platform/upstream/coreclr.git
6 years agoImprove throughput of {U}Int32/64.{Try}Parse for hex (#18952)
Stephen Toub [Tue, 17 Jul 2018 13:00:11 +0000 (06:00 -0700)]
Improve throughput of {U}Int32/64.{Try}Parse for hex (#18952)

* Improve throughput of {U}Int32/64.{Try}Parse for hex

- Copies/modifies the recently added TryParse32/64IntegerStyle routines for hex in order to provide more optimized parsing.
- Since AllowHexSpecifier can't be used with any other styles (other than allowing leading/trailing whitespace), deletes existing code paths for parsing hex.
- As I was copying code around, I kept getting bitten by different names used for the same thing, so I also standardized some parameter names in various helper functions.

* Clean up

- Remove a bunch of unnecessary 'unsafe' keywords
- Rephrase some while(true) loops as do/while loops for consistency and to avoid some repetition
- Remove an unnecessary wrapper function

6 years agoUpdate BuildTools, CoreClr, CoreFx, PgoData to preview1-03013-03, preview1-26716...
dotnet-maestro-bot [Tue, 17 Jul 2018 11:27:23 +0000 (04:27 -0700)]
Update BuildTools, CoreClr, CoreFx, PgoData to preview1-03013-03, preview1-26716-04, preview1-26717-02, master-20180716-0324, respectively (master) (#18927)

* Update BuildTools, CoreClr, CoreFx, PgoData to preview1-03013-03, preview1-26716-04, preview1-26717-02, master-20180716-0324, respectively

* Update EventPipe StackWalker usage to specify ALLOW_INVALID_OBJECTS.

6 years agoDelete redundant broken offsetof definitions (#18947)
Jan Kotas [Tue, 17 Jul 2018 08:51:57 +0000 (01:51 -0700)]
Delete redundant broken offsetof definitions (#18947)

Fixes build breaks with latest Visual Studio Preview.

Fixes #18338

6 years agoApply tiering's call counting delay more broadly (#18610)
Koundinya Veluri [Tue, 17 Jul 2018 05:04:07 +0000 (22:04 -0700)]
Apply tiering's call counting delay more broadly (#18610)

Apply tiering's call counting delay more broadly

Issues
- When some time passes between process startup and first significant use of the app, startup perf with tiering can be slower because the call counting delay is no longer in effect
- This is especially true when the process is affinitized to one cpu

Fixes
- Initiate and prolong the call counting delay upon tier 0 activity (jitting or r2r code lookup for a new method)
- Stop call counting for a called method when the delay is in effect
- Stop (and don't start) tier 1 jitting when the delay is in effect
- After the delay resume call counting and tier 1 jitting
- If the process is affinitized to one cpu at process startup, multiply the delay by 10

No change in benchmarks.

6 years agoJIT: fix value numbering to handle GT_NULLCHECK more generally (#18942)
Andy Ayers [Tue, 17 Jul 2018 01:48:24 +0000 (18:48 -0700)]
JIT: fix value numbering to handle GT_NULLCHECK more generally (#18942)

With the advent of #18819 we may now see GT_NULLCHECK nodes with operands
that also can cause exceptions.

Handle this in value numbering.

Closes #18937.

6 years agoAdding tests for the Avx2 AlignRight, Blend, and ConvertTo* hwintrinsics.
Tanner Gooding [Sun, 15 Jul 2018 15:04:45 +0000 (08:04 -0700)]
Adding tests for the Avx2 AlignRight, Blend, and ConvertTo* hwintrinsics.

6 years agoImplementing the Avx2 AlignRight, Blend, and ConvertTo* hwintrinsics.
Tanner Gooding [Thu, 12 Jul 2018 14:40:30 +0000 (07:40 -0700)]
Implementing the Avx2 AlignRight, Blend, and ConvertTo* hwintrinsics.

6 years agoMerge pull request #18917 from BruceForstall/dev/unix_test_workflow
Bruce Forstall [Mon, 16 Jul 2018 21:41:19 +0000 (14:41 -0700)]
Merge pull request #18917 from BruceForstall/dev/unix_test_workflow

Enable Windows ARM64 corefx jobs in the CI

6 years agoAdd no tiered compilation test jobs (#18864)
Noah Falk [Mon, 16 Jul 2018 20:48:37 +0000 (13:48 -0700)]
Add no tiered compilation test jobs (#18864)

* Add no tiered compilation test jobs

With the upcoming switch to make tiered compilation on by default we want to preserve some JIT testing that will not enable tiered compilation.
1) Add 'no_tiered_compilation_pri0' jit stress job, logically the non-tiered variant of current 'Innerloop' jobs
   - This job will run on all PRs for Checked x (Windows x86, Windows x64, Ubuntu arm) - but not scheduled that way for the moment so it can be tested post-checkin
2) Add 'no_tiered_compilation' jit stress job, logically the non-tiered variant of the current 'normal' job
   - This job has the same periodic schedule as other jit stress jobs
3) Add 'r2r_no_tiered_compilation' r2r stress job, logically the non-tiered variant of the current 'r2r' job
   - This job has the same periodic schedule as other r2r stress jobs

