platform/upstream/coreclr.git
7 years ago[RyuJIT/ARM32] Struct argument passing less than 4 byte
Hyeongseok Oh [Mon, 5 Jun 2017 07:04:26 +0000 (16:04 +0900)]
[RyuJIT/ARM32] Struct argument passing less than 4 byte

Fix assertion when struct argument is less than 4 byte and promoted

7 years agoAdditional fix for ETW failure in 32 process on 64 bit machines. (#12022)
Vance Morrison [Sat, 3 Jun 2017 15:30:36 +0000 (08:30 -0700)]
Additional fix for ETW failure in 32 process on 64 bit machines. (#12022)

* Addtional fix for ETW failure in 32 process on 64 bit machines.

The commit associated with Pull Request
    Fix failures in ETW logging on 4GB aware 32 bit processes #11941

Was incorrect.   Some long were changed to ulong, but the critical casting an IntPtr to ulong still does
sign extention and not zero extension.   Need to cast to a void* first.  This fixes this.

7 years agoMerge pull request #12068 from omajid/master
Matt Ellis [Sat, 3 Jun 2017 05:47:15 +0000 (22:47 -0700)]
Merge pull request #12068 from omajid/master

Fix typo in CRL_CMAKE_WARNINGS_ARE_ERRORS

7 years agoRemove relocations from SECTION_MethodDesc for ngened images for fields accessed...
gbalykov [Sat, 3 Jun 2017 03:31:28 +0000 (06:31 +0300)]
Remove relocations from SECTION_MethodDesc for ngened images for fields accessed from jit code for Linux ARM (#11963)

7 years agoCall sched_yield directly for Sleep(0) (#12063)
Koundinya Veluri [Sat, 3 Jun 2017 00:04:17 +0000 (17:04 -0700)]
Call sched_yield directly for Sleep(0) (#12063)

Related to PR #12057

7 years agoFix typo in CRL_CMAKE_WARNINGS_ARE_ERRORS
Omair Majid [Fri, 2 Jun 2017 23:00:26 +0000 (19:00 -0400)]
Fix typo in CRL_CMAKE_WARNINGS_ARE_ERRORS

7 years agoMerge pull request #12064 from rartemev/fix_bash_script_issues
Roman Artemev [Fri, 2 Jun 2017 22:42:54 +0000 (15:42 -0700)]
Merge pull request #12064 from rartemev/fix_bash_script_issues

Fix path in bash script that runs test

7 years agofix procCount assertion (#12065)
Xiangyang (Mark) Guo [Fri, 2 Jun 2017 21:24:47 +0000 (14:24 -0700)]
fix procCount assertion (#12065)

7 years agoTiny typo in TaskCompletionSource xml comment (#12066)
Bar Arnon [Fri, 2 Jun 2017 21:24:29 +0000 (00:24 +0300)]
Tiny typo in TaskCompletionSource xml comment (#12066)

7 years agoAdd support for not breaking *nix build on warnings (#12039)
Omair Majid [Fri, 2 Jun 2017 20:55:14 +0000 (16:55 -0400)]
Add support for not breaking *nix build on warnings (#12039)

Add a build flag to make -Werror optional and let the build continue
even in the presence of warnings.

This option is very useful for anyone compiling with a different
(version of the) compiler. A different (version of the) compiler may
produce a different set of warnings and a piece of code that compiles
without warnings may emit warnings with a different (version of the)
compiler.

Resolves https://github.com/dotnet/coreclr/issues/8586

7 years agoMerge pull request dotnet/corert#3745 from dotnet/nmirror
Jan Kotas [Wed, 31 May 2017 06:22:43 +0000 (23:22 -0700)]
Merge pull request dotnet/corert#3745 from dotnet/nmirror

Merge nmirror to master

Signed-off-by: dotnet-bot <dotnet-bot@microsoft.com>
7 years agoMerge pull request #11699 from swgillespie/local-gc-branch-merge
Sean Gillespie [Fri, 2 Jun 2017 19:08:18 +0000 (12:08 -0700)]
Merge pull request #11699 from swgillespie/local-gc-branch-merge

Integration from dev/local-gc into master

7 years agoattempt to fix r2r test failures
Roman Artemev [Fri, 2 Jun 2017 18:47:04 +0000 (11:47 -0700)]
attempt to fix r2r test failures

7 years agoMerge pull request #11803 from BruceForstall/FixRuntest
Bruce Forstall [Fri, 2 Jun 2017 18:41:22 +0000 (11:41 -0700)]
Merge pull request #11803 from BruceForstall/FixRuntest

Update runtest.cmd help to reflect reality

7 years agofix SleepEx(0,false) on Linux (#12057)
Xiangyang (Mark) Guo [Fri, 2 Jun 2017 18:28:36 +0000 (11:28 -0700)]
fix SleepEx(0,false) on Linux (#12057)

