platform/upstream/coreclr.git
8 years agoMerge pull request #3150 from krytarowski/netbsd-support-42
Jan Kotas [Fri, 12 Feb 2016 06:08:41 +0000 (22:08 -0800)]
Merge pull request #3150 from krytarowski/netbsd-support-42

Fix build issue on NetBSD: Include <stdlib.h> for alloca(3)

8 years agoMerge pull request #3146 from mikem8361/vso190301
Mike McLaughlin [Fri, 12 Feb 2016 03:44:31 +0000 (19:44 -0800)]
Merge pull request #3146 from mikem8361/vso190301

Fix VSO bug #190301 contract problems in DACRaiseException.

8 years agoFix build issue on NetBSD: Include <stdlib.h> for alloca(3)
Kamil Rytarowski [Fri, 12 Feb 2016 02:46:46 +0000 (03:46 +0100)]
Fix build issue on NetBSD: Include <stdlib.h> for alloca(3)

Header <alloca.h> isn't needed on other supported systems either.

/tmp/pkgsrc-tmp/wip/coreclr-git/work/coreclr/src/pal/src/file/directory.cpp:605:
warning: Warning: reference to the libc supplied alloca(3); this most likely will
not work. Please use the compiler provided version of alloca(3), by supplying the
appropriate compiler flags (e.g. not -std=c89).

$ uname -a
NetBSD chieftec 7.99.25 NetBSD 7.99.25 (GENERIC) #0: Fri Dec 25 20:51:06 UTC 2015  root@chieftec:/tmp/netbsd-tmp/sys/arch/amd64/compile/GENERIC amd64

8 years agoMerge pull request #3076 from krytarowski/netbsd-support-37
Jan Vorlicek [Fri, 12 Feb 2016 01:31:15 +0000 (02:31 +0100)]
Merge pull request #3076 from krytarowski/netbsd-support-37

Permit to inject additional CMake arguments via environment variables

8 years agoFix VSO bug #190301 contract problems in DACRaiseException.
Mike McLaughlin [Thu, 11 Feb 2016 22:25:21 +0000 (14:25 -0800)]
Fix VSO bug #190301 contract problems in DACRaiseException.

Added the proper contracts to the DAC notification functions.

8 years agoMerge pull request #3122 from krytarowski/netbsd-support-40
Jan Vorlicek [Fri, 12 Feb 2016 00:49:35 +0000 (01:49 +0100)]
Merge pull request #3122 from krytarowski/netbsd-support-40

Fix build on NetBSD: Include <cstdarg> for va_list

8 years agoMerge pull request #3136 from sivarv/simdfix2
Sivarv [Fri, 12 Feb 2016 00:22:45 +0000 (16:22 -0800)]
Merge pull request #3136 from sivarv/simdfix2

Fix to issue #2970

8 years agoMerge pull request #3082 from sejongoh/update_corefx_build_steps
Sejong Oh [Thu, 11 Feb 2016 23:32:03 +0000 (15:32 -0800)]
Merge pull request #3082 from sejongoh/update_corefx_build_steps

Fixed corefx build failures

8 years agoMerge pull request #3132 from stephentoub/taskscheduler_tracking
Stephen Toub [Thu, 11 Feb 2016 23:15:17 +0000 (18:15 -0500)]
Merge pull request #3132 from stephentoub/taskscheduler_tracking

Track TaskSchedulers only when debugger attached

8 years agoMerge pull request #3139 from stephentoub/task_id_perf
Stephen Toub [Thu, 11 Feb 2016 23:15:06 +0000 (18:15 -0500)]
Merge pull request #3139 from stephentoub/task_id_perf

Fix access to Task.Id while finishing continuations

8 years agoUpdated build steps to fix corefx build issues
Sejong Oh [Wed, 10 Feb 2016 01:20:06 +0000 (17:20 -0800)]
Updated build steps to fix corefx build issues

8 years agoFix access to Task.Id while finishing continuations
stephentoub [Thu, 11 Feb 2016 21:41:43 +0000 (16:41 -0500)]
Fix access to Task.Id while finishing continuations

Task.Id is lazily-initialized with an interlocked operation and we strive to avoid accessing it anywhere in Task's internals on hot paths, but apparently when some ETW logging was added around continuations, calls were added that access Task.Id even if logging is disabled.  This is accounting for ~30% of the cycles consumed in a TCS Task's lifecycle!  This commit fixes it by making sure these accesses to Id are guarded.

