platform/upstream/coreclr.git
5 years agoFix LONG64 definition
Adeel [Mon, 25 Feb 2019 20:39:13 +0000 (12:39 -0800)]
Fix LONG64 definition

5 years agoAddress janvorli's feedback comments
Adeel [Mon, 25 Feb 2019 12:40:06 +0000 (04:40 -0800)]
Address janvorli's feedback comments

5 years agoAddress CR feedback
Adeel [Mon, 25 Feb 2019 01:25:41 +0000 (17:25 -0800)]
Address CR feedback

5 years agoSuppress conversion
Sinan Kaya [Sat, 23 Feb 2019 02:32:59 +0000 (02:32 +0000)]
Suppress conversion

Suppress warning during hash

add casting

5 years agoFix comparison and narrowing errors reported by GCC
Adeel [Sun, 24 Feb 2019 20:14:06 +0000 (12:14 -0800)]
Fix comparison and narrowing errors reported by GCC

5 years agoRemoved erroneous use of 'PROCESSOR_ARCHITECTURE' Enviroment variable (#22743)
John [Sat, 23 Feb 2019 17:08:00 +0000 (17:08 +0000)]
Removed erroneous use of 'PROCESSOR_ARCHITECTURE' Enviroment variable (#22743)

Removed erroneous use of 'PROCESSOR_ARCHITECTURE' Enviroment variable

Changed all uses of 'Enviroment.GetEnviromentVariable("PROCESSOR_ARCHITECTURE")'
to usage of 'System.Runtime.InteropServices.RuntimeInformation.ProcessArchitecture'
and 'enum System.Runtime.InteropServices.Architecture'.

Fix #22694

5 years agofix BSWAP16 intrinsic handling of signed values (#22666)
Clinton Ingram [Sat, 23 Feb 2019 17:06:16 +0000 (09:06 -0800)]
fix BSWAP16 intrinsic handling of signed values (#22666)

5 years agoMove AssemblyLoadContext to shared partition (#22685)
Marek Safar [Sat, 23 Feb 2019 10:06:32 +0000 (11:06 +0100)]
Move AssemblyLoadContext to shared partition (#22685)

* Move AssemblyLoadContext to shared partition

* Move static initializer to DefaultAssemblyLoadContext

and remove stream copying from lock scope

5 years agoMake System.Runtime.CompilerServices.Unsafe inbox (dotnet/corefx#35526)
Eric StJohn [Sat, 23 Feb 2019 02:20:16 +0000 (18:20 -0800)]
Make System.Runtime.CompilerServices.Unsafe inbox (dotnet/corefx#35526)

* Make System.Runtime.CompilerServices.Unsafe inbox

Fix a couple places where we wanted to use it from other inbox assemblies but couldn't.

Unsafe remains as a package and newer versions of Unsafe will replace the inbox Version,
permitting us to add API over time.

* Update Rune tests to use TryEncodeToUtf8Bytes directly

* Add temporary suppression to package tests

Signed-off-by: dotnet-bot <dotnet-bot@microsoft.com>
5 years agoAdding a missing single-quote (#22809)
Tanner Gooding [Sat, 23 Feb 2019 03:11:32 +0000 (19:11 -0800)]
Adding a missing single-quote (#22809)

5 years agoRemove 'from HRESULT' (#22782)
Dan Moseley [Sat, 23 Feb 2019 02:57:02 +0000 (18:57 -0800)]
Remove 'from HRESULT' (#22782)

5 years agoChange Environment.Version to return product version (#22664)
Jan Kotas [Sat, 23 Feb 2019 01:56:02 +0000 (17:56 -0800)]
Change Environment.Version to return product version (#22664)

* Change Environment.Version to return product version

- Contributes to https://github.com/dotnet/corefx/issues/31099
- Use AssemblyInformationalVersion attribute as fallback

* Add sanity test for Environment.Version

* Disable CodeDom tests

* Fix test assembly name

5 years agoCorrectly build packages for linux-musl-arm64 (#22793)
Jarret Shook [Sat, 23 Feb 2019 00:01:03 +0000 (16:01 -0800)]
Correctly build packages for linux-musl-arm64 (#22793)

* Correctly build packages for linux-musl-arm64

Also add build-package step to PR.

* Correctly add back if

* Fix build-packages

5 years agoEnable FEATURE_BASICFREEZE (#22776)
Mukul Sabharwal [Fri, 22 Feb 2019 23:56:28 +0000 (15:56 -0800)]
Enable FEATURE_BASICFREEZE (#22776)

5 years agoAdd fatal error details printing (#22789)
Jan Vorlicek [Fri, 22 Feb 2019 22:11:14 +0000 (23:11 +0100)]
Add fatal error details printing (#22789)

Before this change, we were only printing details on fatal errors with
error code COR_E_FAILFAST. For other error codes we were just silently
exiting.
This change enables printing fatal error details for all kinds of fatal
errors.

5 years agoUpdate CoreFX Windows Arm32 exclusion list. (#22695)
Sergey Andreenko [Fri, 22 Feb 2019 21:38:50 +0000 (13:38 -0800)]
Update CoreFX Windows Arm32 exclusion list. (#22695)

* Update CoreFX Windows Arm32 exclusion list.

* Exclude System.Globalization.Calendars.Tests.

* Disable System.Net.Sockets.Tests

5 years agoIgnore EINVAL/EBADF errors from FTruncate in FileStream.Init (#22788)
Stephen Toub [Fri, 22 Feb 2019 21:15:31 +0000 (16:15 -0500)]
Ignore EINVAL/EBADF errors from FTruncate in FileStream.Init (#22788)

They indicate we're working with a special file, where failure to truncate is not a fatal condition.

