Pat Gavlin [Mon, 5 Jun 2017 21:33:58 +0000 (14:33 -0700)]
Merge pull request #12091 from pgavlin/TailCallStress11689
Disable tail call stress in GH_11689 if ZapDisable is enabled.
Pat Gavlin [Mon, 5 Jun 2017 21:06:45 +0000 (14:06 -0700)]
Merge pull request #11884 from mikedn/arm-cmp-setcc
[WIP] Lower long compares that are materialized into a register for ARM32
Matt Ellis [Mon, 5 Jun 2017 20:44:29 +0000 (13:44 -0700)]
Merge pull request #12072 from ellismg/update-buildtools
Update BuildTools
Matt Galbraith [Mon, 5 Jun 2017 20:05:43 +0000 (13:05 -0700)]
Merge pull request #12085 from dotnet/Clean_up_properties_sent_for_Publish_builds
Make all publish legs of CoreCLR uniform and match existing ones.
Pat Gavlin [Mon, 5 Jun 2017 19:49:24 +0000 (12:49 -0700)]
Disable tail call stress in GH_11689 if ZapDisable is enabled.
Tail call stress does not mix well with ZapDisable due to the issues
described in #11408.
Matt Galbraith [Mon, 5 Jun 2017 17:34:46 +0000 (10:34 -0700)]
Make all publish legs of CoreCLR uniform and match existing ones. Previously Release - Publish symbols was in the CoreFX-style format and the other three were not. This unifies them.
Xiangyang (Mark) Guo [Mon, 5 Jun 2017 17:16:27 +0000 (10:16 -0700)]
Inline YieldProcessor on Linux (#12058)
More ldar/stlr support
[Arm64] Add emitInsTargetRegSize support for ldar/stlr
Carol Eidt [Mon, 5 Jun 2017 15:34:55 +0000 (08:34 -0700)]
Merge pull request #12025 from hseok-oh/ryujit/remorph_filedlist
Skip argTable assignment for split struct when remorphing
Hyeongseok Oh [Mon, 5 Jun 2017 13:55:49 +0000 (22:55 +0900)]
Update morph.cpp
Hyeongseok Oh [Mon, 5 Jun 2017 03:28:14 +0000 (12:28 +0900)]
Revert and remove isReMorph argument
- Revert commit that is removing nextSlotNum update in remorphing phase
- Remove isReMorph argument in SplitArg by using argsComplete
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.
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
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)
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
Omair Majid [Fri, 2 Jun 2017 23:00:26 +0000 (19:00 -0400)]
Fix typo in CRL_CMAKE_WARNINGS_ARE_ERRORS
Matt Ellis [Fri, 2 Jun 2017 22:59:40 +0000 (15:59 -0700)]
Update BuildTools
This picks up some changes to how the tool runtime is published (we
now publish it as a portable app so we have less random .so's lying
around).
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
Xiangyang (Mark) Guo [Fri, 2 Jun 2017 21:24:47 +0000 (14:24 -0700)]
fix procCount assertion (#12065)
Bar Arnon [Fri, 2 Jun 2017 21:24:29 +0000 (00:24 +0300)]
Tiny typo in TaskCompletionSource xml comment (#12066)
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
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>
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
Roman Artemev [Fri, 2 Jun 2017 18:47:04 +0000 (11:47 -0700)]
attempt to fix r2r test failures
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
Xiangyang (Mark) Guo [Fri, 2 Jun 2017 18:28:36 +0000 (11:28 -0700)]
fix SleepEx(0,false) on Linux (#12057)
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
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
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
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
```
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....
Jonghyun Park [Fri, 2 Jun 2017 13:12:19 +0000 (22:12 +0900)]
Invoke jitShutdown during ProcessShutdownWork (#12054)
Mike Danes [Wed, 24 May 2017 16:58:28 +0000 (19:58 +0300)]
Delete long compare support from ARM32 codegen
Mike Danes [Wed, 24 May 2017 16:53:22 +0000 (19:53 +0300)]
Enable long relop lowering for ARM32
Mike Danes [Wed, 24 May 2017 11:17:28 +0000 (14:17 +0300)]
Add ARM32 support for GT_SETCC/GT_CMP/GTF_SET_FLAGS
Jonghyun Park [Fri, 2 Jun 2017 05:28:27 +0000 (14:28 +0900)]
[x64/Linux] Fix FEATURE_INTERPRETER Release build (#12046)
Hyeongseok Oh [Fri, 2 Jun 2017 04:51:33 +0000 (13:51 +0900)]
Fix formatting and bug
Fix formatting
Fix bug: DEBUG flag usage
Hyeongseok Oh [Fri, 2 Jun 2017 03:18:10 +0000 (12:18 +0900)]
Remove updating nextSlotNum in remorphing phase
Remove updating nextSlotNum in remorphing phase
Remain updating it in remorphing phase for debugging only.