8 years agoFix to issue #2970
sivarv [Thu, 11 Feb 2016 19:39:05 +0000 (11:39 -0800)]
Fix to issue #2970

The repro is a case of one of the method called is returning Vector3/4 type value.
On unix such a struct will gets returned in two registers and hence will
be assigned to a local by importer and marked as multiRegArgOrRet=1.
But it doesn't ensure that the local will not be register allocated.
For the repro case it gets register allocated.

IR: GT_LCL_FLD(tmp1) = GT_CALL returning SIMD12

Codgen routine genStoreRegisterReturnInLclVar() for stroing a struct returned in two registers to a lcl var or lcl_fld
is making the assumption that lcl var will be on stack not in a register.
This may not be true for SIMD structs returned in two registers.

For now to workaround Codegen limitation marking tmp1 as DoNoEnregister.

Fixes #2970

8 years agoMerge pull request #3137 from wtgodbe/disable
William Godbe [Thu, 11 Feb 2016 21:05:13 +0000 (13:05 -0800)]
Merge pull request #3137 from wtgodbe/disable

Disable pri 1 tests failing in the CI

8 years agoDisable pri 1 tests failing in the CI
William Godbe [Thu, 11 Feb 2016 20:33:44 +0000 (12:33 -0800)]
Disable pri 1 tests failing in the CI

8 years agoTrack TaskSchedulers only when debugger attached
stephentoub [Thu, 11 Feb 2016 18:06:41 +0000 (13:06 -0500)]
Track TaskSchedulers only when debugger attached

Today the base TaskScheduler ctor registers every created scheduler in a ConditionalWeakTable.  This is then used in the internal TaskScheduler.GetTaskSchedulersForDebugger() to enable a debugger to find all of the schedulers that are active and then in turn find all tasks scheduled to them.  However, for scenarios that involve creating many, many schedulers, this registration adds up to a non-trivial cost, both in time and space, and is only relevant for debugging.  Plus, since this mechanism was introduced, Visual Studio has additional ways of tracking tasks for display in the Tasks window, e.g. ETW events.

This commit simply makes that registration more pay-for-play, only doing it if the debugger is attached at the time of the scheduler's creation.

8 years agoMerge pull request #3119 from ellismg/add-rhel72
Matt Mitchell [Thu, 11 Feb 2016 16:27:48 +0000 (08:27 -0800)]
Merge pull request #3119 from ellismg/add-rhel72

Add Red Hat Enterprise Linux 7.2 Runs

8 years agoMerge pull request #3126 from jkotas/crossgen
Jan Kotas [Thu, 11 Feb 2016 10:48:05 +0000 (02:48 -0800)]
Merge pull request #3126 from jkotas/crossgen

Change crossgen defaults

8 years agoMerge pull request #3120 from erozenfeld/MultToShiftVN1
Eugene Rozenfeld [Thu, 11 Feb 2016 08:00:26 +0000 (00:00 -0800)]
Merge pull request #3120 from erozenfeld/MultToShiftVN1

Small correction for #3067.

8 years agoMerge pull request #3127 from AndyAyersMS/FlakyTestDisable
Andy Ayers [Thu, 11 Feb 2016 07:29:34 +0000 (23:29 -0800)]
Merge pull request #3127 from AndyAyersMS/FlakyTestDisable

Disable test causing Ubuntu failures

8 years agoDisable test causing Ubuntu failures
Andy Ayers [Thu, 11 Feb 2016 03:53:53 +0000 (19:53 -0800)]
Disable test causing Ubuntu failures

Fails because of mismatched System.IO versions. See #3117 for some
discussion.