7 years agoReturn 0 on success from GC Reliability Framework scripts (#11990)
Sean Gillespie [Fri, 2 Jun 2017 17:17:53 +0000 (10:17 -0700)]
Return 0 on success from GC Reliability Framework scripts (#11990)

* Return 0 on success from GC Reliability Framework scripts

* Don't save standard out, Jenkins already does this

7 years agoMerge pull request #11957 from sjsinju/spill_temp_size
Carol Eidt [Fri, 2 Jun 2017 16:51:37 +0000 (09:51 -0700)]
Merge pull request #11957 from sjsinju/spill_temp_size

[RyuJit/ARM32] Adjust alignment for TYP_DOUBLE in tmpPreAllocateTemps

7 years agoMerge pull request #12003 from sjsinju/reg1_reg2_planB
Carol Eidt [Fri, 2 Jun 2017 16:50:31 +0000 (09:50 -0700)]
Merge pull request #12003 from sjsinju/reg1_reg2_planB

[RyuJit/ARM32] Fix wrong behavior of UINT->CHAR casting

7 years agoAdd support for flat vtables in RyuJIT (#12041)
Michal Strehovský [Fri, 2 Jun 2017 16:25:48 +0000 (09:25 -0700)]
Add support for flat vtables in RyuJIT (#12041)

CoreRT needs one less indirection to get to the virtual method.

On CoreCLR a virtual method call resolution goes like:

```
this -> MethodTable -> chunk(?) -> slot
```

On CoreRT, it's more flat:

```
this -> EEType -> slot
```

7 years agoFix a bug in LoadNativeStringResource to honor the contract properly when the buffer...
Yi Zhang (CLR) [Fri, 2 Jun 2017 16:10:34 +0000 (09:10 -0700)]
Fix a bug in LoadNativeStringResource to honor the contract properly when the buffer is not big enough (#12051)

* Return last error properly

* adjust pcwchUsed if ERROR_INSUFFICIENT_BUFFER

* return hr

* ERROR_INSUFFICIENT_BUFFER is a win32 error....

7 years agoInvoke jitShutdown during ProcessShutdownWork (#12054)
Jonghyun Park [Fri, 2 Jun 2017 13:12:19 +0000 (22:12 +0900)]
Invoke jitShutdown during ProcessShutdownWork (#12054)

7 years ago[x64/Linux] Fix FEATURE_INTERPRETER Release build (#12046)
Jonghyun Park [Fri, 2 Jun 2017 05:28:27 +0000 (14:28 +0900)]
[x64/Linux] Fix FEATURE_INTERPRETER Release build (#12046)

7 years agoModify comments
sjsujinkim [Fri, 2 Jun 2017 01:25:34 +0000 (10:25 +0900)]
Modify comments

7 years agoFix optdata restore functionality (#11935)
Daniel Podder [Thu, 1 Jun 2017 23:41:50 +0000 (18:41 -0500)]
Fix optdata restore functionality (#11935)

Fix optdata restore functionality (#11935)

Bring back optdata restore functionality following the move to 2.0
BuildTools and csproj-based CLI. Disable a harmless warning that broke
the Linux build when consuming PGO counts due to -Werr.

Fix #11796 for master

7 years agoEnsure that platform is considered for LongPlatformName instead of PackagePlatform...
Karthik Rajasekaran [Thu, 1 Jun 2017 22:59:00 +0000 (15:59 -0700)]
Ensure that platform is considered for LongPlatformName instead of PackagePlatform (#11900)

7 years agoSupport msbuildonunsupportedplatform on non-x64 architecture (#11966)
Jonghyun Park [Thu, 1 Jun 2017 22:18:40 +0000 (07:18 +0900)]
Support msbuildonunsupportedplatform on non-x64 architecture (#11966)

7 years agoFix 11991 (#12013)
Sergey Andreenko [Thu, 1 Jun 2017 21:24:39 +0000 (14:24 -0700)]
Fix 11991 (#12013)

Fix the case when the path doesn't have any '\'.
Fix clean_up.

7 years agoFix warnings on 32-bit platforms
Sean Gillespie [Thu, 1 Jun 2017 21:07:19 +0000 (14:07 -0700)]
Fix warnings on 32-bit platforms

7 years agoMerge pull request #11959 from hqueue/arm/ryujit/11779
Carol Eidt [Thu, 1 Jun 2017 20:16:18 +0000 (13:16 -0700)]
Merge pull request #11959 from hqueue/arm/ryujit/11779

[RyuJIT/ARM32][LSRA] Fix restoring TYP_DOUBLE interval in ARM32

7 years agoMerge pull request #11860 from rartemev/fix_tests_build_4
Roman Artemev [Thu, 1 Jun 2017 19:27:23 +0000 (12:27 -0700)]
Merge pull request #11860 from rartemev/fix_tests_build_4