5 years agoDisable IJW tests on Windows 7. (#22779)
Jeremy Koritzinsky [Fri, 22 Feb 2019 20:39:56 +0000 (12:39 -0800)]
Disable IJW tests on Windows 7. (#22779)

Windows 7 loads desktop mscoree.dll when loading System.Private.CoreLib, before we can load our fake mscoree. So, when IJW modules are loaded, they load the real mscoree.dll, which breaks coreclr's IJW loading. Once we have true .NET Core IJW support in the MSVC++ compiler and .NET Core, then we'll be able to reenable these tests on Win7.

Fixes #21796.

cc: @RussKeldorph @echesakovMSFT

5 years agoR2RDump - Gracefully handle inability to parse GC Info
Simon Nattress [Thu, 21 Feb 2019 21:37:50 +0000 (13:37 -0800)]
R2RDump - Gracefully handle inability to parse GC Info

The Amd64 GC Info parser is not complete in its implementation and
sometimes will index to a location beyond the size of the image.

If we fail to parse the GC info for a method, continue dumping the rest
of the image.

Emit a warning when this happens so we don't forget about this.

5 years agoMore GNUC Fixes (#22687)
Sinan Kaya [Fri, 22 Feb 2019 19:11:19 +0000 (14:11 -0500)]
More GNUC Fixes (#22687)

* Replace __sync_swap with __atomic_exchange_n

__sync_swap() is a clang specific function.

* Remove multiline comment

* Add paranthesis around sum

src/md/hotdata/../inc/streamutil.h:73:34: warning: suggest parentheses around ‘+’ in operand of ‘&’ [-Wparentheses]
     UINT32 aligned = *totalBytes + 3 & ~3;

* Define __int64

* Define windows types for tests

* Remove undefined has_builtin defines and define alloca and inline for GNUC

* Remove __clang__ where possible

* Add implicit casting to help compiler find WCHAR* variant

src/binder/assembly.cpp:294:73: error: no matching function for call to ‘SString::SString(SString)’
         return (pAsmName == nullptr ? nullptr : pAsmName->GetSimpleName());
                                                                         ^
In file included from src/inc/sstring.h:1082:0,
                 from src/inc/ex.h:19,
                 from src/inc/stgpool.h:28,
                 from src/inc/../md/inc/metamodel.h:18,
                 from src/inc/../md/inc/metamodelro.h:19,
                 from src/inc/metadata.h:17,
                 from src/binder/../vm/util.hpp:19,
                 from src/binder/../vm/common.h:110,
                 from src/binder/assembly.cpp:14:
src/inc/sstring.inl:73:8: note: candidate: SString::SString(void*, COUNT_T)
 inline SString::SString(void *buffer, COUNT_T size)
        ^
src/inc/sstring.inl:73:8: note:   candidate expects 2 arguments, 1 provided
src/inc/sstring.inl:436:8: note: candidate: SString::SString(SString::tagLiteral, const WCHAR*, COUNT_T)
 inline SString::SString(tagLiteral dummytag, const WCHAR *literal, COUNT_T count)
        ^
src/inc/sstring.inl:436:8: note:   candidate expects 3 arguments, 1 provided
src/inc/sstring.inl:418:8: note: candidate: SString::SString(SString::tagLiteral, const WCHAR*)
 inline SString::SString(tagLiteral dummytag, const WCHAR *literal)
        ^
src/inc/sstring.inl:418:8: note:   candidate expects 2 arguments, 1 provided
src/inc/sstring.inl:401:8: note: candidate: SString::SString(SString::tagUTF8Literal, const UTF8*)
 inline SString::SString(tagUTF8Literal dummytag, const UTF8 *literal)
        ^
src/inc/sstring.inl:401:8: note:   candidate expects 2 arguments, 1 provided
src/inc/sstring.inl:382:8: note: candidate: SString::SString(SString::tagLiteral, const CHAR*)
 inline SString::SString(tagLiteral dummytag, const ASCII *literal)

* Reorder DLLEXPORT and STDAPI

GNUC wants extern "C" <attribute> format.

* Abstract __FUNCSIG__

* Abstract __debugbreak()

* Move common compiler options out of clang and add Wno-unused-value

* Add paranthesis around || and &&

src/gc/gc.cpp:9084:38: warning: suggest parentheses around ‘&&’ within ‘||’ [-Wparentheses]
                     (!chosen_power2) && (i < free_space_count));

* Set Wno-delete-non-virtual-dtor for CXX files only

* Don't warn on unterminated endif labels

* Suppress unused functions

* Use 0x syntax rather than h syntax on GNU asm files

* Correct constructor call directly

src/ToolBox/superpmi/superpmi-shared/logging.cpp:301:27:   required from here
src/inc/clr_std/string:58:9: error: cannot call constructor ‘std::basic_string<char>::basic_string’ directly
         this->basic_string::basic_string(_Ptr, c_len(_Ptr));

* Suppress NULL used in arithmetic warnings

5 years agoDisable tests that fail to build under 3.0.0-preview3 (#22774)
Jarret Shook [Fri, 22 Feb 2019 17:32:38 +0000 (09:32 -0800)]
Disable tests that fail to build under 3.0.0-preview3 (#22774)