8 years agoChange crossgen defaults
Jan Kotas [Thu, 11 Feb 2016 03:35:01 +0000 (19:35 -0800)]
Change crossgen defaults

Change crossgen to generate R2R images by default

8 years agoMerge pull request #3118 from jkotas/buffercopy
Jan Kotas [Thu, 11 Feb 2016 02:47:45 +0000 (18:47 -0800)]
Merge pull request #3118 from jkotas/buffercopy

Add a fast path for byte[] to Buffer.BlockCopy

8 years agoAdd a fast path for byte[] to Buffer.BlockCopy
Jan Kotas [Wed, 10 Feb 2016 17:18:13 +0000 (09:18 -0800)]
Add a fast path for byte[] to Buffer.BlockCopy

8 years agoFix build on NetBSD: Include <cstdarg> for va_list
Kamil Rytarowski [Thu, 11 Feb 2016 00:02:51 +0000 (01:02 +0100)]
Fix build on NetBSD: Include <cstdarg> for va_list

8 years agoSmall correction for #3067.
Eugene Rozenfeld [Wed, 10 Feb 2016 23:50:41 +0000 (15:50 -0800)]
Small correction for #3067.

The code to change multiplication to shift shouldn't be
under LEA_AVAILABLE since the code to process multiplication
by power of 2 isn't.

8 years agoAdd Red Had Enterprise Linux 7.2 Runs
Matt Ellis [Wed, 10 Feb 2016 22:55:34 +0000 (14:55 -0800)]
Add Red Had Enterprise Linux 7.2 Runs

8 years agoAdd new parameter with settable option: cmakeargs
Kamil Rytarowski [Mon, 8 Feb 2016 11:02:20 +0000 (12:02 +0100)]
Add new parameter with settable option: cmakeargs

This will allow to pass user-defined CMake arguments.

One of the users is pkgsrc, which defines it to following example string:
-DPYVERSSUFFIX:STRING=2.7 -DCMAKE_INSTALL_PREFIX:PATH=/usr/pkg \
-DCMAKE_MODULE_PATH:PATH=/tmp/pkgsrc-tmp/wip/coreclr-git/work/.buildlink/cmake-Modules \
-DCMAKE_SKIP_RPATH:BOOL=TRUE -DCMAKE_INSTALL_LIBDIR:PATH=lib \
-DCMAKE_INSTALL_MANDIR:PATH=man

Without these values unmodified build.sh won't integrate with pkgsrc as
expected.

Solution suggested by Jan Vorli (Microsoft)

While there, there was need to alter loop type to handle options passed to
arguments. As a bonus, new loop should be POSIX confirming.

8 years agoMerge pull request #3112 from mikem8361/sosfix
Mike McLaughlin [Wed, 10 Feb 2016 22:37:25 +0000 (14:37 -0800)]
Merge pull request #3112 from mikem8361/sosfix

Fix SOS managed breakpoints when coreclr symbols are stripped.

8 years agoFix SOS managed breakpoints when coreclr symbols are stripped.
Mike McLaughlin [Tue, 9 Feb 2016 01:47:31 +0000 (17:47 -0800)]
Fix SOS managed breakpoints when coreclr symbols are stripped.

Added a SOS DAC interface (ISOSDacInterface4::GetClrNotification) to get the exception
notification arguments instead of using the GetLastExceptionInformation function from the lldb
sosplugin that depends on coreclr symbols being present.

On the coreclr side, the clr notification arguments are saved in a global variable that is DAC
accessible. A critical section was added to protect this global variable while the special
exception is raised.

Setting the internal COMPlus_DebugBreakOnAssert environment variable causes 3 or 4 breaks in
the debugger with no reason.  It was breaking in the function that was determining whether it
should break.  I was using COMPlus_BreakOnEELoad=2 to break after coreclr was loaded and initialized
to set managed breakpoints and on a debug build it generates an assert (on release just a DebugBreak).