Misc adjustments - Added isPri0TestScenario because 'innerloop' is no longer the only scenario name which has this property + eliminated various comments that implied 'innerloop' IFF pri0

6 years agoFixing markdown list formatting (#18936)
Matt Warren [Mon, 16 Jul 2018 18:50:49 +0000 (19:50 +0100)]
Fixing markdown list formatting (#18936)

6 years agoFix disabled test name (#18932)
Jan Kotas [Mon, 16 Jul 2018 18:40:26 +0000 (11:40 -0700)]
Fix disabled test name (#18932)

System.ComponentModel.Composition.Tests use friends visibility. The fixed drop does not work well for testing of the live bits.

6 years agoImprove narrowing of GT_AND nodes. (#18916)
Eugene Rozenfeld [Mon, 16 Jul 2018 18:24:54 +0000 (11:24 -0700)]
Improve narrowing of GT_AND nodes. (#18916)

This is a follow-up to to #18816 which resulted in a 6 byte regression in one of the
desktop SuperPMI methods. This change removes that regression and adds a number
of improved diffs.

If we are narrowing GT_AND to an unsigned type and one of the operands can be narrowed
into that type, the result of the GT_AND will also fit into that type and can be narrowed.
The same is true if one of the operands is an int const and can be narrowed into 'dsst'.

The change also ensures that we don't call optNarrowTree(false) more than once on each of the
GT_AND operands.

6 years agoMark RuntimeEventSourceTest as GCStressIncompatible (#18911)
Egor Chesakov [Mon, 16 Jul 2018 16:41:25 +0000 (09:41 -0700)]
Mark RuntimeEventSourceTest as GCStressIncompatible (#18911)

6 years agoImprove throughput of UInt32/UInt64 integer parsing (#18930)
Stephen Toub [Mon, 16 Jul 2018 13:59:55 +0000 (06:59 -0700)]
Improve throughput of UInt32/UInt64 integer parsing (#18930)

As was recently done for Int32/Int64, ports the Utf8Parser approach to parsing to UInt32/UIn64.{Try}Parse, specifically for NumberStyles.Integer (the default).

Also fixes an issue discovered in the previous Int32/Int64 changes, where if the input both has an overflow and has a formatting error (e.g. Int32.Parse("12345678910blah"), we would end up throwing whichever error was hit first, which is a change from .NET Core 2.1 and netfx.  The FormatException needs to be preferred over the OverflowException, which just means we can't bail early when overflow is detected.

6 years agoCoreFX test fixes (#18929)
Jan Kotas [Mon, 16 Jul 2018 13:45:17 +0000 (06:45 -0700)]
CoreFX test fixes (#18929)

- Unify Windows and Unix exclusion files to avoid duplication
- Disable outdated tests against live CoreFX

6 years agoRemove unused resource strings from System.Private.CoreLib (#18808)
Stephen Toub [Sun, 15 Jul 2018 19:44:54 +0000 (15:44 -0400)]
Remove unused resource strings from System.Private.CoreLib (#18808)

6 years agoExpose SpinWait.SpinOnce(int sleep1Threshold) overload (#18204)
Koundinya Veluri [Sun, 15 Jul 2018 18:18:06 +0000 (11:18 -0700)]
Expose SpinWait.SpinOnce(int sleep1Threshold) overload (#18204)

Expose SpinWait.SpinOnce(int sleep1Threshold) overload

To allow customizing the spin count threshold for Sleep(1) usage, and to allow disabling the use of Sleep(1).

API review: https://github.com/dotnet/corefx/issues/29623
Part of fix for https://github.com/dotnet/corefx/issues/29595

6 years agoTemporarily allow combase.dll!RoGetActivationFactory (#18922)
Stephen Toub [Sun, 15 Jul 2018 17:39:12 +0000 (13:39 -0400)]
Temporarily allow combase.dll!RoGetActivationFactory (#18922)

6 years agoRevert "Fix some failures when using build-test.sh"
Russ Keldorph [Sat, 14 Jul 2018 23:44:49 +0000 (16:44 -0700)]
Revert "Fix some failures when using build-test.sh"

This reverts commit 2bffb76443919b0ad0be8bff398d9a962a3a5f6d.

6 years agoFix some failures when using build-test.sh
russ [Fri, 6 Jul 2018 21:49:47 +0000 (21:49 +0000)]
Fix some failures when using build-test.sh

1) Remove a workaround for broken msbuild RelativePath API that is
   apparently no longer necessary and doesn't seem to work anymore anyway.
2) Fix a casing issue in some test projects.  I'm not sure why this hasn't
   been fixed yet.