5 years agoExclude /Interop/COM/NETClients/Events/NETClientEvent for arm32 Windows. (#22785)
Sergey Andreenko [Fri, 22 Feb 2019 17:27:21 +0000 (09:27 -0800)]
Exclude /Interop/COM/NETClients/Events/NETClientEvent for arm32 Windows. (#22785)

5 years agoStore information about partial ngen in R2R images (#22680)
Michal Strehovský [Fri, 22 Feb 2019 14:44:01 +0000 (15:44 +0100)]
Store information about partial ngen in R2R images (#22680)

5 years agoUpdate ReJIT/MethodDescBackpatchInfoTracker lock relationship (#22741)
David Mason [Fri, 22 Feb 2019 07:59:54 +0000 (23:59 -0800)]
Update ReJIT/MethodDescBackpatchInfoTracker lock relationship (#22741)

5 years agoDelete unused FEATURE flags (#22781)
Jan Kotas [Fri, 22 Feb 2019 06:33:31 +0000 (22:33 -0800)]
Delete unused FEATURE flags (#22781)

5 years agoFix CoreFX build break. (#22773)
Sergey Andreenko [Fri, 22 Feb 2019 05:50:06 +0000 (21:50 -0800)]
Fix CoreFX build break. (#22773)

* Fix CoreFX build break.

Introduced by https://github.com/dotnet/corefx/pull/34385.

5 years agoCleanup unnecessary COUNTER_ONLY (#22759)
Jan Kotas [Fri, 22 Feb 2019 02:16:16 +0000 (18:16 -0800)]
Cleanup unnecessary COUNTER_ONLY (#22759)

5 years agoAdd exception counter (#22734)
Sung Yoon Whang [Fri, 22 Feb 2019 01:48:48 +0000 (17:48 -0800)]
Add exception counter (#22734)

* Add exception counter

* Fix build

* Make FCall return UINT32

* Move GetExceptionCount to Exception

* more PR comments, fix x86 build

* Remove global variable initialization

5 years agoCorrect __CrossBuild in build-packages to be 0 or 1 (#22769)
Jarret Shook [Fri, 22 Feb 2019 00:29:59 +0000 (16:29 -0800)]
Correct __CrossBuild in build-packages to be 0 or 1 (#22769)

5 years agoMerge pull request #22731 from briansull/issue-22436
Brian Sullivan [Thu, 21 Feb 2019 23:37:55 +0000 (15:37 -0800)]
Merge pull request #22731 from briansull/issue-22436

Fix Issue 22436 - noway_assert  'expTyp == cseLclVarTyp' in optcse.cpp

5 years agoDo not use portablebuild=false for linux-musl (#22760)
Jarret Shook [Thu, 21 Feb 2019 22:41:40 +0000 (14:41 -0800)]
Do not use portablebuild=false for linux-musl (#22760)

5 years ago[XUnit] Use ParallelRun==Collections on all platforms. (#22746)
Sergey Andreenko [Thu, 21 Feb 2019 22:38:42 +0000 (14:38 -0800)]
[XUnit] Use ParallelRun==Collections on all platforms. (#22746)

* Use ParallelRun==Collections on all platforms.

* Fix formatting

5 years agoRemove unused perfcounter code (#22742)
Sung Yoon Whang [Thu, 21 Feb 2019 22:19:27 +0000 (14:19 -0800)]
Remove unused perfcounter code (#22742)

* removing references to COUNTER_ONLY, ENABLE_PERF_COUNTERS in src/vm directory

* removing perfcounters* files

* Fix build

* Remove PerfCounterSupportedBuild

* Some more cleanup on the GC side

5 years ago[Unix|*] Rework and share the logic for determining __DistroRid (#22628)
Jarret Shook [Thu, 21 Feb 2019 21:16:33 +0000 (13:16 -0800)]
[Unix|*] Rework and share the logic for determining __DistroRid (#22628)

* [Unix|*] Rework and share the logic for determining __DistroRid

This will consolidate all of our different functions which determine
the __DistroRid into one file. In addition, it will remove the unused
__HostDistroRid file.

Lastly, it corrects the rid choice when crossbuilding musl-arm64 as the
cross-rootfs non-portable build will correctly be picked up.

* Fix typo and use __BuildArch in place of __Arch

* Correct non-portable build decision

* Add error for missing redhat-release

* Fix redhat comparison

* Correct previous rhel or

* Fix cross scenerio

* Fix missed $

* Set __RuntimeId for runtest.proj

* Correct naming convention of file

* Address feedback and support -portablebuild=0

With this change:

1. build.sh and build-test.sh should build portable by default.
2. init-distro-rid.sh checks __PortableBuild=1. If so, then it must force rhel 6 and freebsd to __PortableBuild=0 and use the non-portable build
    - For __PortableBuild=0 init-distro-rid.sh sets the non-portable rid correctly to "${ID}.${VERSION_ID}-${buildArch}"

* Correct arguments

* Fix if to include then

* A few more changes to correct linux-musl-arm64

* Add error for non-cross build and correctly pass ROOTFS_DIR

* Correctly exit instead of echo

* Correct isCrossBuild and use rootfsDir only

* Add documentation

* Address feedback

* Rework rhel if

5 years agoincreasing LocalGC version to 3 for 3.0 (#22751)
Maoni Stephens [Thu, 21 Feb 2019 19:54:27 +0000 (11:54 -0800)]
increasing LocalGC version to 3 for 3.0 (#22751)

5 years agoClean up unused parameters around Assembly::Load (#22716)
Marek Safar [Thu, 21 Feb 2019 18:54:17 +0000 (19:54 +0100)]
Clean up unused parameters around Assembly::Load (#22716)