Fixed tests build on Linux

7 years agoMerge pull request #11876 from hqueue/arm/ryujit/merge_lowerarmarch
Pat Gavlin [Thu, 1 Jun 2017 18:35:33 +0000 (11:35 -0700)]
Merge pull request #11876 from hqueue/arm/ryujit/merge_lowerarmarch

Merge remaining common arm32/arm64 lowering functions

7 years ago[Local GC] Remove uses of g_SystemInfo from the GC (#11628)
Sean Gillespie [Thu, 1 Jun 2017 17:17:35 +0000 (10:17 -0700)]
[Local GC] Remove uses of g_SystemInfo from the GC (#11628)

* [Local GC] Don't use g_SystemInfo to infer the number of processors

* [Local GC] Avoid using g_SystemInfo.dwAllocationGranularity from within
the GC.

In a few places, the GC aligns the size of a request memory reservation
to the allocation granularity before calling
GCToOSInterface::VirtualReserve. Upon inspection, this isn't necessary.

* [Local GC] Remove uses of g_SystemInfo from the handle table

* Add new API on GCToOSInterface for querying the total number of procs on the machine

7 years ago[Local GC] Obtaining configuration information (#11379)
Sean Gillespie [Tue, 30 May 2017 21:20:16 +0000 (17:20 -0400)]
[Local GC] Obtaining configuration information (#11379)

* [Local GC] Skeleton for GC configuration

* Initial tweaks after design feedback:
    1) Use string keys instead of enums. Upon receiving a string key,
    the EE looks at it and, if it's something that comes from startup
    flags, responds using the startup flag information. Otherwise, it
    forwards the string onto CLRConfig.
    2) Add a mechanism for getting string configuration values from the
    EE. This includes adding a RAII wrapper around strings so that they
    are freed correctly.

* Remove uses of g_pConfig from the GC and replace with GCConfig

* Use the GCConfig system for the GC log

* Fix poorly-named parameter

* Add documentation and caching of bool and int configs obtained from the EE

* Remove AppDomainLeaks as dead code

* Remove GC trace configs as dead code

* Repair unix build

* Fix an issue where we started the GC in the wrong latency mode

* Fix the unix build

* Pipe GCRetainVM configuration to the GC

* Dead code removal in the GC sample

* EEConfig -> GCConfig for heap verification constants in the GC

* Populate config information for bools and ints eagerly at startup

* Initialize g_theGCToCLR before initializing GCConfig

* Propegate HoardVM config to the GC

* Fix an incorrect comment

7 years ago[Local GC] Add async pinned handle methods to the handle interface (#11569)
Sean Gillespie [Tue, 16 May 2017 17:19:36 +0000 (10:19 -0700)]
[Local GC] Add async pinned handle methods to the handle interface (#11569)

* [Local GC] Add async pinned handle methods to the handle interface

* Add a callback to HandleAsyncPinnedHandles for handle enumeration

* Rename HandleAsyncPinnedHandles -> EnumerateAsyncPinnedHandles

* Introduce typedef for function pointer type

7 years ago[Local GC] Fix handle table interface violation (#11558)
Sean Gillespie [Sat, 13 May 2017 06:37:21 +0000 (23:37 -0700)]
[Local GC] Fix handle table interface violation (#11558)

7 years ago[Local GC] Fix some handle table interface violations in COMDependentHandle (#11358)
Sean Gillespie [Wed, 17 May 2017 22:41:34 +0000 (15:41 -0700)]
[Local GC] Fix some handle table interface violations in COMDependentHandle (#11358)

7 years agoFix a build break in feature branch caused by obsolete error (#11335)
Sean Gillespie [Wed, 17 May 2017 22:23:04 +0000 (15:23 -0700)]
Fix a build break in feature branch caused by obsolete error (#11335)

7 years ago[Local GC] Scaffolding for loading a standalone GC (#11242)
Sean Gillespie [Wed, 17 May 2017 22:07:34 +0000 (15:07 -0700)]
[Local GC] Scaffolding for loading a standalone GC (#11242)

* Configure the build system to build a CoreCLR capable of loading a standalone GC

* Proto-implementation of dynamic GC loading

* Build the GC with the VM's CMakeLists when doing a non-standalone build of the GC

* [Local GC] Introduce a new feature define, FEATURE_STANDALONE_GC_ONLY, to be used by the CI to explicitly test local GC dynamic loading code paths

* Fix the FEATURE_STANDALONE_GC_ONLY build for unix linkers

* Rebase against master

* Code review feedback: use the existing Unix exports file

7 years agogdbjit symbols exported (#11739)
chunseoklee [Thu, 1 Jun 2017 16:51:31 +0000 (01:51 +0900)]
gdbjit symbols exported (#11739)

* multiple export files as input for generate_exports_file

Now, generate_exports_file is able to process multiple input arguments.
This feature is useful when you want to export additional symbols depending on
compile time dependannt feature(-D option.)