6 years agoRemove version increment from Dictionary<K,V>.Remove overloads (#18854)
Wraith2 [Sat, 14 Jul 2018 19:13:38 +0000 (20:13 +0100)]
Remove version increment from Dictionary<K,V>.Remove overloads (#18854)

* remove version increment from Remove overloads

* add temporary test exclusions

* fixup test exclusion names

* modify test exclusions

6 years agoEnable 'jitnosimd' job for Linux/arm64 (#18890)
Bruce Forstall [Thu, 12 Jul 2018 20:34:06 +0000 (13:34 -0700)]
Enable 'jitnosimd' job for Linux/arm64 (#18890)

Fixes #18174

6 years agoEnable Windows ARM64 corefx jobs
Bruce Forstall [Wed, 11 Jul 2018 18:32:41 +0000 (11:32 -0700)]
Enable Windows ARM64 corefx jobs

Disable various tests that currently fail, tagged with issues
describing the failures.

6 years agoImprove Int32/Int64 {Try}Parse performance for NumberStyles.Integer (#18897)
Stephen Toub [Fri, 13 Jul 2018 21:43:34 +0000 (17:43 -0400)]
Improve Int32/Int64 {Try}Parse performance for NumberStyles.Integer (#18897)

Special-case the default NumberStyles.Integer format to improve the throughput of Int32/Int64.{Try}Parse.  Ports the implementation from Utf8Parser, modifying it to accomodate the additional cases needed for Int32/Int64, e.g. culture impact on positive and negative signs, trailing zeros, leading and trailing whitespace, prohibiting additional characters at the end, etc.

6 years agoJIT: optimize some cases of unused structs (#18819)
Andy Ayers [Fri, 13 Jul 2018 20:21:27 +0000 (13:21 -0700)]
JIT: optimize some cases of unused structs (#18819)

In some cases CSC will use `ldobj; pop` to null check a pointer to
struct. This change avoids copying the struct value for such constructs.

Codegen may still redundantly null check, if there are multiple such checks
in a method.

Fixes #18710

6 years agoRemove unnecessary R2R jobs (#18902)
Bruce Forstall [Fri, 13 Jul 2018 18:02:48 +0000 (11:02 -0700)]
Remove unnecessary R2R jobs (#18902)

6 years agoEnable logs (#18881)
Andon Andonov [Fri, 13 Jul 2018 16:57:38 +0000 (09:57 -0700)]
Enable logs (#18881)

6 years agoRemove System.Security.Cryptography.X509Certificates.Tests from Windows arm exclusion...
Egor Chesakov [Fri, 13 Jul 2018 16:10:25 +0000 (09:10 -0700)]
Remove System.Security.Cryptography.X509Certificates.Tests from Windows arm exclusion list (#18807)

6 years agoAdd prefix to DAC's PAL exports for alpine (#18873)
Mike McLaughlin [Fri, 13 Jul 2018 06:45:39 +0000 (23:45 -0700)]
Add prefix to DAC's PAL exports for alpine (#18873)

Added some cmake logic to create assembly include mapping files. One that maps the
prefixed name (DAC_foo) to the actual name (foo) which is included in the DAC module
and another that maps the actual name to the prefixed name that is included in the SOS,
DBI and createdump modules.

The data exports like IID_IUnknown are not prefixed and don't need to be (immutable static data).

There were some C++ exports functions exported with their decorated names in
the CatchHardwareExceptionHolder and NativeExceptionHolderBase classes. Created
PAL_* style export functions that implements the code.

Fix lldb plugin cmake file to use LLDB_H/LLDB_LIB env vars to build it.

6 years agoDisable 3 long-running tests and 1 failing test (#18896)
Bruce Forstall [Fri, 13 Jul 2018 05:31:48 +0000 (22:31 -0700)]
Disable 3 long-running tests and 1 failing test (#18896)

The following tests exceed the default timeout, especially on Debug builds:
```
GC\Features\LOHCompaction\lohcompactapi\lohcompactapi.cmd
GC\API\NoGCRegion\NoGC\NoGC.cmd
GC\API\GC\GetAllocatedBytesForCurrentThread\GetAllocatedBytesForCurrentThread.cmd
```

Mark them LONG_RUNNING.

The following test asserts in the JIT currently:
```
JIT\HardwareIntrinsics\Arm64\Simd\Simd.cmd
```

Disable it with issue #18895

6 years agoRemove dead/unreachable code from ARM32 and ARM64 emitters (#18878)
Egor Chesakov [Fri, 13 Jul 2018 00:41:17 +0000 (17:41 -0700)]
Remove dead/unreachable code from ARM32 and ARM64 emitters (#18878)

6 years agoMerge pull request #18887 from briansull/atomic-fix
Brian Sullivan [Fri, 13 Jul 2018 00:12:52 +0000 (17:12 -0700)]
Merge pull request #18887 from briansull/atomic-fix

Include the processor atomic operations in the OperIsStore category

6 years agoFixes for bugs in fgMorphCast and optNarrowTree. (#18816)
Eugene Rozenfeld [Thu, 12 Jul 2018 20:42:00 +0000 (13:42 -0700)]
Fixes for bugs in fgMorphCast and optNarrowTree. (#18816)

The fix under NARROW_IND prevents transformation of, e.g.,
CAST      int <- ushort <- int
    CLS_VAR byte

into

CLS_VAR byte.