8 years agoMerge pull request #3107 from wtgodbe/reenable
William Godbe [Wed, 10 Feb 2016 18:10:15 +0000 (10:10 -0800)]
Merge pull request #3107 from wtgodbe/reenable

Re-enable previously disabled tests that are now passing on non-Windows platforms

8 years agoMerge pull request #3095 from krytarowski/netbsd-support-38
Jan Kotas [Wed, 10 Feb 2016 16:43:19 +0000 (08:43 -0800)]
Merge pull request #3095 from krytarowski/netbsd-support-38

Add support for Floating Point registers in ucontext_t on NetBSD

8 years agoMerge pull request #3114 from mmitche/switch-to-auto
Matt Mitchell [Wed, 10 Feb 2016 16:22:10 +0000 (08:22 -0800)]
Merge pull request #3114 from mmitche/switch-to-auto

Switch to auto images

8 years agoSwitch to auto images
Matt Mitchell [Wed, 10 Feb 2016 16:21:19 +0000 (08:21 -0800)]
Switch to auto images

8 years agoMerge pull request #3110 from jkotas/event-tracing
Jan Kotas [Wed, 10 Feb 2016 03:06:50 +0000 (19:06 -0800)]
Merge pull request #3110 from jkotas/event-tracing

Make Unix event tracing conditionally enabled

8 years agoMerge pull request #3108 from sivarv/gcstressfix
Sivarv [Wed, 10 Feb 2016 01:18:21 +0000 (17:18 -0800)]
Merge pull request #3108 from sivarv/gcstressfix

Fix to gc-stress issue #2921.

8 years agoMake Unix event tracing conditionally enabled
Jan Kotas [Wed, 10 Feb 2016 01:15:06 +0000 (17:15 -0800)]
Make Unix event tracing conditionally enabled

Add condition around the expensive event tracing code to enable it conditionally only if the event tracing is enabled

8 years agoMerge pull request #3106 from briansull/issue_3032
Brian Sullivan [Wed, 10 Feb 2016 00:53:57 +0000 (16:53 -0800)]
Merge pull request #3106 from briansull/issue_3032

ARM64: For methods that have very large stack frames we need extra instructions

8 years agoFor methods that have very large stack frames we need an extra instruction
Brian Sullivan [Tue, 9 Feb 2016 22:07:39 +0000 (14:07 -0800)]
For methods that have very large stack frames we need an extra instruction
to load the offset when encoding the prolog/epilog instructions for funclets

Fixes Issue #3032
Incorporates code review feedback and provides a method called
genInstrWithConstant to handle the cases where we need extra instructions
Updated arm64\Tests.lst with additional tags:
MISSING_EXE, ISSUE_3104, ISSUE_3105 and NEED_TRIAGE

8 years agoFix to gc-stress issue #2921.
sivarv [Tue, 9 Feb 2016 22:54:14 +0000 (14:54 -0800)]
Fix to gc-stress issue #2921.

8 years agoMerge pull request #3055 from wtgodbe/serverGC
William Godbe [Tue, 9 Feb 2016 22:46:15 +0000 (14:46 -0800)]
Merge pull request #3055 from wtgodbe/serverGC

Re-enable server GC & lengthen timeouts

8 years agoRe-enable previously disabled tests that are now passing on non-Windows platforms
William Godbe [Tue, 9 Feb 2016 22:45:47 +0000 (14:45 -0800)]
Re-enable previously disabled tests that are now passing on non-Windows platforms

8 years agoRe-enable server GC & lengthen timeouts
William Godbe [Sat, 6 Feb 2016 00:25:51 +0000 (16:25 -0800)]
Re-enable server GC & lengthen timeouts

8 years agoMerge pull request #3100 from AndyAyersMS/FixWarning
Jan Kotas [Tue, 9 Feb 2016 20:24:49 +0000 (12:24 -0800)]
Merge pull request #3100 from AndyAyersMS/FixWarning

Fix const string literal warning

8 years agoMerge pull request #3099 from CarolEidt/x86IntArgs
Carol Eidt [Tue, 9 Feb 2016 19:09:52 +0000 (11:09 -0800)]
Merge pull request #3099 from CarolEidt/x86IntArgs

Constant and Popped Args for x86

8 years agoMerge pull request #3102 from mmitche/add-llitchev-to-allowed
Matt Mitchell [Tue, 9 Feb 2016 18:59:11 +0000 (10:59 -0800)]
Merge pull request #3102 from mmitche/add-llitchev-to-allowed

Add llitchev to allowed

8 years agofixup
Matt Mitchell [Tue, 9 Feb 2016 18:27:11 +0000 (10:27 -0800)]
fixup

8 years agoUpdate arm triggers to be in line with others
Matt Mitchell [Tue, 9 Feb 2016 18:24:13 +0000 (10:24 -0800)]
Update arm triggers to be in line with others

8 years agoConstant and Popped Args for x86
Carol Eidt [Tue, 9 Feb 2016 16:26:08 +0000 (08:26 -0800)]
Constant and Popped Args for x86

Add support for constant arguments,
and handle the case where the callee must pop the stack.