* export file for gdbjit added

src/dlls/mscoree/gdbjit_unixexports.src file added
To enable gdbjit feature, libcoreclr.so should have __jit_xxx
symbols.
Current tizen build system removes .symtab section during rpm debuginfo
build. This patch moves __jit_xx symbols into .dynsym section.

7 years ago[Arm64] Use inner shareable for instGen_MemoryBarrier
Steve MacLean, Qualcomm Datacenter Technologies, Inc [Tue, 16 May 2017 17:37:28 +0000 (17:37 +0000)]
[Arm64] Use inner shareable for instGen_MemoryBarrier

7 years ago[Arm64] Use half barriers for IL Volatile implementation
Steve MacLean [Fri, 12 May 2017 16:44:08 +0000 (12:44 -0400)]
[Arm64] Use half barriers for IL Volatile implementation

7 years agoUpdating version of xUnit-Performance-Api (#11987)
José Rivero [Thu, 1 Jun 2017 15:56:27 +0000 (08:56 -0700)]
Updating version of xUnit-Performance-Api (#11987)

- This update contains a fix for this issue: https://github.com/Microsoft/xunit-performance/issues/230
- Consolidate the Api version in two places: `dependencies.prop` and `PerfHarness.csproj`
- Move performance log files to root directory where Jenkins can archive them
- Fixed the output to console, so it is easier to see on the Jenkins job the sccript steps live and get an exact repro step if needed

7 years agoFixes deserializaing TimeZoneInfo by putting back the IDeserializationCallback.OnDese...
Morgan Brown [Thu, 1 Jun 2017 12:09:07 +0000 (05:09 -0700)]
Fixes deserializaing TimeZoneInfo by putting back the IDeserializationCallback.OnDeserialization method. (#12024)

7 years ago[x64/Linux] Use correct argument registers in InterpreterStub (#12002)
Jonghyun Park [Thu, 1 Jun 2017 08:49:33 +0000 (17:49 +0900)]
[x64/Linux] Use correct argument registers in InterpreterStub (#12002)

* [x64/Linux] Use correct argument registers in InterpreterStub

* Adjust Linux/Windows ABI difference

* Use macros instead of array

* Follow naming convention

7 years agoMerge pull request #11988 from stephentoub/fix_fds
Stephen Toub [Thu, 1 Jun 2017 03:05:06 +0000 (23:05 -0400)]
Merge pull request #11988 from stephentoub/fix_fds

Ensure several runtime file descriptors are CLOEXEC

7 years agoRemove GCX_PREEMP call when allocating a buffer. (#12016)
Brian Robbins [Wed, 31 May 2017 23:48:56 +0000 (16:48 -0700)]
Remove GCX_PREEMP call when allocating a buffer. (#12016)

7 years agoPost-Serialization cleanup fixes to Reflection (#12010)
Atsushi Kanamori [Wed, 31 May 2017 23:40:04 +0000 (16:40 -0700)]
Post-Serialization cleanup fixes to Reflection (#12010)

This is based on the following principles:

- We should not be making changes to instance/virtual methods
in the abstract base types for Reflection. These are apis
in themselves to third party classes that subclass these types.
Even if the Runtime chooses not to make its own implementations
of these type serializable, third party classes have the right
to make that choice for themselves and build off the base class
serialization infrastructure that was shipped before.

RuntimeAssembly and RuntimeModule still override with PNSE
so we'll get the desired exception for those cases.

- The Runtime's own implementations of these types will not be
serializable. Removing all [Serializable] attributes and
serialization interfaces. Runtime types are internal types and
thus have no duty to implement ISerializable if the public
abstract base type does not.

7 years agoDelete useless file (#12021)
mikedn [Wed, 31 May 2017 22:57:46 +0000 (01:57 +0300)]
Delete useless file (#12021)

7 years agoDisable inheritance of several common file descriptors
Stephen Toub [Wed, 31 May 2017 21:18:24 +0000 (17:18 -0400)]
Disable inheritance of several common file descriptors

7 years agoDefault to portable build in dev and official builds (#11985)
Gaurav Khanna [Wed, 31 May 2017 15:53:47 +0000 (08:53 -0700)]
Default to portable build in dev and official builds (#11985)

7 years agoFix issues with RelativePointer instead of RelativeFixupPointer (#12004)
gbalykov [Wed, 31 May 2017 15:39:43 +0000 (18:39 +0300)]
Fix issues with RelativePointer instead of RelativeFixupPointer (#12004)

7 years agoExpose Enumerator members out of EnvironmentAugments. (#11980)
Atsushi Kanamori [Wed, 31 May 2017 15:23:02 +0000 (08:23 -0700)]
Expose Enumerator members out of EnvironmentAugments. (#11980)

Expose Enumerator members out of EnvironmentAugments.