With the fix the CAST is not removed.

The fix under GT_CAST prevents transformation of, e.g.,

CAST      int <- ushort <- long
    CAST      long <- int
       expr short

into

expt short.

With the fix it gets transformed into

CAST      int <- ushort <- int
    expr short

Block cast optimizations in fgMorphCast if the cast expression is an
active CSE candidate.

Update cast expression value numbers when a cast is removed.

Fixes #18238, #18850.

No diffs in frameworks and tests (pmi and crossgen, x64 and x86), except for the added test cases.

6 years agoAdded the repro test case
Brian Sullivan [Thu, 12 Jul 2018 19:49:08 +0000 (12:49 -0700)]
Added the repro test case

6 years agoInclude the processor atomic operations in the OperIsStore category
Brian Sullivan [Thu, 12 Jul 2018 18:11:31 +0000 (11:11 -0700)]
Include the processor atomic operations in the OperIsStore category
THe atomic operations are GT_CMPXCHG, GT_LOCKADD, GT_XCHG, GT_XADD

6 years agoDon't trim compiler-generated anonymous types. (#18835)
Brian Robbins [Thu, 12 Jul 2018 17:26:12 +0000 (10:26 -0700)]
Don't trim compiler-generated anonymous types. (#18835)

6 years agoCoreFX CI Unix (#18753)
Andon Andonov [Thu, 12 Jul 2018 04:39:57 +0000 (21:39 -0700)]
CoreFX CI Unix (#18753)

* Full Unix test support

* Remove Test helper build from build-test

Add msg prefix

Change if condition

* Change Linux URL

* Syntax error

* Fix Merge Conflict

* Address PR Feedback

* Remove duplicate generate_testhost call

* Address PR comments

* Clean up exclusion list

* Disable Client_ReadWriteCancelledToken_Throws_OperationCanceledException

* Disable all flavors of CancelledToken_Throws_OperationCanceledException

* Disable X509StoreTests.Constructor_DefaultStoreName - flaky on OSX

* Update documentation with Linux instructions and workflow

* Disable EventWaitHandleTests.Ctor_InvalidMode - failing on OSX

* Disable Server_ReadWriteCancelledToken_Throws_OperationCanceledException

6 years agoFix VEX.vvvv encoding for AVX.BlendVariable
Fei Peng [Wed, 11 Jul 2018 22:55:37 +0000 (15:55 -0700)]
Fix VEX.vvvv encoding for AVX.BlendVariable

6 years agoRemove TODO comment from CancellationTokenRegistration.Unregister (#18880)
Stephen Toub [Thu, 12 Jul 2018 02:30:56 +0000 (22:30 -0400)]
Remove TODO comment from CancellationTokenRegistration.Unregister (#18880)

6 years agoMerge pull request #18862 from BruceForstall/DisableLongRunningGCStress3Tests
Bruce Forstall [Thu, 12 Jul 2018 00:17:50 +0000 (17:17 -0700)]
Merge pull request #18862 from BruceForstall/DisableLongRunningGCStress3Tests

Disable tests for Windows/arm64 with GCStress=3 that take too long

6 years agoMerge pull request #18859 from hoyosjs/fix-debug-gscookie-check
Juan Hoyos [Wed, 11 Jul 2018 22:36:41 +0000 (15:36 -0700)]
Merge pull request #18859 from hoyosjs/fix-debug-gscookie-check

GS cookie check fix for debugger stackwalks port

6 years agoOptimize Double/Single.IsNegative
Jan Kotas [Wed, 11 Jul 2018 06:46:39 +0000 (23:46 -0700)]
Optimize Double/Single.IsNegative

Signed-off-by: dotnet-bot <dotnet-bot@microsoft.com>
6 years agoMake Number.Parsing.cs BigEndian friendly (#18853)
Egor Bogatov [Wed, 11 Jul 2018 06:04:06 +0000 (09:04 +0300)]
Make Number.Parsing.cs BigEndian friendly (#18853)

* Make Number.Parsing.cs BigEndian friendly

* Get rid of DoubleHelper

6 years agoMerge pull request #18861 from BruceForstall/FixShowLocalloc
Bruce Forstall [Wed, 11 Jul 2018 03:55:39 +0000 (20:55 -0700)]
Merge pull request #18861 from BruceForstall/FixShowLocalloc

Fix ShowLocallocAlignment test for x86

6 years agoMerge pull request #18860 from BruceForstall/ReEnable13731Tests
Bruce Forstall [Wed, 11 Jul 2018 03:55:02 +0000 (20:55 -0700)]
Merge pull request #18860 from BruceForstall/ReEnable13731Tests

Re-enable tests for #13731 which was closed long ago

6 years agoAdjust JIT automated testing for tiered compilation (#18828)
Noah Falk [Wed, 11 Jul 2018 01:33:57 +0000 (18:33 -0700)]
Adjust JIT automated testing for tiered compilation (#18828)

Make it explicit that most JIT testing does not use tiered compilation regardless of the runtime default
Clone jitstress1/2 and r2r_jitstress1/2 test modes both with and without tiering.
Move the existing tieredcompilation test modes to have the same periodic scheduling as other jit test runs.