5 years agoMispelling
Brian Sullivan [Thu, 21 Feb 2019 17:53:40 +0000 (09:53 -0800)]
Mispelling

5 years agoSupport COM events (#22721)
Aaron Robinson [Thu, 21 Feb 2019 17:51:10 +0000 (09:51 -0800)]
Support COM events (#22721)

* Add support for COM events
* Add COM Event tests

5 years agoAdded enclosing brances for if/then stmts
Brian Sullivan [Thu, 21 Feb 2019 17:29:07 +0000 (09:29 -0800)]
Added enclosing brances for if/then stmts

5 years agoJIT: handle preference miss two register xmm return case (#22733)
Andy Ayers [Thu, 21 Feb 2019 16:25:52 +0000 (08:25 -0800)]
JIT: handle preference miss two register xmm return case (#22733)

Fix a bug in codegen when returning a Vector3 result where the target
register is not one of the two return registers.

Re-enable the associated test.

Closes #22401.

5 years agoAdd VS debugging info to our instructions
Noah Falk [Thu, 21 Feb 2019 12:10:41 +0000 (04:10 -0800)]
Add VS debugging info to our instructions

5 years agoEnable the HWIntrinsic extension methods and remove the instance implementations...
Tanner Gooding [Thu, 21 Feb 2019 11:09:40 +0000 (03:09 -0800)]
Enable the HWIntrinsic extension methods and remove the instance implementations (#22705)

* Removing various S.R.I.Vector instance method APIs, since they should now be exposed as extension methods

* Updating the JIT to recognize the S.R.I.Vector extension methods.

* Updating various S.R.I.Vector test templates

* Regenerating the S.R.I tests that are created from a template.

* Fixing the numArgs for Base_Vector256_GetLower

* Fixing the handling for `Base_VectorXXX_As` to normalize the struct type.

* Adding the Base_Vector128_As intrinsics back for arm64

5 years agoMove Thread to shared CoreLib (dotnet/corefx#35462)
Filip Navara [Thu, 21 Feb 2019 02:23:41 +0000 (03:23 +0100)]
Move Thread to shared CoreLib (dotnet/corefx#35462)

Signed-off-by: dotnet-bot <dotnet-bot@microsoft.com>
5 years agoReduce struct copy by 1 in ValueTask.GetAwaiter (#22738)
Ben Adams [Thu, 21 Feb 2019 07:26:06 +0000 (07:26 +0000)]
Reduce struct copy by 1 in ValueTask.GetAwaiter (#22738)

* Reduce struct copy by 1 in ValueTask.GetAwaiter

* Bit more explicit

5 years agoRemoved some dead code
Andrew Au [Wed, 20 Feb 2019 04:39:26 +0000 (20:39 -0800)]
Removed some dead code

5 years agoFix Isuue 22436 - noway_assert 'expTyp == cseLclVarTyp' in optcse.cpp
Brian Sullivan [Thu, 21 Feb 2019 00:26:38 +0000 (16:26 -0800)]
Fix Isuue 22436 - noway_assert  'expTyp == cseLclVarTyp' in optcse.cpp
We now allow TYP_BYREF to be compatible with TYP_I_IMPL when performing CSE substitutions.
Re-enable the failing BestFitMapping interop tests

5 years agoRemove accidentally added Windows.10.Nano.Amd64 from Windows_x86 in platform-matrix...
Egor Chesakov [Wed, 20 Feb 2019 22:43:35 +0000 (14:43 -0800)]
Remove accidentally added Windows.10.Nano.Amd64 from Windows_x86 in platform-matrix.yml (#22723)

5 years agoAdd CoreCLR blobfeed to NuGet.config (#22699)
Juan Hoyos [Wed, 20 Feb 2019 21:23:37 +0000 (13:23 -0800)]
Add CoreCLR blobfeed to NuGet.config (#22699)

CoreCLR packages are not found for versions created after #21947 split the feeds and we stopped publishing to MyGet. This adds the new CoreCLR feed.

5 years agoFix dblarray tests and reenable for arm32 (#22718)
Andy Ayers [Wed, 20 Feb 2019 20:39:02 +0000 (12:39 -0800)]
Fix dblarray tests and reenable for arm32 (#22718)

Run the allocate/gen-check sequence under a forbid GC region, and enable the
tests for arm32.

Closes #22015.

5 years agoUse ThrowHelper in ManualResetValueTaskSourceCore (#22714)
Stephen Toub [Wed, 20 Feb 2019 19:36:56 +0000 (14:36 -0500)]
Use ThrowHelper in ManualResetValueTaskSourceCore (#22714)

Just move the helper being used to ThrowHelper.