8 years agoFix const string literal warning
Andy Ayers [Tue, 9 Feb 2016 16:46:16 +0000 (08:46 -0800)]
Fix const string literal warning

This came up during an FI from LLVM, testing LLILC, see microsoft/llvm#140 .

8 years agoMerge pull request #3096 from gkhanna79/MonoCheck
Gaurav Khanna [Tue, 9 Feb 2016 06:31:27 +0000 (22:31 -0800)]
Merge pull request #3096 from gkhanna79/MonoCheck

Check for Mono presence until CoreCLR 3090 is fixed.

8 years agoCheck for Mono presence until CoreCLR 3090 is fixed.
Gaurav Khanna [Tue, 9 Feb 2016 04:39:51 +0000 (20:39 -0800)]
Check for Mono presence until CoreCLR 3090 is fixed.

8 years agoAdd support for Floating Point registers in ucontext_t on NetBSD
Kamil Rytarowski [Tue, 9 Feb 2016 04:33:02 +0000 (05:33 +0100)]
Add support for Floating Point registers in ucontext_t on NetBSD

This code assumes x86_64 platform and modern fxsave style of FP regs.

i386 Floating Point registers ship with different flavors, what will be
handled once the 64-bit code will be finished.

8 years agoMerge pull request #3086 from gkhanna79/CIMscorlib
Matt Ellis [Tue, 9 Feb 2016 04:23:23 +0000 (20:23 -0800)]
Merge pull request #3086 from gkhanna79/CIMscorlib

Update CI to no longer request mscorlib build for osx/ubuntu on Windows.

8 years agoMerge pull request #3093 from dotnet-bot/from-tfs
Jan Kotas [Tue, 9 Feb 2016 04:00:54 +0000 (20:00 -0800)]
Merge pull request #3093 from dotnet-bot/from-tfs

Merge changes from TFS

8 years agoMerge pull request #3091 from jkotas/revert-workaround
Jan Kotas [Tue, 9 Feb 2016 03:16:50 +0000 (19:16 -0800)]
Merge pull request #3091 from jkotas/revert-workaround

Revert "Add un-prefixed signatures as temporary workaround"

8 years agoMerge pull request #3084 from AndyAyersMS/InlineRefactor2b
Andy Ayers [Tue, 9 Feb 2016 03:07:30 +0000 (19:07 -0800)]
Merge pull request #3084 from AndyAyersMS/InlineRefactor2b

Inline refactor: revise checks in fgFindJumpTargets

8 years agoMerge pull request #2905 from AndyAyersMS/NewILTests
Andy Ayers [Tue, 9 Feb 2016 03:07:12 +0000 (19:07 -0800)]
Merge pull request #2905 from AndyAyersMS/NewILTests

Some tests for caller/callee type mismatches

8 years agoMerge pull request #3083 from AndyAyersMS/InlineRefactor2a
Andy Ayers [Tue, 9 Feb 2016 01:56:43 +0000 (17:56 -0800)]
Merge pull request #3083 from AndyAyersMS/InlineRefactor2a

Inliner refactoring: inline impAbortInline

8 years agoMerge pull request #3087 from sivarv/master
Sivarv [Tue, 9 Feb 2016 01:35:59 +0000 (17:35 -0800)]
Merge pull request #3087 from sivarv/master

Fix to SquareRootUInt16 test failure under JitStress=1/2

8 years agoMerge pull request #3067 from erozenfeld/MultToShiftVN
Eugene Rozenfeld [Mon, 8 Feb 2016 23:37:52 +0000 (15:37 -0800)]
Merge pull request #3067 from erozenfeld/MultToShiftVN

Preserve value numbers when morphing multiplication into shift.