6 years agoDisable tests for Windows/arm64 with GCStress=3 that take too long
Bruce Forstall [Wed, 11 Jul 2018 00:39:37 +0000 (17:39 -0700)]
Disable tests for Windows/arm64 with GCStress=3 that take too long

Fixes #15381

6 years agoFix ShowLocallocAlignment test for x86
Bruce Forstall [Tue, 10 Jul 2018 23:50:54 +0000 (16:50 -0700)]
Fix ShowLocallocAlignment test for x86

The required localloc alignment differs by platform. Teach the test
what the per-platform alignment requirement is.

Fixes #7163

6 years agoRe-enable tests for #13731 which was closed long ago
Bruce Forstall [Tue, 10 Jul 2018 23:22:48 +0000 (16:22 -0700)]
Re-enable tests for #13731 which was closed long ago

6 years agoRemove random.NextDouble() from JIT/HardwareIntrinsics tests (#18838)
Egor Chesakov [Tue, 10 Jul 2018 23:15:15 +0000 (16:15 -0700)]
Remove random.NextDouble() from JIT/HardwareIntrinsics tests (#18838)

* Add ProjectReference to CoreCLRTestLibrary.csproj in the corresponding C# project files
* Replace usages of random.NextDouble() from GenerateTests.csx with CoreCLRTestLibrary.Generator.GetSingle()/GetDouble()
* Regenerate test files

6 years agoInitial implementation of method fixup parser (#18749)
Tomáš Rylek [Tue, 10 Jul 2018 21:48:25 +0000 (23:48 +0200)]
Initial implementation of method fixup parser (#18749)

* Initial implementation of method fixup parser

I'm working on this right now in the CoreRT R2R compiler so I'm
adding the relevant structures and algorithms to the dumper
to let me analyze various R2R files and their method fixups.

I have locally verified that this shows reasonable fixup info
for my Hello World repro compiled both using the CoreRT R2R
compiled and using crossgen.

Thanks

Tomas

* Improvements in signature dumping

Signature dumping was slightly imprecise - the signature field
was named "Section" in the dump and it was displaying a dword,
however signatures are logically arbitrary byte streams.

Sadly I'm aware of no easy way to detect the length of an
arbitrary signature without adding a multitude of dedicated
parsers for the different signature types used by different
fixup / helper types so for now I'm displaying an 8-byte sample.

I have also made the formatting of the import section entries
more compact to make the list easier to read in the presence
of a larger number of fixups and I added support for displaying
the entry offset (as this is what the method fixups indirectly
refer to via the fixup indices).

Thanks

Tomas

* Move NibbleReader into a separate source file

Based on Amy's PR feedback I'm moving NibbleReader into a separate
source file. After all, I have a counterpart NibbleWriter.cs
in the R2R compiler codebase.

Thanks

Tomas

* Remove unnecessary usings from the new NibbleReader.cs file

As an additional bit of cleanup I used the VS refactoring tool
to remove unnecessary usings from the new NibbleReader source file.

Thanks

Tomas

6 years agoFixing encodeXmmRegAsIval to ensure the result meets the "fits in imm8" check
Tanner Gooding [Tue, 10 Jul 2018 04:46:41 +0000 (21:46 -0700)]
Fixing encodeXmmRegAsIval to ensure the result meets the "fits in imm8" check

6 years agoMerge pull request #18805 from CarolEidt/Fix18746
Carol Eidt [Tue, 10 Jul 2018 20:33:02 +0000 (13:33 -0700)]
Merge pull request #18805 from CarolEidt/Fix18746

Don't change PUTARG_REG spill type on arm64

6 years agoFix two issues in Marshal.SecureStringToBSTR (#18851)
Stephen Toub [Tue, 10 Jul 2018 20:02:06 +0000 (16:02 -0400)]
Fix two issues in Marshal.SecureStringToBSTR (#18851)

1. It wasn't synchronized, even though other operations on SecureString are in order to avoid concurrency problems leading to memory corruption issues (and it's synchronized on netfx).