5 years agoUse Build.DefinitionName for Creator when submitting tests to Helix (#22692)
Egor Chesakov [Wed, 20 Feb 2019 19:35:58 +0000 (11:35 -0800)]
Use Build.DefinitionName for Creator when submitting tests to Helix (#22692)

5 years agoFix tizen rootfs building (#22715)
Konstantin Baladurin [Wed, 20 Feb 2019 19:13:35 +0000 (22:13 +0300)]
Fix tizen rootfs building (#22715)

Move tizen-release package from base to unified repo

5 years agoFix bug introduced during cleanup in JapaneseLunisolarCalendar (#22717)
Stephen Toub [Wed, 20 Feb 2019 18:34:32 +0000 (13:34 -0500)]
Fix bug introduced during cleanup in JapaneseLunisolarCalendar (#22717)

5 years agoAllow Task.WaitAny completion to run synchronously (#21245)
Ben Adams [Wed, 20 Feb 2019 18:32:40 +0000 (18:32 +0000)]
Allow Task.WaitAny completion to run synchronously (#21245)

5 years agoAbstract deprecated and selectany for GCC and remove LLVM'ism where possible (#22662)
Sinan Kaya [Wed, 20 Feb 2019 16:59:22 +0000 (11:59 -0500)]
Abstract deprecated and selectany for GCC and remove LLVM'ism where possible (#22662)

* Abstract selectany

* Fix initializer element is not constant

src/corefx/System.Globalization.Native/pal_icushim.c:58:34: error: initializer element is not constant
 static const int MaxICUVersion = MinICUVersion + 20;

* Enable ms extensions

* Apply LLVM patterns to GCC

* Remove deprecated function

* Fix const conversion error

src/corefx/System.Globalization.Native/pal_calendarData.c:390:16: warning: passing argument 1 of ‘ures_close_ptr’ discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers]
ures_close(erasResBundle);

src/corefx/System.Globalization.Native/pal_calendarData.c:419:22: warning: initialization discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers]
char* name = GetCalendarName(calendarId);

* Remove old compiler option

5 years agoFix race condition in ManualResetValueTaskSourceCore (#22711)
Stephen Toub [Wed, 20 Feb 2019 11:54:11 +0000 (06:54 -0500)]
Fix race condition in ManualResetValueTaskSourceCore (#22711)

* Fix race condition in ManualResetValueTaskSourceCore

ManualResetValueTaskSourceCore.GetStatus is used from ValueTaskAwaiter.IsCompleted.  If GetStatus returns Success too early, then IsCompleted will also return true too early, which will result in GetResult being called too early.  This doesn't happen today when an MRVTSC is used sequentially.  But if an instance is pooled by the object its stored on getting put back into a pool as part of a call to the wrapper's GetResult, then we can end up in the following situation:
- Thread 1 starts an await on an instance containing an MRVTSC.  It calls IsCompleted.
- Thread 2 starts to complete that instance, getting as far as calling SignalCompletion and setting _completed to true but not yet setting _continuation.
- Thread 1 sees _completed == true and returns true from IsCompleted.  It calls GetResult, and the wrapper extracts the result from the instance, resets it, and puts it back into the pool.
- Thread 3 takes the object out of the pool and starts using it.
- Thread 2 continues SignalCompletion on that instance, and sets _continuation to the sentinel.
- Now Thread 3's instance's _continuation is the sentinel when it should be null.  If it calls SignalCompletion, it'll erroneously find that _continuation is not null and will queue _continuation/_continuationState for execution, resulting in the sentinel getting executed.  If it calls OnCompleted, it'll find that the _continuation is not null, and will queue the supplied continuation/state to execute immediately even though the operation may not yet actually be done.

The fix is simply to check not just _completed but also _continuation.  The operation is considered pending if either _completed is false, meaning SignalCompletion has not yet been called, or if _continuation is still null, meaning it's neither been set to the supplied delegate nor to the sentinel.  We can't just rely on _completed for the above outlined reasons, and we can't just rely on _continuation because it can be non-null if OnCompleted was called to hook up a callback (if we only cared about the await pattern, then we could just check _continuation and wouldn't need _completed, but we also need to support non-await access).

* Address PR feedback

5 years agoFix CLRDataCreateInstance export. This was breaking SOS. (#22701)
Mike McLaughlin [Wed, 20 Feb 2019 07:16:25 +0000 (23:16 -0800)]
Fix CLRDataCreateInstance export. This was breaking SOS. (#22701)

Fix CLRDataCreateInstance export. This was breaking SOS.

Added the right cmake magic so the function is exported properly.

5 years agoAdd ManagedRuntimeEventSource (#22515)
Sung Yoon Whang [Wed, 20 Feb 2019 05:40:12 +0000 (21:40 -0800)]
Add ManagedRuntimeEventSource (#22515)

* Adding a dummy ManagedRuntimeEventSource

* Add ManagedRuntimeEventSource

* Renaming ManagedRuntimeEventSource to RuntimeEventSource

* some cleanup

* use Condition instead of ifdefing out the whole file

* some cleanups

* remove unused namespaces

* Move m_RuntimeEventSource to RuntimeEventSource from EventPipeController

* Adding some TODO items

* address more pr comments

* Use GUID to initialize RuntimeEventSource

* Exclude BasicEventSourceTests from CoreFX test CI leg

* Addressing rest of the PR feedback

* dont dispose counters on disable command to avoid races

5 years agoEarly out from vectored exception handler on uninitialized threads (#22702)
Jan Kotas [Wed, 20 Feb 2019 04:48:19 +0000 (20:48 -0800)]
Early out from vectored exception handler on uninitialized threads (#22702)

Fixes #22698

5 years agoUse out arguments in UmAlQuraCalendar (#22704)
Jan Kotas [Wed, 20 Feb 2019 04:38:15 +0000 (20:38 -0800)]
Use out arguments in UmAlQuraCalendar (#22704)

5 years agoCleanup CoreLib makefiles (#22706)
Jan Kotas [Wed, 20 Feb 2019 04:36:46 +0000 (20:36 -0800)]
Cleanup CoreLib makefiles (#22706)