7 years ago[RyuJit/ARM32] Fix wrong behavior of UINT->CHAR casting
sjsujinkim [Wed, 31 May 2017 09:39:08 +0000 (18:39 +0900)]
[RyuJit/ARM32] Fix wrong behavior of UINT->CHAR casting

'JIT\Regression\CLR-x86-JIT\V1-M11-Beta1\b44879\b44879\b44879.cmd'

When UINT cast to CHAR, MOV instruction had been emitted with R0 to R0 for extending the type.
It is wrong behavior to cast.

In case of INT to BOOL casting, that use uxtb instruction to extend BOOL type.

So this PR would use a uxth instruction too insted of wrong mov instruction.

@dotnet/arm64-contrib @dotnet/arm32-contrib Please make sure this code is okay.

7 years ago[RyuJIT/ARM32] Update restoring TYP_DOUBLE interval
Hyung-Kyu Choi [Wed, 31 May 2017 05:14:59 +0000 (14:14 +0900)]
[RyuJIT/ARM32] Update restoring TYP_DOUBLE interval

- Introduce findAnotherHalfRegRec()
- Restore TYP_DOUBLE interval to valid double register

Signed-off-by: Hyung-Kyu Choi <hk0110.choi@samsung.com>
7 years ago[RyuJIT/ARM32] Applying review feedback
Hyung-Kyu Choi [Wed, 31 May 2017 05:56:25 +0000 (14:56 +0900)]
[RyuJIT/ARM32] Applying review feedback

- Group cases for each target to improve readability

Signed-off-by: Hyung-Kyu Choi <hk0110.choi@samsung.com>
7 years ago[RyuJIT/ARM32] Applying review feedback
Hyung-Kyu Choi [Wed, 31 May 2017 03:24:19 +0000 (12:24 +0900)]
[RyuJIT/ARM32] Applying review feedback

- Use canRestorePreviousInterval() for all target
- Update condition for restoring TYP_DOUBLE interval

Signed-off-by: Hyung-Kyu Choi <hk0110.choi@samsung.com>
7 years agoMerge pull request #11995 from ViktorHofer/SerializationFollowUp
Krzysztof Wicher [Wed, 31 May 2017 03:02:47 +0000 (20:02 -0700)]
Merge pull request #11995 from ViktorHofer/SerializationFollowUp

Internal hashtable serialization attribute removed

7 years agoAllowing larger segment size for SOH, in preparation for enabling it for NoGC region...
Maoni Stephens [Wed, 31 May 2017 01:29:29 +0000 (18:29 -0700)]
Allowing larger segment size for SOH, in preparation for enabling it for NoGC region. (#11923)

This is just so that I can test with larger SOH seg size to make sure we are not getting AVs.
More perf changes will happen with the NoGC region changes.

7 years agoRemove relocations from SECTION_MethodDesc for ngened images (#11394)
gbalykov [Wed, 31 May 2017 01:25:04 +0000 (04:25 +0300)]
Remove relocations from SECTION_MethodDesc for ngened images (#11394)

7 years agoInternal hashtable serialization attribute removed
Viktor Hofer [Wed, 31 May 2017 00:46:43 +0000 (02:46 +0200)]
Internal hashtable serialization attribute removed

7 years agoCoretype variables renamed back to netfx counterpart for reflection based serializati...
Viktor Hofer [Wed, 31 May 2017 00:16:32 +0000 (02:16 +0200)]
Coretype variables renamed back to netfx counterpart for reflection based serialization (#11910)

* Variables renamed for reflection based serialization

* Make EqualityComparers serialize like desktop

* add missing interfaces

* TimeZone serializable added

* Internal hashtable serializable

* Removed TimeZone as serializable type

* Remove Lazy<T>'s [Serializable] attribute for 2.0

For performance, Lazy was completely rewritten for .NET Core 2.0 and has an entirely different format than desktop; trying to get it to match the desktop serialization format would require either reverting or providing a complicated custom serialization/deserialization implementation to try to match.  Lazy can also wrap an Exception that occurred from trying to instantiate the object, and the only exception types that are serializable as of now in core are the base Exception and AggregateException.  As such, we're cutting it from the list of supported types in 2.0.  An easy workaround is simply to do what the implementation does: serialize lazy.Value instead of lazy.