2. It wasn't throwing an ObjectDisposedException if the instance was disposed.

6 years agoMerge pull request #18812 from briansull/fix_610014
Brian Sullivan [Tue, 10 Jul 2018 20:01:41 +0000 (13:01 -0700)]
Merge pull request #18812 from briansull/fix_610014

Fix 610014

6 years agoMerge pull request #18653 from acmyu/build
Amy [Tue, 10 Jul 2018 18:10:08 +0000 (11:10 -0700)]
Merge pull request #18653 from acmyu/build

R2RDump - Build integration

6 years agoGS cookie check fix for debugger stackwalks port
Juan Sebastian Hoyos Ayala [Mon, 9 Jul 2018 21:28:57 +0000 (14:28 -0700)]
GS cookie check fix for debugger stackwalks port

This bug fix is a port from the equivalent fix in framework. The
debugger tried performing a stackwalk in the epilog due to the JIT
incorrectly reporting epilogue information. This caused an invalid
GS cookie to be checked and caused the debugger to crash. A flag was
added to allow debug stackwalks to skip the cookie check.

6 years agoRegenerating the x86 HWIntrinsic tests to cover struct field scenarios
Tanner Gooding [Tue, 10 Jul 2018 01:42:01 +0000 (18:42 -0700)]
Regenerating the x86 HWIntrinsic tests to cover struct field scenarios

6 years agoUpdating the x86 HWIntrinsic test templates to test struct field scenarios
Tanner Gooding [Tue, 10 Jul 2018 01:41:28 +0000 (18:41 -0700)]
Updating the x86 HWIntrinsic test templates to test struct field scenarios

6 years agoMove BinaryReader to shared CoreLib partition (#18845)
Jan Kotas [Tue, 10 Jul 2018 11:13:04 +0000 (04:13 -0700)]
Move BinaryReader to shared CoreLib partition (#18845)

6 years agoExclude failing CoreFx test (#18844)
Amy [Tue, 10 Jul 2018 08:49:54 +0000 (01:49 -0700)]
Exclude failing CoreFx test (#18844)

6 years agoUpdated fix for VSO 610014
Brian Sullivan [Fri, 6 Jul 2018 21:33:36 +0000 (14:33 -0700)]
Updated fix for VSO 610014

6 years agoFixing a call to emitIns_SIMD_R_R_AR_R to pass arguments in the correct order
Tanner Gooding [Sat, 7 Jul 2018 20:02:32 +0000 (13:02 -0700)]
Fixing a call to emitIns_SIMD_R_R_AR_R to pass arguments in the correct order

6 years agoExclude failing CoreFx test
Amy Yu [Mon, 9 Jul 2018 23:36:39 +0000 (16:36 -0700)]
Exclude failing CoreFx test

6 years agoUpdating InstructionSet_BMI1/2 to be controlled by the COMPlus_EnableAVX flag and...
Tanner Gooding [Sat, 7 Jul 2018 14:11:51 +0000 (07:11 -0700)]
Updating InstructionSet_BMI1/2 to be controlled by the COMPlus_EnableAVX flag and fixing formatting issues

6 years agoAdding tests for the ParallelBitDeposit and ParallelBitExtract Bmi2 HWIntrinsics
Tanner Gooding [Fri, 6 Jul 2018 16:29:33 +0000 (09:29 -0700)]
Adding tests for the ParallelBitDeposit and ParallelBitExtract Bmi2 HWIntrinsics

6 years agoImplementing the ParallelBitDeposit and ParallelBitExtract Bmi2 HWIntrinsics
Tanner Gooding [Fri, 6 Jul 2018 15:31:10 +0000 (08:31 -0700)]
Implementing the ParallelBitDeposit and ParallelBitExtract Bmi2 HWIntrinsics

6 years agoAdding tests for the AndNot, ExtractLowestSetBit, GetMaskUpToLowestSetBit, and ResetL...
Tanner Gooding [Fri, 6 Jul 2018 06:36:30 +0000 (23:36 -0700)]
Adding tests for the AndNot, ExtractLowestSetBit, GetMaskUpToLowestSetBit, and ResetLowestSetBit x86 HWIntrinsics

6 years agoImplementing the AndNot, ExtractLowestSetBit, GetMaskUpToLowestSetBit, and ResetLowes...
Tanner Gooding [Fri, 6 Jul 2018 06:34:58 +0000 (23:34 -0700)]
Implementing the AndNot, ExtractLowestSetBit, GetMaskUpToLowestSetBit, and ResetLowestSetBit x86 HWIntrinsics

6 years agoAdding tests for the x86 TrailingZeroCount HWintrinsic
Tanner Gooding [Wed, 4 Jul 2018 23:32:55 +0000 (16:32 -0700)]
Adding tests for the x86 TrailingZeroCount HWintrinsic

6 years agoImplementing the x86 TrailingZeroCount HWIntrinsic
Tanner Gooding [Wed, 4 Jul 2018 23:32:27 +0000 (16:32 -0700)]
Implementing the x86 TrailingZeroCount HWIntrinsic