5 years agoAdding tannergooding to the arm64Users list. (#22665)
Tanner Gooding [Wed, 20 Feb 2019 02:07:44 +0000 (18:07 -0800)]
Adding tannergooding to the arm64Users list. (#22665)

5 years agoMerge pull request #22656 from briansull/fix_arm64_assert
Brian Sullivan [Wed, 20 Feb 2019 01:31:27 +0000 (17:31 -0800)]
Merge pull request #22656 from briansull/fix_arm64_assert

ARM64: Correct the assert to add R8 as a potential valid arg register

5 years agoFix formatting
Brian Sullivan [Tue, 19 Feb 2019 23:30:09 +0000 (15:30 -0800)]
Fix formatting

5 years agoUse RunContinuationsAsynchronously in SemaphoreSlim.WaitAsync (#22686)
Stephen Toub [Tue, 19 Feb 2019 22:35:06 +0000 (17:35 -0500)]
Use RunContinuationsAsynchronously in SemaphoreSlim.WaitAsync (#22686)

SemaphoreSlim.Release shouldn't invoke arbitrary continuations as part of its invocation, so when it dequeues a task waiter and goes to complete it, rather than just calling TrySetResult, it queues the task to the thread pool to have TrySetResult invoked there.  Now that we have RunContinuationsAsynchronously, though, we can just use that functionality instead.  This has two main benefits:
1. We avoid queueing a work item entirely if there are no continuations from the task.  This might happen, for example, if the semaphore is released so quickly after waiting on it that the WaitAsync caller hasn't yet hooked up a continuation, in which case the await on the WaitAsync task will just see IsCompleted as true and continue running.
2. We avoid queueing a work item when the task represents a synchronous Wait, which happens if there's already a pending WaitAsync when Wait goes to block.  The main benefit here is avoiding potential thread pool starvation, if threads in the pool are blocked in such Waits, and previously another thread pool thread would have been needed to run the queued work item to complete the synchronous Wait.

5 years agoCorrect the assert to add R8 as a potential valid register
Brian Sullivan [Sat, 16 Feb 2019 02:06:38 +0000 (18:06 -0800)]
Correct the assert to add R8 as a potential valid register
Also fix the calls to RemoveElemD to use lvVarIndex
Re-enable the interop tests that were failing on ARM64

5 years agoCleanup Calendars code (#22582)
Hugh Bellamy [Tue, 19 Feb 2019 21:33:26 +0000 (21:33 +0000)]
Cleanup Calendars code (#22582)

5 years agoFixing extension method overload resolution for some S.R.I.Vector methods (#22696)
Tanner Gooding [Tue, 19 Feb 2019 21:24:30 +0000 (13:24 -0800)]
Fixing extension method overload resolution for some S.R.I.Vector methods (#22696)

5 years agoUpdate dependencies from https://github.com/dotnet/core-setup build 20190218.4 (...
dotnet-maestro[bot] [Tue, 19 Feb 2019 18:42:20 +0000 (10:42 -0800)]
Update dependencies from https://github.com/dotnet/core-setup build 20190218.4 (#22689)

This change updates the following dependencies
- Microsoft.NETCore.App - 3.0.0-preview4-27418-4

5 years agoUse testGroup for defining test jobs (#22608)
Egor Chesakov [Tue, 19 Feb 2019 18:12:43 +0000 (10:12 -0800)]
Use testGroup for defining test jobs (#22608)

This adds a mapping between testGroup string and values of the following parameters:

* scenarios
* priority
* timeoutTotalInMinutes
* timeoutPerTestCollectionInMinutes
* timeoutPerTestInMinutes

and uses testGroup on a level of azure-pipeline.yml to set these values.

5 years agoEnable telemetry for official builds
Russ Keldorph [Fri, 15 Feb 2019 23:47:01 +0000 (15:47 -0800)]
Enable telemetry for official builds

5 years agoOptimize Vector128/256<T>.Get/WithElement (#22353)
Fei Peng [Tue, 19 Feb 2019 15:50:15 +0000 (07:50 -0800)]
Optimize Vector128/256<T>.Get/WithElement (#22353)

5 years agoAdding an update-dependencies.cmd script, like CoreFX has. (#22672)
Tanner Gooding [Tue, 19 Feb 2019 15:29:07 +0000 (07:29 -0800)]
Adding an update-dependencies.cmd script, like CoreFX has. (#22672)

5 years agoFail loading types with default interface methods for fragile ngen (#22603)
Michal Strehovský [Tue, 19 Feb 2019 12:47:50 +0000 (13:47 +0100)]
Fail loading types with default interface methods for fragile ngen (#22603)

We are not generating the right data structures for this. Since fragile crossgen is no longer a mainstream scenario, it's not worth the effort to add default interface methods support for it.

This avoids `!"Precode::GetPrecodeFromEntryPoint: Unexpected code in precode"` and similar issues.

5 years agoAdjust gc stress compatibility settings for two tests (#22633)
Andy Ayers [Tue, 19 Feb 2019 07:43:46 +0000 (23:43 -0800)]
Adjust gc stress compatibility settings for two tests (#22633)

* enable eventsourcetrace for gc stress. Closes #17188.
* disable GitHub_12224 for gc stress (relies on timeout). Closes #19086.

