrect "fgMakeMultiUse" usage

* update op1

* Address feedback

* Add opts.OptimizationEnabled()

4 years agoFix Linux x86 build error related to #35552 (#35892)
Gleb Balykov [Wed, 6 May 2020 18:35:03 +0000 (21:35 +0300)]
Fix Linux x86 build error related to #35552 (#35892)

4 years agoFix GC hole in ReadyToRun code on ARM64 (#35858)
Anton Lapounov [Wed, 6 May 2020 18:24:13 +0000 (11:24 -0700)]
Fix GC hole in ReadyToRun code on ARM64 (#35858)

TransitionBlock.IsRetBuffPassedAsFirstArg must return false for ARM64.

4 years agoAdd basic OpenSSL support to Android (#35893)
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.

4 years agoMoved NoWriteNoSeekStreamContent to WinHttpHandler. (#35898)
Marie Píchová [Wed, 6 May 2020 18:06:19 +0000 (20:06 +0200)]
Moved NoWriteNoSeekStreamContent to WinHttpHandler. (#35898)

4 years agoEnable more formerly-failing-on-mono tests. (#35849)
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.

4 years agoImprove Crossgen2 performance via sorting improvements (#35870)
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

4 years agoMore SuperPmi Tweaks (#35875)
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

4 years agoRe-enable xunit.console code coverage (#35894)
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.

4 years ago[mono] Reintroduce an internal Marshal class method used by the runtime (#34900)
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

4 years ago[mono] Enable XHarness CLI for Android tests (#35786)
Egor Bogatov [Wed, 6 May 2020 12:43:57 +0000 (15:43 +0300)]
[mono] Enable XHarness CLI for Android tests (#35786)

4 years agoFix nullable warnings in HttpStress (#35862)
Stephen Toub [Wed, 6 May 2020 11:55:15 +0000 (07:55 -0400)]
Fix nullable warnings in HttpStress (#35862)

4 years agoRemove duplicate tests from System.Linq.Tests (#35840)
Stephen Toub [Wed, 6 May 2020 11:54:17 +0000 (07:54 -0400)]
Remove duplicate tests from System.Linq.Tests (#35840)

4 years agoRemove duplicate tests from System.IO.FileSystem.Tests (#35842)
Stephen Toub [Wed, 6 May 2020 11:53:42 +0000 (07:53 -0400)]
Remove duplicate tests from System.IO.FileSystem.Tests (#35842)

4 years agoChange netcoreapp5.0 to net5.0 (#35176)
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

4 years agoUpdate arcade from build 20200505.6 (#35871)
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>
4 years ago[interpreter] Fix incorrect arg storage for arm softfp interpreter (#35643)
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

4 years agoFixes to EtwCallback and EtwCallbackCommon: (#35793)
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.

4 years agoRevert "Restore Seperate test builds, but fix broken pipe (#35378)" (#35868)
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

4 years agoAdding a clarifying comment as to why HWIntrinsicInfo::lookupIval returns an inverted...
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)

4 years agoBuild an apphost with hostfxr and hostpolicy linked in (#35368)
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>
4 years agoCreate CI Lane for Mono in Interpreter Mode running the runtime tests (#35648)
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.

4 years agoARM64 intrinsic support for Vector64.Create() and Vector128.Create() (#35590)
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

4 years agoDisable NonDisposedSocket_SafeHandlesCollected test (#35847)
Santiago Fernandez Madero [Tue, 5 May 2020 20:36:55 +0000 (13:36 -0700)]
Disable NonDisposedSocket_SafeHandlesCollected test (#35847)

4 years agoReduce unnecessary Regex match attempts for expressions beginning with atomic loops...
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

4 years agoUnmanagedCallersOnlyAttribute in load_assembly_and_get_function_pointer (#35763)
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.

4 years agoMigrate CBOR classes to System.Formats.Cbor namespace (#35808)
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

4 years agoRemove duplicate tests from System.Linq.Parallel.Tests (#35838)
Stephen Toub [Tue, 5 May 2020 17:22:46 +0000 (13:22 -0400)]
Remove duplicate tests from System.Linq.Parallel.Tests (#35838)

4 years agoDisable preEnumerateTheories for running tests (#35837)
Stephen Toub [Tue, 5 May 2020 16:58:29 +0000 (12:58 -0400)]
Disable preEnumerateTheories for running tests (#35837)

4 years ago[master] Update dependencies from mono/linker dotnet/xharness (#35833)
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>
4 years agoFix duplicate theories warning in S.S.C.Algorithms
Kevin Jones [Tue, 5 May 2020 14:42:20 +0000 (10:42 -0400)]
Fix duplicate theories warning in S.S.C.Algorithms

4 years agoMerge pull request #35813 from monojenkins/sync-pr-19704-from-mono
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

4 years agoremoved unused vars from System.DirectoryServices (#34161)
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

4 years agoUpdate dependencies from https://github.com/dotnet/llvm-project build 20200504.1...
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>
4 years agoSynchronize SR.cs (#35830)
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.

4 years agoEnable test for MemoryBarrierProcessWide on arm64. Update comments. (#35817)
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

4 years agoMove test results into TestResults directory (#35829)
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.

4 years agoPreserve fields of AssemblyLoadContext since its accessed by runtime code through...
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)

4 years agoAdd soft debugger support to s390x (#35781)
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

4 years agoMerge pull request #35794 from janvorli/mark-event-test-unloadability-incompatible
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

4 years agoBundleProbe: Disable a test to keep rolling builds green. (#35828)
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.

4 years agoCompile SERP (#35729)
Simon Nattress [Tue, 5 May 2020 05:55:27 +0000 (22:55 -0700)]
Compile SERP (#35729)

Add R2RTest support for compiling specific larger applications.

4 years agoFix error with superpmi.py if `-output_mch_path` not used (#35822)
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)

4 years agoDon't retype struct as primitive types in import. (#33225)
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.

4 years agoAddress some new Roslyn nullability warnings (#35782)
Levi Broderick [Tue, 5 May 2020 04:59:46 +0000 (21:59 -0700)]
Address some new Roslyn nullability warnings (#35782)

4 years agoFix a couple of Windows createdump issues (#35820)
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.

4 years agoSeparate registration for global ComWrappers instance for tracker support and marshal...
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)

4 years agoBundleProbe: Disable a test to keep rolling builds green. (#35811)
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.

4 years agoReduce usage of Win7 and Win8 queues, move to CI and add CI runs in Win arm/arm64...
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

4 years agoMove FreeBSD builds to DncEng Pool provider builds (#35537)
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.

4 years agoAdding basic support for recognizing and handling SIMD intrinsics as HW intrinsics...
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

4 years agofix a regression introduced by Refactor handling of non-SOH generations (#1688)
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.

4 years agoRemove dead field in Uri (#33044)
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

4 years agoDisable failing stack overflow test on mono. (#35812)
Nathan Ricci [Mon, 4 May 2020 22:18:45 +0000 (18:18 -0400)]
Disable failing stack overflow test on mono. (#35812)

4 years agoCultureInfo and CultureData use OS information to initialize culture routines on...
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