* tiny fixes to equalitycomparer.cs

7 years agoMerge pull request #11977 from mskvortsov/FixDoubleRegArgNext
Carol Eidt [Tue, 30 May 2017 23:28:05 +0000 (16:28 -0700)]
Merge pull request #11977 from mskvortsov/FixDoubleRegArgNext

[RyuJIT/arm32] Fix double argument passing

7 years agoMerge pull request #11872 from hqueue/arm/ryujit/dodge_lowering_long_cmp
Pat Gavlin [Tue, 30 May 2017 23:06:23 +0000 (16:06 -0700)]
Merge pull request #11872 from hqueue/arm/ryujit/dodge_lowering_long_cmp

[RyuJIT/ARM32] Dodge lowering long compares

7 years agoSkip cast nodes for the tail call check. (#11986)
Sergey Andreenko [Tue, 30 May 2017 23:02:20 +0000 (16:02 -0700)]
Skip cast nodes for the tail call check. (#11986)

Fix #11983

7 years agoFix build break
Jan Kotas [Tue, 30 May 2017 16:21:04 +0000 (09:21 -0700)]
Fix build break

7 years agoMoves windows specific field to Debug.Windows.cs file (dotnet/corert#3733)
Marek Safar [Tue, 30 May 2017 06:29:57 +0000 (08:29 +0200)]
Moves windows specific field to Debug.Windows.cs file (dotnet/corert#3733)

Signed-off-by: dotnet-bot <dotnet-bot@microsoft.com>
7 years agoUse FD_CLOEXEC instead of equivalent 1 in InternalCreatePipe
Stephen Toub [Tue, 30 May 2017 19:49:42 +0000 (15:49 -0400)]
Use FD_CLOEXEC instead of equivalent 1 in InternalCreatePipe

7 years agoRemove Debian Armel (#11978)
Gaurav Khanna [Tue, 30 May 2017 19:15:05 +0000 (12:15 -0700)]
Remove Debian Armel (#11978)

7 years agoRyuJIT/arm32: Fix double argument passing
Mikhail Skvortcov [Tue, 30 May 2017 15:47:54 +0000 (18:47 +0300)]
RyuJIT/arm32: Fix double argument passing

7 years agoFix failures in ETW logging on 4GB aware 32 bit processes (#11941)
Vance Morrison [Tue, 30 May 2017 16:09:29 +0000 (09:09 -0700)]
Fix failures in ETW logging on 4GB aware 32 bit processes (#11941)

We incorrectly cast a 32 bit pointer to a 64 bit poitner using as a SIGNED integer.
If this is a 32 bit process that is using more than 2GB of memory this can result in
sign rather than zero extension.  This makes the poitner invalid at the OS level
and causes the OS API to fail.

We disovered this in Visual Studio when debugging large (Rosyln) scenarios.
There were numerous failures which causes sever slowdowns becasue the
EventSource logged OutputDebugString events when the OS API failed.

The fix is to use unsigned extension.   Note that I have confirmed that casting
from a IntPtr or a void* to a ulong does zero extension (that is it uses the target type
to determine whether to use sign or zero extension).

To be useful for Visual Studio, this needs to be ported to the desktop runtime.

7 years agoMerge pull request #11972 from hseok-oh/ryujit/issue_11971
Carol Eidt [Tue, 30 May 2017 15:31:47 +0000 (08:31 -0700)]
Merge pull request #11972 from hseok-oh/ryujit/issue_11971

[RyuJIT/ARM32] Fix offset of struct argument when using stack

7 years agoMerge pull request #11973 from hseok-oh/ryujit/issue_11837
Carol Eidt [Tue, 30 May 2017 15:30:15 +0000 (08:30 -0700)]
Merge pull request #11973 from hseok-oh/ryujit/issue_11837

[RyuJIT/ARM32] Fix assertion failed 'remainingSize == TARGET_POINTER_SIZE'

7 years ago[RyuJIT/ARM32] Fix assertion failed 'remainingSize == TARGET_POINTER_SIZE'
Hyeongseok Oh [Tue, 30 May 2017 10:05:35 +0000 (19:05 +0900)]
[RyuJIT/ARM32] Fix assertion failed 'remainingSize == TARGET_POINTER_SIZE'

Fix assertion failed by 'remainingSize == TARGET_POINTER_SIZE'
Fix to check register allocation correctly

7 years ago[RyuJIT/ARM32] Fix offset of struct argument when using stack
Hyeongseok Oh [Tue, 30 May 2017 09:58:33 +0000 (18:58 +0900)]
[RyuJIT/ARM32] Fix offset of struct argument when using stack

Fix offset value when struct argument is passed on stack

7 years ago[RyuJIT/ARM32] Fix restoring TYP_DOUBLE interval in ARM32
Hyung-Kyu Choi [Mon, 29 May 2017 11:38:01 +0000 (20:38 +0900)]
[RyuJIT/ARM32] Fix restoring TYP_DOUBLE interval in ARM32

- Fix segmentation fault by defining nextRegRec when nextRegRec is not defined yet.

- Introduce canRestorePreviousInterval()
  This helper check whether previous interval can be restored considering TYP_DOUBLE.