5 years agoUse correct version for Microsoft.NETCore.Platforms (#22683)
Jan Kotas [Tue, 19 Feb 2019 00:38:20 +0000 (16:38 -0800)]
Use correct version for Microsoft.NETCore.Platforms (#22683)

5 years agoRemove resetting of unmanaged thread locale from thread pool APIs (#22682)
Jan Kotas [Tue, 19 Feb 2019 00:37:19 +0000 (16:37 -0800)]
Remove resetting of unmanaged thread locale from thread pool APIs (#22682)

The cost of doing this is showing up in the TechEmpower profiles. There is a lot of thread locale state in unmanaged libraries and OS APIs that the threadpool friendly code has to be aware off. It does not make sense to keep reseting just the Locale.

5 years agoUpdate dependencies from https://github.com/dotnet/corefx build 20190218.1 (#22678)
dotnet-maestro[bot] [Mon, 18 Feb 2019 19:04:44 +0000 (11:04 -0800)]
Update dependencies from https://github.com/dotnet/corefx build 20190218.1 (#22678)

This change updates the following dependencies
- Microsoft.NETCore.Platforms - 3.0.0-preview4.19118.1
- Microsoft.Private.CoreFx.NETCoreApp - 4.6.0-preview4.19118.1

5 years agoFix constrained call corner cases (#22464)
Michal Strehovský [Mon, 18 Feb 2019 16:19:55 +0000 (17:19 +0100)]
Fix constrained call corner cases (#22464)

Fixes #22423. I'm still unclear on how JitStress manages to get itself into that situation, but I was able to write a repro that triggers the `!pMDAfterConstraintResolution->IsInterface()` assert using regular IL, so we need to handle that either way. The repro for that is constrained3.il.

While figuring out the repro, I wrote a bunch of other test code and found another bug (constrained2), where we would box in a situations that doesn't require boxing (canonically ambiguous situation where there's a suitable default interface implementation and a valuetype implementation of the constrained method that does not requires boxing once we no longer deal with __Canon).

5 years agoFix warning causing error in warning as error (Windows) (#22660)
Ben Adams [Mon, 18 Feb 2019 08:17:13 +0000 (08:17 +0000)]
Fix warning causing error in warning as error (Windows) (#22660)

BOOL -> bool conversion

5 years agoUpdate dependencies from https://github.com/dotnet/arcade build 20190215.1 (#22669)
dotnet-maestro[bot] [Mon, 18 Feb 2019 05:16:42 +0000 (21:16 -0800)]
Update dependencies from https://github.com/dotnet/arcade build 20190215.1 (#22669)

This change updates the following dependencies
- Microsoft.DotNet.Arcade.Sdk - 1.0.0-beta.19115.1
- Microsoft.DotNet.Helix.Sdk - 2.0.0-beta.19115.1

5 years agoUpdate dependencies from https://github.com/dotnet/corefx build 20190217.1 (#22670)
dotnet-maestro[bot] [Mon, 18 Feb 2019 05:16:20 +0000 (21:16 -0800)]
Update dependencies from https://github.com/dotnet/corefx build 20190217.1 (#22670)

This change updates the following dependencies
- Microsoft.NETCore.Platforms - 3.0.0-preview4.19117.1
- Microsoft.Private.CoreFx.NETCoreApp - 4.6.0-preview4.19117.1

5 years agoFix Integer Rounding issue and duplicate definition of __deref in sal.h (#22645)
Sinan Kaya [Mon, 18 Feb 2019 05:12:28 +0000 (00:12 -0500)]
Fix Integer Rounding issue and duplicate definition of __deref in sal.h (#22645)

* Fix large integer truncation

pal/pal.h:4988:43:
warning: large integer implicitly truncated to unsigned type
[-Woverflow]
     static const SIZE_T NoTargetFrameSp = SIZE_MAX;
                                                ^~~~~~~~

* Fix compilation error with GCC

/usr/include/c++/7.3.0/debug/functions.h:452:15: error: expected unqualified-id before ')' token
       __deref();
               ^
/usr/include/c++/7.3.0/debug/functions.h:455:34: error: expected primary-expression before '<' token
       typename = decltype(__deref<_It>() < __deref<_It>())>
                                  ^
/usr/include/c++/7.3.0/debug/functions.h:455:38: error: expected primary-expression before '>' token
       typename = decltype(__deref<_It>() < __deref<_It>())>
                                      ^
/usr/include/c++/7.3.0/debug/functions.h:455:40: error: expected primary-expression before ')' token
       typename = decltype(__deref<_It>() < __deref<_It>())>
                                        ^

5 years agoDelete RtlMoveMemory and tests (#22668)
Adeel Mujahid [Mon, 18 Feb 2019 04:29:23 +0000 (06:29 +0200)]
Delete RtlMoveMemory and tests (#22668)

5 years agoFix the build of Loader/NativeLib test (#22648)
Swaroop Sridhar [Sun, 17 Feb 2019 21:27:41 +0000 (13:27 -0800)]
Fix the build of Loader/NativeLib test (#22648)

The native test assets were not build correctly on Unix platforms.
The native library exports were generated as private symbols by clang.

This change fixes the export declaration so that the global symbols are
correctly generated for exported symbols.

Fixes #22549

5 years agoOnly import the RoslynPropsFile if it exists. (#22657)
Tanner Gooding [Sun, 17 Feb 2019 03:38:04 +0000 (19:38 -0800)]
Only import the RoslynPropsFile if it exists. (#22657)

5 years agoFix visibility and signed comparison issues for GCC (#22586)
Adeel Mujahid [Sun, 17 Feb 2019 01:50:17 +0000 (03:50 +0200)]
Fix visibility and signed comparison issues for GCC (#22586)

* Fix visibility and signed comparison issues for GCC

* Fix hidden _CLRDataCreateInstance warned by macOS

* Fix indentation in vswprintf/test1

* Change void* to PVOID in implementation files

5 years agoAvoid pulling in Lazy<T> during startup (#22659)
Jan Kotas [Sun, 17 Feb 2019 01:49:06 +0000 (17:49 -0800)]
Avoid pulling in Lazy<T> during startup (#22659)

Environment constructor runs on every startup path. Remove dependency on Lazy<T> from it.