sjsujinkim [Fri, 2 Jun 2017 01:25:34 +0000 (10:25 +0900)]
Modify comments
Hyeongseok Oh [Fri, 2 Jun 2017 01:12:47 +0000 (10:12 +0900)]
Fix formatting
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
Karthik Rajasekaran [Thu, 1 Jun 2017 22:59:00 +0000 (15:59 -0700)]
Ensure that platform is considered for LongPlatformName instead of PackagePlatform (#11900)
Jonghyun Park [Thu, 1 Jun 2017 22:18:40 +0000 (07:18 +0900)]
Support msbuildonunsupportedplatform on non-x64 architecture (#11966)
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.
Sean Gillespie [Thu, 1 Jun 2017 21:07:19 +0000 (14:07 -0700)]
Fix warnings on 32-bit platforms
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
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
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
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
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
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
Sean Gillespie [Sat, 13 May 2017 06:37:21 +0000 (23:37 -0700)]
[Local GC] Fix handle table interface violation (#11558)
Sean Gillespie [Wed, 17 May 2017 22:41:34 +0000 (15:41 -0700)]
[Local GC] Fix some handle table interface violations in COMDependentHandle (#11358)
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)
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
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.
[Arm64] Use inner shareable for instGen_MemoryBarrier
Steve MacLean [Fri, 12 May 2017 16:44:08 +0000 (12:44 -0400)]
[Arm64] Use half barriers for IL Volatile implementation
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
Morgan Brown [Thu, 1 Jun 2017 12:09:07 +0000 (05:09 -0700)]
Fixes deserializaing TimeZoneInfo by putting back the IDeserializationCallback.OnDeserialization method. (#12024)
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
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
Hyeongseok Oh [Thu, 1 Jun 2017 01:19:16 +0000 (10:19 +0900)]
[RyuJIT/ARM32] Skip argTable info assignment for split struct when remorphing
- Skip argTable no. of register & stack slot assignment for split struct when remorphing
- Modify assertion check condition for structs morphed to GT_FIELD_LIST
Brian Robbins [Wed, 31 May 2017 23:48:56 +0000 (16:48 -0700)]
Remove GCX_PREEMP call when allocating a buffer. (#12016)
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.
mikedn [Wed, 31 May 2017 22:57:46 +0000 (01:57 +0300)]
Delete useless file (#12021)
Stephen Toub [Wed, 31 May 2017 21:18:24 +0000 (17:18 -0400)]
Disable inheritance of several common file descriptors
Gaurav Khanna [Wed, 31 May 2017 15:53:47 +0000 (08:53 -0700)]
Default to portable build in dev and official builds (#11985)
gbalykov [Wed, 31 May 2017 15:39:43 +0000 (18:39 +0300)]
Fix issues with RelativePointer instead of RelativeFixupPointer (#12004)
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.
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.
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>
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>
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>
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
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.
gbalykov [Wed, 31 May 2017 01:25:04 +0000 (04:25 +0300)]
Remove relocations from SECTION_MethodDesc for ngened images (#11394)
Viktor Hofer [Wed, 31 May 2017 00:46:43 +0000 (02:46 +0200)]
Internal hashtable serialization attribute removed
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
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
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
Sergey Andreenko [Tue, 30 May 2017 23:02:20 +0000 (16:02 -0700)]
Skip cast nodes for the tail call check. (#11986)
Fix #11983
Jan Kotas [Tue, 30 May 2017 16:21:04 +0000 (09:21 -0700)]
Fix build break
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>
Stephen Toub [Tue, 30 May 2017 19:49:42 +0000 (15:49 -0400)]
Use FD_CLOEXEC instead of equivalent 1 in InternalCreatePipe
Gaurav Khanna [Tue, 30 May 2017 19:15:05 +0000 (12:15 -0700)]
Remove Debian Armel (#11978)
Mikhail Skvortcov [Tue, 30 May 2017 15:47:54 +0000 (18:47 +0300)]
RyuJIT/arm32: Fix double argument passing
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.
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
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'
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
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
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>
Roman Artemev [Tue, 23 May 2017 22:34:53 +0000 (15:34 -0700)]
Make linux test build worked
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
Buyduck [Mon, 29 May 2017 11:26:41 +0000 (14:26 +0300)]
added DWARF exprloc for static value class fields (#11911)
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.
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>
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.