8 years agoUpdate CI to no longer request mscorlib build for osx/ubuntu on Windows.
Gaurav Khanna (CLR) [Mon, 8 Feb 2016 20:16:01 +0000 (12:16 -0800)]
Update CI to no longer request mscorlib build for osx/ubuntu on Windows.

8 years agoRevert "Add un-prefixed signatures as temporary workaround"
Jan Kotas [Mon, 8 Feb 2016 22:32:32 +0000 (14:32 -0800)]
Revert "Add un-prefixed signatures as temporary workaround"

This reverts commit fb80bad2ed19970472ddefe539520abef42a52d0.

8 years agoMerge pull request #2981 from kouvel/Fixes
Koundinya Veluri [Mon, 8 Feb 2016 22:13:12 +0000 (14:13 -0800)]
Merge pull request #2981 from kouvel/Fixes

Miscellaneous fixes

8 years agoPreserve value numbers when morphing multiplication into shift.
Eugene Rozenfeld [Sat, 6 Feb 2016 07:09:43 +0000 (23:09 -0800)]
Preserve value numbers when morphing multiplication into shift.

This change fixes a morpher transformation of multiplication to shift
to preserve the value number of the tree since the new shift expression
will compute the same value as the old multiplication expression.
Without that change we were getting  asserts in fgMoveOpsLeft,
which expects to see value numbers on trees it transforms.

Closes #2920.

8 years agoFix to SquareRootUInt16 test failure under JitStress=1/2
sivarv [Mon, 8 Feb 2016 21:03:11 +0000 (13:03 -0800)]
Fix to SquareRootUInt16 test failure under JitStress=1/2

JitStress=1/2 is inducing GS check as a result of which the Vector argument
gets shadow copied.  But on the shadow copy lcl var SIMD specific state
is not set and as a result of which rationalization doesn't happen
properly leading to silent bad codegen.

8 years agoInline refactor: revise checks in fgFindJumpTargets
Andy Ayers [Mon, 8 Feb 2016 20:23:57 +0000 (12:23 -0800)]
Inline refactor: revise checks in fgFindJumpTargets

In this method, the inliner will check for unsupported opcodes and
similar things. Refactor these checks to directly bail out if an
issue is discovered, rather than going to a common handler snippet.

There are two functionality changes:

First, if an unsupported opcode is found (currently: switch, jmp,
calli, localloc, cpblk, mkrefany, rethrow) mark the inlinee as
never inlinable, rather than just failing the current attempt.
This should not change externally visible behavior but will improve
jit TP marginally if such cases are common.

Second, for NETCF quirk handling, make sure to guard some of the
observations with a check that we're reading the IL for inlining.

8 years agoMerge pull request #2898 from adityamandaleeka/remove_sigprocmask
Aditya Mandaleeka [Mon, 8 Feb 2016 19:45:39 +0000 (11:45 -0800)]
Merge pull request #2898 from adityamandaleeka/remove_sigprocmask

Use pthread_sigmask to change the mask of blocked signals.

8 years agoInliner refactoring: inline impAbortInline
Andy Ayers [Mon, 8 Feb 2016 19:42:42 +0000 (11:42 -0800)]
Inliner refactoring: inline impAbortInline

Inline this helper method at its 3 call sites, and delete it.

8 years agoFix build breaks in the CTP JIT.
Pat Gavlin [Mon, 8 Feb 2016 18:53:02 +0000 (10:53 -0800)]
Fix build breaks in the CTP JIT.

These breaks were due to unguarded usage of constants that are not present in the CTP JIT/EE interface.