Signed-off-by: Hyung-Kyu Choi <hk0110.choi@samsung.com>
7 years agoMake linux test build worked
Roman Artemev [Tue, 23 May 2017 22:34:53 +0000 (15:34 -0700)]
Make linux test build worked

7 years agoImplement GetCurrentProcessCpuCount for Unix and merge NumberOfProcesors implementati...
Tom Deseyn [Mon, 29 May 2017 13:46:10 +0000 (15:46 +0200)]
Implement GetCurrentProcessCpuCount for Unix and merge NumberOfProcesors implementations. (#11742)

* Implement GetCurrentProcessCpuCount for Unix and merge NumberOfProcessors implementations.

* Split GetCurrentProcessCpuCount in PAL and non-PAL parts

* gcenv.unix.cpp: GetCurrentProcessAffinityMask: revert, GetCurrentProcessCpuCount: add implementation

* bugfix: #if instead of #ifdef HAVE_SCHED_GETAFFINITY

* Remove PAL_GetCurrentProcessCpuCount

* gcenv.unix: implement GetCurrentProcessAffinityMask, GetCurrentProcessCpuCount

* numa.cpp: initialize g_cpuToAffinity[i].Number when numa.h is missing

* cleanup

* mscorwks_unixexports.src: Add GetProcessAffinityMask

* PR feedback

* Add GetProcessAffinityMask to mscordac_unixexports.src

7 years agoadded DWARF exprloc for static value class fields (#11911)
Buyduck [Mon, 29 May 2017 11:26:41 +0000 (14:26 +0300)]
added DWARF exprloc for static value class fields (#11911)

7 years ago[RyuJit/ARM32] Adjust alignment for TYP_DOUBLE in tmpPreAllocateTemps
sjsujinkim [Mon, 29 May 2017 06:39:38 +0000 (15:39 +0900)]
[RyuJit/ARM32] Adjust alignment for TYP_DOUBLE in tmpPreAllocateTemps

In #11782,

Assertion failed 'spillTempSize <= lvaGetMaxSpillTempSize()' has been occurred in code generation phase.

Adjusting alignment for TYP_DOUBLE in tmpPreAllocateTemps is needed.

7 years agoFix build breaks
Morgan Brown [Sat, 27 May 2017 02:13:06 +0000 (19:13 -0700)]
Fix build breaks

* Remove UnitySerializationHolder references
* Remove the fields of StrongNameKeyPair
* Initialize BadImageFormatException._fusionLog
* Remove MemberInfoSerializationHolder references

Signed-off-by: dotnet-bot <dotnet-bot@microsoft.com>
7 years agoIDeserializationCallback cleanup
Morgan Brown [Sun, 28 May 2017 04:07:32 +0000 (21:07 -0700)]
IDeserializationCallback cleanup

Throws PlatformNotSupportedException from IDeserializationCallback and IObjectReference implementations on types that are no longer serializable.

7 years agoFix tool initialization detection (#11912)
mikedn [Sat, 27 May 2017 23:09:11 +0000 (02:09 +0300)]
Fix tool initialization detection (#11912)

init-tools.cmd attempts to create a file in a directory that doesn't exist. File creation fails and next time the script is run initialization will be performed again. That wastes a lot of time.

7 years agoUpdate CoreClr to preview1-25327-01 (#11949)
dotnet bot [Sat, 27 May 2017 23:07:59 +0000 (16:07 -0700)]
Update CoreClr to preview1-25327-01 (#11949)

7 years agoUpdate test coverage statements in workflow doc
Rich Lander [Sat, 27 May 2017 18:46:13 +0000 (11:46 -0700)]
Update test coverage statements in workflow doc

7 years agoUpdate license links
Rich Lander [Sat, 27 May 2017 05:01:25 +0000 (22:01 -0700)]
Update license links

7 years agoUpdate CoreClr, CoreFx to preview1-25326-02, preview1-25324-02, respectively (#11932)
dotnet bot [Sat, 27 May 2017 05:00:51 +0000 (22:00 -0700)]
Update CoreClr, CoreFx to preview1-25326-02, preview1-25324-02, respectively (#11932)

7 years agoUpdate contributor docs (#11944)
Rich Lander [Sat, 27 May 2017 04:56:27 +0000 (21:56 -0700)]
Update contributor docs (#11944)

* Update contributor docs

* Update

* Added up-for-grabs and ports

* Update binary licensing

* Update\

* Simplify developer workflow doc

* Update per feedback

7 years agoEnable linux perf runs (#11905)
José Rivero [Sat, 27 May 2017 01:01:05 +0000 (18:01 -0700)]
Enable linux perf runs (#11905)

- Removing `[assembly: MeasureInstructionsRetired]` from performance tests.
- Due to this issue: https://github.com/Microsoft/xunit-performance/issues/231, we are blocked from running on performance tests on Linux.
  The tag is not used by infrastructure and it is passed during runtime when profiling on Windows.
- Adding the option to execute CoreRun with the stability prefix.
- Moving file to be archived by Jenkins and updating groovy file.
- Setting the maximum number of iterations.