6 years agoRenaming ScalarUnOpTest to ScalarSimdUnOpTest
Tanner Gooding [Wed, 4 Jul 2018 14:55:28 +0000 (07:55 -0700)]
Renaming ScalarUnOpTest to ScalarSimdUnOpTest

6 years agoFix serialization type in NotFiniteNumberException (#18833)
Viktor Hofer [Mon, 9 Jul 2018 21:41:07 +0000 (23:41 +0200)]
Fix serialization type in NotFiniteNumberException (#18833)

6 years agoFix test errors
Amy Yu [Tue, 3 Jul 2018 19:59:22 +0000 (12:59 -0700)]
Fix test errors

6 years agoUse BuildManagedTools argument instead of environment variable, removed unneeded...
Amy Yu [Tue, 3 Jul 2018 19:13:51 +0000 (12:13 -0700)]
Use BuildManagedTools argument instead of environment variable, removed unneeded flags

6 years agoAdd R2RDump to build.proj, running dotnet restore before target build
Amy Yu [Tue, 3 Jul 2018 00:45:00 +0000 (17:45 -0700)]
Add R2RDump to build.proj, running dotnet restore before target build

6 years agoCoreFX CI Unix Staging commit (#18750)
Andon Andonov [Mon, 9 Jul 2018 17:45:24 +0000 (10:45 -0700)]
CoreFX CI Unix Staging commit (#18750)

* Modify netci.groovy

* Add script exit codes

* Ad PR Triggers for Ubuntu and OSX10.12

* Remove release PR triggers

* Rename arch

Remove IsJitStressTestScenario assert

* Add correct asserts

* Reformat conditions and add triggers for Release/Checked/Debug

* Change %WORKSPACE% to \${WORKSPACE}

6 years agoFix a comment in the string.ToLower() method (#18832)
HumanEquivalentUnit [Mon, 9 Jul 2018 15:22:38 +0000 (16:22 +0100)]
Fix a comment in the string.ToLower() method (#18832)

Line 307 is lowercasing the current character, but the comment says "store the current character, upper-cased", fixed comment to say "store the current character, lower-cased".
(Equivalent and correct "upper-cased" comment is on line 252)

6 years agoUndo unnecessary BinaryReader::ReadOneChar/InternalReadOneChar split (#18830)
Jan Kotas [Mon, 9 Jul 2018 13:50:13 +0000 (06:50 -0700)]
Undo unnecessary BinaryReader::ReadOneChar/InternalReadOneChar split (#18830)

This method had misleading comment that this split is required for performance. It was the case when
InternalReadOneChar was called from other places in CoreLib. These calls do not exist anymore.

6 years agoMoved ManualResetEventSlim to shared (#18799)
Anirudh Agnihotry [Mon, 9 Jul 2018 09:21:55 +0000 (02:21 -0700)]
Moved ManualResetEventSlim to shared (#18799)

* Moved ManualResetEventSlim to shared

* covert System.int to System.Int32

6 years agoCorrect AssemblyLoadContext comment (#18809)
Rich Lander [Sun, 8 Jul 2018 21:56:41 +0000 (14:56 -0700)]
Correct AssemblyLoadContext comment (#18809)

6 years agoLastIndexOf corner case fix when span is empty (#18826)
Viktor Hofer [Sun, 8 Jul 2018 17:41:40 +0000 (19:41 +0200)]
LastIndexOf corner case fix when span is empty (#18826)

6 years agoFix Span LastIndexOf empty value handling
Viktor Hofer [Sun, 8 Jul 2018 13:41:26 +0000 (15:41 +0200)]
Fix Span LastIndexOf empty value handling

6 years agoTargetFrameworkName property initialized (#18814)
Anirudh Agnihotry [Sat, 7 Jul 2018 01:47:46 +0000 (18:47 -0700)]
TargetFrameworkName property initialized (#18814)

6 years agoImprove DateTime{Offset}.ParseExact performance for "O"/"o" roundtrip format (#18800)
Stephen Toub [Fri, 6 Jul 2018 16:12:54 +0000 (12:12 -0400)]
Improve DateTime{Offset}.ParseExact performance for "O"/"o" roundtrip format (#18800)

Ports the code used by Utf8Parser, modified to support things Utf8Parser doesn't but DateTime{Offset{.ParseExact do, such as single-digit offset hours.