[tfs-changeset: 1573367]

8 years agoMerge pull request #3050 from sejongoh/add_more_stress_modes_to_corefx_windows_testing
Matt Mitchell [Mon, 8 Feb 2016 18:50:27 +0000 (10:50 -0800)]
Merge pull request #3050 from sejongoh/add_more_stress_modes_to_corefx_windows_testing

Add more JIT stress modes to corefx testing

8 years agoMerge pull request #3043 from gkhanna79/FixNugetPkg
Gaurav Khanna [Mon, 8 Feb 2016 07:17:02 +0000 (23:17 -0800)]
Merge pull request #3043 from gkhanna79/FixNugetPkg

 Generate Microsoft.NETCore.Runtime.CoreCLR nuget packages

8 years agoChange to support Microsoft.NetCore.Runtime.CoreCLR nuget package generation.
Gaurav Khanna (CLR) [Fri, 5 Feb 2016 00:22:06 +0000 (16:22 -0800)]
Change to support Microsoft.NetCore.Runtime.CoreCLR nuget package generation.

8 years agoMerge pull request #3073 from krytarowski/netbsd-support-36
Jan Kotas [Sun, 7 Feb 2016 21:19:14 +0000 (13:19 -0800)]
Merge pull request #3073 from krytarowski/netbsd-support-36

Make missing libunwind fatal on NetBSD

8 years agoMake missing libunwind fatal on NetBSD
Kamil Rytarowski [Sun, 7 Feb 2016 17:37:33 +0000 (18:37 +0100)]
Make missing libunwind fatal on NetBSD

8 years agoMerge pull request #2973 from krytarowski/netbsd-support-25
Jan Kotas [Sun, 7 Feb 2016 15:54:15 +0000 (07:54 -0800)]
Merge pull request #2973 from krytarowski/netbsd-support-25

Disable FEATURE_HIJACK in NetBSD as unimplemented

8 years agoMerge pull request #3072 from krytarowski/netbsd-support-35
Jan Kotas [Sun, 7 Feb 2016 15:40:38 +0000 (07:40 -0800)]
Merge pull request #3072 from krytarowski/netbsd-support-35

Fix build on NetBSD: Support different versions of "struct reg"

8 years agoFix build on NetBSD: Support different versions of "struct reg"
Kamil Rytarowski [Sun, 7 Feb 2016 12:32:18 +0000 (13:32 +0100)]
Fix build on NetBSD: Support different versions of "struct reg"

FreeBSD style: ".r_rax"
NetBSD x86_64 style: ".regs[_R_RAX]"

NetBSD i386 style: "._r_eax" will be defined once the amd64 version will be
finished.

8 years agoDisable FEATURE_HIJACK in NetBSD as unimplemented
Kamil Rytarowski [Tue, 2 Feb 2016 01:37:06 +0000 (02:37 +0100)]
Disable FEATURE_HIJACK in NetBSD as unimplemented

Solution suggested by Jan Kotas (Microsoft).

8 years agoMerge pull request #3071 from dotnet/revert-3062-instrumentation
Jan Kotas [Sun, 7 Feb 2016 05:52:43 +0000 (21:52 -0800)]
Merge pull request #3071 from dotnet/revert-3062-instrumentation

Revert "Temporary instrumentation in native image loading"

8 years agoMerge pull request #3070 from jkotas/crossgen-fix
Jan Kotas [Sun, 7 Feb 2016 05:52:35 +0000 (21:52 -0800)]
Merge pull request #3070 from jkotas/crossgen-fix

Fix loading of native images on Unix

8 years agoRevert "Temporary instrumentation in native image loading"
Jan Kotas [Sun, 7 Feb 2016 03:48:33 +0000 (19:48 -0800)]
Revert "Temporary instrumentation in native image loading"

8 years agoFix loading of native images on Unix
Jan Kotas [Sun, 7 Feb 2016 03:40:58 +0000 (19:40 -0800)]
Fix loading of native images on Unix

R2R images that were compiled from architecture-specific IL images failed to load with bad image format exception on Unix. The fix is to take IMAGE_FILE_MACHINE_NATIVE_OS_OVERRIDE into account for the platform check.