7 years agoISerializable cleanup (#11873)
Morgan Brown [Sat, 27 May 2017 00:45:37 +0000 (17:45 -0700)]
ISerializable cleanup (#11873)

Changes to throw PlatformNotSupportedException from ISerializable.GetObjectData and serialization constructors on non-Serializable types. Also removes private serialization constructors and some unneeded code that was used to support serializing non-serializable types. A few tests testing GetObjectData implementations are also removed. For Exception types, we call base instead of throwing from GetObjectData to be consistent with the many Exceptions that don't override GetObjectData

7 years agoMove rundown to the same code-site where we do ETW rundown before EEShutdown is marke...
Brian Robbins [Fri, 26 May 2017 21:39:42 +0000 (14:39 -0700)]
Move rundown to the same code-site where we do ETW rundown before EEShutdown is marked as started. (#11937)

7 years agoMerge pull request #11920 from briansull/crossgen-tuning
Brian Sullivan [Fri, 26 May 2017 18:08:01 +0000 (11:08 -0700)]
Merge pull request #11920 from briansull/crossgen-tuning

Remove the check that requires an instrumented flags for SPC when collecting IBC data

7 years agoFix BitSetUint64, clean and improve the surrounding code. (#11721)
Sergey Andreenko [Fri, 26 May 2017 16:13:08 +0000 (09:13 -0700)]
Fix BitSetUint64, clean and improve the surrounding code. (#11721)

1. Add the functions that were not implemented in bitsetasuint64inclass/
2. BitSetUint64: make copy_constructor public
3. delete VARSET_INIT_NOCOPY and ALLVARSET_INIT_NOCOPY
4. delete VARSET_INIT and ALLVARSET_INIT.
5. Use the correct const type for VARSET_TP.

7 years agoMerge pull request #11823 from hqueue/arm/ryujit/lsra_gc_helper
Carol Eidt [Fri, 26 May 2017 15:13:27 +0000 (08:13 -0700)]
Merge pull request #11823 from hqueue/arm/ryujit/lsra_gc_helper

[RyuJIT/ARM][LSRA] Update register mask for GC helper

7 years agoEETypeHashTable lookups can deadlock with GC during comparison with unrestored persis...
Jan Kotas [Fri, 26 May 2017 14:17:51 +0000 (07:17 -0700)]
EETypeHashTable lookups can deadlock with GC during comparison with unrestored persisted TypeHandles (#11922)

7 years agoFix compile with disabled FEATURE_READYTORUN option (#11875)
Petr Bred [Fri, 26 May 2017 14:17:05 +0000 (17:17 +0300)]
Fix compile with disabled FEATURE_READYTORUN option (#11875)

- just remove FEATURE_READYTORUN definitions
in clrdefinitions.cmake to turn off option

Signed-off-by: Petr Bred <bredpetr@gmail.com>
7 years ago[RyuJIT/ARM32] Update RMB for helper function
Hyung-Kyu Choi [Fri, 26 May 2017 07:12:33 +0000 (16:12 +0900)]
[RyuJIT/ARM32] Update RMB for helper function

Make RBM_CALLEE_TRASH be consistent with CORINFO_HELP_STOP_FOR_GC helper
which is JIT_RareDisableHelper defined in vm/arm/asmhelpers.S

Signed-off-by: Hyung-Kyu Choi <hk0110.choi@samsung.com>
7 years agoMerge pull request dotnet/corert#3709 from dotnet/nmirror
Michal Strehovský [Fri, 26 May 2017 01:24:34 +0000 (18:24 -0700)]
Merge pull request dotnet/corert#3709 from dotnet/nmirror

Merge nmirror to master

Signed-off-by: dotnet-bot <dotnet-bot@microsoft.com>
7 years agoMerge pull request #11917 from JosephTremoulet/RationalCall
Joseph Tremoulet [Fri, 26 May 2017 05:41:38 +0000 (01:41 -0400)]
Merge pull request #11917 from JosephTremoulet/RationalCall

Clear GTF_CALL from non-calls in rationalize

7 years agoApply review feedback to increase readability
Hyung-Kyu Choi [Fri, 26 May 2017 01:39:25 +0000 (10:39 +0900)]
Apply review feedback to increase readability

Signed-off-by: Hyung-Kyu Choi <hk0110.choi@samsung.com>
7 years agoRemove the check that requires an instrumented flags for mscorlib when collecting...
Brian Sullivan [Thu, 25 May 2017 22:27:15 +0000 (15:27 -0700)]
Remove the check that requires an instrumented flags for mscorlib when collecting IBC data
This is not needed for CoreCLR and blocks some valuable usage cases.