5 years agoMerge pull request #22528 from mikedn/idx-addr
Carol Eidt [Sat, 16 Feb 2019 15:06:45 +0000 (07:06 -0800)]
Merge pull request #22528 from mikedn/idx-addr

Fix genCodeForIndexAddr

5 years ago[master] Update dependencies from dotnet/corefx (#22595)
dotnet-maestro[bot] [Sat, 16 Feb 2019 07:50:21 +0000 (23:50 -0800)]
[master] Update dependencies from dotnet/corefx (#22595)

* Update dependencies from https://github.com/dotnet/corefx build 20190213.23

This change updates the following dependencies
- Microsoft.NETCore.Platforms - 3.0.0-preview4.19113.23
- Microsoft.Private.CoreFx.NETCoreApp - 4.6.0-preview4.19113.23

* Update dependencies from https://github.com/dotnet/corefx build 20190215.1

This change updates the following dependencies
- Microsoft.NETCore.Platforms - 3.0.0-preview4.19115.1
- Microsoft.Private.CoreFx.NETCoreApp - 4.6.0-preview4.19115.1

* Fixing up the tests dependent on the S.R.I.Vector API changes

* Regenerating the tests dependent on the S.R.I.Vector API changes

* Disabling an outdated CoreFX test.

5 years agoSet the 'PLATFORM_OSX' constant during OSX build (#22639)
Aaron Robinson [Sat, 16 Feb 2019 05:28:50 +0000 (21:28 -0800)]
Set the 'PLATFORM_OSX' constant during OSX build (#22639)

* Set the 'PLATFORM_OSX' constant during OSX build

* Remove PLATFORM_OSX define from SPCL

5 years agoFix bug in FindDispatchSlot usage
Aaron Robinson [Sat, 16 Feb 2019 05:25:08 +0000 (21:25 -0800)]
Fix bug in FindDispatchSlot usage

5 years agoUpdate BuildTools, IbcData, PgoData to 3.0.0-preview1-03715-01, master-20190208.1...
Tanner Gooding [Sat, 16 Feb 2019 05:11:56 +0000 (21:11 -0800)]
Update BuildTools, IbcData, PgoData to 3.0.0-preview1-03715-01, master-20190208.1, master-20190208.1, respectively (master) (#22647)

5 years agoUpdating some local functions to be 'static' (captureless). (#22635)
Tanner Gooding [Fri, 15 Feb 2019 23:17:52 +0000 (15:17 -0800)]
Updating some local functions to be 'static' (captureless). (#22635)

5 years agoJIT: modify how jit determines when to update a type (#22618)
Andy Ayers [Fri, 15 Feb 2019 22:39:33 +0000 (14:39 -0800)]
JIT: modify how jit determines when to update a type (#22618)

For single-def locals, the type of a reference seen at the assignment to the
local may be a more specific type than the local's declared type. If so the jit
would prefer to use the assignment type to describe the local's value, as this
will lead to better optimization. For instance in
```
    object x = "a string";  // only assignment to x
```
the jit can optimize better if it models the type of `x` as `string`.

Instead of relying on `mergeClasses` plus some jit-side screening to decide if
the assignment type is a more specific type, implement a new jit interface
method `isMoreSpecificType` that tries to answer this question more directly.

Added a test case with type equivalence that hit asserts.

Closes #22583.

5 years agoDisable arm64 contracts. (#22605)
Sergey Andreenko [Fri, 15 Feb 2019 22:03:47 +0000 (14:03 -0800)]
Disable arm64 contracts. (#22605)

* Fix ifdef for `g_DbgSuppressAllocationAsserts`.

It needs to be defined under `FEATURE_INTEROP_DEBUGGING`.

* Unify contracts disabling for UNIX/ARM.

It also disables some parts that were disabled for ARM, but not for UNIX.
It fixes the difference that we see between ARM32 and ARM64 Unix Checked.

* Disable contracts on arm64.

FIxes the difference that we see between windows arm32 and arm64 in checked/debug builds.

* Revert CoreFX arm64 timeout change.

That is no longer necessary.

* Reenable arm64 corefx jobs.

They should not fail with timeouts anymore.

* Don't allow `g_DbgSuppressAllocationAsserts` defnition in release builds.

* delete AssertAllocationAllowed

5 years agoPort profiler howtos from David Broman's blog to the documentation folder (#22363)
David Mason [Fri, 15 Feb 2019 22:02:31 +0000 (14:02 -0800)]
Port profiler howtos from David Broman's blog to the documentation folder (#22363)

5 years agoPublish version to dotnet/versions
Steve MacLean [Mon, 11 Feb 2019 21:39:44 +0000 (16:39 -0500)]
Publish version to dotnet/versions

5 years agoFix UpdatePublishedVersions.ps1
Steve MacLean [Thu, 14 Feb 2019 21:04:47 +0000 (16:04 -0500)]
Fix UpdatePublishedVersions.ps1

5 years agoWiden index to a temporary register
Mike Danes [Fri, 15 Feb 2019 18:56:59 +0000 (20:56 +0200)]
Widen index to a temporary register