8 years agoMerge pull request #3066 from gkhanna79/Fix3049
Jan Kotas [Sun, 7 Feb 2016 00:33:29 +0000 (16:33 -0800)]
Merge pull request #3066 from gkhanna79/Fix3049

Fix for CoreCLR issue 3049

8 years agoFix for CoreCLR issue 3049
Gaurav Khanna [Sat, 6 Feb 2016 22:10:46 +0000 (14:10 -0800)]
Fix for CoreCLR issue 3049

8 years agoMerge pull request #3062 from jkotas/instrumentation
Jan Kotas [Sat, 6 Feb 2016 21:37:32 +0000 (13:37 -0800)]
Merge pull request #3062 from jkotas/instrumentation

Temporary instrumentation in native image loading

8 years agoTemporary instrumentation for PE file loading
Jan Kotas [Sat, 6 Feb 2016 07:16:30 +0000 (23:16 -0800)]
Temporary instrumentation for PE file loading

8 years agoMerge pull request #3059 from jkotas/intrinsics
Jan Kotas [Sat, 6 Feb 2016 18:14:11 +0000 (10:14 -0800)]
Merge pull request #3059 from jkotas/intrinsics

Improve robustness of intrinsic fallback path

8 years agoMerge pull request #3063 from dotnet/revert-3027-remove_security_checks
Stephen Toub [Sat, 6 Feb 2016 18:03:11 +0000 (13:03 -0500)]
Merge pull request #3063 from dotnet/revert-3027-remove_security_checks

Revert "Remove useless security checks"

8 years agoMerge pull request #3061 from stephentoub/update_index
Jan Kotas [Sat, 6 Feb 2016 16:51:57 +0000 (08:51 -0800)]
Merge pull request #3061 from stephentoub/update_index

Tweak Enum.TryParse whitespace skipping slightly

8 years agoRevert "Remove useless security checks"
Jan Kotas [Sat, 6 Feb 2016 16:45:25 +0000 (08:45 -0800)]
Revert "Remove useless security checks"

8 years agoMerge pull request #3060 from dotnet-bot/from-tfs
Jan Kotas [Sat, 6 Feb 2016 15:45:09 +0000 (07:45 -0800)]
Merge pull request #3060 from dotnet-bot/from-tfs

Merge changes from TFS

8 years agoTweak Enum.TryParse whitespace skipping slightly
stephentoub [Sat, 6 Feb 2016 15:04:43 +0000 (10:04 -0500)]
Tweak Enum.TryParse whitespace skipping slightly

As a simple addendum to my previous Enum.TryParse change, if the string
begins with whitespace, we already know where the whitespace ends, so we
don't need to retraverse it.

8 years agoMerge pull request #3048 from sejongoh/add_unit_test_for_issue_2580
Sejong Oh [Sat, 6 Feb 2016 10:59:01 +0000 (02:59 -0800)]
Merge pull request #3048 from sejongoh/add_unit_test_for_issue_2580

Add unit test for issue 2580

8 years agoAdd corefx baseline
Sejong Oh [Sat, 6 Feb 2016 10:58:17 +0000 (02:58 -0800)]
Add corefx baseline

8 years agoAdd unit test for issue 2580
Sejong Oh [Fri, 5 Feb 2016 21:07:57 +0000 (13:07 -0800)]
Add unit test for issue 2580

- The test checks if code generation for ushort is correct.

8 years agoDelete mdil support from coreclr
Jan Kotas [Sat, 6 Feb 2016 08:10:40 +0000 (00:10 -0800)]
Delete mdil support from coreclr

[tfs-changeset: 1573006]

8 years agoImprove robustness of intrinsic fallback path
Jan Kotas [Sat, 6 Feb 2016 07:09:29 +0000 (23:09 -0800)]
Improve robustness of intrinsic fallback path

- The fallback path for intrinsics nodes is assumed to be direct call. Verify this assumption using no_way assert instead of regular assert.
- Make VM to always tell the JIT to use direct calls for intrinsics, even in version resilient code

8 years agoChange corefx jit stress scenario names due to git complaining too long
Sejong Oh [Sat, 6 Feb 2016 05:19:39 +0000 (21:19 -0800)]
Change corefx jit stress scenario names due to git complaining too long
file name

8 years agoMerge pull request #3056 from krytarowski/netbsd-support-34
Jan Kotas [Sat, 6 Feb 2016 04:06:47 +0000 (20:06 -0800)]
Merge pull request #3056 from krytarowski/netbsd-support-34

Fix NetBSD build: Define virtual destructors when needed