6 years agoMerge pull request #18795 from BruceForstall/RemoveArmemJobs
Bruce Forstall [Fri, 6 Jul 2018 16:00:49 +0000 (09:00 -0700)]
Merge pull request #18795 from BruceForstall/RemoveArmemJobs

Remove Ubuntu/Ubuntu16.04 emulator CI jobs

6 years agoDon't change PUTARG_REG type on arm64
Carol Eidt [Fri, 6 Jul 2018 14:00:03 +0000 (07:00 -0700)]
Don't change PUTARG_REG type on arm64

Fix #18746

6 years ago[ARM] Mention a clang-5.0 issue in the documentation (#18803)
nkaretnikov2 [Fri, 6 Jul 2018 12:54:28 +0000 (15:54 +0300)]
[ARM] Mention a clang-5.0 issue in the documentation (#18803)

6 years agoMerge pull request #18775 from briansull/add-noway-assert
Brian Sullivan [Thu, 5 Jul 2018 23:03:18 +0000 (16:03 -0700)]
Merge pull request #18775 from briansull/add-noway-assert

Add noway asserts in gtNewZeroConNode and gtNewOneConNode

6 years agoRemove Ubuntu/Ubuntu16.04 emulator CI jobs
Bruce Forstall [Thu, 5 Jul 2018 22:07:14 +0000 (15:07 -0700)]
Remove Ubuntu/Ubuntu16.04 emulator CI jobs

Now that we have Ubuntu arm hardware in the lab, these are unnecessary.

The jobs have also been failing for months.

6 years agoUpdate certain Marshal APIs to match exception thrown on .NET Framework (#18791)
Aaron Robinson [Thu, 5 Jul 2018 21:29:27 +0000 (14:29 -0700)]
Update certain Marshal APIs to match exception thrown on .NET Framework (#18791)

Update certain Marshal APIs to match exception thrown on Desktop

6 years agoAdded noway asserts in gtNewZeroConNode and gtNewOneConNode
Brian Sullivan [Wed, 20 Jun 2018 23:51:15 +0000 (16:51 -0700)]
Added noway asserts in gtNewZeroConNode and gtNewOneConNode
Added source code change that was needed to service 4.7.2 in fgMorphRecursiveFastTailCallIntoLoop
Set zero/one to null after noway_assert

6 years agoMerge pull request #18793 from BruceForstall/UpdateLibUnwindIssue
Bruce Forstall [Thu, 5 Jul 2018 17:51:42 +0000 (10:51 -0700)]
Merge pull request #18793 from BruceForstall/UpdateLibUnwindIssue

Update Linux ARM libunwind instructions

6 years agoUpdate Linux ARM libunwind instructions
Bruce Forstall [Thu, 5 Jul 2018 17:50:06 +0000 (10:50 -0700)]
Update Linux ARM libunwind instructions

6 years agoEnable genFnCalleeRegArgs for Arm64 Varargs (#18714)
Jarret Shook [Thu, 5 Jul 2018 16:46:51 +0000 (09:46 -0700)]
Enable genFnCalleeRegArgs for Arm64 Varargs (#18714)

* Enable genFnCalleeRegArgs for Arm64 Varargs

Before the method would early out and incorrectly expect the usage
of all incoming arguments to be their homed stack slots. It is
instead possible for incoming arguments to be homed to different
integer registers.

The change will mangle the float types for vararg cases in the same
way that is done during lvaInitUserArgs and fgMorphArgs.

* Apply format patch

* Account for softfp case

* Address feedback

* Apply format patch

* Use standard function header for mangleVarArgsType

* Remove confusing comment

6 years agoDelete dead code (#18783)
Jan Kotas [Thu, 5 Jul 2018 06:04:02 +0000 (08:04 +0200)]
Delete dead code (#18783)

- Leftover from unobserved exception desktop quirk
- Unnecessary calls to AppDomain.IsFinalizingForUnload(). They were actually never required even with AppDomain support because of Environment.HasShutdownStarted returns true during AppDomain unload.

6 years agoImprove DateTime{Offset}.ParseExact{Multiple} performance for RFC1123 ("r") (#18771)
Stephen Toub [Thu, 5 Jul 2018 01:38:05 +0000 (21:38 -0400)]
Improve DateTime{Offset}.ParseExact{Multiple} performance for RFC1123 ("r") (#18771)

* Improve DateTime{Offset}.ParseExact{Multiple} performance for RFC1123 ("r")

Significantly improves the performance by porting and adapting the Utf8Parser code from corefx.  This optimizes for the (default) case of a DateTimeStyles.None; specifying any other style falls back to the normal parsing support, as that requires handling things such as arbitrary whitespace anywhere in the string.

* Address PR feedback

6 years agoRe-enable jit regression test DevDiv_590771 (#18724)
Andy Ayers [Wed, 4 Jul 2018 20:33:36 +0000 (13:33 -0700)]
Re-enable jit regression test DevDiv_590771 (#18724)

The issue the test was hitting was worked around in #18292.

Fixes #17967.

Also add GitHub_18522_8 to the arm/arm64 lists (from #18708).

6 years agoAdd LastIndexOf compareoptions overload (#18652)
Viktor Hofer [Wed, 4 Jul 2018 18:56:25 +0000 (20:56 +0200)]
Add LastIndexOf compareoptions overload (#18652)

* Add LastIndexOf compareoptions overload

6 years agoAdding tests for the byte overload of Ssse3.Shuffle
Tanner Gooding [Sun, 1 Jul 2018 15:06:46 +0000 (08:06 -0700)]
Adding tests for the byte overload of Ssse3.Shuffle